- Email: [email protected]
Reinforcing and other behavioral effects of nicotine
Neuroscience & BiobehavioralReviews, Vol. 5, pp. 487--495, 1981. Printed in the U.S.A.
Reinforcing and Other Behavioral Effects of Nicotine I JOHN DOUGHERTY,
DIANNE
MILLER,
GLENN
T O D D A N D H A R R Y B. K O S T E N B A U D E R
Departments o f Psychiatry and Pharmacology, University o f Kentucky Medical Center Lexington, K Y 40536 Veterans Administration Medical Center, Lexington, K Y 40511 and College o f Pharmacy, University o f Kentucky, Lexington, K Y
DOUGHERTY, J., D. MILLER, G. TODD AND H. B. KOSTENBAUDER. Reinforcing and other behavioral effects of nicotine. NEUROSCI. BIOBEHAV. REV. 5(4) 487--495, 1981.--Published findings of intravenous nicotine self-injection indicate that the reinforcing properties of nicotine are weak when the drug is made available according to continuous reinforcement (CRF) or fixed-ratio (FR) schedules. CRF self-injection rates are generally only 2-3 times saline control levels and self-injection frequency is largely insensitive to changes in unit dose. In contrast, drugs of the psychomotor stimulant, opiate, and sedative-hypnotic classes, with similar pharmacokinetic parameters, maintain much higher selfinjection rates and show systematic changes in rate with unit dose variations. Recent studies using interval and secondorder schedules of nicotine presentation have been more successful in maintaining higher rates of self-administration behavior. Systematic dose-response functions have also been found under these conditions. Food-deprivation, species and strain differences, circadian rhythms, and duration of exposure to the drug also appear to be important variables in determining self-injection rate. Finally, the rapid development of tolerance to the effects of nicotine may account for changes in the pattern of self-administration within daily sessions and the differential sensitivity of those patterns to nicotine pretreatment. Nicotine
Reinforcement
Fixed-ratio
Self-administration
N I C O T I N E has been considered the most active compound in tobacco and the major factor in the maintenance of tobacco use. According to the operant conditioning-addiction model of drug self-administration, the internal stimuli produced by the drug serve to reinforce or maintain the behavior leading to its presentation ]311. However, the role of nicotine as a maintaining stimulus in human smoking behavior has been difficult to define. The finding that smokers will not smoke non-nicotine cigarettes and will often decrease smoking rates after nicotine pretreatment offers support for nicotine as a reinforcer [ 15,19l. However, nicotine pretreatment does not always affect subsequent smoking rates, and may be aversive under some conditions [17, 24, 25]. The situation is further complicated by indications that cigarette smoking can be a strong habit in people who do not inhale the smoke, suggesting that external stimuli associated with smoking and social-learning factors are also important in the maintenance of the behavior. The possible interplay of pharmacologic and psychologic factors in smoking make the evaluation o f nicotine's reinforcing effects in humans a difficult task. If nicotine is a reinforcer, the drug should be selfadministered by animals in a manner analogous to the selfadministration by animals o f the opiate, psychomotor stimulant, and sedative-hypnotic drug classes commonly
Tolerance
used by man [11,27]. There are common features of drug self-administration behavior that indicate that the stimulus properties of the self-administered drug, and not other effects, are of primary importance in the maintenance of the behavior. These commonalities are: (1) Systematic changes in injection or response rate occurring with changes in unit dose size. (2) An injection or response rate maintained by the drug that is greater than that maintained by the vehicle alone over a substantial portion o f the dose-response function. (3) Systematic changes in injection or response rate occurring after pretreatment with the same drug or with pharmacologic agonists or antagonists. This review will critically examine the animal oral and intravenous (where the amount o f nicotine ingested or absorbed is known) nicotine self-administration literature and examine the proposition that nicotine serves as a stimulus maintaining behavior leading to its administration by determining the extent to which nicotine self-administration behavior shares the above commonalities with other drug classes. In addition, the influence that tolerance development may exert on nicotine self-administration will be discussed. The manner in which drug availability is scheduled has an important influence on self-administration behavior. Thus, this review will distinguish between studies in which
1Paper presented at a symposium entitled CNS Sites and Mechanisms of Nicotine's Psychopharmacological Effects, Meeting of the American Society for Pharmacology and Experimental Therapeutics, Rochester, MN, August 18, 1980.
487
488 one response per drug injection {continuous reinforcement schedule--CRF) was required and those in which multiple responses (intermittent reinforcement schedules) were used. CONTINUOUSREINFORCEMENTSCHEDULE(CRF) STUDIES Using CRF schedules of drug availability, the reinforcing properties of many self-injected drugs have been readily demonstrable. For example, with psychomotor stimulants, self-injection rates are substantially higher than saline, and alterations in unit dose size produce orderly biphasic (inverted U-shaped) or monophasic changes in injection rate in rats and monkeys [21, 22, 32]. However, showing that nicotine has reinforcing stimulus properties under these schedule conditions has been a difficult and frustrating task. Nicotine self-administration rates, using CRF schedules, that are significantly higher than saline have been difficult to produce, and systematic changes in injection rate with increases or decreases in unit dose have not been found with any consistency. Deneau and Inoki were the first investigators to report intravenous nicotine self-injection [4]. Rhesus monkeys received hourly automatic injections of nicotine at a unit dose of 25/~g/kg for 2-10 days before the animals began to selfinject the drug on a CRF schedule between non-contingent injections. Eventually, the automatic injections were stopped and the monkeys continued to self-administer between 28 to 68 injections per 24-hr period, usually during the light phase of the 12-hr light-dark cycle. Day-to-day variability in the frequency of injections was often as much as 100%. Although dose-response curves were not presented, Deneau and Inoki reported that injections declined to 7 per day at a unit dose of 2.0 mg/kg. Their report suggested that nicotine would be self-injected in large amounts by some of their monkeys. They did not compare nicotine self-injection rates with that of the saline vehicle. Yanagita also reported intravenous nicotine self-injection by rhesus monkeys, and that it occurred only during the daytime with 24-hr access to 20/~g/kg/inj on a CRF schedule [33,34]. One monkey increased self-injection rates about 5 times when nicotine was substituted for saline. However, when access was limited to 4 hrs daily and nicotine selfinjection rates were again compared to those of saline and to SPA (a psychomotor stimulant), nicotine self-injection rates were not different from self-injection rates of 0.25 ml/inj physiological saline, and were less than 20% of the SPA selfinjection rates. Furthermore, the nicotine injection rate did not change even when the unit dose was varied from 2.5 to 640/zg/kg. In comparison, SPA was self-injected at a rate as much as 7 times greater than saline. Nicotine self-administration studies with rats have also been difficult to interpret. Although quantitative data were not presented, Clark reported that rats preferred nicotine when given a choice between oral or IV nicotine (1-10 t~g,/kg/inj) and odal water or IV saline [3]. Hanson and coworkers have reported extensive studies of CRF intravenous nicotine self-injection in rats 112,13]. Daily differences between animals self-injecting nicotine and those self-injecting saline were small (12 nicotine vs 5 saline inj in one representative 24-hr period) but consistent, and cumulation of the number of injections across 5 days o f continuous access to the drug revealed that the average number of injections in the nicotine rats was about 0.5 per hr at a unit dose of 60 ~g/kg (base) and about 0.2 per hr in the saline group. In another experiment, 4 rats were allowed to self-inject at 60 ~.g/kg/inj
DOUGHERTY E1 A I for 150 days. During the last 30 days, the median intake was about 1.2 inj/hr, and the rate of nicotine intake increased gradually in all 4 animals over the 150-day period. Interestingly, nicotine self-injection rates in the nocturnal rats were much higher during the dark phase of the 12-hr lightdark cycle than in the light. Hanson et al. [13] also evaluated the effects of substituting saline for the nicotine injection. In 12 rats, the number of injections declined from about 0.4/hr during nicotine availability, to about 0.1/hr after 10 days of saline. The above differences between the nicotine and saline group injection rates, the decline after saline substitution, and the gradual but steady increase in nicotine intake over a 150-day period all suggest that the stimulus properties of the drug were important in maintaining self-administration behavior. However, when different unit doses were examined, the rats took the same number of injections at 7.5 ~tg/kg as they did at 60 /xg/kg. The first administration of mecamylamine (a central and peripheral nicotine antagonist) clearly altered the rate of nicotine self-injection, as anticipated. However, this pretreatment produced little effect on 5 subsequent treatments. (A study in which mecamylamine was used to antagonize the discriminative stimulus properties of nicotine [23] found no decrease in effectiveness after multiple injections.) Lang et al. [18] and Smith and Lang 126] have studied intravenous and oral nicotine self-injection in food deprived and non-deprived rats. In their 1977 study, they compared a continuous 90-hour session of CRF nicotine selfadministration (50-100 tzg/kg/inj) with the same schedule of saline availability in non-deprived animals and found no difference between the nicotine and saline groups. Fooddepriving the rats to 80% of free-feeding weight, however. increased the rate of nicotine self-administration to about 2.3 times that of saline. With limited access (2 hr/day) to nicotine, food-deprived rats self-administered the drug {100 /~g/kg/inj, presumably as the salt) at 2.7 times the saline rate in food-deprived rats. Smith and Lang's 126] later paper reported similar results when rats were allowed 28 days of continuous access to nicotine. Comparing the results obtained by Lang et al. [ 18], Smith and Lang [26], and Hanson et al. [131, a similarity in selfadministration behavior is evident. Hanson [12,13], using a roughly comparable unit dose of nicotine base 160 p.g/kg), found a 2-3-fold greater intake rate for nicotine rats relative to saline rats (both groups non-food deprived). This agrees with the 2.7 factor reported by Lang and Smith 118,26]. In nicotine self-injection studies performed in our laboratory (discussed more fully below), we found a nicotine-saline ratio of about 1.5 to 2.0 within individual non-food deprived rats given access to nicotine or saline on different days. In general, absolute rates of nicotine self-injection are highest in food-deprived animals. However. rats self-administer nicotine at a unit dose of about 100/xg/kg (salt) about 2-3 times more frequently than saline controls across a variety of experimental conditions, regardless of the deprivational state. Studies of intravenous nicotine self-injection in our laboratory have yielded results similar to that found above. Wistar and F-344 male albino rats were allowed limited access to nicotine on a CRF schedule in overnighl I2-hr sessions. Overnight sessions were selected because little nicotine was self-administered during the daylight phase of the light-dark cycle. Differences between rat strains and among individuals of the same strain were evident. Figure I
BEHAVIORAL EFFECTS OF NICOTINE
30 I
$
.
.
.
.
.
.
RAT NW~I02 . . . .
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
100z ÷l
~
60.
~
120
RAT NW-402 ~ liT ( )=NO. OF SESSIONS ~ INJECTIONS
RAT NW-409
$ ....
60
489
,
.
.
.
.
,
.
,
.
.
,
.
.
.
.
.
.
.
.
.
.
.
.
80
.
<
,
~ 3°f .................................
O Z
D
60
RAT NF-405 ~
"60 "60
,
.
,
,
,
,
*90
2200 HRS
i
S = SALINE
04OO HRS.
* NO FOOD AVAILABLE
1000 HRS
2 HRS.
FIG. 1. Event records from intravenous saline and nicotine selfinjection sessions in 2 Wistar (NW-409 and NW-402) and one F-344 (NF-405) rat. Each horizontal line is a separate 12-hr session. Vertical deflections indicate self-injections. Food was removed from the self-injection chamber for three sessions in Rat NF-405.
presents sample event records from nicotine sessions in two Wistar and one F-344 rat. Rat NW-409 shows a higher injection rate for one unit dose of nicotine than for saline, but the pattern of injections across time is random. Rat NW-402, however, shows higher self-injection rates and relatively evenly spaced single injections at unit doses of 60-120/zg/kg (as the base), but not at 30 and 160 #.g/kg. The duration of the pause between injections (inter-injection time: liT) also appears to increase with increases in unit dose from 30 to 120 /tg/kg. The F-344 rats, exemplified by Rat NF-405, selfinjected nicotine only at very low rates except when temporarily food deprived by removing food from the chamber for the 12-hr overnight session. Some evenly spaced responding is evident in parts of the nicotine sessions in this rat also. The intravenous self-injection of cocaine and amphetamine by rats is characterized by single injections separated by relatively uniform periods of time, and these interinjection intervals change in a systematic fashion with unit dose size, leading to somewhat inverted U-shaped or negatively sloped unit dose-injection frequency functions that are also seen in rhesus monkeys [8, 21, 22, 32, 35]. The finding of a similar regular pattern of nicotine injections in rats, along with self-injection at rates higher than saline, lends further support to the idea that the drug serves as a reinforcer for behavior leading to its injection. While the spaced responding in these 2 nicotine rats suggested control over lever pressing by the drug, we also found that the dose-response functions were quite flat (Fig. 2) when several consecutive sessions at each injection dose were averaged together. Variability in the number of injections per 12-hr session and in the l i T was high and there was no systematic effect of injection dose on either dependent variable. The only instance of a systematic effect of injection dose was when the injection dose was increased by 30 #.g/kg each night (Fig. 3). The inter-inject/on interval increased and
I
I
1
I
I
30 (3)
60 (7)
90 (6)
120 (10)
160 (3)
NICOTINE INJECTION DOSE (MCG/KG/BASE) FIG. 2. Intravenous nicotine serf-injection rate and inter-injection times at various unit doses of nicotine in Rat NW-402. Each point represents the mean and standard deviation of several 12-hr sessions at each unit dose. Saline self-injection rates averaged about 14 inj/session in this rat.
RAT
NW- 402 S" SALIN(
I00
Z
o---.o
lit
e,,,--e
INJECTION8
NS • NO SFSSION
80
[
~ - j_ 6o < •
40
H ~
20
o 0 z
S
30 S0 901L)O SO SO 3 0 ~ I O N 8
30 Z l O S O S O
SOS0110SO
S
S
S
S
S
NICOTINE INJECTION DOSE ( M C G / K C / E A S E ) SUCCESSIVE DAILY SESSIONS
FIG. 3. Intravenous saline and nicotine self-injection rate and interinjection times, at various unit doses, in successive nightly 12-hr sessions in Rat NW-402.
the number o f injections per sessions decreased as the injection dose was increased in successive sessions. When the dose was subsequently decreased, however, few systematic changes in behavior were found, and an attempt in sessions 10-12 to reproduce the decrease in injection rate by again increasing unit dose resulted in no change in behavior. Evidently a change occurred during the dosing sequence which
490
DOUGHERTY E 1 A L .
rendered the behavior less sensitive to alterations in drug dose. Thus, nicotine self-administration rates in excess of saline intake have been obtained in rats and monkeys under CRF schedule conditions, but the nicotine/saline ratio of 2-3 suggests that the preference for nicotine is quite weak when compared with that for the psychomotor stimulants or opiates, where the drug/saline ratio may be 20 or more. The duration of exposure necessary to show differences between nicotine and saline self-injection rates is also quite long (e.g. cumulated injections over 5 days in Hanson's study [13]) compared to other classes of self-injected drugs. Large differences between cocaine and saline self-injoction may be seen within hours of the first cocaine availability [8]. In addition, great variability in the pattern and the rate of nicotine self-injection occurs both within and across species. Lang et al. [18], in 1977, reported that differences were evident between nicotine and saline rats within 90 hr. However, there were no differences in rate of self-injection between nicotine and saline rats during the first 14 days of the Smith and Lang 1980 experiment [26], even though the two studies appear to have been conducted under identical conditions. Apparently there can be considerable variability in the time necessary for the acquisition of nicotine self-injection even when the experiments are directly replicated in the same laboratory using the same strain (Lister) of rats. The continuous reinforcement conditions under which most drugs of abuse have been studied do not engender high rates of nicotine self-administration. Indeed, two strains of rats (Lister, F-344) showed significant self-injection behavior only when food-deprived, indicating that nicotine is at best a very weak reinforcer under CRF conditions. The variables that may be responsible for this weakness are unclear. There is some evidence that insensitivity to unit dose size develops over time, and we would agree with Hanson's assessment that "feedback from this reinforcement system is poor" [13]. However, when scheduled on a CRF basis, the direct suppressant effects of nicotine on responding [6, 9, 20] may interfere with self-injection responding supported by the reinforcing effects of the drug. Scheduling nicotine injections on an intermittent basis, however, can minimize the contribution of direct drug effects. Recent studies using intermittent schedules of reinforcement have yielded more positive results. INTERMITTENTREINFORCEMENTSCHEDULESTUDIES On intermittent schedule~, of drug self-administration, multiple responses are generally emitted for each drug injection. The number of responses or the rate or responding supported by each drug injection can vary over a wide range and can thereby more clearly reflect the ability of a drug to maintain behavior leading to its ingestion with less interference from direct (e.g. suppressant) effects of the drug. Yanagita was the first investigator to employ an intermittent schedule of nicotine self-injection 133,34]. He employed a progressive-ratio schedule in which the number of responses per injection, starting at a ratio of 100:1, doubled every t6 injections until the monkey failed to take more than o n e injection in 24 hr. In two out of 4 monkeys, a unit dose of 200 v,g&g supported 4 and 8 times the response ratio maintained by IV saline injections, clearly showing the effectiveness of nicotine in maintaining substantial amounts of responding. However, the weakness of nicotine relative to other self-administered drugs was evident by the finding that
480 p.g/kg cocaine supported between 8 and 64 times the saline ratio value in all four monkeys. In addition, Yanagita [34] reported that repeated use of the progressive-ratio procedure led to high rates of saline-maintained responding. Griffiths et al. Ill] use a fixed-ratio schedule (FR 160), rather than a progressive-ratio schedule, in baboons given the opportunity to earn up to 8 nicotine injections in 24 hr. Although they studied unit doses of 1 to 3000 p.g/kg for 12-day exposure periods at each unit dose, nicotine was not self-injected at a rate exceeding saline control levels. In a subsequent study, Ator and Griffiths [l] switched baboons from cocaine or saline baselines to nicotine (10-320/xg/kg) available according to a fixed-ratio 2 schedule in daily 2-hr sessions. Distinctive inverted U-shaped close-response functions were found immediately after substitution off the cocaine baseline, but continued exposure to nicotine produced flatter functions similar to those obtained after switching from saline to nicotine. Although one or more unit doses of nicotine maintained responding at rates greater than saline, the investigators characterized nicotine as a -marginally reinforcing drug." Perhaps the most convincing demonstration of the reinforcing properties of nicotine was recently obtained in squirrel monkeys by Goldberg and Spealman [9] and Goldberg, et al. [10]. They used two kinds of intermittent reinforcement schedules: a 5-min fixed-interval schedule (FI 5) in which the first response after a 5-rain interval produced an IV nicotine injection, and a second-order schedule in which the FI 5 was modified to include a 2-see light at the completion of every 10th response and a light plus nicotine injection upon completion of the first 10-response unit after the end of the 5-min interval. Employing unit doses of from 3 to 560 v.g/kgin daily sessions during which up to 10 injections were allowed, they found that nicotine could maintain response rates equal to that engendered by food or cocaine, and at least 10 times greater than response rates supported by saline injections. In addition, unit dose-response rate functions obtained under the FI procedure showed the characteristic inverted U-shape, and saline substitution or injection of mecamylamine consistently produced systematic decreases in response rate and in the number of injections obtained under the FI and second-order schedules. These results therefore fulfill the three general criteria for identifying reinforcing stimulus properties of drugs as outlined in the introduction. Goldberg et al. [ 10] also found that response rates within the 5-min fixed interval were higher when the brief stimuli were presented after every 10th response than when they were omitted. Presumably, the conditioned reinforcing properties of the brief stimuli, acquired by being associated with the nicotine injections, account for the higher response rates. The existence of conditioned reinforcers is further evidence for the positive reinforcing properties of nicotine injections. Studies of intermittent schedules of IV nicotine selfinjection in rhesus monkeys conducted in our laboratory have been less successful. While Goldberg and Spealman's [9] drug naive squirrel monkeys required a series of automatic injections and from 20--40 daily sessions to show response rates of about 10 resp/min, our rhesus monkeys required from 2 to 9 months of daily 2-hr sessions on FR and FI schedules before 3 of 4 monkeys began to self-inject nicotine above saline control levels. The acquisition of similar levels of cocaine or codeine self-injection under similar experimental conditions usually requires hours or at most a few
BEHAVIORAL EFFECTS OF NICOTINE
BENSON
491
FI 60 (FR 4)
GUNTHER
FRI FI 16 SEC
_z =E ¢f) bJ u3 Z 0 ,.~ u') I~J r,,
•
,
."I,
5
rP. I
I
Z
g laJ u3
50
Q p¢j
I-
Z
I
I
I
I
I
7' I
3
f
o
~z ~ i ~_D :S
eo
R AL P IO D W S
i 15
I ~lO
o
~-o I 60
I 90
I
I
I
I
I
I
SLOW o--no RAPID
Z
I SALINE
I
15
~'~o--'~.~
~::~:::::~~/
I
I 120
I 150
I 220
NICOTINEDOSE(.ug/ kg ) FIG. 4. Unit dose-response functions for one rhesus monkey, Benson. Data points in"Siow" functions represent the mean-SEM from 5-10 consecutive daily sessions at each unit dose. Points in "Rapid" curves are from a procedure in which the unit dose was progressively increased or decreased in successive dally sessions. This monkey obtained self-injections according to a second-order schedule in which a 0.5-sec tone was presented after each 4th response, and in which the first unit of 4 responses completed after a 60-sec fixed-interval resulted in a tone presentation and drug injection. The nicotine dose is calculated as the base.
days of access. As with rats, nicotine self-injection did not develop in all monkeys. After the eventual acquisition of self-injection behavior, we also studied dose-response functions in these monkeys using second-order and interval schedules o f intermittent reinforcement. We used two different procedures to study the effects o f nicotine unit dose on self-injection rate. In the " s l o w " procedure, the animal was allowed to self-inject at each unit dose level for .5-10 sessions (3-hour daily sessions) before the level was changed. In the " r a p i d " procedure, the unit dose was changed daily, first progressively increased and then decreased o v e r successive daily sessions. Figures 4 and 5 show the dose-effect curves generated by both procedures in two monkeys. In Figure 4, Benson's response rates and injection rates are clearly greater than saline in the lower dosage range and the dose-response function derived from the rapid procedure is shifted to the left of that from the slow procedure in both monkeys (Figs. 4 and 5). Gunther's performance illustrates the advantage of using intermittent reinforcement schedules. His nicotine injection rates in the slow
or) O3 I¢I O3
I0
X
=E I
t
SALINE I0
!
:50 90 180 NICOTINE DOSE(~g/kg)
I
270
FIG. 5. Unit dose-response functions for one rhesus monkey, Gunther. Procedure for obtaining data points in "Slow" and "Rapid" functions are as described in Fig. 4 legend. This monkey obtained self-injections according to a tandem interval schedule in which the first response after a drag injection started a 16-sec fixedinterval. The first response after 16 sec produced the nicotine injection.
procedure appear to be not significantly different from saline self-injection values, but the amount o f responding maintained by nicotine is greater than supported by saline and the unit dose-response rate function has the characteristic inverted U-shape. Gunther's nicotine injection rates in the slow procedure are also responsive to unit dose changes, but the high rate of saline self-injection appears to decrease the significance of this finding. Figure 5 shows nicotine and saline injection rates totalled across the 3-hr session in this animal. However, an examination o f Gunther's individual nicotine and saline sessions shows that the pattern of responding within nicotine and saline sessions is markedly different. Figure 6 shows cumulative response records from a single nicotine self-injection session and several subsequent saline selfinjection sessions. In the 30 pg/kg nicotine session, the record shows a very high rate o f drug intake at the beginning of
D O U G H E R T Y E'I AL
492 GUNTHER 50
J .•
GUNTHER UG/KG
BENSON
NICOTINE
f
SAL
. . . . .
SAL
f
SALINE . ,:,e~.
EXT.
,!
~
3o
30
/ ,
/
.........!
2
4
.
9O
I 6 • _~_~_~_o.~--~
~<~_~--~+---~-~ •
10 M I N
-'-~
__~.
i
+
150
180
210
210
J,
I
FIG. 6. Cumulative records from nicotine and saline self-injection sessions in one rhesus monkey, Gunther. The vertical excursion of the pen indicates cumulative responses. The pen resets after 300 responses and at the end of the 3-hr session. Pen offset indicates injections. On the day after the 30 tts/kg/inj (base) session, saline was substituted for nicotine for six days, of which records from days 1, 2, 4, and 6 are shown.
~'~
1
• 1 0 MIeN
the session, a long pause in responding, and then a low rate of self-in~mctionin the latter half of the session. Saline was substitut~lfor nicotine over the next six sessions. The rapid self-injection at the start of the session gradually extinguished, but injection rates increased in the last half of the sesssion to the extent that the total number of injections in tlm 6~h saline session was higlmr than the number of 30/a4~g
.
,
FIG. 7. Cumulative records from the initial 22 rains of3-hr sadine and nicotine s d f - i ~ sessions at various unit doses in two mo~eys, Gunther and k n s o n . The vert!cal excursion of the p ~ im~:a~es cumulative responses. Gunther s response pen res~ ~ 300 responses and Be~son's. after I8 rain. Pen Offset ~ m ~ . Nicotine unit doses, in #41/k4(base), are listed+immide~ record. Arrows indicate occurrence of emesis in Gunther.
~ s
in the preced/n8 nicotine session. Because o f this s ~ in t h e distribution o f responses over the session between nicotine ~ saline sessions, we examined the period o f rapid s e l f - i ~ c t i o n at the ~ i n S o f the session to determine if this ~ r e r e s t ~ ~ of b~mvior was sensitive to unit d o s e changes. We found ~ b,~vior in the early part o f the session c ~ sy-~emtically with variations in unit dose, and t h a t Gum.her's rates were much higher than that s u p ~ r t e d : b y ~ (Fill. 7). It is apparent that an examination o f the ~ pattern o f responding in nicotine self-injection can yield important
informmion that m i ~ t be obscured by a v e r a s i n s injection frequency over entire sessions o r days, and m a y lWlpexplain why flat unit dose-injection rate functions h a v e L~quently been found. Intravenous n/cotinc injections are more effective in maintaining self-administratien behavior when ~ d intermittently than when continuously available. The reasons
BEHAVIORAL EFFECTS OF NICOTINE
493
30
UJ
•
FIRST D&ILY I N J E C T I O N
0
SECOND DALLY I N J E C T I O N
=E
ONE H O U R INTERINJECTION INTERVAL
n-6
IZ
20
o~
o i-U
~
o~ C--'~
I0
I---
,
I I I I I i t
3 DAYS
9 DAYS
I 0 DAYS
12 DAYS
4 DAYS
SAL
200
:350
500
650
NICOTINE DOSE (#G/KG/INJ)
FIG. 8. Duration of response suppression after twice-daily saline or nicotine injections given IP one hr apart over 38 successive days in 6 Wistar rats. The first ratio completion time is measured from the beginning of the session to the completion of the first ratio on a FR 50 schedule of water-reinforced lever pressing. Saline or nicotine injections were given 1 rain before the start of 30-rain lever pressing sessions.
for this difference are not clear. Intermittent scheduling of drug injections may decrease the direct response rate suppressant effects of the drug. However, if response suppression were a major variable, then one would have expected far more responding on fixed- and progressive-ratio schedules of nicotine presentation than was actually found. The number of external stimuli correlated with nicotine administration has also been thought to contribute to the ability of the drug to maintain behavior, and the higher response rates obtained by Goldberg et al. [10], under second-order schedules support this idea. Yet without those brief stimuli, response rates generated by fixed-interval nicotine selfinjection in squirrel monkeys were substantially higher than fixed-ratio responding in rhesus monkeys or baboons. Perhaps some aspect of interval scheduling is critical for engendering high rates of nicotine self-administration. From the pattern of self-injections within sessions in our monkeys, it is clear that changes in the rates of self-administration behavior may change over time independently of schedule conditions. Tolerance to the effects of nicotine may be involved. T O L E R A N C E TO THE BEHAVIORAL EFFECTS OF NICOTINE
A number of the studies reviewed above have suggested that a change in responsiveness of the animals to nicotine occurred over time. Nicotine self-injection frequency in rats was systematically related to unit dose size only when the dose was increased in successive sessions. In monkeys, the unit dose-injection rate function derived from the "rapid" procedure was shifted to the left of the function obtained from the "slow" procedure. Other aspects of rat and monkey behavior suggest that tolerance may develop: 90/~g/kg
IV nicotine injections initially produced convulsions in naive rats, but twice that unit dose failed to elicit even motor weakness after a few days of exposure to the drug. The emesis that occurred in Gunther and other monkeys at doses of 90 ~tg/kg and above in initial sessions disappeared within a few days, and amounts of nicotine that would elicit emesis in Gunther at the beginning of a 3-hr session would fail to produce that effect just after the session. Tolerance to the behavioral effects of nicotine has frequently been observed in non-self--administration studies. Tolerance to the effects of nicotine on behavior may develop gradually over days with daily injections or more rapidly, within a few hours, ff multiple daily injections are given [5, 6, 14, 16, 20, 23, 28, 29]. Because multiple nicotine selfinjections were taken each day over successive days, we decided to study the tolerance that may develop when twice-daily injections were administered on successive days. With this design, the temporal properties of the rapidly (within hours) and slowly (over days) developing tolerance to nicotine can be evaluated. Male albino rats were trained to lever-press on a fixedratio 50 (FR 50) schedule of water reinforcement. On the first day of the study, the first IP injection of 200 ~g/kg (base) nicotine, administered just prior to the 30-min session, produced a complete suppression of responding for about 15 rain, after which the behavior abruptly returned to pre-drug rates. However, the 2nd daily injection (given l, 2, or 4 hr later) produced a much smaller effect (Fig. 8). When the sequence of twice-daily nicotine injections was continued, tolerance to the effects of the first daily injection occurred over a 9-day period until neither injection produced an effect. However, when the injection dose was increased to 350, 500, and 650/zg/kg, the differences between the 2 daily injections reappeared, even though the effect of both injections was increased as a function of dose. Note also that there was virtually no further development of tolerance to the effects of either injection across days at the three higher doses. If one can generalize from the suppressant effects of nicotine on water-reinforced lever pressing to patterns of nicotine self-injection (there is a direct relationship between pausing in cocaine self-injection and suppression of lever pressing behavior by cocaine given non-contingently [2, 7, 22]), it is possible that the reinforcing effects of nicotine may also change as a function of time since the first injection. Tentative support for this hypothesis is provided by preliminary results from our laboratory. In one study, 1440/zg/kg nicotine was injected SC either 0.5, 1, or 2 hr prior to IV nicotine self-injection sessions in rhesus monkeys. The 0.5-hr and the 1-hr nicotine pretreatments suppressed responding more than the 2-hr pretreatment, but all reduced only the rapid responding at the beginning of the session and had little effect on injection rate in the remainder of the 3-hour session. Saline pretreatment had no effect. Thus, the injection rate characteristic of most of the session was not suppressed by nicotine pretreatment, and is relatively independent of the amount of nicotine taken in the beginning of the session. This suggests that the factors that maintain responding in the latter part of the session are different from the effects that engender rapid responding at the beginning of the session and that the self-injection responding at the end of the session may not be sensitive to the amount of nicotine in the body. The development of tolerance to the effects of nicotine within 1 or 2 hr may play a role in this pattern of self-injection.
494
DOUGHERTY E7 AI..
If tolerance develops to the behavioral suppressant effects of nicotine, then why didn't self-injection rates increase over the session rather than decrease? This question cannot be answered at the present time. However, one hypothesis that may be proposed is that tolerance also occurs to the positively reinforcing stimulus effects of nicotine. The rapid development of tolerance to a variety of CNS effects of nicotine, including reinforcing effects, would be consistent with the findings pertaining to tolerance reported by other investigators participating in this symposium and congruent with the depolarization-blockade mechanism of action of nicotine at the receptor level. Thus, the behavior in the last half of the 3-hr self-injection sessions may be functionally equivalent to saline-maintained responding or may possibly be maintained by negative reinforcement contengencies. More conclusive information would be provided by an experiment in which a large, non-contingent injection or saline substitution was given a f t e r the period of rapid nicotine intake. The results of such a study may contribute to an understanding of nicotine self-administration in animals as well as smoking behavior in humans.
restricted than the situations in which drugs of the psychomotor stimulant, opiate, and sedative-hypnotic classes can be shown to be reinforcers. The manner in which the drug injections are scheduled with reference to response and temporal contingencies appears to be of considerable importance. It has been difficult to demonstrate that nicotine is a reinforcer under continuous reinforcement schedule (CRF) conditions or on fixed-ratio schedules. The most successful conditions appear to involve interval schedules, with or without supplemental external stimuli correlated with drug presentation (second-order schedule). The reinforcing properties of nicotine are more easily demonstrated in nonhuman primates than in rats. Food deprivation, strain and individual differences, circadian rhythms, and duration of exposure to nicotine are also important variables that influence the rate of nicotine self-administration. The choice of dependent variable (e.g., injection rate vs response rate) and the manner in which the data are organized or averaged may also have an influence on the interpretation of the results. Finally, the very rapid development of tolerance to some of the behavioral effects of nicotine may account for variations in temporal pattern of nicotine self-injection.
SUMMARY Published studies of intravenous and oral nicotine selfadministration have been reviewed for evidence that nicotine serves as a reinforcing stimulus for behavior leading to its presentation. The range of conditions under which nicotine demonstrates reinforcing stimulus properties is much more
ACKNOWLEDGEMENTS This research was supported in part by USPHS grants DA01778 and BRSG RR 05374-15, and by the Veterans Administration Research Program.
REFERENCES
1. Ator, N. A. and R. R. Griffiths. Intravenous self-administration of nicotine in the baboon. Fedn Proc. 40: 298, 1981. 2. Balster, R. L. and C. R. Schuster. Fixed-interval schedule of cocaine reinforcement: Effect of dose and infusion duration. J. exp. Analysis Behav. 20: 119-129, 1973. 3. Clark, M. S. G. Self-administered nicotine solutions preferred to placebo by the rat. Br. J. Pharmac. 35: 376, 1969. 4. Deneau, G. A. and R. Inoki. Nicotine serf-administration in monkeys. Ann. N. Y. Acad. Sci. 142: 277-279, 1967. 5. Domino, E. F. Electroencephalographic and behavioral arousal effects of small doses of nicotine: A neuropsychopharmacological study. Ann. N.Y. Acad. Sci. 142: 216-244, 1967. 6. Domino, E. F. and M. P. Lutz. Tolerance to the effects of daily nicotine on rat bar pressing behavior for water reinforcement. Pharmac. Biochem. Behav. 1:445 ~S, 1973. 7. Dougberty, J. and R. Pickens. Fixed-interval schedules of intravenous cocaine presentation in rats. J. exp. Analysis Behav. 20: 111-118, 1973. 8. Dougherty, J. and R. Pickens. Pharmacokinetics of intravenous cocaine self-injection. In: Cocaine: Chemical, Biological. Clinical, Social and Treatment Aspects, edited by S. J. Muir. Cleveland: CRC Press, 1976, pp. 105--120. 9. Goldberg, S. R. and R. D. Spealman. Maintenance and suppression of behavior by intravenous nicotine injections in squirrel monkeys. Fedn Proc., in press. 10. Ooidberg, S. R., R. D. Sl~mlman and D. M. Goldberg. Persistent behavior at high rates maintained by intravenous selfadministration of nicotine, Science 214: 573--575, 1981. 11. Griffiths, R. R., J. V. Brady and L. D, Bradford. Predicting the abuse liability of drugs with animal drug self-administration procedures: Psychomotor stimulants and hallucinogens. In: Advances in Behavioral Pharmacology, Vol. 2, edited by T. Thompson and P. B. Dews. New York: Academic Press, 1979, pp. 163--208. 12. Hanson, H. M., C. A. lvester and B. R. Morton. The effects of selected compounds on the self-administration of nicotine in rats. Fedn Proc. 36: 1040, 1977.
13. Hanson, H. M.. C. A. lvester and B. R. Morton. Nicotine selfadministration in rats. In: Cigarette Smoking as a Dependence Process. edited by N A Krasnegor. Washington. DC: U.S. Government Printing Office. 1979, pp. 70-90. 14. Hatchell. P. C. and A. C. Collins. Influence ofgenotyp¢ and sex on behavioral tolerance to nicotinein mice. Pharmac. Biochem. Behav. 6: 25--30. 1977. 15. Jaffe, J. H. and M. E. Jarvik. Tobacco use and tobacco use disorder. In: Psychoph~lrmacology : A Generution o f Progress. edited by M. A. Lipton. A. DiMascio and K. F. Kitlam. New York: Raven Press. 1978. pp, 1665--t676. 16. Jones, R. T., T. R. Farr©li. IIl and R. 1. Herning. Tobacco smoking and nicotine tolerance, in: SelJ:Administrution o f Abused Substances: Methods j~r Study. National Institute on Drug Abuse Research Monograph 20. edited by N. Krasnegor. Washington. DC: U.S. Government Printing Office, 1978, pp, 202-208. 17. Kumar. R.. E. C. Cooke. M. H. Lacier and M. A. H. Russell. ls nicotine important in tobacco-smoking? Clin. Pharmac. Ther. 21: 520--529. 1977. 18. Lang, W. J.. A. A. Latiff, A. McQueen and G. Singer. Selfadministration of nicotine with and without a food delivery schedule. Pharmac. Biochem. Behav. 7: 65-70. 1977. 19. Lucchesi, B. R.. C. R. Schuster and G. S. Emley. The role of nicotine as a determinant of cigarette smoking in man with observations of certain cardiovascular effects associated with the tobacco alkaloid. Clin. Pharmac. Ther. g: 789-796. 1967. 20. Morrtson, C. F. and J. A. Stcphenson. The occurrence of tolerance to a central depressant effect of nicotine. Br. J. Pharmac. 45: 151-156, 1972. 21. Pickenso R, and W. C. Harris. Self,administration of d-amphetamine by rats. Psychopharmacologia 12: 158-163. 1968.
22. Pickens, R, and T. Thompson. Cocaine reinforced behavior in rats: Effects of reinforcement magnk-'-ud¢andfixed-ratio size. J. Pharmac. exp. Ther. 161: 122-129. 1968.
BEHAVIORAL EFFECTS OF NICOTINE 23. Rosecrans, J. A. Nicotine as a discriminative stimulus to behavior: its characterization and relevance to smoking behavior. In: Cigarette Smoking as a Dependence Process, edited by N. A. Krasnegor. Washington, D.C.: U.S. Government Printing Office, 1979, pp. 58-69. 24. Russell, M. A. H. Tobacco smoking and nicotine dependence. In: Research Advances in Alcohol and Drug Problems, Vol. 3, edited by R. J. Gibbins, Y. Israel, H. Kalant, R. E. Popham, W. Schmidt and R. G. Smart. New York: Wiley and Sons, 1976, pp. 1--47. 25. Russell, M. A. H. Tobacco dependence: Is nicotine rewarding or aversive? In: Cigarette Smoking as a Dependence Process, edited by N. A. Krasnegor. Washington, DC: U.S. Government Printing Office, 1979, pp. 100-122. 26. Smith, L. A. and W. J. Lang. Changes occurring in selfadministration of nicotine by rats over a 28-day period. Pharmac. Biochem. Behav. 13: 215--220, 1980. 27. Spealman, R. D. and S. R. Goldberg. Drug self-administration by laboratory animals: Control by schedules of reinforcement. A. Rev. Pharmac. Toxicol. 18: 313-339, 1978. 28. Stolerman, I. P., P. Bunker and M. E. Jarvik. Nicotine tolerance in rats: Role of dose and dose interval. Psychopharmacologia 34: 317-324, 1974.
495 29. Stolerman, 1. P., R. Fink and M. E. Jarvik. Acute and chronic tolerance to nicotine measured by activity in rats. Psychopharmacologia 30: 329-342, 1973. 30. Todd, G. D. and J. Dougherty. Dose-dependent tolerance to a behavioral effect of nicotine. Soc. Neurosci. A bstr. 6: 544, 1980. 31. Wikler, A. Dynamics of drug dependence. Implications of a conditioning theory for research and treatment. Archs gen. Psychiat. 28: 611--616, 1973. 32. Wilson, M. C., M. Hitomi and C. R. Schuster. Psychomotor stimulant self-administration as a function of dosage per injection in the rhesus monkey. Psychopharmacologia 22: 271-281, 1971. 33. Yanagita, T. An experimental framework for evaluation of dependence liability of various types drugs in monkeys. Bull. Narcotics 25: 57-64, 1972. 34. Yanagita, T. Brief review on the use of self-administration techniques for predicting drug abuse potential. In: Predicting Dependence Liability of Stimulant and Depressant Drugs. edited by T. Thompson and K. Unna. Baltimore: University Park Press, 1977, pp. 231-242. 35. Yokel, R. A. and R. Pickens. Self-administration of optical isomers of amphetamine and methylamphetamine by rats. J. Pharmac. exp. Ther. 187: 27-33, 1973.
Reinforcing and Other Behavioral Effects of Nicotine I JOHN DOUGHERTY,
DIANNE
MILLER,
GLENN
T O D D A N D H A R R Y B. K O S T E N B A U D E R
Departments o f Psychiatry and Pharmacology, University o f Kentucky Medical Center Lexington, K Y 40536 Veterans Administration Medical Center, Lexington, K Y 40511 and College o f Pharmacy, University o f Kentucky, Lexington, K Y
DOUGHERTY, J., D. MILLER, G. TODD AND H. B. KOSTENBAUDER. Reinforcing and other behavioral effects of nicotine. NEUROSCI. BIOBEHAV. REV. 5(4) 487--495, 1981.--Published findings of intravenous nicotine self-injection indicate that the reinforcing properties of nicotine are weak when the drug is made available according to continuous reinforcement (CRF) or fixed-ratio (FR) schedules. CRF self-injection rates are generally only 2-3 times saline control levels and self-injection frequency is largely insensitive to changes in unit dose. In contrast, drugs of the psychomotor stimulant, opiate, and sedative-hypnotic classes, with similar pharmacokinetic parameters, maintain much higher selfinjection rates and show systematic changes in rate with unit dose variations. Recent studies using interval and secondorder schedules of nicotine presentation have been more successful in maintaining higher rates of self-administration behavior. Systematic dose-response functions have also been found under these conditions. Food-deprivation, species and strain differences, circadian rhythms, and duration of exposure to the drug also appear to be important variables in determining self-injection rate. Finally, the rapid development of tolerance to the effects of nicotine may account for changes in the pattern of self-administration within daily sessions and the differential sensitivity of those patterns to nicotine pretreatment. Nicotine
Reinforcement
Fixed-ratio
Self-administration
N I C O T I N E has been considered the most active compound in tobacco and the major factor in the maintenance of tobacco use. According to the operant conditioning-addiction model of drug self-administration, the internal stimuli produced by the drug serve to reinforce or maintain the behavior leading to its presentation ]311. However, the role of nicotine as a maintaining stimulus in human smoking behavior has been difficult to define. The finding that smokers will not smoke non-nicotine cigarettes and will often decrease smoking rates after nicotine pretreatment offers support for nicotine as a reinforcer [ 15,19l. However, nicotine pretreatment does not always affect subsequent smoking rates, and may be aversive under some conditions [17, 24, 25]. The situation is further complicated by indications that cigarette smoking can be a strong habit in people who do not inhale the smoke, suggesting that external stimuli associated with smoking and social-learning factors are also important in the maintenance of the behavior. The possible interplay of pharmacologic and psychologic factors in smoking make the evaluation o f nicotine's reinforcing effects in humans a difficult task. If nicotine is a reinforcer, the drug should be selfadministered by animals in a manner analogous to the selfadministration by animals o f the opiate, psychomotor stimulant, and sedative-hypnotic drug classes commonly
Tolerance
used by man [11,27]. There are common features of drug self-administration behavior that indicate that the stimulus properties of the self-administered drug, and not other effects, are of primary importance in the maintenance of the behavior. These commonalities are: (1) Systematic changes in injection or response rate occurring with changes in unit dose size. (2) An injection or response rate maintained by the drug that is greater than that maintained by the vehicle alone over a substantial portion o f the dose-response function. (3) Systematic changes in injection or response rate occurring after pretreatment with the same drug or with pharmacologic agonists or antagonists. This review will critically examine the animal oral and intravenous (where the amount o f nicotine ingested or absorbed is known) nicotine self-administration literature and examine the proposition that nicotine serves as a stimulus maintaining behavior leading to its administration by determining the extent to which nicotine self-administration behavior shares the above commonalities with other drug classes. In addition, the influence that tolerance development may exert on nicotine self-administration will be discussed. The manner in which drug availability is scheduled has an important influence on self-administration behavior. Thus, this review will distinguish between studies in which
1Paper presented at a symposium entitled CNS Sites and Mechanisms of Nicotine's Psychopharmacological Effects, Meeting of the American Society for Pharmacology and Experimental Therapeutics, Rochester, MN, August 18, 1980.
487
488 one response per drug injection {continuous reinforcement schedule--CRF) was required and those in which multiple responses (intermittent reinforcement schedules) were used. CONTINUOUSREINFORCEMENTSCHEDULE(CRF) STUDIES Using CRF schedules of drug availability, the reinforcing properties of many self-injected drugs have been readily demonstrable. For example, with psychomotor stimulants, self-injection rates are substantially higher than saline, and alterations in unit dose size produce orderly biphasic (inverted U-shaped) or monophasic changes in injection rate in rats and monkeys [21, 22, 32]. However, showing that nicotine has reinforcing stimulus properties under these schedule conditions has been a difficult and frustrating task. Nicotine self-administration rates, using CRF schedules, that are significantly higher than saline have been difficult to produce, and systematic changes in injection rate with increases or decreases in unit dose have not been found with any consistency. Deneau and Inoki were the first investigators to report intravenous nicotine self-injection [4]. Rhesus monkeys received hourly automatic injections of nicotine at a unit dose of 25/~g/kg for 2-10 days before the animals began to selfinject the drug on a CRF schedule between non-contingent injections. Eventually, the automatic injections were stopped and the monkeys continued to self-administer between 28 to 68 injections per 24-hr period, usually during the light phase of the 12-hr light-dark cycle. Day-to-day variability in the frequency of injections was often as much as 100%. Although dose-response curves were not presented, Deneau and Inoki reported that injections declined to 7 per day at a unit dose of 2.0 mg/kg. Their report suggested that nicotine would be self-injected in large amounts by some of their monkeys. They did not compare nicotine self-injection rates with that of the saline vehicle. Yanagita also reported intravenous nicotine self-injection by rhesus monkeys, and that it occurred only during the daytime with 24-hr access to 20/~g/kg/inj on a CRF schedule [33,34]. One monkey increased self-injection rates about 5 times when nicotine was substituted for saline. However, when access was limited to 4 hrs daily and nicotine selfinjection rates were again compared to those of saline and to SPA (a psychomotor stimulant), nicotine self-injection rates were not different from self-injection rates of 0.25 ml/inj physiological saline, and were less than 20% of the SPA selfinjection rates. Furthermore, the nicotine injection rate did not change even when the unit dose was varied from 2.5 to 640/zg/kg. In comparison, SPA was self-injected at a rate as much as 7 times greater than saline. Nicotine self-administration studies with rats have also been difficult to interpret. Although quantitative data were not presented, Clark reported that rats preferred nicotine when given a choice between oral or IV nicotine (1-10 t~g,/kg/inj) and odal water or IV saline [3]. Hanson and coworkers have reported extensive studies of CRF intravenous nicotine self-injection in rats 112,13]. Daily differences between animals self-injecting nicotine and those self-injecting saline were small (12 nicotine vs 5 saline inj in one representative 24-hr period) but consistent, and cumulation of the number of injections across 5 days o f continuous access to the drug revealed that the average number of injections in the nicotine rats was about 0.5 per hr at a unit dose of 60 ~g/kg (base) and about 0.2 per hr in the saline group. In another experiment, 4 rats were allowed to self-inject at 60 ~.g/kg/inj
DOUGHERTY E1 A I for 150 days. During the last 30 days, the median intake was about 1.2 inj/hr, and the rate of nicotine intake increased gradually in all 4 animals over the 150-day period. Interestingly, nicotine self-injection rates in the nocturnal rats were much higher during the dark phase of the 12-hr lightdark cycle than in the light. Hanson et al. [13] also evaluated the effects of substituting saline for the nicotine injection. In 12 rats, the number of injections declined from about 0.4/hr during nicotine availability, to about 0.1/hr after 10 days of saline. The above differences between the nicotine and saline group injection rates, the decline after saline substitution, and the gradual but steady increase in nicotine intake over a 150-day period all suggest that the stimulus properties of the drug were important in maintaining self-administration behavior. However, when different unit doses were examined, the rats took the same number of injections at 7.5 ~tg/kg as they did at 60 /xg/kg. The first administration of mecamylamine (a central and peripheral nicotine antagonist) clearly altered the rate of nicotine self-injection, as anticipated. However, this pretreatment produced little effect on 5 subsequent treatments. (A study in which mecamylamine was used to antagonize the discriminative stimulus properties of nicotine [23] found no decrease in effectiveness after multiple injections.) Lang et al. [18] and Smith and Lang 126] have studied intravenous and oral nicotine self-injection in food deprived and non-deprived rats. In their 1977 study, they compared a continuous 90-hour session of CRF nicotine selfadministration (50-100 tzg/kg/inj) with the same schedule of saline availability in non-deprived animals and found no difference between the nicotine and saline groups. Fooddepriving the rats to 80% of free-feeding weight, however. increased the rate of nicotine self-administration to about 2.3 times that of saline. With limited access (2 hr/day) to nicotine, food-deprived rats self-administered the drug {100 /~g/kg/inj, presumably as the salt) at 2.7 times the saline rate in food-deprived rats. Smith and Lang's 126] later paper reported similar results when rats were allowed 28 days of continuous access to nicotine. Comparing the results obtained by Lang et al. [ 18], Smith and Lang [26], and Hanson et al. [131, a similarity in selfadministration behavior is evident. Hanson [12,13], using a roughly comparable unit dose of nicotine base 160 p.g/kg), found a 2-3-fold greater intake rate for nicotine rats relative to saline rats (both groups non-food deprived). This agrees with the 2.7 factor reported by Lang and Smith 118,26]. In nicotine self-injection studies performed in our laboratory (discussed more fully below), we found a nicotine-saline ratio of about 1.5 to 2.0 within individual non-food deprived rats given access to nicotine or saline on different days. In general, absolute rates of nicotine self-injection are highest in food-deprived animals. However. rats self-administer nicotine at a unit dose of about 100/xg/kg (salt) about 2-3 times more frequently than saline controls across a variety of experimental conditions, regardless of the deprivational state. Studies of intravenous nicotine self-injection in our laboratory have yielded results similar to that found above. Wistar and F-344 male albino rats were allowed limited access to nicotine on a CRF schedule in overnighl I2-hr sessions. Overnight sessions were selected because little nicotine was self-administered during the daylight phase of the light-dark cycle. Differences between rat strains and among individuals of the same strain were evident. Figure I
BEHAVIORAL EFFECTS OF NICOTINE
30 I
$
.
.
.
.
.
.
RAT NW~I02 . . . .
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
100z ÷l
~
60.
~
120
RAT NW-402 ~ liT ( )=NO. OF SESSIONS ~ INJECTIONS
RAT NW-409
$ ....
60
489
,
.
.
.
.
,
.
,
.
.
,
.
.
.
.
.
.
.
.
.
.
.
.
80
.
<
,
~ 3°f .................................
O Z
D
60
RAT NF-405 ~
"60 "60
,
.
,
,
,
,
*90
2200 HRS
i
S = SALINE
04OO HRS.
* NO FOOD AVAILABLE
1000 HRS
2 HRS.
FIG. 1. Event records from intravenous saline and nicotine selfinjection sessions in 2 Wistar (NW-409 and NW-402) and one F-344 (NF-405) rat. Each horizontal line is a separate 12-hr session. Vertical deflections indicate self-injections. Food was removed from the self-injection chamber for three sessions in Rat NF-405.
presents sample event records from nicotine sessions in two Wistar and one F-344 rat. Rat NW-409 shows a higher injection rate for one unit dose of nicotine than for saline, but the pattern of injections across time is random. Rat NW-402, however, shows higher self-injection rates and relatively evenly spaced single injections at unit doses of 60-120/zg/kg (as the base), but not at 30 and 160 #.g/kg. The duration of the pause between injections (inter-injection time: liT) also appears to increase with increases in unit dose from 30 to 120 /tg/kg. The F-344 rats, exemplified by Rat NF-405, selfinjected nicotine only at very low rates except when temporarily food deprived by removing food from the chamber for the 12-hr overnight session. Some evenly spaced responding is evident in parts of the nicotine sessions in this rat also. The intravenous self-injection of cocaine and amphetamine by rats is characterized by single injections separated by relatively uniform periods of time, and these interinjection intervals change in a systematic fashion with unit dose size, leading to somewhat inverted U-shaped or negatively sloped unit dose-injection frequency functions that are also seen in rhesus monkeys [8, 21, 22, 32, 35]. The finding of a similar regular pattern of nicotine injections in rats, along with self-injection at rates higher than saline, lends further support to the idea that the drug serves as a reinforcer for behavior leading to its injection. While the spaced responding in these 2 nicotine rats suggested control over lever pressing by the drug, we also found that the dose-response functions were quite flat (Fig. 2) when several consecutive sessions at each injection dose were averaged together. Variability in the number of injections per 12-hr session and in the l i T was high and there was no systematic effect of injection dose on either dependent variable. The only instance of a systematic effect of injection dose was when the injection dose was increased by 30 #.g/kg each night (Fig. 3). The inter-inject/on interval increased and
I
I
1
I
I
30 (3)
60 (7)
90 (6)
120 (10)
160 (3)
NICOTINE INJECTION DOSE (MCG/KG/BASE) FIG. 2. Intravenous nicotine serf-injection rate and inter-injection times at various unit doses of nicotine in Rat NW-402. Each point represents the mean and standard deviation of several 12-hr sessions at each unit dose. Saline self-injection rates averaged about 14 inj/session in this rat.
RAT
NW- 402 S" SALIN(
I00
Z
o---.o
lit
e,,,--e
INJECTION8
NS • NO SFSSION
80
[
~ - j_ 6o < •
40
H ~
20
o 0 z
S
30 S0 901L)O SO SO 3 0 ~ I O N 8
30 Z l O S O S O
SOS0110SO
S
S
S
S
S
NICOTINE INJECTION DOSE ( M C G / K C / E A S E ) SUCCESSIVE DAILY SESSIONS
FIG. 3. Intravenous saline and nicotine self-injection rate and interinjection times, at various unit doses, in successive nightly 12-hr sessions in Rat NW-402.
the number o f injections per sessions decreased as the injection dose was increased in successive sessions. When the dose was subsequently decreased, however, few systematic changes in behavior were found, and an attempt in sessions 10-12 to reproduce the decrease in injection rate by again increasing unit dose resulted in no change in behavior. Evidently a change occurred during the dosing sequence which
490
DOUGHERTY E 1 A L .
rendered the behavior less sensitive to alterations in drug dose. Thus, nicotine self-administration rates in excess of saline intake have been obtained in rats and monkeys under CRF schedule conditions, but the nicotine/saline ratio of 2-3 suggests that the preference for nicotine is quite weak when compared with that for the psychomotor stimulants or opiates, where the drug/saline ratio may be 20 or more. The duration of exposure necessary to show differences between nicotine and saline self-injection rates is also quite long (e.g. cumulated injections over 5 days in Hanson's study [13]) compared to other classes of self-injected drugs. Large differences between cocaine and saline self-injoction may be seen within hours of the first cocaine availability [8]. In addition, great variability in the pattern and the rate of nicotine self-injection occurs both within and across species. Lang et al. [18], in 1977, reported that differences were evident between nicotine and saline rats within 90 hr. However, there were no differences in rate of self-injection between nicotine and saline rats during the first 14 days of the Smith and Lang 1980 experiment [26], even though the two studies appear to have been conducted under identical conditions. Apparently there can be considerable variability in the time necessary for the acquisition of nicotine self-injection even when the experiments are directly replicated in the same laboratory using the same strain (Lister) of rats. The continuous reinforcement conditions under which most drugs of abuse have been studied do not engender high rates of nicotine self-administration. Indeed, two strains of rats (Lister, F-344) showed significant self-injection behavior only when food-deprived, indicating that nicotine is at best a very weak reinforcer under CRF conditions. The variables that may be responsible for this weakness are unclear. There is some evidence that insensitivity to unit dose size develops over time, and we would agree with Hanson's assessment that "feedback from this reinforcement system is poor" [13]. However, when scheduled on a CRF basis, the direct suppressant effects of nicotine on responding [6, 9, 20] may interfere with self-injection responding supported by the reinforcing effects of the drug. Scheduling nicotine injections on an intermittent basis, however, can minimize the contribution of direct drug effects. Recent studies using intermittent schedules of reinforcement have yielded more positive results. INTERMITTENTREINFORCEMENTSCHEDULESTUDIES On intermittent schedule~, of drug self-administration, multiple responses are generally emitted for each drug injection. The number of responses or the rate or responding supported by each drug injection can vary over a wide range and can thereby more clearly reflect the ability of a drug to maintain behavior leading to its ingestion with less interference from direct (e.g. suppressant) effects of the drug. Yanagita was the first investigator to employ an intermittent schedule of nicotine self-injection 133,34]. He employed a progressive-ratio schedule in which the number of responses per injection, starting at a ratio of 100:1, doubled every t6 injections until the monkey failed to take more than o n e injection in 24 hr. In two out of 4 monkeys, a unit dose of 200 v,g&g supported 4 and 8 times the response ratio maintained by IV saline injections, clearly showing the effectiveness of nicotine in maintaining substantial amounts of responding. However, the weakness of nicotine relative to other self-administered drugs was evident by the finding that
480 p.g/kg cocaine supported between 8 and 64 times the saline ratio value in all four monkeys. In addition, Yanagita [34] reported that repeated use of the progressive-ratio procedure led to high rates of saline-maintained responding. Griffiths et al. Ill] use a fixed-ratio schedule (FR 160), rather than a progressive-ratio schedule, in baboons given the opportunity to earn up to 8 nicotine injections in 24 hr. Although they studied unit doses of 1 to 3000 p.g/kg for 12-day exposure periods at each unit dose, nicotine was not self-injected at a rate exceeding saline control levels. In a subsequent study, Ator and Griffiths [l] switched baboons from cocaine or saline baselines to nicotine (10-320/xg/kg) available according to a fixed-ratio 2 schedule in daily 2-hr sessions. Distinctive inverted U-shaped close-response functions were found immediately after substitution off the cocaine baseline, but continued exposure to nicotine produced flatter functions similar to those obtained after switching from saline to nicotine. Although one or more unit doses of nicotine maintained responding at rates greater than saline, the investigators characterized nicotine as a -marginally reinforcing drug." Perhaps the most convincing demonstration of the reinforcing properties of nicotine was recently obtained in squirrel monkeys by Goldberg and Spealman [9] and Goldberg, et al. [10]. They used two kinds of intermittent reinforcement schedules: a 5-min fixed-interval schedule (FI 5) in which the first response after a 5-rain interval produced an IV nicotine injection, and a second-order schedule in which the FI 5 was modified to include a 2-see light at the completion of every 10th response and a light plus nicotine injection upon completion of the first 10-response unit after the end of the 5-min interval. Employing unit doses of from 3 to 560 v.g/kgin daily sessions during which up to 10 injections were allowed, they found that nicotine could maintain response rates equal to that engendered by food or cocaine, and at least 10 times greater than response rates supported by saline injections. In addition, unit dose-response rate functions obtained under the FI procedure showed the characteristic inverted U-shape, and saline substitution or injection of mecamylamine consistently produced systematic decreases in response rate and in the number of injections obtained under the FI and second-order schedules. These results therefore fulfill the three general criteria for identifying reinforcing stimulus properties of drugs as outlined in the introduction. Goldberg et al. [ 10] also found that response rates within the 5-min fixed interval were higher when the brief stimuli were presented after every 10th response than when they were omitted. Presumably, the conditioned reinforcing properties of the brief stimuli, acquired by being associated with the nicotine injections, account for the higher response rates. The existence of conditioned reinforcers is further evidence for the positive reinforcing properties of nicotine injections. Studies of intermittent schedules of IV nicotine selfinjection in rhesus monkeys conducted in our laboratory have been less successful. While Goldberg and Spealman's [9] drug naive squirrel monkeys required a series of automatic injections and from 20--40 daily sessions to show response rates of about 10 resp/min, our rhesus monkeys required from 2 to 9 months of daily 2-hr sessions on FR and FI schedules before 3 of 4 monkeys began to self-inject nicotine above saline control levels. The acquisition of similar levels of cocaine or codeine self-injection under similar experimental conditions usually requires hours or at most a few
BEHAVIORAL EFFECTS OF NICOTINE
BENSON
491
FI 60 (FR 4)
GUNTHER
FRI FI 16 SEC
_z =E ¢f) bJ u3 Z 0 ,.~ u') I~J r,,
•
,
."I,
5
rP. I
I
Z
g laJ u3
50
Q p¢j
I-
Z
I
I
I
I
I
7' I
3
f
o
~z ~ i ~_D :S
eo
R AL P IO D W S
i 15
I ~lO
o
~-o I 60
I 90
I
I
I
I
I
I
SLOW o--no RAPID
Z
I SALINE
I
15
~'~o--'~.~
~::~:::::~~/
I
I 120
I 150
I 220
NICOTINEDOSE(.ug/ kg ) FIG. 4. Unit dose-response functions for one rhesus monkey, Benson. Data points in"Siow" functions represent the mean-SEM from 5-10 consecutive daily sessions at each unit dose. Points in "Rapid" curves are from a procedure in which the unit dose was progressively increased or decreased in successive dally sessions. This monkey obtained self-injections according to a second-order schedule in which a 0.5-sec tone was presented after each 4th response, and in which the first unit of 4 responses completed after a 60-sec fixed-interval resulted in a tone presentation and drug injection. The nicotine dose is calculated as the base.
days of access. As with rats, nicotine self-injection did not develop in all monkeys. After the eventual acquisition of self-injection behavior, we also studied dose-response functions in these monkeys using second-order and interval schedules o f intermittent reinforcement. We used two different procedures to study the effects o f nicotine unit dose on self-injection rate. In the " s l o w " procedure, the animal was allowed to self-inject at each unit dose level for .5-10 sessions (3-hour daily sessions) before the level was changed. In the " r a p i d " procedure, the unit dose was changed daily, first progressively increased and then decreased o v e r successive daily sessions. Figures 4 and 5 show the dose-effect curves generated by both procedures in two monkeys. In Figure 4, Benson's response rates and injection rates are clearly greater than saline in the lower dosage range and the dose-response function derived from the rapid procedure is shifted to the left of that from the slow procedure in both monkeys (Figs. 4 and 5). Gunther's performance illustrates the advantage of using intermittent reinforcement schedules. His nicotine injection rates in the slow
or) O3 I¢I O3
I0
X
=E I
t
SALINE I0
!
:50 90 180 NICOTINE DOSE(~g/kg)
I
270
FIG. 5. Unit dose-response functions for one rhesus monkey, Gunther. Procedure for obtaining data points in "Slow" and "Rapid" functions are as described in Fig. 4 legend. This monkey obtained self-injections according to a tandem interval schedule in which the first response after a drag injection started a 16-sec fixedinterval. The first response after 16 sec produced the nicotine injection.
procedure appear to be not significantly different from saline self-injection values, but the amount o f responding maintained by nicotine is greater than supported by saline and the unit dose-response rate function has the characteristic inverted U-shape. Gunther's nicotine injection rates in the slow procedure are also responsive to unit dose changes, but the high rate of saline self-injection appears to decrease the significance of this finding. Figure 5 shows nicotine and saline injection rates totalled across the 3-hr session in this animal. However, an examination o f Gunther's individual nicotine and saline sessions shows that the pattern of responding within nicotine and saline sessions is markedly different. Figure 6 shows cumulative response records from a single nicotine self-injection session and several subsequent saline selfinjection sessions. In the 30 pg/kg nicotine session, the record shows a very high rate o f drug intake at the beginning of
D O U G H E R T Y E'I AL
492 GUNTHER 50
J .•
GUNTHER UG/KG
BENSON
NICOTINE
f
SAL
. . . . .
SAL
f
SALINE . ,:,e~.
EXT.
,!
~
3o
30
/ ,
/
.........!
2
4
.
9O
I 6 • _~_~_~_o.~--~
~<~_~--~+---~-~ •
10 M I N
-'-~
__~.
i
+
150
180
210
210
J,
I
FIG. 6. Cumulative records from nicotine and saline self-injection sessions in one rhesus monkey, Gunther. The vertical excursion of the pen indicates cumulative responses. The pen resets after 300 responses and at the end of the 3-hr session. Pen offset indicates injections. On the day after the 30 tts/kg/inj (base) session, saline was substituted for nicotine for six days, of which records from days 1, 2, 4, and 6 are shown.
~'~
1
• 1 0 MIeN
the session, a long pause in responding, and then a low rate of self-in~mctionin the latter half of the session. Saline was substitut~lfor nicotine over the next six sessions. The rapid self-injection at the start of the session gradually extinguished, but injection rates increased in the last half of the sesssion to the extent that the total number of injections in tlm 6~h saline session was higlmr than the number of 30/a4~g
.
,
FIG. 7. Cumulative records from the initial 22 rains of3-hr sadine and nicotine s d f - i ~ sessions at various unit doses in two mo~eys, Gunther and k n s o n . The vert!cal excursion of the p ~ im~:a~es cumulative responses. Gunther s response pen res~ ~ 300 responses and Be~son's. after I8 rain. Pen Offset ~ m ~ . Nicotine unit doses, in #41/k4(base), are listed+immide~ record. Arrows indicate occurrence of emesis in Gunther.
~ s
in the preced/n8 nicotine session. Because o f this s ~ in t h e distribution o f responses over the session between nicotine ~ saline sessions, we examined the period o f rapid s e l f - i ~ c t i o n at the ~ i n S o f the session to determine if this ~ r e r e s t ~ ~ of b~mvior was sensitive to unit d o s e changes. We found ~ b,~vior in the early part o f the session c ~ sy-~emtically with variations in unit dose, and t h a t Gum.her's rates were much higher than that s u p ~ r t e d : b y ~ (Fill. 7). It is apparent that an examination o f the ~ pattern o f responding in nicotine self-injection can yield important
informmion that m i ~ t be obscured by a v e r a s i n s injection frequency over entire sessions o r days, and m a y lWlpexplain why flat unit dose-injection rate functions h a v e L~quently been found. Intravenous n/cotinc injections are more effective in maintaining self-administratien behavior when ~ d intermittently than when continuously available. The reasons
BEHAVIORAL EFFECTS OF NICOTINE
493
30
UJ
•
FIRST D&ILY I N J E C T I O N
0
SECOND DALLY I N J E C T I O N
=E
ONE H O U R INTERINJECTION INTERVAL
n-6
IZ
20
o~
o i-U
~
o~ C--'~
I0
I---
,
I I I I I i t
3 DAYS
9 DAYS
I 0 DAYS
12 DAYS
4 DAYS
SAL
200
:350
500
650
NICOTINE DOSE (#G/KG/INJ)
FIG. 8. Duration of response suppression after twice-daily saline or nicotine injections given IP one hr apart over 38 successive days in 6 Wistar rats. The first ratio completion time is measured from the beginning of the session to the completion of the first ratio on a FR 50 schedule of water-reinforced lever pressing. Saline or nicotine injections were given 1 rain before the start of 30-rain lever pressing sessions.
for this difference are not clear. Intermittent scheduling of drug injections may decrease the direct response rate suppressant effects of the drug. However, if response suppression were a major variable, then one would have expected far more responding on fixed- and progressive-ratio schedules of nicotine presentation than was actually found. The number of external stimuli correlated with nicotine administration has also been thought to contribute to the ability of the drug to maintain behavior, and the higher response rates obtained by Goldberg et al. [10], under second-order schedules support this idea. Yet without those brief stimuli, response rates generated by fixed-interval nicotine selfinjection in squirrel monkeys were substantially higher than fixed-ratio responding in rhesus monkeys or baboons. Perhaps some aspect of interval scheduling is critical for engendering high rates of nicotine self-administration. From the pattern of self-injections within sessions in our monkeys, it is clear that changes in the rates of self-administration behavior may change over time independently of schedule conditions. Tolerance to the effects of nicotine may be involved. T O L E R A N C E TO THE BEHAVIORAL EFFECTS OF NICOTINE
A number of the studies reviewed above have suggested that a change in responsiveness of the animals to nicotine occurred over time. Nicotine self-injection frequency in rats was systematically related to unit dose size only when the dose was increased in successive sessions. In monkeys, the unit dose-injection rate function derived from the "rapid" procedure was shifted to the left of the function obtained from the "slow" procedure. Other aspects of rat and monkey behavior suggest that tolerance may develop: 90/~g/kg
IV nicotine injections initially produced convulsions in naive rats, but twice that unit dose failed to elicit even motor weakness after a few days of exposure to the drug. The emesis that occurred in Gunther and other monkeys at doses of 90 ~tg/kg and above in initial sessions disappeared within a few days, and amounts of nicotine that would elicit emesis in Gunther at the beginning of a 3-hr session would fail to produce that effect just after the session. Tolerance to the behavioral effects of nicotine has frequently been observed in non-self--administration studies. Tolerance to the effects of nicotine on behavior may develop gradually over days with daily injections or more rapidly, within a few hours, ff multiple daily injections are given [5, 6, 14, 16, 20, 23, 28, 29]. Because multiple nicotine selfinjections were taken each day over successive days, we decided to study the tolerance that may develop when twice-daily injections were administered on successive days. With this design, the temporal properties of the rapidly (within hours) and slowly (over days) developing tolerance to nicotine can be evaluated. Male albino rats were trained to lever-press on a fixedratio 50 (FR 50) schedule of water reinforcement. On the first day of the study, the first IP injection of 200 ~g/kg (base) nicotine, administered just prior to the 30-min session, produced a complete suppression of responding for about 15 rain, after which the behavior abruptly returned to pre-drug rates. However, the 2nd daily injection (given l, 2, or 4 hr later) produced a much smaller effect (Fig. 8). When the sequence of twice-daily nicotine injections was continued, tolerance to the effects of the first daily injection occurred over a 9-day period until neither injection produced an effect. However, when the injection dose was increased to 350, 500, and 650/zg/kg, the differences between the 2 daily injections reappeared, even though the effect of both injections was increased as a function of dose. Note also that there was virtually no further development of tolerance to the effects of either injection across days at the three higher doses. If one can generalize from the suppressant effects of nicotine on water-reinforced lever pressing to patterns of nicotine self-injection (there is a direct relationship between pausing in cocaine self-injection and suppression of lever pressing behavior by cocaine given non-contingently [2, 7, 22]), it is possible that the reinforcing effects of nicotine may also change as a function of time since the first injection. Tentative support for this hypothesis is provided by preliminary results from our laboratory. In one study, 1440/zg/kg nicotine was injected SC either 0.5, 1, or 2 hr prior to IV nicotine self-injection sessions in rhesus monkeys. The 0.5-hr and the 1-hr nicotine pretreatments suppressed responding more than the 2-hr pretreatment, but all reduced only the rapid responding at the beginning of the session and had little effect on injection rate in the remainder of the 3-hour session. Saline pretreatment had no effect. Thus, the injection rate characteristic of most of the session was not suppressed by nicotine pretreatment, and is relatively independent of the amount of nicotine taken in the beginning of the session. This suggests that the factors that maintain responding in the latter part of the session are different from the effects that engender rapid responding at the beginning of the session and that the self-injection responding at the end of the session may not be sensitive to the amount of nicotine in the body. The development of tolerance to the effects of nicotine within 1 or 2 hr may play a role in this pattern of self-injection.
494
DOUGHERTY E7 AI..
If tolerance develops to the behavioral suppressant effects of nicotine, then why didn't self-injection rates increase over the session rather than decrease? This question cannot be answered at the present time. However, one hypothesis that may be proposed is that tolerance also occurs to the positively reinforcing stimulus effects of nicotine. The rapid development of tolerance to a variety of CNS effects of nicotine, including reinforcing effects, would be consistent with the findings pertaining to tolerance reported by other investigators participating in this symposium and congruent with the depolarization-blockade mechanism of action of nicotine at the receptor level. Thus, the behavior in the last half of the 3-hr self-injection sessions may be functionally equivalent to saline-maintained responding or may possibly be maintained by negative reinforcement contengencies. More conclusive information would be provided by an experiment in which a large, non-contingent injection or saline substitution was given a f t e r the period of rapid nicotine intake. The results of such a study may contribute to an understanding of nicotine self-administration in animals as well as smoking behavior in humans.
restricted than the situations in which drugs of the psychomotor stimulant, opiate, and sedative-hypnotic classes can be shown to be reinforcers. The manner in which the drug injections are scheduled with reference to response and temporal contingencies appears to be of considerable importance. It has been difficult to demonstrate that nicotine is a reinforcer under continuous reinforcement schedule (CRF) conditions or on fixed-ratio schedules. The most successful conditions appear to involve interval schedules, with or without supplemental external stimuli correlated with drug presentation (second-order schedule). The reinforcing properties of nicotine are more easily demonstrated in nonhuman primates than in rats. Food deprivation, strain and individual differences, circadian rhythms, and duration of exposure to nicotine are also important variables that influence the rate of nicotine self-administration. The choice of dependent variable (e.g., injection rate vs response rate) and the manner in which the data are organized or averaged may also have an influence on the interpretation of the results. Finally, the very rapid development of tolerance to some of the behavioral effects of nicotine may account for variations in temporal pattern of nicotine self-injection.
SUMMARY Published studies of intravenous and oral nicotine selfadministration have been reviewed for evidence that nicotine serves as a reinforcing stimulus for behavior leading to its presentation. The range of conditions under which nicotine demonstrates reinforcing stimulus properties is much more
ACKNOWLEDGEMENTS This research was supported in part by USPHS grants DA01778 and BRSG RR 05374-15, and by the Veterans Administration Research Program.
REFERENCES
1. Ator, N. A. and R. R. Griffiths. Intravenous self-administration of nicotine in the baboon. Fedn Proc. 40: 298, 1981. 2. Balster, R. L. and C. R. Schuster. Fixed-interval schedule of cocaine reinforcement: Effect of dose and infusion duration. J. exp. Analysis Behav. 20: 119-129, 1973. 3. Clark, M. S. G. Self-administered nicotine solutions preferred to placebo by the rat. Br. J. Pharmac. 35: 376, 1969. 4. Deneau, G. A. and R. Inoki. Nicotine serf-administration in monkeys. Ann. N. Y. Acad. Sci. 142: 277-279, 1967. 5. Domino, E. F. Electroencephalographic and behavioral arousal effects of small doses of nicotine: A neuropsychopharmacological study. Ann. N.Y. Acad. Sci. 142: 216-244, 1967. 6. Domino, E. F. and M. P. Lutz. Tolerance to the effects of daily nicotine on rat bar pressing behavior for water reinforcement. Pharmac. Biochem. Behav. 1:445 ~S, 1973. 7. Dougberty, J. and R. Pickens. Fixed-interval schedules of intravenous cocaine presentation in rats. J. exp. Analysis Behav. 20: 111-118, 1973. 8. Dougherty, J. and R. Pickens. Pharmacokinetics of intravenous cocaine self-injection. In: Cocaine: Chemical, Biological. Clinical, Social and Treatment Aspects, edited by S. J. Muir. Cleveland: CRC Press, 1976, pp. 105--120. 9. Goldberg, S. R. and R. D. Spealman. Maintenance and suppression of behavior by intravenous nicotine injections in squirrel monkeys. Fedn Proc., in press. 10. Ooidberg, S. R., R. D. Sl~mlman and D. M. Goldberg. Persistent behavior at high rates maintained by intravenous selfadministration of nicotine, Science 214: 573--575, 1981. 11. Griffiths, R. R., J. V. Brady and L. D, Bradford. Predicting the abuse liability of drugs with animal drug self-administration procedures: Psychomotor stimulants and hallucinogens. In: Advances in Behavioral Pharmacology, Vol. 2, edited by T. Thompson and P. B. Dews. New York: Academic Press, 1979, pp. 163--208. 12. Hanson, H. M., C. A. lvester and B. R. Morton. The effects of selected compounds on the self-administration of nicotine in rats. Fedn Proc. 36: 1040, 1977.
13. Hanson, H. M.. C. A. lvester and B. R. Morton. Nicotine selfadministration in rats. In: Cigarette Smoking as a Dependence Process. edited by N A Krasnegor. Washington. DC: U.S. Government Printing Office. 1979, pp. 70-90. 14. Hatchell. P. C. and A. C. Collins. Influence ofgenotyp¢ and sex on behavioral tolerance to nicotinein mice. Pharmac. Biochem. Behav. 6: 25--30. 1977. 15. Jaffe, J. H. and M. E. Jarvik. Tobacco use and tobacco use disorder. In: Psychoph~lrmacology : A Generution o f Progress. edited by M. A. Lipton. A. DiMascio and K. F. Kitlam. New York: Raven Press. 1978. pp, 1665--t676. 16. Jones, R. T., T. R. Farr©li. IIl and R. 1. Herning. Tobacco smoking and nicotine tolerance, in: SelJ:Administrution o f Abused Substances: Methods j~r Study. National Institute on Drug Abuse Research Monograph 20. edited by N. Krasnegor. Washington. DC: U.S. Government Printing Office, 1978, pp, 202-208. 17. Kumar. R.. E. C. Cooke. M. H. Lacier and M. A. H. Russell. ls nicotine important in tobacco-smoking? Clin. Pharmac. Ther. 21: 520--529. 1977. 18. Lang, W. J.. A. A. Latiff, A. McQueen and G. Singer. Selfadministration of nicotine with and without a food delivery schedule. Pharmac. Biochem. Behav. 7: 65-70. 1977. 19. Lucchesi, B. R.. C. R. Schuster and G. S. Emley. The role of nicotine as a determinant of cigarette smoking in man with observations of certain cardiovascular effects associated with the tobacco alkaloid. Clin. Pharmac. Ther. g: 789-796. 1967. 20. Morrtson, C. F. and J. A. Stcphenson. The occurrence of tolerance to a central depressant effect of nicotine. Br. J. Pharmac. 45: 151-156, 1972. 21. Pickenso R, and W. C. Harris. Self,administration of d-amphetamine by rats. Psychopharmacologia 12: 158-163. 1968.
22. Pickens, R, and T. Thompson. Cocaine reinforced behavior in rats: Effects of reinforcement magnk-'-ud¢andfixed-ratio size. J. Pharmac. exp. Ther. 161: 122-129. 1968.
BEHAVIORAL EFFECTS OF NICOTINE 23. Rosecrans, J. A. Nicotine as a discriminative stimulus to behavior: its characterization and relevance to smoking behavior. In: Cigarette Smoking as a Dependence Process, edited by N. A. Krasnegor. Washington, D.C.: U.S. Government Printing Office, 1979, pp. 58-69. 24. Russell, M. A. H. Tobacco smoking and nicotine dependence. In: Research Advances in Alcohol and Drug Problems, Vol. 3, edited by R. J. Gibbins, Y. Israel, H. Kalant, R. E. Popham, W. Schmidt and R. G. Smart. New York: Wiley and Sons, 1976, pp. 1--47. 25. Russell, M. A. H. Tobacco dependence: Is nicotine rewarding or aversive? In: Cigarette Smoking as a Dependence Process, edited by N. A. Krasnegor. Washington, DC: U.S. Government Printing Office, 1979, pp. 100-122. 26. Smith, L. A. and W. J. Lang. Changes occurring in selfadministration of nicotine by rats over a 28-day period. Pharmac. Biochem. Behav. 13: 215--220, 1980. 27. Spealman, R. D. and S. R. Goldberg. Drug self-administration by laboratory animals: Control by schedules of reinforcement. A. Rev. Pharmac. Toxicol. 18: 313-339, 1978. 28. Stolerman, I. P., P. Bunker and M. E. Jarvik. Nicotine tolerance in rats: Role of dose and dose interval. Psychopharmacologia 34: 317-324, 1974.
495 29. Stolerman, 1. P., R. Fink and M. E. Jarvik. Acute and chronic tolerance to nicotine measured by activity in rats. Psychopharmacologia 30: 329-342, 1973. 30. Todd, G. D. and J. Dougherty. Dose-dependent tolerance to a behavioral effect of nicotine. Soc. Neurosci. A bstr. 6: 544, 1980. 31. Wikler, A. Dynamics of drug dependence. Implications of a conditioning theory for research and treatment. Archs gen. Psychiat. 28: 611--616, 1973. 32. Wilson, M. C., M. Hitomi and C. R. Schuster. Psychomotor stimulant self-administration as a function of dosage per injection in the rhesus monkey. Psychopharmacologia 22: 271-281, 1971. 33. Yanagita, T. An experimental framework for evaluation of dependence liability of various types drugs in monkeys. Bull. Narcotics 25: 57-64, 1972. 34. Yanagita, T. Brief review on the use of self-administration techniques for predicting drug abuse potential. In: Predicting Dependence Liability of Stimulant and Depressant Drugs. edited by T. Thompson and K. Unna. Baltimore: University Park Press, 1977, pp. 231-242. 35. Yokel, R. A. and R. Pickens. Self-administration of optical isomers of amphetamine and methylamphetamine by rats. J. Pharmac. exp. Ther. 187: 27-33, 1973.