Neuroscience & Biobehavioral Reviews, Vol. 6, pp. 77-83, 1982. Printed in the U.S.A.

Schedule-Induced Self-Injection of Drugs I

G E O R G E S I N G E R , T I A N P. S. O E I 2 A N D M E R E D I T H W A L L A C E

Department of Psychology, La Trobe University, Bundoora, Australia 3083

R e c e i v e d 19 A u g u s t 1980

SINGER, G., T. P. S. OEI AND M. WALLACE. Schedule-induced self-injection of drugs. NEUROSCI. BIOBEHAV. REV. 6(1) 77-83, 1982.--Studies of acquisition and maintenance patterns of drug intake, including dependence, necessi- tate voluntary intake of drugs. Voluntary intake in animal studies is difficult to achieve because of aversive taste factors associated with most drugs, and involuntary or forced choice methods obscure the behavior which is the object of study. The schedule-induced polydipsia paradigm has been used to induce oral ingestion of large volumes of alcohol, barbiturate and other drug solutions. We have developed a method of schedule-induced self-injection which allows the study of acquisition and maintenance of drug intake behavior with changing environments free from the interference of taste factors or imbalances due to excessive water intake. In this paper we review our findings on the acquisition and maintenance patterns of amphetamine, methadone, heroin, alcohol, nicotine, cocaine, Ag-THC and haloperidol. For all drugs except amphetamine, the combination of schedule and nutritional deprivation leads to the highest rates of drug intake as compared to controls. The schedule does not appear to be a potent factor at 90% and free feeding weight, and drug.i_ntake is the result of interaction of environmental factors and pharmacological properties of the drugs, rather than the efl'ect~ ofdrug or environmental factors separately. The maintenance patterns of nicotine, cocaine, heroin and alcohol are also discussed and the advantages of schedule-induced self-injection over schedule-induced polydipsia methods are presented.

Schedule Self-administration Heroin Methadone Ethanol Amphetamine A g - T H C Cocaine

MOST animal studies of drug intake have been concerned with dependence, tolerance and withdrawal symptoms in order to learn something about the properties of a variety of psychoactive drugs and to understand drug-taking behavior. Interest in the behavioral aspects of human drug dependence has increased the need for an animal model incorporating voluntary intake of drugs. Initial research with animals uti- lized involuntary injections or inhalation, and semi-voluntary (forced choice) methods. Schuster [18] notes an early belief that "only under the physical distress of withdrawal would animals self-administer a drug" (p. 68).

Schuster 's quotation highlights the fact that hypotheses about dependence and addiction were inferred from the occurrence of withdrawal symptoms as it is impossible to study dependence behavior directly with involuntary intake methods. These inferences were based on the assumption that cessation of drug intake is always followed by with- drawal, which is contradicted by some recent evidence [9,10]. A problem with voluntary ingestion is that drugs may have an aversive taste for naive animals. This is borne out by the fact that when the flavor is disguised with saccharin or fruity essences the animal will usually drink moderate quantities of drug solutions [2, 5, 17]. In these cases, dependencies are diffi- cult to distinguish from the learning of new taste preferences.

Concern with acquisition of drug-taking behavior and with psychological dependence, as well as with the basic mechanisms underlying physical dependence, tolerance and withdrawal has led to the development of self-administration techniques [26]. Since drug intake and drug dependence are behavioral responses, a direct test of dependence and drug intake is impossible without voluntary administration of drugs. Unfortunately, for many types of drugs, intake through

self-injection is too low to produce physical dependency [1]. Another method of voluntary drug intake which has been successfully used to induce the intake of large quantities of drugs is schedule-induced polydipsia (SIP) (e.g. [4, 5, 8]).

In our laboratory, we have combined the self-injection technique with a food delivery schedule and developed the method of schedule-induced self-injection (SISI). This al- lows us to observe the acquisition and maintenance patterns of dependence for most psychoactive drugs, as the SISI paradigm seems to overcome the reluctance of animals to self-inject some of these drugs. In this procedure, rats re- duced below free feeding weight (FFW) and implanted with venous cannulae are placed in operant chambers with the drug available intravenously through bar pressing. A noncon- tingent fixed-time 60 sec (FT-60) food delivery schedule oper- ates throughout the test session.

The general experimental design for the studies that we are reporting here is shown in Table 1. This design allows the comparison of a drug with a placebo and the observation of nutritional factors and schedule effects on drug intake as well as their interactions. Prior to reviewing the results of the schedule-induced self-injection experiments using a number of drugs, we present a brief account of the general methodol- ogy. Detailed procedures can be found in the individual papers. For these series of experiments male rats only were used. They were housed individually in single wire-mesh cages in a room with a 12 hour light/dark cycle with tempera- ture thermostatically controlled at 21-+1 °C. Animals re- quired at 80% free feeding weight were reduced to this weight over a period of 5 days by restricting food intake and were maintained at this 80% F F W throughout the experiment by limiting their food supply. They were normally tested for

tThis is one of a series of invited Theoretical Review articles on the general topic of adjunctive behavior. ZPresent address: Department of Psychology, University of Otago, Dunedin, New Zealand.

C o p y r i g h t © 1982 A N K H O In t e rna t i ona l Inc . - -0149-7634/82/010077-07503.00/0

78 SINGER, OEI AND WA LLACE

1 hour/day at almost the same time each day, with the ex- ception of the Lang et al. [7] experiment where they were tested for 2 hours/day. An initial priming dose of drug or saline solution was administered prior to each testing ses- sion. No liquid was available in the test chamber, but water was freely available in the home cages. Prior to each test ses- sion, animals were weighed and the catheters were flushed with a heparinized saline solution to prevent blood clotting in them.

Sodium pentobarbital and chloral hydrate (1.7 ml/kg) was generally used as an anaesthetic. Implantation of the catheter involved removing hair around the neck and exposing the left jugular vein. A polythene catheter (CSP 28 Dural plastic) was passed through the skin at shoulder level and drawn subcu- taneously towards the exposed vein. The vein was slit and the catheter inserted, terminating above the left auricle of the heart. The catheter was anchored to the midline neck mus- cles and the wound closed. A leather jacket worn by each animal was used to maintain the catheter in position. The catheter was connected to a flexible swivel system allowing each animal unrestricted movement. All animals were al- lowed 3 days to recover from surgery before being assigned to an experimental group.

A modified operant chamber made of clear Plexiglas was used. It contained a bar and a food pellet dispensing unit at- tached to one of the side walls. The bar was situated 5 cm and the pellet dispensing unit 3 cm above the grid floor and located 10 cm from the corner. The bar operated a syringe infusion pump which delivered 0.7 ml of solution when trig- gered. A timing device set for a fixed interval of 5 sec was incorporated into the drug delivery system so that any fur- ther bar presses during the 5 sec infusions were not rewarded with drug or saline injections. Cumulative recorders were used to record the number of bar presses and infusions dur- ing test sessions. When a fixed time 1 rain schedule was operating, Noyes food pellets (4.5 gm) were delivered regu- larly, one each minute, to the animals.

In general each drug was freshly prepared for intra- venous administration prior to each test session by dis- solving it in 0.9% sterile saline. However, for Ag-THC, a solution of Tween 80 in physiological saline was used. The following drugs have been tested and are reported on here:

Diacetyimorphine 0.1 mg/kg/infusion Victorian hydrochloride Health

Department Nicotine hydrogen (+) 0.1 mg/kg/infusion B.D.H. Ltd.

tartrate Methadone 0.1 mg/kg/infusion Wellcome D-amphetamine sulfate 0.5, 0.2, Sigma

0.8 mg/kg/infusion Chemical 0.05 mg/kg/infusion CSR Chemical O. I mg/kg/infusion Johnson O. 1 mg/kg/infusion Mcfarlan

Smith Ltd. 6.25, 12.5, 25.50 NIDA #.g/kg/infusion

Ethanol (ethyl ethanol) Haloperidol Cocaine hydrochloride

Ag-THC

Analysis of variance was the main statistical analysis used for these series of studies. Post hoc analyses were also used to detect specific differences.

We have found that some drugs which are not readily self-infused when the animal is at free feeding body weight

TABLE 1 EXPERIMENTAL PARADIGM USED FOR SIS1 STUDIES

(INVOLVING 8 INDEPENDENT GROUPS)

Drug Saline

100% FFW 81)% FFW 100% FFW 80% FFW and FT-60 and FT-60

no schedule no schedule schedule schedule

are taken in significantly increased amounts at 80cA F F W (e.g. amphetamine and cocaine). With most other drugs, self- infusion is significantly increased in an 80% F F W animal by the addition of a FT-60 food delivery schedule. The design used in these studies not only pehnits the observation of acquisition patterns for self-infusion of different drugs but, by manipulating body weight or varying the schedule after acquisition, patterns of maintenance and voluntary extinc- tion of the self-injection behavior can also be studied.

In this review, we report observations on the acquisition patterns for self-infusion of nicotine, alcohol, amphetamine, heroin, methadone, haloperidol, Ag-THC and cocaine, and the maintenance patterns of nicotine, alcohol and heroin self- infusion using the SISI procedure.

ACQUISITION PATTERNS OF DRUG INTAKE

Alcohol

The acquisition pattern of alcohol self-infusion for four groups of animals is presented in Fig. 1. The data show that the level of self-infusion of alcohol for 80% F F W rats under the FT-60 condition was greater than for rats at 80% F F W with no schedule and rats at 100% with or without a schedule. Body weight reduction alone (80%) produced a higher level of self-infusion than m free-feeding rats (100%) with and without a FT-60 schedule. As can be seen from the figure, the 80% FT-60 group showed a steep increase in alco- hol intake by Day 3 and remained at that level to the end of the experiment on Day 10. This pattern was not observe in the other 3 groups [14].

Amphetamine

The acquisition pattern of d-amphetamine for 5 groups of animals under 80% and 90% F F W with or without a food delivery schedule and 100% with no schedule is shown in Fig. 2 (top). As can be seen from Fig. 2, the highest level of amphetamine self-infusion occurs with 80% F F W animals without a FI'-60 schedule and the lowest with 100% FFW. Animals at 80%+FT-60, and 90%+FT-60 and 90% no sched- ule fall between the 80% and 100% F F W conditions [22]. This pattern of intake is maintained for three dose levels of the drug as shown in Fig. 2 (bottom) although the overall level of drug intake varies with the dose. The lowest dose used (0.05 mg/kg) showed the highest self-infusion level, with decreases as dose level increased. This is a puzzling exception in the SISI patterns when compared to other drugs tested so far and may be related to the anorectic properties of amphetamine.

Heroin

The level of heroin acquisition is shown in Fig. 3. The

SELF-INJECTION OF DRUGS 79

o ,-~o B0%+ FT- 60

- - - - . 4 80% NO SCHEDULE

o o 100%+ FT-60 - - - ,-e 100% NO SCHEDULE

30

_ _ . . , + _ . _ _ _

DAYS

FIG. 1. Mean number of alcohol self-injections (0.05 mg/kg/infusion) for the four treatment groups over 10 days (modified from [14]).

60"

SO'

~ 3 0

z 20

10

8 0 * / ~ F T - 6 0

m . - - + - - 4 1 80% NO SCHEDULE

¢ o 100%o F T - 6 0

e - - - - 4 1 0 0 % NO SCHEDULE

(3

1 2 3 z. 5 6 7 8 9 10 DAYS

70

6O

o s0

o ~ . 0

z

z< 3o

2O

l

10'

o o B0%+FT- 60

I - - - - -41 80% NO SCHEDULE

,~, A 90% + FT- 60

A - - - - -& 90% NO SCHEDULE

e- . . . . ,.e 100"/* NO SCHEDULE

] •

I /

/ / /

/

i !

/ /

1 2 3 4 5 6 DAYS

÷ ~ G Id

2L

~ 2 2

~ 8

~ 6 I [

r. 1 (8)

1 J [ ] 8 0 % + F T - l r n , n

!iii!i!i I , l ,00 %

L 0-05 O? 0'I~ 1 SALINE d-AMPHETAMINE

FIG. 2. (Top) Mean number of d-amphetamine self-injections (0.05 mg/kg/infusion) for the live treatment groups over 6 days. (Bottom) Mean number of d-amphetamine (0.05, 0.2, 0.8 mg/kg/infusion) and saline self-injections for three treatment groups (from [22]). *Refers to significant difference (p<0.05). ( ) is number of animals in each group.

FIG. 3. Mean number of heroin self-injections (0.1 mg/kg/infusion) for the four treatment groups over 10 days (modified from [13] and [151).

level of heroin self-infusion for the 80%+FT-60 group was significantly higher than the 80% no schedule group, which in turn self-infused more heroin than rats at 100% FFW with or without a FT schedule. The difference in heroin intake be- tween the 100% and 100%+Fr-60 groups was not significant. Overall, animals at 100% and 80% FFW with or without a schedule self-infused significantly more heroin (see Fig. 3), when compared with the appropriate saline control groups (Fig. 4) [13,15].

Methadone

The patterns of methadone self-administration are shown in Fig. 5. The highest level of self-infusion of methadone was in the 80%+FT-60 group. Although the mean intake for the 80% FFW group was considerably higher than for the two 100% FFW groups, there were no significant differences in the amount of methadone self-infused among the 80% with no schedule and the two 100% groups [13,15].

Nicotine and Aa-Tetrahydrocannabinol

The pattern of nicotine intake is shown in Fig. 6a and 6b and is similar to methadone, with self-infusion levels at 80% and 100% FFW not different from those of the three placebo groups. However, the 80%+FT-60 self-injected significantly more nicotine than the controls [7]. Of the four doses of Ag-Tetrahydrocannabinol (Ag-THC) studied, only the two lowest concentrations of 6.25 and 12.5 /~g/kg, respectively show significantly higher self-infusion levels than saline con- trols when made available to animals at 80% FFW with a FT-60 food delivery schedule (see Fig. 7). None of the four concentrations was self-infused more than saline at 100% FFW with a FT-60 schedule. The acquisition pattern for AY-THC seems similar to nicotine and methadone, but a proper comparison is not possible since no data are available for 80% FFW without a schedule. The findings that the mean number of infusions was not as high as seen with narcotic analgesics and d-amphetamine using the same design sug-

80 SINGER, OEI AND WALLACE

TABLE 2 SUMMARY OF THE LEVELS OF DRUG INTAKE USING A SCHEDULE-INDUCED

SELF-INJECTION PARADIGM

80% FFBW 100% FFBW Drugs References 80% FFBW + FT-60 100% FFBW + FT-60

Alcohol 14 I"T ~']'~" NS NS (0.05 mg/kg)

D-Amphetamine 22 TTT ]'T TT NS (0.05 mg/kg)

Cocaine 16 TTT TTT NS NS (0.1 mg/kg)

Haloperidol 11 NS NS NS NS Heroin 13, 15 TT]' TTT TT TT

(0. l mg/kg) Nicotine 7 NS ~'TT NS N S

(0.01 mg/kg) Methadone 13, 15 NS I']'T NS NS

(0.1 mg/kg) Ag-THC 21 NS T~T NA NS

(12.5/~g/kg)

TT=Significantly higher than saline control (o<0.05). l'l'l'=Significantly higher than saline control (p<0.01). NS=Similar to that of saline control. FFBW=Free feeding body weight. FT-60=Fixed time 60 minutes food delivery schedule. NA=No data available.

~ ~ ~-~ ~ _ 3° 12O~o

o a 8 0 % * FT. 60

• . . . . - I 8 0 % NO S C H E D U L E

o o 100% • FT - 60

. . . . • 100% NO S C H E D U L E

I 2 3 ~ 5 6 7 B 9 10 Ot tYS

FIG. 4. Mean number of saline self-injections for the four treatment groups over 10 days (modified from [13] and ]15]).

50

o~ ~O

m 30 + z 20

13 O B 0 % ÷ FT- 60

I - . . . . 41 8 0 % NO SCHEDULE

I - . . . . .o 1 0 O % * F T - 60

~ _ . . ~ _ ~ . - . -- --+ . . . . . + . . . . J

1 2 3 ~ 5 5 7 8 9 10 DAYS

FIG. 5. Mean number of methadone self-injections (0.1 mg/kg/infu- sion) for the four treatment groups over 10 days (modified from [13] and [15]).

gests that Aa-THC in rats may have weak reinforcing proper- ties [20,21].

Haloperidol

The mean intakes of haloperidoi at a dose level of 0.1 mg/ kg for the four treatment groups are presented in Fig. 8. Analysis of variance with repeated measures showed that the levels of self-injection were not significantly different from those of the saline group.

Cocaine

The self-infusion patterns of this drug for the four groups are presented in Fig. 9. The highest levels of self-infusions were the 80% FFW animals with and without a schedule. The presence of an FT-60 food delivery schedule did not change significantly the amount of cocaine intake in either the 80% or the 100% FFW animals [16].

CHARACTERISTICS OF DRUG INTAKE LEVELS AND PATTERNS OF ACQUISITION

The levels of drug intake for the drugs discussed above are summarized in Table 2. There are interesting correspond- ences and differences in the patterns of acquisition for am- phetamine, cocaine and heroin. Amphetamine and heroin are similar in that the schedule is not essential for the acquisition of this behavior. There is a dissimilarity between stimulant and narcotic analgesics in that for amphetamine and cocaine, reduced body weight is sufficient to produce very high self- infusion rates, whereas for heroin, the three-way interaction between deprivation, schedule and pharmacological properties of the drug is necessary. Similarly, a three-way interaction was also necessary for high intake of alcohol and Ag-THC.

SELF- INJECTION OF DRUGS 81

5 0 -

1,o.

'~ 30'

20'

1o. i-

5 0 ~

3 0

2o

10" o

A NICOTINE

O n 8 0 % * F T - 6 0

. . . . . . . . . • 80% NO SCHEDULE - -e 100% NO SCNEOULE 16

l Z ,

12.

i ....... [ I, ~ ..... " ,1, ~10

' ....... t ....... ~ ..... ' ..... ':ff:::""r ~8 1 2 3 1, 5 6

DAYS _z

B: SALINE B

6,

o a 8 0 % * FT- 60

• . . . . . . . . • 80% NO SCHEDULE ~ /." z

. . . . . . . 4 100% NO SCHEDULE z

~2-

I 2 3 1, 5 6 DAYS

FIG. 6. (A) Mean number of nicotine self-injections by three treat- ment groups over six days (for 2 hours per day). Vertical bars refer to standard deviation. (B) Mean number of saline self-injections by three treatment groups over six days (for 2 hours per day) (from [7]).

The levels of responding for methadone and nicotine (Table 2) show almost identical patterns; for these drugs the schedule is a necessary condition if the animals are at 100% FFW. Alcohol intake is also highest at 80%+FT-60 and the levels of alcohol intake at 100% F F W with or without the schedule was not significantly different from those of saline controls, but alcohol resembles amphetamine, heroin and cocaine in that a state of deprivation alone will lead to in- creased drug intake. For methadone and nicotine, the nutri- tional state of the animal is important only in interaction with the drug and environmental factors (in this case, the sched- ule). Rats, however, did not self-inject haloperidol under the conditions studied.

The data obtained from the SISI paradigm on the devel- opment of drug self-infusion strongly suggest that an inter- action exists between pharmacological properties of the drug, the nutritional state of the animal, the presence of a schedule, and that environmental conditions affect the level of drug intake in different ways.

In summary, the findings show that: (1) for alcohol, cocaine, amphetamine and heroin, re-

duced body weight alone is sufficient to induce self-infusion of significant quantities of the drugs when compared to con- trol groups;

8 0 % + F T - 6 0

( 8 )

z m ~625 125 25 50,

"~ ~,g - T H C u1

100% * FT-60

( 8 )

zw 16-2512"5 2S 50

A 9 -THC u')

Dose (.I.~g t k g)

FIG. 7. The overall mean infusions for the control and Ag-THC solutions under conditions of 80% and 100% free feeding body weight in the presence of a fixed time 60 sec food delivery schedule. *p<0.05, **p<0.01 student's t-test, one tail compared to control group. ( ) denotes the number of animals in each group.

1 =o

~- 20 o

2 _z 10

o 43 80% * FT- 60

m- . . . . 4 80% NO SCHEDULE

o -o 100%*FT- 60

o - . . . . -o100% NO SCHEDULE

- • " ' - - " l . . . . . . . " t ' " ~. . ' 7 . . . . . . . . ~ . . . . . t o ° -

I 2 3 z, S 6 7 B 9 10 DAY5

FIG. 8. Mean number of haloperidol self-injections (0.1 mg/kg/infu- sion) for the four treatment groups over 10 days.

(2) for all drugs except amphetamine, the co-occurrence of schedule and deprivation leads to the highest rates of drug intake, whereas the schedule does not appear to be a potent factor at 100% F F W ;

(3) the environmental conditions interact differently with different drugs to produce different levels of drug intake [6, 13, 14, 15, 22]; and

(4) drug intake is usually the result of interaction of envi- ronmental factors and pharmacological properties of drugs rather than the effects of drug or environmental factors sepa- rately [6,16].

82 SINGER, OEI AND WALLACE

gO

7 0

60 ¸

g o = 50

~ 4()'

z_

~ 30 '

2 0 '

10.

o o 8 0 % * FT- 60

II~ II.- . . . . . • 80% NO SCHEDULE

, , 0 0 100°lo*FT-60

/ i l l I - . . . . . " -e100% NO SCHEDbLE

f , / ,

f

I 2 3 ~ 5 6 7 8 9 10 DAYS

FIG. 9. Mean number of cocaine self-injections (0.1 mg/kg/infusion) for the four treatment groups over 10 days (from [16]).

MAINTENANCE PAI-rI~RNS OF DRUG INTAKE

The study of maintenance patterns of drug intake is only in its infancy and we report here four studies as an illustra- tion of the methodology, which can be expanded in many directions. Maintenance is used here to refer to drug intake behavior once dependence has developed. Our work on nicotine [19] shows that once nicotine intake is established, the effect of body weight is minimal, i.e., the reversal of body weight from 80% to 100~ does not change the rate and amount of nicotine intake. However, in the case of cocaine [16], the effect of body weight on cocaine self-infusion is crucial. The change of body weight has a significant impact on the rate and the amount of cocaine self-infused. Interest- ingly, however, neither an FT-60 food delivery schedule nor the order of presentation of schedule influences the mainte- nance rates of cocaine self-administration [16]. The removal of the schedule does change the level of alcohol intake in animals previously induced to self-infuse high levels of alco- hol [14]. Rats dropped to the control level of alcohol intake approximately 2 days after removal of the schedule. Similar results were observed with heroin self-infi~ion: cessation of a FF-60 food delivery disrupted, but did not extinguish, her- oin self-infusion behavior [9,10]. When saline was substituted for heroin following a 21 day period of schedule induced heroin self-injection the rats continued to self-inject saline at a level similar to the heroin self- injection [9]. This is further strong confirmation of the environment drug interaction hy- pothesis proposed in this paper, and has important implica- tions for the clinical management of heroin addicts.

DISCUSSION

The data from schedule-induced self-injection experi- ments lead us to re-evaluate the concepts of physical and psychological dependence. These data allow an independent conceptualization of environmental factors which interact with pharmacological factors in the acquisition and mainte- nance of drug intake. This model has been fully described by

Jefferys et al. [6] and should serve as a starting point for the exploration of other drug acquisition and maintenance pat- terns. It has the distinct advantage of allowing more precise specification of environmental factors instead of the present reliance on vaguely defined concepts such as psychological and physical dependence. Furthermore, the possibility of di- rect observation of drug seeking behavior makes it unneces- sary to rely on an assumption that dependence results from the alleviation of withdrawal symptoms.

It appears that, in general, a FT-60 food delivery schedule accelerates the acquisition of drug self-injection behavior except in the case of amphetamines and cocaine. These are anoretic drugs and therefore the presence of the food is ir- relevant and in fact may serve as a distractor in the acquisi- tion of the behavior. What is it about food delivery schedules which makes them so potent in the acquisition of drug-taking behavior? One may speculate that the schedule introduces a stress factor and that both water intake as well as intake of drugs are behaviors which relieve stress. This is supported by an experiment of Falk et al. [4] which showed that scheduled delivery of food leads to an increase in the blood pressure response of mononephrectomized rats after 12 days.

SIS I Advantages

The SISI technique has methodological advantages over the schedule-induced polydipsia paradigm in the investiga- tion of the development of dependence upon psychoactive drugs. Although the noxious taste of such drugs is apparently overridden by the schedule in the SIP paradigm, the effect of the taste factor persists in all non-scheduled conditions, lim- iting the environmental conditions which can be investi- gated. The elimination of the inhibition caused by the taste factor with the SISI technique has made it possible to show that dependence on psychoactive compounds can develop under a variety of environmental conditions. The results of studies using the SISI method suggest that environmental conditions which enhance voluntary drug intake interact with the pharmacological properties of the drug and may provide necessary preconditions for this enhancement. Fu- ture studies may show that under different environmental conditions, different pharmacological properties of the psy- choactive drug operate to enhance drug intake.

Since it has been shown that schedule-induced behaviors also occur in human subjects (e.g. [23, 24, 25]), the findings from SISI animal work are likely to be relevant to the analy- sis of the problem of drug abuse in man. The acquisition and maintenance patterns reported so far could be used for pre- diction of the abuse potential of new psychoactive drugs. Specific drug/environmental interactions could form the basis for treatment programs. For example, SlSI findings on nicotine intake show that mere availability of the drug is not sufficient to lead to the acquisition and maintenance of this behavior. This may account for the failure of treatment pro- grams involving either nicotine-free cigarettes or nicotine- containing chewing gum or pads (instant smokes). In one case the drug has been removed, in the other case the envi- ronmental factors have been changed and the interaction which is crucial for the maintenance of the behavior may require different substitutions. The findings of differences in acquisition and maintenance patterns of drug use may re- quire reconsideration of legal and administrative attitudes to drug intake behavior. Well-known deficiencies such as the equal legal treatment of soft and hard drug offenders and the

SELF-INJECTION OF DRUGS 83

basis of the classification into 'soft ' and 'hard' need to be re-examined.

The SISI paradigm also allows for expansion into a num- ber of areas of research which have presented difficulties in the past. In our laboratories we have commenced a number of projects which go beyond the determination of acquisition and maintenance patterns of different drugs. These include determination of the reinforcement value of drugs using a period of acquisition of the drug self-injection behavior and switching the animal to saline self-injections. The subse- quent increase of saline self-injections over the base level provided by a saline control group can be used as an index of reinforcement value of the drug [9].

Through pretreatment prior to self-injection with a range of putative transmitter agonists or antagonists, it is possible to determine the substrate involved in the drug self-injection behavior more precisely [12]. In particular, where a drug has

multiple pharmacological properties, the one involved in the acquisition and maintenance of the drug taking behavior can be specified through these experiments.

Thus Takahashi has shown that the substrate necessary for the maintenance of Ag-THC intake is dopaminergic and that a and/3 noradrenergic and serotonergic substrates seem of little importance [20]. Finally, by using a crossover de- sign, the facilitatory effect of one drug on the self-injection pattern of another one can be determined. Preliminary work in our laboratory shows only minimal interaction between Ag-THC and amphetamine.

In summary, SISI provides a new model for the investi- gation of drug intake behavior, and may prove useful in advancing the understanding of the complex interaction be- tween environmental and pharmacological factors that is in- volved in this behavior.

REFERENCES

I. Deneau, G. A. and R. Inoki. Nicotine self-administration in monkeys. Ann. N.Y. Acad. Sci. 142: 227-229, 1967.

2. Everett, P. B. and R. A. King. Schedule-induced alcohol inges- tion. Psychon. Sci. 18: 278--279, 1970.

3. Falk, J. L., H. H. Samson and G. Winger. Behavioral mainte- nance of high concentrations of blood ethanol and physical de- pendence in the rat. Science 177: 811-813, 1972.

4. Falk, J. L., M. Tang and S. Forman. Schedule induced chronic hypertension. Psychosom. Med. 39: 252-263, 1977.

5. Falk, J. L. and H. H. Samson. Schedule-induced physical de- pendence on ethanol. Pharmac. Rev. 27: 449--464, 1976.

6. Jefferys, D., T. P. S. Oei and G. Singer. A reconsideration of the concept of drug dependence. Neurosci. Biobehav. Rev. 3: 149-153, 1979.

7. Lang, W. J., A. A. Latiff, A. McQueen and G. Singer. Self administration of nicotine with and without a food delivery schedule. Pharmac. Biochem. Behav. 7: 65-70, 1977.

8. Leander, J. O. and D. E. McMillan. Schedule induced narcotic ingestion. Pharmac. Rev. 27: 475--487, 1976.

9. Madden, C., T. P. S. Oei and G. Singer. The effect of schedule removal on the maintenance of heroin self-injection. Pharmac. Biochem. Behav. 12: 983--986, 1980.

10. Madden, C., G. Singer and T. P. S. Oei. The involvement of interoceptive factors in the maintenance of heroin-seeking be- havior. Pharmac. Biochem. Behav. 11:445 448, 1979.

11. Oei, T. P. S. Schedule induced self-injection of haloperidol. Unpublished manuscript, Department of Psychology, La Trobe University, 1980.

12. Oei, T. P. S. Reversal of schedule-induced self-injection of her- oin by naloxone. Pharmac. Biochem. Behav. 13: 457--459, 1980.

13. Oei, T. P. S., G. Singer, D. Jefferys, W. Lang and A. Latiff. Schedule induced self injection of nicotine, heroin and metha- done by naive animals. In: Stimulus Properties o f Dru,~,s: Ten Years o f Progress, edited by F. C01paert and J. Rosecrans. Amsterdam: Elsevier, North Holland, 1978. pp. 503--516.

14. Oei, T. P. S. and G. Singer. The effects of a fixed time schedule and body weight on ethanol self administration. Pharmac. Bio- chem. Behav. 10: 767-770, 1979.

15. Oei, T. P. S., G. Singer and D. Jefferys. The interaction of a fixed-time food delivery schedule and body weight on self-ad- ministration of narcotic analgesics. Psychopharmacology 67: 171-176, 1980.

16. Papasava, M., T. P. S. Oei and G. Singer. Low dose cocaine self-administration by naive rats: Effects of body weight and a fixed-time one minute food delivery schedule. Pharmac. Bio- chem. Behav. 15: 485--488, 1981.

17. Samson, H. H. and J. L. Falk. Alteration of fluid preference in ethanol-dependent animals. J. Pharmac. exp. Ther. 190: 365- 376, 1974.

18. Schuster, C. R. Self-administration of drugs. In: Bayer- Symposium IV, Psychic Dependence. edited by L. Golderg and F. Hoffmeister. Bedim Springer-Verlag, 1973, pp. 68-74.

19. Singer, G., F. Simpson and W. J. Lang. Schedule induced self injections of nicotine with recovered body weight. Pharmac. Biochem. Behav. 9: 387-389, 1978.

20. Takahashi, R. N. Behavioural and pharmacological effects of AS-tetrahydrocannabinol self-administration in animals. Unpub- lished Doctor of Philosophy Thesis, La Trobe University, 1980.

21. Takahashi, R. N. and G. Singer. Self administration of A,~. tetrahydrocannabinol by rats. Pharmac. Biochem. Behav. 11: 737-741, 1979.

22. Takahashi, R. N., G. Singer and T. P. S. Oei. Schedule induced self-injection of D-amphetamine by naive animals. Pharmac. Biochem. Behav. 9: 857-862, 1978.

23. Wallace, M. and G. Singer. Schedule induced behavior: A re- view of its generality, determinants and pharmacological data. Pharmac. Biochem. Behav. 5: 483--490, 1976.

24. Wallace, M. and G. Singer. Adjunctive behavior and smoking induced by a maze solving schedule in humans. Physiol. Behav. 17: 849-852, 1976.

25. Wallace, M., A. Sanson and G. Singer. Adjunctive behavior in humans on a food delivery schedule. Physiol. Behav. 20: 203- 204, 1978.

26. Weeks, J. R. Experimental morphine addiction: Method for automatic intravenous injections in unrestrained rats. Science 138: 143-144, 1962.