Effects of acute and repeated cocaine administration on fixed-interval responding and schedule-induced polydipsia

MISSING IMAGE

Material Information

Title:
Effects of acute and repeated cocaine administration on fixed-interval responding and schedule-induced polydipsia
Physical Description:
vii, 169 leaves : ill. ; 29 cm.
Language:
English
Creator:
Anderson, Karen G., 1967-
Publication Date:

Subjects

Subjects / Keywords:
Psychology thesis, Ph.D   ( lcsh )
Dissertations, Academic -- Psychology -- UF   ( lcsh )
Genre:
bibliography   ( marcgt )
non-fiction   ( marcgt )

Notes

Thesis:
Thesis (Ph.D.)--University of Florida, 1998.
Bibliography:
Includes bibliographical references (leaves 163-168).
Statement of Responsibility:
by Karen G. Anderson.
General Note:
Typescript.
General Note:
Vita.

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 029040433
oclc - 40059435
System ID:
AA00014287:00001

Full Text













EFFECTS OF ACUTE AND REPEATED COCAINE ADMINISTRATION ON
FIXED-INTERVAL RESPONDING AND SCHEDULE-INDUCED POLYDIPSIA










By

KAREN G. ANDERSON













A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE
REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY

UNIVERSITY OF FLORIDA

1998















ACKNOWLEDGMENTS


For first exposing me to behavior analysis in the introductory course and through

undergraduate research, I would like to thank Christine Hughes and Kevin Jackson,

respectively. A very special and deeply felt thank you is due Ed Malagodi who helped

set my life on a path greatly influenced by radical behaviorism. His memory and

teachings will never be forgotten. My experiences in graduate school would also not

be so warmly remembered if it were not for the support and camaraderie I felt with my

fellow students, especially David Stafford, Diana Walker, Manish Vaidya, Cynthia

Pietras, Nikki Keefer, and Shonnie Turner.

Throughout all the good times and bad times, I am incredibly grateful to have had

the support and confidences of Matthew Athan and Peg Gratton. My parents Jon and

Jean Anderson have also been wonderful with providing support (financial and

emotional) throughout my graduate training. I also thank Bryan Anderson, my

brother, for always having a joke ready and always making me laugh.

For all their help with managing incredible amounts of data for this dissertation, I

thank all of my undergraduate assistants, especially Joseph Ciaravella, Crystal Lewis,

and Amy O'Quinn. Of course, none of this could have been accomplished without the

opportunities and support given to me by my advisor and committee chair, Frans van

Haaren. Lastly, for their time, teachings, and thoughtful comments, not only on this
ii









dissertation, but also throughout my graduate training, I extend a tremendous amount

of appreciation to my committee members: Marc Branch, Timothy Hackenberg, Brian

Iwata, Mark Lewis, Henry Pennypacker, and Donald Stehouwer.













































iii














TABLE OF CONTENTS


A CK N O W LED G M EN TS ......................... .......................................................... ii

ABSTRACT ....................... .. ........................ ......... vi

GENERAL INTRODUCTION ........... ...... .................................................. 1
Description of Fixed-Interval Responding............ ....... ............... .... 2
Effects of Acute Administration of Stimulant Drugs
on Fixed-Interval R esponding................................................ ........... ........... 5
O verall R esponse R ates ....... ................................................ ................. 5
Rate D ependency .............................................. 6
Tem poral Patterning ...... ....................... .................. ..... 7
Effects of Repeated Administration of Stimulant Drugs
on Fixed-Interval Responding...................... .................... ............... .......... 8
B ehavioral Tolerance............. ............................................................... 8
Reinforcement-Loss Hypothesis .................................................................. 9
Tolerance and Fixed-Interval Schedules of Reinforcement ............................. 11
Schedule-Induced Behavior ...................................... .............. 13
Drugs and Schedule-Induced Behavior ....................................... .. ....... ... 16
Effects of Drugs on Schedule-Induced Polydipsia...................................... 16
Effects of Amphetamine on Fixed-Interval Behavior in a
Schedule-Induced Polydipsia Context .......................................... 23
Purpose for the Following Experiments .............................................. 23

EXPERIMENT ONE
Introduction ............. ... ......... .......... ......... .............. 25
Method
Subjects ............... .. ... .... ............... ................ .......... 27
Apparatus ................. .... ................... ................... 27
P rocedu re ...... .................................. ...... .. ..... .. .. ..... .......... .. 28
R e su lts .... ........ .................................................................... ....................... 3 3
D iscussion ................... ............ ...... .. .. ........ ... ........ ... ...... 79





iv









EXPERIMENT TWO
Intro du ctio n ......................................... .. .... ..... .... ....... ......... 8 5
Method
S u b je cts ...... .. ........ .. .. .. .. .. .. .. .... ....... .. .... .. ........... ..................... 8 6
A pparatus ..... .................. ............ 86
Procedure ................ ........... .... .................... 86
R esu lts ............................................................................................ ................. 8 9
D iscussion................. .. .......... ...... 148

G EN ER A L D ISCU SSIO N ........................................... ..................................... 153

R E FER EN C E S ..................................................... ... 163

BIOGRAPHICAL SK ETCH ............ .................................. .................... 169




































v














Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy

EFFECTS OF ACUTE AND REPEATED COCAINE ADMINISTRATION ON
FIXED-INTERVAL RESPONDING AND SCHEDULE-INDUCED POLYDIPSIA

By

Karen G. Anderson

May 1998

Chairman: Frans van Haaren
Major Department: Psychology

Six rats were exposed to a fixed-interval 60-second schedule of food reinforcement

(Experiment 1) with a water drinking-tube available at all times. All subjects

developed schedule-induced polydipsia, i.e., excessive drinking, at the beginning of

the interfood interval. Acutely, cocaine dose-dependently decreased licking for all rats

and decreased lever pressing for four rats. The response distributions for both

activities revealed earlier initiation of lever pressing and a restriction of licking to the

earlier segments of the interval. Examination of the response rates after repeated

exposure to cocaine revealed no development of tolerance to its rate-decreasing effects

for either activity. However, a more detailed analysis of response distributions

throughout the interval and average post-reinforcement pauses indicated that for many

subjects temporal organization had returned to near control (pre-drug) levels for lever

vi









pressing and for licking, but to a lesser extent. Similar results were obtained when six

other rats were exposed to a multiple fixed-interval 60-second fixed-time 60-second

schedule of reinforcement (Experiment 2). In this experiment, food was sometimes

delivered response-independently and therefore, cocaine's effects on licking were not

indirect effects due to changes in lever pressing.







































vii














INTRODUCTION

Fixed-interval schedules (FI) are time-based schedules of intermittent

reinforcement that are commonly thought of as one of the basic laboratory schedules

of reinforcement. As such, they may be considered "fundamental determinants of

behavior" (Zeiler, 1984, p. 487) in that they are likely to maintain predictable behavior

patterns characterized by a period of time immediately after reinforcer delivery when

no responding is recorded, followed by an increase in response rate (Ferster &

Skinner, 1957). Thus, such behavior is schedule-controlled. Providing the subject

with an opportunity to drink during the FI is also likely to result in reliable patterns of

responding. Specifically, excessive drinking usually occurs at the beginning of the

interval when the schedule-controlled response is less probable (cf. Falk 1971). Since

schedule-induced drinking occurs most often during the first half of a fixed interfood

interval, and the schedule-controlled response (e.g., lever pressing) occurs most often

during the second half of the fixed interfood interval, changes in the probability of one

activity may result in changes in the probability of the other activity.

Administration of a drug is one method by which a change in responding may be

initiated. Specifically, amphetamine, when given acutely, has been shown to decrease

schedule-induced licking and/or to shift the distribution of licks to earlier portions of

the interval (e.g., Sanger, 1978) and also to produce increases in lever pressing (often

1







2
at the beginning of the interval). Although there is some discussion as to whether or

not the drug effect on licking is direct or indirect (cf. Mittleman, Rosner, & Schaub,

1994; Sanger, 1978), it may be possible to compare changes in one activity with those

in another to learn more about the behavioral effects of drug administration in a

context where the two activities are recorded. Although many studies have

documented drug effects on schedule-controlled and schedule-induced behaviors, few

have included an analysis of drug effects on dependent measures other than response

rates, e.g., response patterning or distribution. Also, to date and to the author's

knowledge, no study has been reported in which a drug's effects on schedule-

controlled and schedule-induced behavior were evaluated in the same experimental

context and over an extended period of time (repeated administration) with a wide

range of doses. Such research may provide information regarding the development of

behavioral tolerance (or sensitization) in the two classes of behavior, as well as a

further characterization of a very robust behavioral phenomenon (schedule-induced

behavior) that is not well understood. The present dissertation begins to address these

issues.

Description of Fixed-Interval Responding

An FI schedule of reinforcement specifies that the first response emitted after a

fixed period of time has elapsed will be reinforced. Other responses are recorded, but

do not result in any scheduled consequence. The overall response pattern typically

reported as a result of scheduling reinforcement according to an FI schedule has been

described as a period of no responding followed by a gradual acceleration in response







3
rate until the end of the interval and presentation of the reinforcer (Ferster & Skinner,

1957). This pattern can also be described by various other measures that include

quarter-life, the time in the interval for one-quarter of the responses to be emitted (the

larger the number the steeper the slope of the curve) and the index of curvature (Fry,

Kelleher, & Cook, 1960). When the index of curvature is a large, positive number, a

steep slope for a positively-accelerated curve is indicated.

Even though this positively accelerated function or "scallop" is frequently reported,

another characterization of the response pattern has also been suggested (cf. Schneider,

1969). Schneider showed that within the majority of individual intervals during

conditions with various FI schedules (16 512 s), response patterns of pigeons

consisted of a period of little or no responding followed by a period of high, constant

responding that terminated with reinforcer delivery. Since the time to response

initiation varied among individual intervals, Schneider suggested that the characteristic

"scallop" was often an averaging artifact that was produced when the patterns within

each interval during the session were averaged to form one interval. Schneider (1969)

also showed that the duration of the post-reinforcement pause (state one) and the rate

during the period of high, constant responding (state two) were a function of the FI

parameters tested (16 512 s). For example, when the FI value was relatively large,

the post-reinforcement pause was longer and the response rate in the second state was

lower, than for smaller FI values.

A more detailed analysis of pigeons' FI responding suggested that Schneider's

"two-state" hypothesis be modified (Branch & Gollub, 1974). Like Schneider, Branch








4
and Gollub noted that the intermediate rates often seen in the common "FI scallop" did

not accurately characterize key pecking during the middle portions of the interval (30

s, 100 s, or 300 s). In fact, this pattern was largely a result of "averaging periods of

responding and not responding" (p. 531). However, these authors did observe some

positive acceleration of responding in the majority of intervals. Therefore, the authors

suggested (p. 538) "an amended version of the "two-state" hypothesis...is that

performance under fixed intervals by pigeons can be characterized by two states, a

period of very low or zero rate followed by a period of slightly, but reliably, positively

accelerated responding."

A study by Baron and Leinenweber (1994) was designed to compare molecular

(individual intervals) and molar (an average interval over entire session) analyses of

rats' lever pressing maintained by an FI 30-s schedule of reinforcement. They found

that after repeated exposure to this schedule, the response patterns changed from being

positively accelerated to one best characterized as a pause followed by a run of

responding (similar to what was found with pigeons) or as a longer pause followed by

a single response (a pattern not typically observed with pigeons). (Although

qualitative behavioral differences may exist between species, it would be premature to

conclude such a distinction without careful consideration of other variables operating

in each experimental condition, e.g., the nature of the response, cf. Lowe & Harzem,

1977.) It should be noted however, that the pattern obtained from averaging all

intervals within the session, i.e., the "scallop" or positively-accelerated pattern of

responding, observed at the molar level, has generally been reported and used in data








5
analysis with orderly outcomes. In addition, the across-interval average "can reveal

aspects of performance that are not immediately apparent within the individual

intervals [i.e., overall response probabilities]" (Baron & Leinenweber, 1994, p. 18).

The level of analysis should, therefore, be taken into account when evaluating FI

responding.

Effects of Acute Administration of Stimulant Drugs on Fixed-Interval Responding

Overall Response Rates

Three possible effects on overall response rate may be observed after drug

administration. The drug may produce an increase or decrease in the control rate of

responding or may have no noticeable effect on this pre-drug response rate. All three

effects have been reported when administration of stimulant drugs (e.g., amphetamine,

cocaine, methamphetamine, caffeine) precedes exposure to an FI schedule of

reinforcement. A number of studies have been reported with various species, schedule

parameters, and types of reinforcers, in which low or moderate doses of a stimulant

either did not affect or increased response rates and higher doses dose dependently

decreased responding (Branch & Gollub, 1974; Logan, Carney, Holloway, & Seale,

1989; McMillan, 1968; Schama & Branch, 1989, 1994; van Haaren, 1992a; van

Haaren & Anderson, 1994; Woolverton & Balster, 1983).

The effect a dose of drug will have on response rates often is predictable from the

control (non-drug) rate of responding, which may be influenced by the schedule of

reinforcement (Clark & Steele, 1966) or duration of the interfood interval in an FI-

schedule context (McMillan, 1969). Since longer intervals usually engender lower








6
response rates, it is expected and often observed that stimulant drugs administered

under these situations result in a greater increase in responding than with higher

response rates engendered by a shorter interval duration (cf. McMillan, 1969; Schama

& Branch, 1989). Therefore, it has been suggested that the overall control rate of

responding is one of the determinants of stimulant drug action.

Rate Dependency

The notion that control rates of responding can predict the stimulant drug's effects

has come to be known as the "rate-dependency hypothesis" (see Dews & Wenger,

1977; Sanger & Blackman, 1978, for reviews). In general, it has been shown that low

to moderate doses of amphetamine will result in an increase in low-rate behavior while

decreasing high-rate behavior (cf. Dews & Wenger, 1977). In the case of responding

during an FI, which is often characterized as positively accelerated, it is expected and

has been shown that responding is increased at the beginning of each interval

following drug administration (Dews, 1958; Logan et al., 1989; McMillan, 1968).

This has often been seen as an increase in response rates in the early segments of the

interval, a decrease in quarter-life, a decrease in index of curvature, and/or a shift to

the left in a response-distribution curve.

It was often suggested that the increase in responding at the beginning of the

interval was a result of the drug's effects on the low rates of responding. Similarly, a

decrease in responding towards the end of the interval was suggested to be due to the

drug's effects on high rates of responding (cf. Sanger & Blackman, 1976). Thus, FI

schedules were believed to be useful for examining rate-dependent effects of drugs








7
within a single experimental session (for a more detailed discussion, see Branch &

Gollub, 1974; Sanger & Blackman, 1976). However, this approach has not been

fruitful after more molecular analyses of FI responding (cf. Branch, 1984; Branch &

Gollub, 1974; McAuley & Leslie, 1986; Schneider, 1969), which revealed that the

positively accelerated curve (an averaging artifact) is not likely to represent the

response pattern in any individual interval.

Despite the fact that it has been misapplied in the context of FI schedules, the rate-

dependency hypothesis has often proven useful in predicting/describing the effects of

stimulants on overall rates of behavior maintained by positive or negative

reinforcement (McKearney, 1981). However, it should be noted that behavior under

strong stimulus control and behavior that occurs at low rates due to punishment have

not been shown to increase following stimulant administration (cf. McKearney, 1981;

McMillan, 1969; Sanger & Blackman, 1976). Thus, the rate-dependency hypothesis

may have some utility in restricted situations, but its use, especially in the context of

within-interval responding in FI schedules, has been shown to be problematic.

Temporal Patterning

As mentioned above, when low doses of stimulant drugs increase response rates

under FI schedules of reinforcement, this is often due to an increase in the probability

of responding in early portions in the fixed interval (Branch & Gollub, 1974; Dews,

1958; McAuley & Leslie, 1986; Shelnutt, Hudzik, Lattin, & McMillan, 1992). Under

non-drug conditions, few responses are emitted in the early segments of the FI, relative

to the number recorded in the later portions of the interval. A leftward shift in this








8
distribution of responses, indicating an increase in responding early in the interval,

often follows stimulant administration. This change in response patterning may also

be captured by a decrease in the quarter-life, even when there is no overall change in

rate of responding (Logan et al., 1989).

Since FI responding within each interval is usually characterized as a period of no

responding followed by responding at a constant or slightly increasing rate, the

increase at the beginning of the interval observed after stimulant administration is due

to a decreased post-reinforcement pause and thus, disruption of temporal control.

McAuley and Leslie (1986) showed that this effect (shortening of the post-

reinforcement pause) varied systematically with the dose of d-amphetamine with rats

as subjects. Specifically, as the dose of amphetamine was increased, the duration of

the post-reinforcement pause decreased. They concluded (p. 218) that "increases in

low rates are best described as changes in temporal control--the probability of

initiating responding changes, rather than the response rate." Thus, the results from

their study supports and further extends the generality of the findings from Branch and

Gollub (1974) by including data from another species with a different response

topography (lever pressing) maintained by different FI schedules (30 s, 60 s, or 120 s).

Effects of Repeated Administration of Stimulant Drugs on Fixed-Interval

Responding

Behavioral Tolerance

When a drug is presented repeatedly to an organism, behavioral tolerance or

behavioral sensitization may develop. The assessment of this potential outcome








9
involves several steps. After selection of an aspect of behavior to be measured

(usually rate), experimental sessions are conducted until a stable baseline is achieved.

Once responding is judged reliable in the absence of a drug, various doses of a

particular drug and its vehicle are administered occasionally (usually once or twice per

week) and the specific behavioral effect(s) are noted. From the obtained dose-

response function, a dose is chosen for repeated administration and is administered

prior to every experimental session for some extended period of time. This dose

usually has been selected because it reduces, but does not eliminate responding. When

responding is once again observed to be stable, another dose-response function is

obtained by substituting other doses and the vehicle for the repeated dose on an

occasional basis.

Behavioral tolerance is assessed by comparing the second (chronic) dose-response

function with the first (acute) dose-response function. If the second curve is shifted to

the right of the initial one, then tolerance is said to have resulted. Behavioral

sensitization occurred if the second function shifted to the left of the first curve, i.e., a

lower dose now produces the effect that a higher one used to produce. Although the

factors that determine the development of behavioral tolerance or sensitization are not

fully understood, some important variables have been identified (see Branch, 1993 for

a review).

Reinforcement-Loss Hypothesis

Schuster, Dockens, and Woods (1966) have offered one hypothesis for why

behavioral tolerance may result from repeated administration of a drug. In their study,








10
rats were exposed to a multiple (MULT) FI 30-s differential-reinforcement-of-low-rate

(DRL) 30-s schedule of food reinforcement. It was found (for two of the three rats)

that one acute effect of d-amphetamine was to increase lever pressing in both

components of the MULT schedule. Increased responding during the FI schedule had

no effect on overall amount of food reinforcement obtained. However, increased lever

pressing during the DRL schedule resulted in a loss of reinforcers. After repeated d-

amphetamine administration, tolerance developed to the rate-increasing effects of d-

amphetamine only during the DRL-schedule component, not during the FI schedule-

component. In addition, for the rat with a decreased number of responses during the

FI schedule component and decreased numbers of reinforcers, tolerance was observed.

This rat's number of lever presses and rate of reinforcement increased to baseline

levels. Thus, the authors concluded that reinforcement loss is an important

determinant of the development of tolerance.

Behavioral tolerance will develop in those aspects of the organism's
behavioral repertoire where the action of the drug is such that it disrupts
the organism's behavior in meeting the environmental requirement for
reinforcements. Conversely, where the actions of the drug enhance, or do
not affect the organism's behavior in meeting reinforcement requirements
we do not expect the development of behavioral tolerance. (p. 181)

Although more recent studies have shown that reinforcement loss is not always a

determinant or predictor of the development of behavioral tolerance (cf. Wolgin,

1989), the hypothesis continues to have merit in some circumstances (cf. Branch,

1993).








11
Tolerance and Fixed-Interval Schedules of Reinforcement

In the Schuster, Dockens, and Woods (1966) study, the rat whose behavior showed

tolerance to the rate-decreasing effect of 1.0 mg/kg d-amphetamine during the FI

schedule may have developed this tolerance as a result of reinforcement loss. Similar

findings of tolerance with rats (Smith, 1990) and pigeons (Schama & Branch, 1989)

have been reported elsewhere. Therefore, when the initial effect of the drug is to

decrease responding such that reinforcement rate decreases, the development of

tolerance to these rate-decreasing effects is likely to be observed.

However, it is possible that during an FI schedule, overall response rates or number

of responses emitted may decrease as a result of drug administration, but that no

reinforcement loss will be experienced. This may occur because, in an FI schedule

only one response is needed for reinforcement to be presented. If the reinforcement-

loss hypothesis is to hold in this context, then one would not expect to see tolerance to

the rate-decreasing effects of a drug unless the rate of reinforcement was to decrease as

well.

Likewise, since an increase in responding under an FI schedule is not likely to

result in a loss of reinforcement, tolerance to the rate-increasing effects also is not

likely to occur, according to the reinforcement-loss hypothesis. This notion is

supported by the findings from a study by Schuster et al. (1966). In their experiment,

a rat that received 1.0 mg/kg amphetamine showed an initial response-rate increase.

No tolerance to this rate increase was observed after 30 days of repeated

administration of this dose. Although there has been little work conducted in the area








12
of tolerance with FI schedules with rats, there are exceptions to this general finding of

no tolerance to the rate-increasing effects of a stimulant drug reported with other

species. For example, Schama and Branch (1994) repeatedly administered a dose of

cocaine that was shown to produce initial (acute) response-rate increases in squirrel

monkeys' lever pressing under an FI schedule (either 5 or 8 min). Following repeated

administration of the rate-increasing dose of cocaine, tolerance was observed, i.e.,

response rates decreased to near baseline levels. It is also interesting to note that

through examination of cumulative records, response-distribution graphs, and indices

of curvature, it was shown that cocaine increased responding throughout the interval.

Tolerance to the rate-increasing effects typically was accompanied by a decrease in

responding in the early portions of the interval and an increase in index of curvature.

As in the case with a drug's acute effects, it should be noted that tolerance to either

the rate-increasing or rate-decreasing effects of a stimulant on the schedule-controlled

response may be influenced indirectly by the drug's effects on other behavior that may

be occurring in the experimental context. For instance, if the effect of the drug was to

decrease the frequency of some other type of (competing) behavior during the interval,

particularly behavior occurring at the beginning of the interval, then an increase could

be observed in schedule-controlled responding. Therefore, tolerance to the initial rate

changes may be seen in both activities, but in different directions. The effects of drug

administration on the rates and patterns of different behaviors, particularly behaviors

under control of the contingencies arranged by an FI schedule, have not been

extensively investigated.








13
Schedule-Induced Behavior

The behavioral phenomenon that is now known as schedule-induced behavior was

first reported by Falk in 1961. He reported that rats developed a characteristic

drinking pattern when food pellets (45 mg) were presented intermittently in a standard

operant conditioning situation. Falk exposed rats to a variable-interval (VI) 1-min

schedule of reinforcement with water continuously available in the experimental

setting. In this arrangement, the first lever press after an average time of 1 minute had

elapsed was followed by presentation of one food pellet. The rats were food-restricted

so that their weights equaled approximately 70% to 80% of their free-feeding weights,

but they were never deprived of water.

During the 3.17-hour sessions, the rats consumed excessive amounts of water (3.43

times the amount they drank, on average, during a 24-hour period assessed prior to an

experimental session). Falk noted that this behavior was especially peculiar due to

reports that food-restricted animals typically drink less than free-feeding animals

(Bolles, 1961). In addition, through examination of cumulative records of responding,

this excessive drinking or polydipsia was observed to occur at reliable times during the

sessions, i.e., immediately after pellet presentation during the post-reinforcement

period (see Falk, 1971 and Wetherington, 1982 for reviews).

Since Falk's first report of this behavioral phenomenon, many experiments have

been conducted to further investigate the intermittency of reinforcer delivery and food

deprivation as two of the determinants of schedule-induced behavior. Many

experiments (e.g., Dove, 1976; Falk, 1966a; Flory, 1969; Gentry, 1968; Maygar &








14
Malagodi, 1980; Roper, 1978) have shown that the behavior is largely determined by

the scheduling of reinforcers used for maintaining schedule-controlled behavior (or by

the intermittency of food presentation as seen in response-independent schedules),

regardless of the species, response topography, or schedule (e.g., FI, VI, DRL). For

instance, Falk (1966b), systematically manipulated the value of an FI schedule across

conditions in a within-subject design. He observed an inverted U-shaped (bitonic)

function when milliliters of water consumed was plotted against the FI value. The

maximum point on the function varied between subjects (rats), but the overall shape of

the function was consistent.

Eliminating the intermittency of food-pellet presentation has been suggested as a

control procedure against which to assess the excessiveness of the schedule-induced

behavior (cf. Roper, 1981). When the same number of pellets that are delivered

according to a reinforcement schedule in a regular session are presented all at once at

the beginning of a session of equal length (massed-pellet procedure), subjects drink

less than under control conditions (Falk, 1967). In addition, extinction sessions of

durations equal to baseline sessions have been recommended (Roper, 1981) as a

control for food presentation in the baseline sessions. If in both control (probe)

conditions, schedule-induced behavior occurs at a lower overall rate than when the

food is presented on an intermittent basis, then it is reasonable to call the behavior

schedule-induced in non-probe conditions.

Falk (1969) has also shown that level of food deprivation is a major determinant of

schedule-induced drinking. He showed that as rats' percent of free-feeding body








15
weight was increased from 80% to 105%, schedule-induced water intake decreased.

The rats also drank more as their weights were systematically decreased to 80% of

their free-feeding weights. This change in drinking was observed even when there was

no change in lever pressing on an FI 90-s schedule of reinforcement.

It has also been shown that many of the characteristics (e.g., excessiveness,

temporal patterning) that help to define schedule-induced polydipsia apply to behavior

with different topographies, e.g., wheel-running, pica, air-licking, drug self-

administration, defecation, attack, escape (see Falk, 1971, and Wetherington, 1982, for

reviews). Because many kinds of behavior appear to be schedule-induced, share

similar characteristics, and have been reported in many different species with different

schedules of reinforcement, there has been some discussion (see Falk, 1971; Staddon,

1977; Wetherington, 1982 for extensive reviews) as to whether or not it should be

classified as operant, respondent, or a whole new class of behavior, i.e., adjunctive.

Schedule-induced behavior has also been used as a model for certain human

behavioral disorders, e.g., anxiety (Tung, Wu, Tseng, & Yin, 1994), obsessive-

compulsive disorder (Woods, Smith, Szewczak, Dunn, Cornfeldt, & Corbett, 1993),

drug abuse (Falk & Tang, 1989; Falk, Zhang, Chen, & Lau, 1994). As such a model,

schedule-induced behavior has been evaluated after administration of many drugs, but

the effects of extended repeated administration on schedule-controlled and schedule-

induced behaviors have not been examined in detail.








16
Drugs and Schedule-Induced Behavior

Some time after the initial report on schedule-induced behavior (Falk, 1961),

researchers focused attention on how drug administration would affect this new

phenomenon. Sanger and Blackman (1978), continued to encourage this line of

research by raising three issues that should be considered by behavioral

pharmacologists investigating schedule-induced behavior. First they noted, "if the

argument is to be considered that a new category is required in behavioral analysis to

supplement those of respondent and operant behavior, it is important to ask whether

the effects of drugs on adjunctive behavior conform to the general principles which

have emerged in studies of other behavior" (p. 253). They also suggest investigation

of drug effects on patterns of schedule-induced behavior for comparison to behavior

that is similar in form but "which is generated and maintained by other means" (p.

254). For instance, one could determine if a drug's effects on schedule-induced

polydipsia are similar to the effects on drinking brought about through water

deprivation or a salt load. Third, the authors suggest studying drug effects on this

behavioral phenomenon because it is a "striking component of an animal's total

behavioral repertoire in an experimental situation (regardless of how the behavior is to

be interpreted)" (p.254). What follows are some of the general findings of drug effects

on schedule-induced behavior.

Effects of Drugs on Schedule-Induced Behavior

During the past 30 years, many investigations of drug effects on schedule-induced

behavior have been published. The majority of these studies targeted schedule-








17
induced drinking, and the overall findings suggest that a pharmacological challenge

results in a decrease in this behavior (cf. Mittleman et al., 1994; Sanger & Blackman,

1978). However, benzodiazepines have been reported to increase schedule-induced

drinking (Bacotti & Barrett, 1976; McKearney, 1973) while other anxiolytics, e.g.,

barbiturates, usually decrease schedule-induced drinking or have no observed effect at

the doses tested (Segal, Oden, & Deadwyler, 1965). Given these mixed findings with

anxiolytics, and the many procedural and measurement variations in the experiments,

it would be premature to conclude anything definitively about this particular class of

drugs and their possible effects on schedule-induced drinking.

Effects of amphetamine on overall drinking. It has been shown in a variety of

experimental procedures that acute amphetamine administration suppresses schedule-

induced drinking at doses that will increase operant lever-pressing rates (cf. Sanger &

Blackman, 1978). In one of the earliest studies (Segal et al., 1965) involving d,l-

amphetamine and schedule-induced polydipsia produced by a response-independent

schedule (FT 30 to 480 s), it was reported that amphetamine (0.5 2.0 mg/kg) given

intraperitoneally (IP) to rats immediately before the start of an experimental session

decreased the average duration of each drinking bout as well as the percent of intervals

containing a drink. Work with response-dependent schedules has resulted in similar

findings with respect to schedule-induced drinking, and often the same dose that

suppresses drinking will increase lever pressing. Wayner, Greenberg, and Trowbridge

(1973) conducted a study, in which two rats' lever pressing was reinforced according

to an FI 60-s schedule, to investigate the effects of IP d-amphetamine administration








18
(0.05 2.0 mg/kg) on total number of both responses (licking and pressing) and total

amount of water consumed. Amphetamine dose-dependently increased the mean

number of lever presses for both rats and also decreased both measures of drinking for

one rat in a similar dose-dependent fashion. For the other rat, there was a slight

increase in the number of licks and water consumed only at the lowest doses of

amphetamine. The authors concluded (p. 110) that (given their experimental

procedures, parameters, and measurements), "d-amphetamine affects schedule-

dependent and schedule-induced behavior differently." Findings similar to these and

others (e.g., Smith & Clark, 1975) have led to a substantial amount of work where the

differential drug effects on schedule-controlled and schedule-induced behavior have

been investigated.

That d-amphetamine suppressed schedule-induced drinking at doses that often

increased lever pressing raised an important question which was addressed a few years

later. Sanger (1978) designed an experiment to study d-amphetamine's effects on

schedule-induced drinking in the context of a MULT FI 60-s FT 60-s schedule. The

purpose behind using the MULT schedule was to assess whether or not drug-induced

suppression of drinking was a direct or an indirect effect. If d-amphetamine produced

increases in lever pressing, it is possible that these responses could compete with

drinking and, hence, reduce drinking by an indirect, rather than direct, action. To

investigate this possibility, Sanger administered IP d-amphetamine (0.25 2.0 mg/kg)

to rats in 60-min sessions with alternating 10-min components of either the FI 60-s or








19
the FT 60-s schedule. During the FT 60-s schedule, the lever was retracted so that

lever presses could not occur.

Sanger (1978) found that there were no differences in schedule-induced drinking in

either schedule component (FI or FT) prior to or during drug administration. In

general, lever-pressing rates in the FI increased when the lower doses of d-

amphetamine were administered. Schedule-induced drinking (rate of licking and total

water intake) was decreased in a dose-dependent manner in both components of the

MULT schedule. The similar results with drinking induced by a response-dependent

and a response-independent schedule led the author to conclude that the drug's effects

on schedule-induced drinking were direct.

An additional, but contradictory, explanation is suggested by Mittleman et al.

(1994). These researchers also reported that rats' schedule-induced drinking

engendered by an FT 60-s schedule of sucrose-pellet presentation decreased dose-

dependently following IP d-amphetamine administration (0.125 2.0 mg/kg).

However, these authors suggest that the decrease in drinking was an indirect effect,

based on concurrent recordings of photobeam breaks in the rear of the chamber

(opposite the drinking tube). Rats' overall activity increased in a dose-dependent

manner as drinking decreased. Therefore, the authors conclude (p. 647) that it is

"most parsimonious to suggest that some drugs [d-amphetamine] reduced SIP

[schedule-induced polydipsia] as an indirect consequence of increases in locomotion."

It should be mentioned, however, that the authors' conclusions were based on the use

of one photobeam, the results were presented as overall averages across 23 subjects,








20
and it was not clear if each subject received all of the test doses or how many times

each dose was tested per rat. In addition, at what points in the interval the breaks in

the beam were recorded were not presented. Increased locomotor activity would only

compete with schedule-induced drinking if it occurred in the immediate post-pellet

period. If the breaks in the photobeam occurred toward the end of the interval and just

before delivery of the food pellet, it is possible that the increase may be a result of

adventitious reinforcement.

Sanger (1978) reported that when food pellets were presented contingent on lever

pressing, amphetamine increased lever pressing at the beginning of the interval. This

responding would likely compete with locomotor activity and therefore, any increases

in locomotor activity may be more relevant when there is no explicitly arranged

operant contingency. Thus, although increased locomotor activity may be a viable

account for the reduced drinking in some cases and with some drugs, a more detailed

analysis is warranted. An investigation into the temporal patterning or organization of

the post-pellet behaviors is one such analysis that may be useful in this and other

characterizations of drug effects on schedule-induced behavior.

Effects of amphetamine on temporal patterning of drinking. As previously

mentioned, most schedule-induced behaviors occur at the beginning of the interfood

interval and decrease in probability as the interval progresses. Thus, this reliable

pattern may serve as an additional behavioral measure to characterize further the

effects of drugs on schedule-induced activity. However, few researchers report data

regarding the temporal patterning of behavior. Of those who have, a consistent finding








21
concerning amphetamine's effects on schedule-induced drinking is that it increases the

probability of licking at the beginning of the interval, while often decreasing the

overall rate. Williams and White (1984) reported such an outcome on rats' water

drinking induced by an FI 90-s schedule with a mixture of salt (7.5% NaC1) in a 1:1

dilution of sweetened condensed milk with tap water as the reinforcer for lever

pressing. (It is sometimes difficult to induce drinking with a liquid reinforcer, but

such may be accomplished with manipulation of water and salt content, see also Falk,

1969.) They found that the probability of licks in the earlier parts of the interval

increased while the probability of licks in the later portions decreased as the dose of IP

d,1-amphetamine was increased (0.3 10.0 mg/kg). This dose-dependent shift in the

probability of drinking to earlier in the interval was correlated with a dose-dependent

decrease in overall licking rates and in total amount of water consumed.

McKearney (1973) examined methamphetamine's effects on schedule-controlled

and schedule-induced licking in the same animal. Licking was reinforced by food

pellet presentation according to an FI 3-min schedule. McKearney reported high rates

of licking shortly after food pellet presentation consistent with schedule-induced

drinking. Following this bout of drinking and a subsequent period of no licking, the

rats began to lick again towards the end of the interval in a pattern consistent with FI-

like performance. After methamphetamine (IP) administration (0.03 1.7 mg/kg),

there was a dose-dependent decrease in licking in the later portions of the interval (the

schedule-controlled responses) while a leftward shift in licking was observed in the

early parts of the interval (the schedule-induced responses). The bitonic function at








22
the beginning of the interval and the positively-accelerated pattern observed later in

the interval were still evident, but the peak of the bitonic function, indicating schedule-

induced drinking, was shifted slightly to the left.

A more recent study (Pellon & Blackman, 1992) yielded results that support these

earlier findings. These investigators also found that IP d-amphetamine administration

(0.25 2.0 mg/kg) increased rats' schedule-induced drinking in the early parts of the

interfood interval (FT 60-s schedule) and decreased drinking in later parts, indicating a

shift to the left for the response-distribution curve. The authors concluded that the

drug effects on schedule-induced drinking were direct, since there was no schedule-

controlled response required on the FT schedule that could interfere with drinking.

However, locomotor activity was not recorded and therefore, changes in general

activity can not be eliminated as a possibility for the changes in drinking (cf.

Mittleman et al., 1994). Pellon and Blackman (1992) also pointed out that at some

doses of d-amphetamine the overall licking rates were similar, which might lead one to

surmise that there was no effect of drug administration on drinking. However, at these

same doses they noted a distinct shift to the left in the distribution of licks in the

interval. Thus, roughly the same number of licks were recorded in the FT-interfood

interval for the different doses, but the temporal organization within the interval was

altered.








23
Effects of Amphetamine on Fixed-Interval Behavior in a Schedule-Induced

Polydipsia Context.

As stated earlier, amphetamine often produces effects on schedule-induced drinking at

doses where no obvious effect on the schedule-controlled response is observed. When

a change in the schedule-controlled response, e.g., lever pressing, is noted, it is usually

that low doses of amphetamine increase overall lever-press rates. This increase in rate

is often a result of an increase in lever pressing occurring at the beginning of the

interfood interval. Following interval-by-interval analyses of FI responding, this

change in the overall distribution of responding is now thought to be a result of earlier

response initiation, rather than a direct increase in low-rate behavior, per se (cf.

Branch, 1984; Branch & Gollub, 1974; McAuley & Leslie, 1986). When

amphetamine's effects are evaluated on lever pressing and schedule-induced drinking

in the same context, the overall response distributions indicate that the drug produces

an increase in lever pressing at the beginning of the FI and also increases licking at the

beginning of the FI, which is often accompanied with a leftward shift in the curve

(Pellon & Blackman, 1992). By looking at overall response distributions, it would

seem that amphetamine administration results in response initiation (lever pressing and

licking) earlier in the interfood interval than during control (non-drug) conditions.

Purpose for the Following Experiments

The effects of acute administration of stimulant-type drugs have been reported for

lever pressing and water-tube licking in a variety of circumstances. However, no

study known to this experimenter, has been designed to evaluate the effects of








24
repeated administration of stimulants (e.g., amphetamine, cocaine) on schedule-

controlled and schedule-induced behaviors in the same experimental context. The

purpose of the present experiments was to determine the effects of acute and repeated

cocaine administration on rats' schedule-controlled lever pressing and schedule-

induced licking. Specifically, response rates were analyzed for rate-increasing or rate-

decreasing effects and overall response patterns across the interfood intervals were

analyzed for changes in the temporal patterning/organization of the behaviors. These

analyses were made following each determination of the acute and chronic dose-

response functions, and comparisons between the two functions, which were separated

by an intervening drug history of repeated exposure to the same dose of cocaine, were

evaluated for the development of behavioral tolerance or sensitization.














EXPERIMENT ONE

Introduction

This experiment was designed to evaluate the effects of acute and repeated cocaine

administration on rates and distributions of lever pressing and licking engendered by

an FI 60-s schedule of food presentation. When rats are exposed to an FI schedule of

reinforcement and stable responding has been established, lever pressing is not likely

to occur immediately following food-pellet presentation, but the probability of

responding increases as the interval elapses. In contrast, rats will often drink excessive

amounts of water during the immediate post-pellet period if a drinking tube is present,

although there are no programmed consequences for this behavior (Falk, 1961). Acute

administration of low doses of amphetamine has been shown to disrupt both lever

pressing and drinking in different ways (e.g., Wayner, Greenberg, & Trowbridge,

1973) usually by increasing lever pressing rates and decreasing licking rates. The

increase in overall lever pressing rates is often a result of increased responding earlier

in the interval. Thus, an interaction may exist between the two behaviors following

acute drug administration during the early portions of the interval.

The present experiment serves to extend the findings from studies involving

amphetamine administration to a situation where cocaine is administered. (Both drugs

are indirect dopamine agonists in that they block reuptake of dopamine, which results

25








26
in increased concentrations of extracellular dopamine. Amphetamine also increases

dopamine release and has a much longer duration of action than cocaine (cf. Hoffman

& Lefkowitz, 1990).) In addition, after repeated administration of cocaine, behavioral

tolerance may develop (cf. Branch, 1993). Lever-pressing rates and distributions may

then return to near baseline (pre-drug) levels and if they do, it is possible that this

change may also be accompanied by tolerance to the rate-disrupting and/or pattern-

disrupting effects for licking.

Response distributions for lever pressing and licking were evaluated for any

changes in temporal organization or patterning of behavior following repeated

exposure to cocaine in the experimental context. Since the purpose of this experiment

was to determine the effects of cocaine on schedule-controlled lever pressing and

schedule-induced drinking the following questions were asked: Would there be an

interaction in the drug's effects on the two behaviors (rates and/or distribution of

responses)? Would tolerance develop to the rate-disrupting and pattern-disrupting

effects of cocaine?

In this experiment, lever pressing was maintained by presentation of food pellets

according to an FI 60-s schedule of reinforcement and a drinking tube allowed

continuous access to distilled water. Thus, any recorded effects of cocaine

administration could be observed for the schedule-controlled and schedule-induced

behaviors and any potential interaction between the two behaviors might also be

observed. This experiment included two chronic drug phases. Because no tolerance

was observed in lever-press and lick rates when 17.0 mg/kg cocaine (a dose that








27
initially decreased lever pressing and licking rates) was administered repeatedly, the

dose was reduced to 10.0 mg/kg and the dose-response functions were redetermined.


Methods

Subjects

Six experimentally naive, male Wistar rats obtained from a commercial supplier

(Charles River Laboratories, Wilmington, MA) served as subjects. They were

approximately 100 days old at the start of the experiment and weighed between 304

and 326 g. They were housed in two cages (three rats per cage) and were exposed to a

reversed 12-hour light-dark cycle (lights on at 7 p.m. and off at 7 a.m.).

Approximately 16 g of Purina Rat Chow was placed in each of the cages after the

experimental session (and on weekends). Subjects usually consumed this food in

about 1.5 hours. Subjects had free access to tap water in the home cage at all times,

and room temperature conditions remained constant.

Apparatus

The experiment was conducted in two identical Coulbourn Instruments rodent

operant conditioning chambers that were 24 cm deep, 30 cm wide, and 29 cm high.

The walls of the chamber were made of translucent Plexiglas except for the aluminum

intelligence (front) panel. The floor consisted of 16 stainless steel rods 6 mm in

diameter spaced 2 cm apart (center to center). The houselight was located 2 cm below

the ceiling in the middle of the intelligence panel. A retractable lever was located 7

cm above the floor of the chamber and 8.5 cm to the left of the pellet tray, which was








28
in the center of the panel. The lever, when extended, protruded 1.75 cm into the

chamber from the intelligence panel and required a force of at least 0.2 N to be

operated. A lever press produced auditory feedback from a clicker located directly

behind the intelligence panel underneath the pellet tray. Three stimulus lights (red,

yellow, and green) were located immediately above the lever. Only the red and green

lights were illuminated during experimental sessions. Access to a sipper tube was

available through an aperture 8.5 cm to the right of the pellet tray. Individual licks

were recorded whenever the subject interrupted an infrared photobeam with the tip of

its tongue. The photobeam (manufactured by Coulbour Instruments) passed

immediately in front of the tip of the drinking tube. Each experimental chamber was

housed in a sound-attenuating cabinet and connected to a PDP 11-23 microcomputer

located in the experimental room itself. Experimental contingencies and data

acquisition procedures were programmed in SKED-11 (Snapper & Inglis, 1985).

Procedure

Lever press acquisition and initial training. All sessions were conducted during

the dark phase of the light-dark cycle (usually between 8 a.m. and 11 a.m.). Subjects

were placed in a darkened chamber at the beginning of an experimental session. The

houselight and the lever lights were illuminated after 10 min, and food pellets were

presented according to a random-time (RT) 60-s schedule with a probability of 0.17

every second. In addition, food pellets were presented following each lever press (a

fixed-ratio (FR) 1 schedule of pellet presentation). Sessions ended following 45 min

or after the delivery of 40 food pellets, whichever came first (cf. van Haaren, 1992b).








29
Once the subjects pressed the lever reliably, as determined by visual inspection of

the data, the schedule of reinforcement was changed to an FI 5-s schedule for one

experimental session. This session lasted for 45 min or until 50 food pellets had been

delivered. An FI 10-s schedule (maximum: 50 food pellets) replaced the FI 5-s

schedule for the next 3 sessions. The schedule contingency was changed to FI 30 s for

the next 8 sessions. During this condition, sessions ended after 40 min or when 30

food pellets had been delivered, whichever came first. The FI schedule was changed

to its terminal value (FI 60 s) when reliable responding was observed, and all

subsequent experimental sessions were terminated following 45 min or after the

delivery of 30 food pellets, whichever came first. The drinking tube that allowed

access to distilled water was present during all experimental sessions, which were

conducted five days per week (Monday through Friday).

Acute cocaine administration and dose-response function determination. Once

lever-pressing and licking rates were stable on the FI 60-s schedule (93 sessions), as

determined by visual inspection of the data, saline was IP administered prior to an

experimental session. If no disruption in response rates (lever pressing or licking) was

detected, then the acute dose-response function was assessed. Saline or cocaine

hydrochloride obtained from the National Institute on Drug Abuse (1.0, 3.0, 5.6, 10.0,

17.0, or 30.0 mg/kg) was administered via IP injection in a volume of 1 mg/ml.

Injections were given immediately prior to the start of the experimental session, which

involved a 10-min blackout period before illumination of chamber lights and initiation

of the schedule contingencies. Drug or vehicle injections were given twice a week on








30
Tuesdays and Fridays if the control data were stable. Data from the immediately

preceding sessions were used as control data. Each dose was administered at least

twice (range 2 6) in descending order. The series of injections began with saline

administration.

Repated saline administration I. Following determination of the acute dose-

response function and prior to repeated (chronic) administration of cocaine, all

subjects were exposed to six consecutive experimental sessions, each of which was

preceded by a saline injection. After observing no disruption in lever-pressing or

licking rates, by visual inspection of the data, a dose of cocaine was selected to be

administered before each experimental session (Monday through Friday).

Repeated cocaine administration I and dose-response function determination I.

After determination of the acute dose-response functions for each subject and repeated

saline administration, a dose of cocaine was selected for daily administration. This

dose of cocaine was chosen on the basis of the individual dose-response functions.

The criterion for repeated (chronic) dose selection was that the dose had to alter

(increase or decrease) responding (lever-pressing and licking rates) by at least 25% of

the control rates, but not completely eliminate responding. For all six subjects, this

dose was 17.0 mg/kg cocaine. (Subject 77-4 is the only exception to the above

criterion. Although licking was reduced by more than 25%, lever-pressing rates were

not greatly affected by 17.0 mg/kg administered. However, given that not all available

reinforcers were obtained during the sessions in which 17.0 mg/kg was administered,

it was chosen as the repeated dose for this subject as well.) Repeated (chronic)








31
administration of 17.0 mg/kg cocaine involved injecting each subject immediately

prior to the experimental session. After the subjects had experienced the drug dose for

at least 40 experimental sessions and the response rates (lever pressing and licking)

were stable, as determined by visual inspection of the data, another dose-response

function was obtained (chronic dose-response function I). On Tuesdays and Fridays of

each week the chronic dose of cocaine (17.0 mg/kg) was replaced by either saline or

other doses of cocaine (1.0, 3.0, 5.6, 10.0, or 30.0 mg/kg). Again, each dose was

administered IP in descending order and the effects were assessed at least twice (range

2-3).

Repeated saline administration II. Once the chronic dose-response function I was

determined, all subjects received saline injections (IP) for the next 20 experimental

sessions. During this phase, response rates (lever pressing and licking) were allowed

to stabilize, as determined by visual inspection of the data. Due to minimal changes in

lever-press and lick rates relative to rates prior to the repeated administration of 17.0

mg/kg cocaine, and given the advanced age of the rats, another dose-response function

was not assessed prior to the repeated administration of 10.0 mg/kg and a reassessment

of the chronic dose-response function.

Repeated cocaine administration II and dose-response function determination II.

Immediately following the repeated-saline-administration-II phase, the repeated dose

of cocaine was reduced from 17.0 mg/kg to 10.0 mg/kg for all subjects. After at least

25 sessions of exposure to 10.0 mg/kg cocaine and when response rates (lever pressing

and licking) were observed to be stable (by visual inspection of the data), the second








32
dose-response function was assessed. Saline or cocaine (1.0, 3.0, 5.6, 17.0, or 30.0

mg/kg) was injected (IP) immediately prior to the start of the experimental session.

Each dose in the series was administered at least twice (range: 2 4) and was

presented in descending order after beginning with saline.

Extinction and massed-pellet presentation probes. Following the final dose-

response assessment, all injections were terminated. After 20 sessions and when

responding had stabilized, probe sessions were conducted on Tuesdays and Fridays to

provide a comparison for the intermittency of food-pellet delivery arranged by the FI

60-s schedule. A total of four probe sessions was conducted (two extinction sessions

and two with massed-pellet presentation). During the extinction sessions, lever

pressing had no scheduled consequences. Mass-pellet presentation sessions consisted

of the delivery of 30 food pellets at the start of the session (following the 10-min

blackout) and lever presses had no arranged consequences. In both types of probe

sessions, the houselight and the lights above the lever were illuminated. In addition,

the drinking tube was present throughout all sessions, which ended after 45 min.

Data collection. Throughout the experiment, the number of lever presses and the

number of licks were recorded and converted to rate measures (responses per minute)

by using overall session time (excluding the time it took to deliver food pellets). The

overall distributions of lever pressing and licking within the interval were also

collected. The number of milliliters of water consumed, the total number of food

pellets delivered, and the mean post-reinforcement pause were also recorded and

examined.








33
Results

Effects of acute cocaine administration on lever-press rates. Figure 1 shows lever-

press rates after acute cocaine administration and following repeated exposure to 17.0

mg/kg (filled circles) and 10.0 mg/kg (filled triangles) cocaine. Lever-press rates in

the absence of drug administration ranged between 20.56 to 55.71 lever presses per

minute. For two of the subjects (77-1 and 77-2), acute cocaine administration (open

symbols) resulted in a dose-dependent increase in lever-press rates when compared to

control (non-drug) rates. Lever-press rates for Subject 77-4 also increased, but only

after the medium doses (5.6 and 10.0 mg/kg). It should be noted that control rates of

lever pressing for these three subjects were lower than those of the other three

subjects. For two of the subjects (77-3 and 77-5), cocaine decreased rates of lever

pressing in a dose-dependent manner and for one subject (77-6) no dose of cocaine

tested completely eliminated lever pressing and any decreases in rates were minimal.

Figures 2 and 3 present average lever-press rates as a percent of the average control

(pre-drug) and saline rate, respectively. Such figures allow for analysis of relative or

proportional changes in rates after cocaine administration when compared to rates

obtained without drug administration or following saline injections. Points above the

dashed line indicate an increased rate of lever pressing relative to the control (or

saline) rate. Also evident from these figures is that acute cocaine administration either

increased lever-press rates (open symbols) at the medium or high doses for some

subjects (77-1, 77-2, 77-4) or dose-dependently decreased rates (77-3, 77-5).








34
The mean number of food pellets delivered during the experimental sessions are

presented Table 1. All six subjects earned the maximum number of pellets during the

determination of the acute dose-response function when doses lower than 17.0 mg/kg

were administered. Only Subject 77-1 obtained all possible pellets at all doses tested.

Effects of acute cocaine administration on lever-press distributions and post-

reinforcement pauses. Figure 4 (a-f) presents the distribution of lever pressing

throughout the 60-s interval following acute and repeated cocaine administration.

Control (non-drug) patterns are consistent with overall FI-type responding, i.e., as time

in the interval passes, more lever presses are emitted. When the total number of

responses in each 3-s bin of each interval is collapsed across all intervals in the

session, the resulting pattern is a positively-accelerated curve. Thus, there are few

lever presses immediately after reinforcer delivery and the majority of presses are

recorded just prior to reinforcer presentation. The absence of responding immediately

following reinforcer delivery under control conditions is also indicated by the open

symbols in Figure 5, which show that post-reinforcement pauses, on average, are

longer than half of the interval.

The two subjects (77-1 and 77-2) that showed dose-dependent increases in lever-

pressing rates during the acute dose-response determination, also showed dose-

dependent increases in responding at the beginning portions of the intervals (Figure 4,

open circles). For one subject (77-2), the increases in responding at the start of the

interval were accompanied by slight decreases in the later parts of the interval, relative

to the control conditions. This change resulted in a relatively equal distribution of








35
responses throughout the interval, which is evident at the largest doses (17.0 and 30.0)

of cocaine. The subject (77-4) with moderate increases in lever-press rates following

administration of the medium doses (5.6 and 10.0 mg/kg) also showed increases in

responding at relatively earlier portions of the interval (i.e., at mid-interval) but the

magnitude of this change was not large. For two other subjects (77-3 and 77-6), the

largest doses of cocaine (17.0 and 30.0 mg/kg) decreased the overall number of lever

presses and slightly increased lever pressing in earlier segments of the interval. No

overall change in the distribution of lever presses was recorded for Subject 77-5 when

lever pressing occurred. For all subjects, when cocaine increased responding in earlier

portions of the interval, dose-dependent decreases were also observed for post-

reinforcement pauses (see Figure 5, open circles).

Effects of repeated cocaine administration (17.0 mg/kg) on lever-press rates. The

dose-response functions for lever presses per minute following exposure to repeated

administration of 17.0 mg/kg (Figure 1, filled circles) generally indicated that response

rates either did not change or decreased, when compared to the functions following

acute cocaine administration (open circles). It should be noted however, that a shift in

response rate when saline was administered in this later dose-response function is

evident for a couple of subjects (77-1 and 77-6). In one case, the effect of repeated

administration of 17.0 mg/kg cocaine was to decrease the response rate after saline

injections relative to the acute effect (77-6) and in the other case (77-1), repeated

exposure to cocaine was followed by increased rates under saline conditions. The

corresponding dose-response functions shifted in a manner consistent with the change








36
in rates after saline administration. It is interesting to note that the subject for whom

lever pressing rates increased, following both acute and repeated cocaine

administration, was the one with the lowest non-drug rates.

Figure 2 presents average response rates as a percent of the average control (pre-

drug) rate. Points above the dashed line indicate an increased rate of responding

relative to the control rate. Lever-press rates after repeated administration of 17.0

mg/kg cocaine (Figure 2, filled circles), when plotted as a percent of control (non-

drug) rates, either did not change or decreased, when compared to rates assessed

during acute cocaine administration (open circles). When the change in lever-press

rates following saline injections was considered and drug effects were plotted as a

percent of saline (Figure 3), any changes observed between the acute (open circles)

and repeated (filled circles) cocaine dose-response functions were generally shown to

be decreases.

The average number of reinforcers delivered during sessions (see Table 1) when the

repeated dose (17.0 mg/kg) was administered, increased relative to the acute phase for

the two subjects that experienced the largest initial reinforcement loss (77-3 and 77-5).

These increases in number of obtained reinforcers were not accompanied by large

increases in lever-press rates (see Figure 1). For Subject 77-5, an increase in the

number of reinforcers was also observed following administration of 30.0 mg/kg

during this chronic dose-response assessment, when compared to the acute dose-

response assessment. No decrease in reinforcement occurred after repeated exposure

to cocaine for the two subjects (77-1 and 77-6) that earned all the possible food pellets








37
during acute cocaine administration. However, the two subjects (77-2 and 77-4) that

lost an average of 0.3 and 1.2 reinforcers per session following acute administration of

17.0 mg/kg, respectively, lost more reinforcers after repeated administration of 17.0

mg/kg cocaine at doses equal to or greater than this dose.

Effects of repeated cocaine administration (17.0 mg/kg) on lever-press

distributions and post-reinforcement pauses. Following repeated administration of

17.0 mg/kg cocaine, the overall shapes of the lever-press distribution curves (see

Figure 4 a-f, filled circles) were relatively unchanged when compared to the patterns

observed during assessment of the acute dose-response function (open circles). One

notable difference, however, was a slight decrease in responding during the early and

middle segments of the interval. The subjects often waited until later in the interval to

begin responding and when they did, they typically emitted fewer responses, but the

shape of the curve was still positively accelerated. This overall delay to response

initiation following exposure to repeated administration of 17.0 mg/kg (filled circles)

is also evident from the increased post-reinforcement pauses, relative to those recorded

during acute administration of cocaine (open circles), presented in Figure 5. For

subject 77-2, responding changed from the relatively constant and even distribution of

lever presses observed during the acute dose-response phase to more control-like

patterns following injections of 10.0 and 17.0 mg/kg during the chronic dose-response

phase doses (see Figure 4b).

Effects of repeated cocaine administration (10.0 mg/kg) on lever-press rates.

When the dose-response function was redetermined after repeated administration of








38
10.0 mg/kg cocaine (Figure 1, filled triangles) little or no change from the 17.0 mg/kg

dose-response function was observed for two of the subjects (77-2 and 77-3). For

Subject 77-1, lever-press rates decreased to control (pre-drug) levels. Lever-press

rates also decreased for Subject 77-4 during this assessment of the chronic dose-

response function for all doses of cocaine and saline, relative to both the acute and the

first chronic (17.0 mg/kg) dose-response functions. Lever-press rates decreased

slightly relative to the first chronic dose-response function only after administration of

the largest doses of cocaine for Subject 77-6. Subject 77-5's rates increased relative to

the 17.0 mg/kg cocaine dose-response function and returned to the levels seen

following acute cocaine administration, but no lateral shift in the function was

observed.

The change in lever-press rates following saline administration, relative to that

obtained in the acute dose-response phase, is observed during the second chronic dose-

response assessment for Subjects 77-4 and 77-6 and is evident when the rates are

plotted as a percent of control (see Figure 2, filled triangles). When the dose-response

functions are compared, relative to the rates after saline administration (Figure 3, filled

triangles), most doses of cocaine result in lever-press rates that are approximately

equal to or exceed rates following acute cocaine injections. This is especially notable

for the subjects whose lever-press rates changed after saline administration relative to

the acute dose-response function (77-4 and 77-6). Thus, when the change in rate of

lever pressing is considered, the drug effect appears to be attenuated.








39
The number of reinforcers delivered did not change after administration of any dose

lower than 17.0 mg/kg for the subjects that were earning the maximum number (30)

either during the acute dose-response assessment or after repeated exposure to 17.0

mg/kg cocaine (see Table 1). The number of food pellets earned after administration

of 17.0 mg/kg cocaine during the second chronic (10.0 mg/kg) dose-response

assessment either increased slightly or remained similar to the number earned after

repeated administration of 17.0 mg/kg cocaine for the remaining subjects. After

repeated exposure to 10.0 mg/kg cocaine, the number of reinforcers increased

following administration of 30.0 mg/kg for two subjects (77-5 and 77-6), remained

constant for one subject (77-1), and decreased for the other three subjects (see Table

1).

Effects of repeated cocaine administration (10.0 mg/kg) on lever-press distributions

and post-reinforcement pauses. For most subjects, no change was observed in the

shape of lever-press distributions after repeated exposure to 10.0 mg/kg (see Figure 4

a-f, filled triangles) compared to after 17.0 mg/kg administration (filled circles).

Subject 77-1's response patterns are an exception. Fewer lever presses were emitted at

the beginning of the interval, when compared to the first chronic dose-response

function, after saline and several doses of cocaine (1.0, 10.0, 17.0 mg/kg) were tested.

The patterns during determination of the second dose-response function more closely

resemble saline patterns from that same experimental phase, particularly when two of

the largest doses (10.0 and 17.0 mg/kg) were administered. For Subject 77-4, a

decrease in the number of lever presses emitted during the middle and later portions of


























Figure 1. Average lever presses per minute (+/- 1 SEM) as a function of the dose of
cocaine. Open circles represent control (non-drug) rates (C) and rates after acute
administration of various doses of cocaine and saline (S). Filled circles represent rates
following repeated exposure to 17.0 mg/kg cocaine. Filled triangles represent rates
following repeated exposure to 10.0 mg/kg cocaine.








41
150 50
77-1 T 77-2
125 125 acute
chronic 17.0 mg/kg
A chronic 10.0 mg/kg
100 100

75 75





00 I I
Z7 C S 1.0 3.0 5.6 10.0 17.0 30.0 C S 1.0 3-0 5.6 10,0 17,0 30.0

150r 150
r 5 77-3 77-4
L 1255 -
0L
100 100
C/o
75 75

Li 50 50 -


n 25 25 O

00
ELi C S 1.0 3.0 5.6 10.0 17.0 30.0 C S 1.0 3.0 5.6 10.0 17.0 30.0
150 150
LU 77-5 77-6
125 125

100 100

75 75
50- 50


25 25

0 0
C 5 1.0 3.0 56 10.0 17.0 30.0 C S 1.0 3.0 56 10.0 17.0 30.0


COCAINE (MG/KG)
Figure 1. Average lever presses per minute as a function of dose of cocaine.


























Figure 2. Average lever press rates plotted as a percent of control (non-drug) rates across
various doses of cocaine and saline (S). Open circles represent rates after acute
administration of various doses of cocaine. Filled circles represent rates following
repeated exposure to 17.0 mg/kg cocaine. Filled triangles represent rates following
repeated exposure to 10.0 mg/kg cocaine. The dashed line indicates 100 percent of
average control rates. Note the different ordinate scales.








43
800 300
_1 7 77-1 77-2 c2
700 0 acute
0 700 250 chronic 17.0 mg/kg
Elf 600 -A chronic 10.0 mg/kg
Z500 / 200 0- 0
Z soo 200
0
400 150


S3200 -
loo -
020 i o ---~-I-0 ------------ o ----------

S 1.0 3.0 5.6 10.0 17.0 30.0 S 1.0 3.0 5.6 17.0 0 30.0
0
200 200
J 1 77-3 77-4


(f) 150 150
<

I-- 100 -^ ... .... .. 100 ... . . .. ....
loo Io

- 50 A 50


0 \
SS 1.0 3.0 5.6 10.0 17,0 30.0 S 1.0 3.0 5.6 10.0 17.0 30.0
200 200 77-6
( 7) I 777-6


w 0
S150 150


100 100


L 50 50 -----



S 10 3.0 5.6 10.0 17,0 300 5 1 0 3,0 5-6 10.0 17.0 300
COCAINE (MG/KG)

Figure 2. Average lever-press rates as a function of percent of control (nondrug) rates.


























Figure 3. Average lever press rates plotted as a percent of rates following saline
administration. Open circles represent rates after acute administration of various doses of
cocaine. Filled circles represent rates following repeated exposure to 17.0 mg/kg cocaine.
Filled triangles represent rates following repeated exposure to 10.0 mg/kg cocaine. Bars
indicate SEM. The dashed line indicates 100 percent of average rates after administration
of saline. Note the different ordinate scales.








45
800oo 0300
8L 077-1 77-2
S7o00 / 2 acute
S25* chronic 17.0 mg/kg
J /00 A chronic 10.0 mg/kg
< 200 -
400 so50 /
0 13oo00 *
1020


0 200 200



77-3 77-4
DI oo l o o
< 0 150 __I A _
, S 1.0 3.0 5 0. 0 17.0 30.0 S 1,0 3.0 5.6 10.0 17.0 30.0
,LJ
U) 200 r 200


i 77-3 77-4


100 100

S150o 15so












5 0
) 100 lO ... . ..... ...... 100 .....0 *3 -<_,.. e V




S 50 ,50
n o < i 0
5 1.0 3.0 5.6 10.0 17.0 30.0 S 1.0 3.0 5.6 10 .0 170 300
f) 200 r- 200 -
LJ 77-5 77-6
CE

i 150 150
ry



ELJ 5 .




S 10 30 5.6 10.0 17.0 30. S 1.0 3.0 5.6 10.0 17.0 30.0

COCAINE (MG/KG)
Figure 3. Average lever-press rates as a percent of rates after saline administration.








46



Table 1. Average number of reinforcers earned as a function of dose of cocaine during acute
cocaine administration, the first repeated-administration/chronic condition (17.0 mg/kg),
and the second repeated-adminstrationichronic condition (10.0 mg/kg) for each subject.



Subject condition control saline 1.0 3.0 5.6 10.0 17.0 30.0
77-1 acute 30 30 30 30 30 30 30 30
chron 1 30 30 30 30 30 30 30
chron 2 30 30 30 30 30 30 29.8

77-2 acute 30 30 30 30 30 30 29.7 27.8
chron 1 30 30 30 30 30 22.1 7.3
chron 2 30 30 30 30 30 24.3 4

77-3 acute 30 30 30 30 30 30 23.5 17.5
chron 1 30 30 30 30 30 27.1 14
chron 2 30 30 30 30 30 28.3 13.7

77-4 acute 30 30 30 30 30 30 28.8 19
chron 1 30 30 30 30 30 26.3 15.7
chron 2 30 30 30 30 30 25.7 5.3

77-5 acute 30 30 30 30 30 30 25.2 0
chron 1 30 30 30 30 30 30 9.7
chron 2 30 30 30 30 30 30 13.3

77-6 acute 30 30 30 30 30 30 30 23.3
chron 1 30 30 30 30 30 30 22
chron 2 30 30 30 30 30 30 27


























Figure 4. Overall distribution of lever presses within the 60-second fixed-interval. Open
circles represent control (non-drug) distributions and distributions after acute
administration of various doses of cocaine. Filled circles and filled triangles represent
distributions at various doses of cocaine and saline following repeated exposure to 17.0
mg/kg cocaine and repeated exposure to 10.0 mg/kg cocaine, respectively. Bars indicate
SEM. Parts a-f correspond to each subject numbered 77-1 through 77-6. Note the
different ordinate scales.









400 48
40 CONTROL SALINE 48
o acute
300 chronic 17.0 mg/kg
A chronic 10.0 mg/kg

200






400
3 1.0 mg/kg 3.0 mg/kg



200



El 100
n


> 5.6 mg/kg 10.0 mg/kg
LJ
300 -

I-
0 200

^t 100^


400
4 17.0 mg/kg 30.0 mg/kg

300 -


200



A Ao I -.^ ^~ 1 1 ,I1 I- t 1 1 1 ,
10-A
-oo 2- A TA


0 2 4 6 8 10 12 14 16 18 20 0 2 4 6 8 10 12 14 16 18 20

3-S BIN NUMBER


Figure 4a. Overall distribution of lever presses in the 60-s interval for Subject 77-1.









400 49
CONTROL 77-2) SALINE acute
Si | chronic 17.0 mg/kg
A chronic 10.0 mg/kg
I T
200


100



400
1.0 mg/kg 3.0 mg/kg

J 300


(/ 200





L 400
S -5.6 mg/kg 10.0 mg/kg

LL 300
-J

LL 200







S F17.0 mg/kg 30.0 mg/kg
zoo
300


200





C I -f ~ A A A
100 i -


0 2 4 6 8 10 12 14 16 18 20 0 2 4 6 8 10 12 14 16 18 20


3-S BIN NUMBER


Figure 4b. Overall distribution of lever presses in the 60-s interval for Subject 77-2.









600- 5
CONTROL SALINE 5
500 7 acute
*- chronic 17.0 mg/kg
A chronic 10.0 mg/kg





300 0>a
200 -




100
600




( 0) 3oo0 T ,
1J If T

soo

Lii 400




300


200




C1 600

t 17.0 mg/kg 30.0 mg/kg



300

200

100


0 2 4 6 8 10 12 14 16 18 20 0 2 4 6 8 10 12 14 16 18 20


5-S BIN NUMBER


Figure 4c. Overall distribution of lever presses in the 60-s interval for Subject 77-3.









soo 51
FCONTROL SALINE
500o 77-4 acute
Schronic 17.0 mg/kg
400 A chronic 10.0 mg/kg

300

20o L

100


600
1.0 mg/kg 3.0 mg/kg
500



300





rJ 0 - -- i
200




LLJ 600 6
S 50.6 mg/kg 10.0 mg/kg
1 500

-- 400 -

LI. 300
0
200



0 F- A
600
D 17.0 mg/kg 30.0 mg/kg
Z 500 K

400 K

300 -

200 -




0 2 4 6 8 10 12 14 16 18 20 0 2 4 6 8 10 12 14 16 18 20


3-S BIN NUMBER



Figure 4d. Overall distribution of lever presses in the 60-s interval for Subject 77-4.









o00 52
500 CONTROL SALINE o acute
CO chronic 17.0 mg/k
00 -A chronic 10.0 mg/kg
400

300


100



S1.0 mg/kg 3.0 mg/kg
500 -

400 -
300 -

2 300 -







S100 -

LL- 300
0 o ^ 8 ^ ^ *.**| |'| I I. .
LJ r







200






5100

400
300











0 2 4 6 8 10 12 14 16 18 20 0 2 4 6 8 10 12 14 16 18 20


3-S BIN NUMBER


Figure 4e. Overall distribution of lever presses in the 60-s interval for Subject 77-5.









600 CONTROL SALINE

077 6' o acute
40* chronic 17.0 mg/kg

300 -






600
1.0 mg/kg 3.0 mg/kg
S500 -


L(J
(f) 300 -

200



sLJ oo -
100 -i






> 5.6 mg/kg 10.0 mg/kg








z 17.0 mg/kg 30.0 mg/kg
11 500 ~






400 -









200
300







0 2 4 6 8 10 12 14 16 18 20 0 2 4 6 8 10 12 14 16 18 20



3-S BIN NUMBER

Figure 4f. Overall distribution of lever presses in the 60-s interval for Subject 77-6.



























Figure 5. Average post-reinforcement pause as a function of dose of cocaine (+/- 1
SEM). Open circles represent control (non-drug) rates (C) and rates after acute
administration of various doses of cocaine and saline (S). Filled circles represent rates
following repeated exposure to 17.0 mg/kg cocaine. Filled triangles represent rates
following repeated exposure to 10.0 mg/kg cocaine.








55
80 r- 80 -
S77-1 0 77-2 0 acute
70 70 chronic 17.0 mg/kg
60 60 chronic 10.0 mg/kg
T


S40 LT 40
Z 30 30 O
2
20 20 -
0o 10

C S 1.0 3.0 5,6 10.0 17.0 30.0 C S 1.0 3.0 5.6 10.0 17.0 30.0
LJ
S80 80
) 77-3 77-4
< 70 70 -


F- 1
na
60 I 60 -

S50 50
LL40 40 -






o
10 10 -
Li
Z-0O '0 0 I---------I
C S 1.0 3.0 5.6 10.0 17.0 30.0 C S 1.0 3.0 5.6 10.0 17.0 30.0

so r80
80 77-5 77-6
70 70
-60 F- 60
0 50 5- 50

30 30 -


20 20 -
10 10 -
0 I i i I i-- i I i i i
C 5 1.0 3.0 5.6 10.0 17.0 30.0 C S 1.0 3.0 5.6 10.0 17.0 30.0

COCAINE (MG/KG)
Figure 5. Average post-reinforcement pauses as a function of dose of cocaine.








56
the interval was observed following administration of all doses, but was most evident

after injections of saline and the lowest dose of cocaine (1.0 mg/kg). Lever pressing

was also completely eliminated after 30.0 mg/kg cocaine. The decreased number of

lever presses in the later portions of the FI are accompanied by a decrease in the

number of reinforcers delivered following 17.0 and 30.0 mg/kg cocaine

administration. For most subjects and for most doses of cocaine, post-reinforcement

pauses did not change after repeated exposure to 10.0 mg/kg cocaine (Figure 5, filled

triangles) when compared to the first chronic dose-response function (filled circles).

Relative to control durations (open circles), pauses were unchanged or were slightly

longer following repeated exposure to 10.0 mg/kg cocaine.

Effects of acute cocaine administration on lick rates and amount consumed. Figure

6 shows the effects of various doses of cocaine on licking rates. Control rates of

licking (range: 29.59 91.11 licks per minute) were more variable between subjects

than lever- pressing rates. Acute cocaine administration (open circles) resulted in a

dose-dependent decrease in the rates of licking for all six subjects with complete

suppression observed for all but two subjects (77-4 and 77-6) when the largest dose,

30.0 mg/kg, of cocaine was administered.

Figures 7 and 8 (open circles) present lick rates as a percent of control rates and

rates following saline administration, respectively, during the acute dose-response

determinations. Lick rates decreased more (as a percent of control) following

administration of the largest doses of cocaine than lever-press rates (see Figure 2, open








57
circles). Similar effects are noted when lick rates are plotted as a percent of rates

following saline injections (Figure 8, open circles).

The average number of milliliters consumed for each subject are presented in

Figure 9. Milliliters consumed ranged from approximately 7.5 to 19 under control

(non-drug) conditions. This quantity also decreased dose-dependently and these

changes generally correspond to the decreases in lick rates.

Effects of acute cocaine administration on lick distributions. Figure 10 (a-f) shows

that for all subjects, licking occurred predominantly in the first half of the interval.

Licking usually commenced immediately after reinforcer delivery and continued until

about a third or a half of the 60-s interval had elapsed. One subject (77-5) licked the

drinking tube during later portions of the interval, but the peak of the distribution

curve was still at the beginning of the interval. During the acute dose-response

determination (open circles), as the dose of cocaine increased the total number of licks

decreased, and for two of the subjects (77-3 and 77-5), responding was also dose-

dependently restricted to earlier segments of the interval. This effect is most

pronounced in 77-5's licking and can be seen by comparing the effects of the medium

and larger doses (5.6 17.0 mg/kg) of cocaine to those of saline and the smaller doses

(1.0 3.0 mg/kg) of cocaine.

Effects of repeated cocaine administration (17.0 mg/kg) on lick rates and amount

consumed. When 17.0 mg/kg cocaine was administered repeatedly and the dose-

response function was reassessed (Figure 6, filled circles), it was found that licking

rates either matched those seen with the acute dose-response function or were








58
decreased. In two instances there was a slight leftward shift of the curve (77-3 and 77-

5). Only one subject (77-2) showed any rate increases following cocaine (1.0 and 5.6

mg/kg) administration, but these increases were also seen when saline was injected. A

change in licking following saline administration was also noted for Subject 77-1. In

addition, after exposure to repeated administration of 17.0 mg/kg cocaine, licking for

this subject was greatly reduced at all doses relative to the acute effects (open circles).

Lick rates for Subject 77-6 also decreased relative to the acute effects at doses of

cocaine larger than 1.0 mg/kg. These decreases were observed without a large change

in responding after saline injections.

Figure 7 shows these lick-rate changes as a percent of control (non-drug) rates.

Filled symbols above the dashed line indicate that lick rates following repeated

exposure to 17.0 mg/kg were greater than control rates. Again, the increase in licking

for Subject 77-2 and the decreases in licking for the other subjects are evident along

with any change in lick rates following saline administration. When the dose-effects

are plotted as a percent of these altered rates (Figure 8, filled circles), the relative

increases after administration of 3.0 and 5.6 mg/kg for Subject 77-1 appear large, even

though overall lick rates are extremely low.

The total milliliters consumed during assessment of the chronic dose-response

function (Figure 9, filled circles), generally equaled what was consumed during the

acute dose-response function (open circles) or they were slightly decreased. The

changes in amount consumed were similar to the changes in lick rates for all subjects.








59
Effects of repeated cocaine administration (17.0 mg/kg) on lick distributions.

When acute cocaine administration restricted licking to the earlier parts of the interval,

the effects of these same doses after repeated exposure to 17.0 mg/kg cocaine

enhanced this effect (see Figure 10 a-f, filled circles). This is evident for subject 77-3

at doses of 1.0 5.6 mg/kg and for subject 77-5 at 1.0 3.0 mg/kg cocaine. It is also

noted that for these two subjects a leftward shift in the chronic dose-response function

for licks per minute was observed. For subject 77-4 an overall decrease in licking

resulted from repeated 17.0 mg/kg cocaine exposure, but the shape of the distribution

curve remained intact. An increase in the number of licks seen at the beginning of the

interval was noted for Subject 77-2 not only for the low to medium doses of cocaine,

but also for saline. For most doses of cocaine, licking was nearly or completely

suppressed for Subjects 77-1 and 77-6.

Effects of repeated cocaine administration (10.0 mg/kg) on lick rates and amount

consumed. Figure 6 shows that the dose-response functions obtained following

repeated exposure to 10.0 mg/kg (filled triangles) generally were lower or shifted to

the left of the initial (acute) dose-response curve (open circles) for five of six subjects.

For the exception (Subject 77-6), lick rates were increased at the lower doses of

cocaine (1.0 5.6 mg/kg), relative to the acute dose-response function, but this change

was accompanied by a small change in responding following saline administration. In

general, for this subject, lick rates following cocaine injections were substantially

increased relative to the dose-response function obtained after repeated exposure to








60
17.0 mg/kg (filled circles). A shift in lick rates after saline administration was also

observed for Subjects 77-4 and 77-5.

The relative change from control rates after repeated administration of 10.0 mg/kg

(Figure 7, filled triangles) is similar to the change following repeated exposure to 17.0

mg/kg (filled circles). Again, Subject 77-6 is the exception. Lick rates after

administration of the low to medium doses returned to control rates. When the change

in lick rates after repeated exposure to 10.0 mg/kg (Figure 8, filled triangles) was

plotted as a percent of the changed rates following saline administration, further

relative decreases were observed for Subject 77-1 while relative increases were noted

for Subject 77-4. Rates approached 100% of rates after saline injections or the

respective rates following the acute dose-response function (open symbols) for the

remaining subjects.

During the second chronic dose-response function (Figure 9, filled triangles), the

amounts consumed either did not vary from the first chronic dose-response function

(filled circles) or decreased. The total milliliters consumed was in good

correspondence with the overall lick rates for all subjects.

Effects of repeated cocaine administration (10.0 mg/kg) on lick distributions.

Figure 10 (a-f) shows that lick distributions returned to the initial (acute) dose-

response patterns (open circles) for Subjects 77-2 and 77-6 after 10.0 mg/kg cocaine

(filled triangles) was presented repeatedly. Also, for Subject 77-6, the pattern after

administration of the low to medium doses resembled that seen during control

distributions. For Subjects 77-3, 77-4, and 77-5 the patterns did not differ greatly


























Figure 6. Average licks per minute as a function of dose of cocaine (+/- 1 SEM). Open
circles represent control (non-drug) rates (C) and rates after acute administration of
various doses of cocaine and saline (S). Filled circles represent rates following repeated
exposure to 17.0 mg/kg cocaine. Filled triangles represent rates following repeated
exposure to 10.0 mg/kg cocaine.









62
150 r- 150
77-1 77-2
125 125 0- acute
6 chronic 17.0 mg/kg
100 100 A chronic 10.0 mg/kg

75 75 -

50 -50 -

25 25

0 T 0 I
C S 1.0 3.0 5.6 10.0 17.0 30.0 C S 1.0 3.0 5.6 10.0 17.0 30.0

L.J 150 150 r-
77-3 77-4
125 125

S100 100

75 A 75
0 L

50 50 -

25 \ 25 V


-I C S 1.0 3.0 5.6 10.0 17.0 30.0 C S 1.0 3.0 ..6 10.0 17.0 30.0


S77-5 77-6
125 125

100- 100

75- 75 -

50 1 "\ 50

25 1. .\. 25


C S 1.0 3.0 5.6 10.0 17.0 300 C S 1.0 3.0 5.6 10.0 17.0 30.0

COCAINE (MG/KG)

Figure 6. Average licks per minute as a function of dose of cocaine.



























Figure 7. Average lick rates plotted as a percent of control (non-drug) rates across
various doses of cocaine and saline (S). Open circles represent rates after acute
administration of various doses of cocaine. Filled circles represent rates following
repeated exposure to 17.0 mg/kg cocaine. Filled triangles represent rates following
repeated exposure to 10.0 mg/kg cocaine. The dashed line indicates 100 percent of
average control rates. Note the different ordinate scales.








64
250 250
77-1 I 77-2
20 200 0 acute
0 chronic 17.0 mg/kg
A chronic 10.0 mg/kg
0 150 150 -
O iso


Z 100 .100 -
so so

S 0 50
LL_ t" -. ,
0 I_____--t_-_tt_ 0
S 1 0 3.0 5.6 10.0 17,0 30.0 S 1 0 3.0 5.6 10.0 17.0 30.0
F--
S200 250
LJ 77-3 77-4

1 200
LLJ 150

5 o150
) oo00 ..... ..........
< u 100
Li

-- 50 5\ s

Z \. 0 A 0
S 1.0 3.0 56 10.0 17.0 30.0 S 1.0 3.0 5.6 10.0 17,0 30.0

f 200 r 200
L 77-5 77-6
CL
150 150 -
(.,

o100 100

A 0
50 A 50-


0-------L-------- o----
S 1.0 3.0 5.6 10.0 170 30,0 S 1.0 3.0 5.6 10.0 170 30.0

COCAINE (MG/KG)
Figure 7. Average lick rates as a percent of control (nondrug) rates.


























Figure 8. Average lick rates plotted as a percent of rates following saline administration.
Open circles represent rates after acute administration of various doses of cocaine. Filled
circles represent rates following repeated exposure to 17.0 mg/kg cocaine. Filled
triangles represent rates following repeated exposure to 10.0 mg/kg cocaine. The dashed
line indicates 100 percent of average control rates. Note the different ordinate scales.








66
250 250 0
77-1 77-2
2 acute
200 200 e ocute
Schronic 17.0 mg/kg
chronic 10.0 mg/kg


100 100
< yA
150 AA 50
t
0 7o. 0o- 300-0-S.0 3.0-- 0
S .0 3.0 5.6 10.0 7.0 30.0 S 10 30 5 10.0 17.0 30.0
F-
200 250
.. 77-3 77-4
ScA
r 200 A
LL 150

150
(1) 100 ... 2 -A ...... ........ ......
< 0 -

50 \ 50 -

z
------------A^-- o i----------^^







S0 b 0
S 1.0 3.0 5.6 10.0 17.0 300 S 1.0 3.0 5.6 10.0 17.0 30.0
ry 200 200
Lj 77-5 77-6
n

150 150
U')



AA,


0 0
--------------i o i---- ^ ______ L J-







5 1.0 3.0 5.6 10.0 17.0 30.0 S 1.0 3.0 5.6 10.0 17.0 30.0

COCAINE (MG/KG)
Figure 8. Average lick rates as a percent of rates following saline administration.


























Figure 9. Total number of milliliters of water consumed as a function of dose of cocaine.
Open circles represent control (non-drug) amounts (C) and amounts after acute
administration of various doses of cocaine and saline (S). Filled circles represent
amounts following repeated exposure to 17.0 mg/kg cocaine. Filled triangles represent
amounts following repeated exposure to 10.0 mg/kg cocaine. Bars around the mean
indicate standard error.








68
25 25
77-1 2 77-2 acute
Schronic 17.0 mg/kg
20 20 chronic 10.0 mg/kg




10 10 -


L L

0 0 0 I-L y
C S 1.0 3.0 5.6 10,0 17.0 30.0 C S 1,0 3.0 5.6 10,0 17.0 30.0

() 25r- 25
Z 77-3 77-4
2


-- 20 20






C S 1.0 3.0 5.6 10.0 17.0 30.0 C 5 1.0 3.0 5.6 10.0 17.0 30.0
-J
-) 5 15 -








-- 25 25



T
10



0 5
20 T 8-
















C S 1.0 3.0 5.6 10.0 17.0 30.0 C S 1.0 3.0 5.6 10.0 17.0 30.0

COCAINE (MG/KG)
Figure 9. Average milliliters consumed as a function of dose of cocaine.



























Figure 10. Overall distribution of licks within the 60-second fixed-interval. Open circles
represent control (non-drug) distributions and distributions after acute administration of
various doses of cocaine. Filled circles and filled triangles represent distributions at
various doses of cocaine and saline following repeated exposure to 17.0 mg/kg cocaine
and repeated exposure to 10.0 mg/kg cocaine, respectively. Parts a-f correspond to each
subject numbered 77-1 through 77-6. Note the different ordinate scales.








400 CONTROL FSALINE 70

7 i-1) o Lacute
I; t> L T !: o
300 T chronic 17.0 mg/kg
I A chronic 10.0 mg/kg|
200 -

iao -t i


400 *
1.0 mg/kg T- 3.0 mg/kg
300

200~






0 5.6 mg/kg 10.0 mg/kg
300

m 200





100
17.0 mg/kg 30.0 mg/kg
300

200

100


0 2 4 6 8 10 12 14 16 18 200 2 4 6 8 10 12 14 16 18 20

3-S BIN NUMBER


Figure 10a. Overall distribution of licks in the 60-s interval for Subject 77-1.









700 71
CONTROL J SALINEo acute
o-- I i* chronic 17.0 mg/kg
5o00 chronic 10.0 mg/kg

400


200 i1

I 0

700
1.0 mg/kg 3.0 mg/kg
600


400 -

300 -

200 -



L_
-J 100o



S 5.6 mg/kg 10.0 mg/kg
600 -

Lio 50o
4D 00

300




SI^ 1- I i I .A i .1 *..i. 1 A
200


17.0 mg/kg -30.0 mg/kg
500

400 -

300 -

200 -

100 -


0 2 4 6 8 10 12 14 16 18 200 2 4 6 8 10 12 14 16 18 20



3-S BIN NUMBER


Figure 10b. Overall distribution of licks in the 60-s interval for Subject 77-2.









700 72
CONTROL --- SALINE o acute
soo 77- t- chronic 17.0 mg/kg
SrA chronic 10.0 mg/kg



300

200
100


700
1.0 mg/kg 3.0 mg/kg
600

500




200

U 1001 I, _,_


70oo
S 5.6 mg/kg 10.0 mg/kg

500 -

400



300
700



7 170 mg/kg 30.0 mg/kg
600















500

400

3 00

200

100 -

0 2 4 6 8 10 12 14 16 18 200 2 4 6 8 10 12 14 16 18 20



3-S BIN NUMBER


Figure 10c. Overall distribution of licks in the 60-s interval for Subject 77-3.









600 73
SCONTROL SALINE
5oo00 (774) acute
'" I chronic 17.0 mg/kg
00- A chronic 10.0 mg/kg

300

200 I-




600 -
1 .0 mg/kg i3.0 mg/kg












300
200

500





\,oo






10 2 m g/kg 2 4 0.0 mg/kg
500





200




0 2 4 6 8 10 12 14 16 18 200 2 4 6 8 10 12 14 16 18"20






Figure 10d. Overall distribution of licks in the 60-s interval for Subject 77-4.
500S -I UM


40ur 0d vrl itiuino ik nth 0sitra o ujc 71








600 74
CONTROL SALINE I acute
500 77-5 ch roni c 17 0 kg
400 chronic 17.0 mg/kg

400
oo -$



2 0 m0



600
1.0 mg/kg 3.0 mg/kg
500




200 I-T







400 -

m 300

200

100 K
o .. ._ _.. _ __ _

00
17.0 mg/kg 30.0 mg/kg
500
S200 i t I .Lbt r







400






0 2 4 6 10 2 4 16 18 200 2 4 6 8 10 12 14 16 18 20



3-S BIN NUMBER


Figure lOe. Overall distribution of licks in the 60-s interval for Subject 77-5.
3- BN UM E

Fiue1e vrl dsrbto flcs nte6- neva o ujc 75









7-CONTROL -SALINE N acute
00 7 77-6 T- chronic 17.0 mg/kg
500 _" -J- _A chronic 10.0 mg/kg


300 -
200 -



700
7 1.0 mg/kg 3.0 mg/kg
600 -

too



400 L
300
200
U 100K


i_ 5.6 mg/kg 10.0 mg/kg




7 600-




7oo-17.0 mg/kg 30.0 mg/kg
500 -


400
300
S200 -/

z







17.0 mg/kg 6 012 1618 0 mg/kg2 4 10 14 18







3-S BIN NUMBER


Figure 10f. Overall distribution of licks in the 60-s interval for Subject 77-6.



























Figure 11. Overall lever press rates (upper panel), overall lick rates (center panel), and
total amount consumed (lower panel) during control and probe sessions (+/- S.E.M.).
Filled bars indicate rates and milliliters consumed during the FI 60-s schedule, double-
hatched bars represent rates and milliliters consumed during the two extinction sessions,
and single-hatched bars represent rates and milliliters consumed during the massed
pellet/extinction sessions.










80 77
70 control Fl 60 s
.c extinction
E 60 massed pellet/ext

C 0 50

U 40 -

3 30-

S20-

10


77-1 77-2 77-3 77-4 77-5
100 -
loo
so
a so
S70

E 60o
- 50


.2
S40

- 30

-J 20
10

0
77-1 77-2 77-3 77-4 77-5

21

O 18

E i



Ur

fn 9







77-1 77-2 77-3 77-4 77-5

Subject


Figure 11. Average response rates and milliliters consumed during probe sessions.








78
from the first chronic dose-response function (filled circles) at the low or medium

doses, or from the most recent pattern following saline administration. Subject 77-1's

licking remained almost completely suppressed.

Nearly complete suppression of licking at all doses tested was recorded for the

subject (77-1) with the lowest overall lick rate following repeated exposure to both

doses of cocaine (17.0 and 10.0 mg/kg). Also regardless of the repeated dose, all

subjects' licking was either completely suppressed, or more greatly suppressed,

following administration of the largest doses (17.0 and/or 30.0 mg/kg) of cocaine.

Extinction and massed-pellet-presentation probes. Figure 11 shows the rates of

lever pressing and licking, as well as the total milliliters of water consumed, for the

two types of probe sessions and their control sessions. Subject 77-6 died before this

phase of the experiment was conducted. For the five remaining subjects, lever-press

rates during the FI 60-s schedule of reinforcement were similar to or lower than the

control (non-drug) rates assessed at the beginning of the experiment (see Figure 1).

Subject 77-5 is an exception as lever-press rates were increased relative to initial (pre-

drug) rates. Lick rates and milliliters consumed under control conditions revealed less

of a decrease compared to pre-drug control data (see Figures 6 and 9) than those

observed for lever-press rates. In the probe sessions, when the intermittency of food

pellet delivery was eliminated, lever-press and lick rates, as well as milliliters

consumed, decreased relative to their respective control values (filled bars).








79
Discussion

The results from the present experiment yielded several interesting findings. First,

it was found that for most subjects, cocaine administration dose-dependently produced

decreases in lever pressing maintained by an FI 60-s schedule of food presentation as

well as in schedule-induced licking, and that little or no tolerance to these rate-

decreasing effects was observed for either activity. However, post-reinforcement

pausing and temporal patterning of lever pressing did approximate control values for

some subjects after repeated exposure to cocaine, but these changes did not always

correspond to an increase in reinforcers delivered. Second, although no leftward shifts

in lick distributions were observed, there was a tendency for licking in the middle

portions of the intervals to be suppressed dose-dependently for some of the subjects

and again, little or no tolerance was observed for these effects. Third, no systematic

interaction between lever pressing and licking following cocaine administration was

noted in this experiment.

Little or no tolerance was observed to cocaine's rate-decreasing effects on lever

pressing for any of the subjects for either repeated dose of cocaine (10.0 or 17.0

mg/kg). For one subject (77-1), tolerance to the rate-increasing effects of cocaine was

noted after repeated exposure to 10.0 mg/kg cocaine as lever-press rates returned close

to control values. (This effect is also observed after repeated dosing with 17.0 mg/kg

cocaine if the rates are plotted as a percent of rates following saline administration.)

In addition, the lever-press distributions that characterized the rate increases at the

largest doses, i.e., increased responding at the beginning of the interval, began to








80
resemble those distributions recorded during non-drug conditions, i.e., a positively-

accelerated curve. This return to control patterning was also seen for Subject 77-2 but

was accompanied by a decrease in lever-press rates and number of reinforcers

obtained.

For some of the other subjects, lever-press rates decreased after an extended history

with the drug and/or the FI schedule in the experimental context. This change in

baseline response rates was evident from the changes in control (non-drug) lever-press

rates from the beginning of the acute dose-response function to the rates observed at

the end of the experiment with the extinction and massed-pellet presentation probes.

However, this decrease in lever-press rates did not systematically affect the number of

response-contingent food pellets presented. In some cases after repeated dosing,

although lever-press rates did not change, relative to cocaine's acute effects, the post-

reinforcement pauses returned to control durations and the number of obtained

reinforcers increased (e.g., 77-3 and 77-5 after administration of 17.0 mg/kg). In other

words, although the number of responses within the intervals did not vary, on average,

their temporal organization did change slightly when some reinforcers were no longer

obtained.

The subjects for whom no tolerance to decreased lever-press rates were observed,

generally maintained the same pattern of response distributions, but fewer numbers of

responses were emitted. It might therefore be argued that exposure to the rate-

reducing effects of cocaine allowed the behavior to contact the FI contingency more

efficiently. Since only one response is needed for reinforcement to occur in an FI-








81
schedule context, decreased response rates would not necessarily result in decreased

reinforcement frequency. In fact, under the drug conditions with reduced response

rates, the same number of reinforcers may be obtained with less response effort (i.e.,

fewer lever presses emitted). Therefore, the lack of tolerance to the rate-decreasing

effects for some of the subjects under the present schedule contingencies is not

surprising given that reinforcement frequency was not greatly altered, except by the

largest dose of cocaine (30.0 mg/kg).

A slight shift toward earlier segments of the FI was noted for many subjects' lever

pressing patterns following acute cocaine administration. These small increases in

lever pressing during the middle and early portions of the FI (and decreases during the

later portions of the intervals) are evident from the overall distributions and usually

occurred following administration of the medium or large doses of cocaine.

Consistent with other findings with amphetamine (e.g., McAuley & Leslie, 1986), it is

likely that the increased lever pressing observed during the middle portions of the

interval is due to earlier response initiation, rather than a rate-dependent effect, per se.

Although some changes in acceleration of rats' lever-pressing rates were reported by

McAuley and Leslie (1986) they suggested that this was most likely related to the

decreased post-reinforcement pause (i.e., disruption of temporal control) rather than a

direct effect on response rate. The shortened time to response initiation is suggested in

the present experiment by the decreases in the post-reinforcement pauses, which

occurred dose-dependently for most subjects. Although not evaluated in the present

work, the degree of this disruption may be related to the schedule parameter used.








82
McAuley and Leslie (1986) found that the increase in responding at the beginning of

the interval was greater for rats exposed to an FI 120-s schedule than for rats exposed

to an FI 60-s schedule of reinforcement. This effect of the schedule parameter may

have contributed to the minimal changes in overall patterning observed in the present

experiment.

The overall rates and patterns of schedule-controlled lever pressing and schedule-

induced drinking within the FI 60-s schedule of reinforcement were consistent with

those reported in other studies (cf. Sanger & Blackman, 1978). Drinking was shown

to fit the general characterization of schedule-induced behavior, not only by its

temporal patterning, but also because elimination of the intermittency of food-pellet

presentation resulted in decreased licking (cf. Falk, 1971). The present results also

suggest that schedule-induced licking was easier to disrupt following cocaine

administration than schedule-controlled lever pressing maintained by an FI 60-s

schedule of reinforcement. Although licking generally occurred at a greater overall

frequency than lever pressing, there were no explicitly arranged consequences for this

behavior, which may have contributed to the differential effects of the large and

medium doses of cocaine.

These results further extend previous findings with d-amphetamine's effects on FI

responding and schedule-induced drinking (cf. Sanger & Blackman, 1978; Williams &

White, 1984) to cocaine administration. One difference, however, concerns the

leftward shift in the distribution of licks following acute stimulant administration.

Although these effects are reported frequently with amphetamine (Flores & Pellon,








83
1997; Sanger, 1978; Williams & White, 1984), a similar decrease in licking during the

middle segments of the interval, but not a leftward shift, was observed in the present

experiment for only two subjects. However, this effect may be due to the FI schedule

parameters and not the specific drug employed. It has been reported (Flores & Pellon,

1997) that the shift to the left in lick distributions is influenced by the duration of the

FI. In the Flores and Pellon (1997) study, the peak of the distribution of licks was

recorded further into the interval when the schedule was an FI 120-s, as compared to

an FI 60-s schedule of reinforcement. Thus, if longer FI values had been programmed

in the present experiment, it may have been possible to record such a shift in

patterning for more of the subjects following cocaine administration.

Licking decreased in a dose-dependent manner following acute cocaine

administration consistent with earlier findings (cf. Sanger & Blackman, 1978). For

some of the subjects, drinking became more restricted to the early segments of the

interval. The subjects initiated drinking about the same time, but stopped drinking

sooner than under non-drug conditions. Thus, this change cannot be characterized as a

leftward shift. This earlier cessation of drinking does not appear to be influenced by

lever pressing as lever presses did not change greatly with respect to temporal

patterning with the exceptions of 77-1 and 77-2 during assessment of the acute dose-

response function.

It is possible that the decrease in licking after administration of the large doses of

cocaine may be a result of longer interfood intervals. Schedule-induced licking has

been shown to be sensitive to the interfood interval (Falk, 1966b) and may be








84
attenuated by an increase in FI duration. Since the intermittency of food pellet

presentation is an important determinant of schedule-induced drinking (cf. Falk,

1966a) and large doses of cocaine can decrease the number of food pellets delivered, it

may be of no surprise that drinking is reduced. In such an instance, it would be

appropriate to suggest that cocaine's effects on schedule-induced drinking were

indirect as a result of its direct effects on lever pressing. However, this explanation

seems unlikely given that decreases in licking were observed at doses that did not

decrease lever-press rates or the number of reinforcers delivered (30 reinforcers in a

maximum of a 35-min session).














EXPERIMENT TWO

Introduction

In this experiment, the effects of acute and repeated cocaine administration on

responding (lever pressing and licking) were examined in a multiple-schedule context

with response-dependent (FI 60-s) and response-independent (FT 60-s) schedules of

food presentation. This arrangement was chosen for several reasons. First, since food

pellet (reinforcer) presentation is one of the determinants of schedule-induced drinking

and cocaine can suppress the response required for food pellet delivery, it is possible

that any changes in licking may be due to an indirect effect (e.g., longer interfood

intervals) rather than a direct action of the drug. Second, any change in either

response during the FI components may be due to a drug effect on a competing

response. For instance, if cocaine administration produces increased lever pressing in

the early segments of the interval, licking may decrease as an indirect result of this

increase in pressing. Thus, by using this MULT FI 60-s FT 60-s schedule

arrangement, the acute and repeated drug effects on licking (rates and distributions)

could be assessed in a context where no schedule-controlled response is required and

compared with one in which lever pressing is necessary for food presentation.






85








86
Method

Subjects

Six male Wistar rats with no prior experimental history served as subjects. They

were obtained from a commercial supplier (Charles River Laboratories, Wilmington,

MA) and were approximately 100 days old with weights ranging from 344 440 g at

the start of the experiment. Subjects were housed in two cages (three rats per cage)

under a 12-hour light-dark cycle (lights on at 7 p.m. and off at 7 a.m.). Approximately

16 g of Purina Rat Chow was placed in each of the cages after each experimental

session (and on weekends). Subjects usually consumed this food in about 1.5 hours.

Free access to tap water in the home cage was provided at all times and temperature

conditions remained constant.

Apparatus

Experiments were conducted in the same two Coulbourn Instruments rodent

operant conditioning chambers that were used in Experiment 1.

Procedure

Lever-press acquisition and initial training. Subjects were placed in a darkened

chamber at the beginning of an experimental session, which was conducted during the

dark phase of the subjects' light-dark cycle. After 10 min, the houselight and the lever

lights were illuminated and food pellets were presented according to an RT 60-s

schedule. Food presentation was determined every second with a probability of 0.17.

In addition, food pellets were presented following each recorded lever press (FR 1








87
schedule of pellet presentation). Sessions ended following 45 min or after the delivery

of 40 food pellets, whichever came first (cf. van Haaren, 1992b).

Once subjects reliably pressed the lever, the schedule of reinforcement was

changed to a multiple fixed-interval 30-s fixed-time 30-s (MULT FI 30-s FT 30-s)

schedule. These schedule contingencies were in effect for five experimental sessions.

The terminal schedule of reinforcement was a MULT FI 60-s FT 60-s schedule of

reinforcement. Each component of the multiple schedule was presented twice in

simple alternation. The schedule component to start each experimental session was

randomly determined at the start of each session. Each component of the multiple

schedule was in effect for eight minutes, and a maximum of seven food pellets could

be delivered during each component. During the FI 60-s schedule, the lever was

extended into the experimental chamber and the stimulus lights above the lever were

illuminated. When the FT 60-s schedule was in effect, the response lever was

retracted from the experimental chamber and the lever lights were extinguished. The

houselight was illuminated during all schedule components. Experimental sessions

were terminated after the completion of the fourth component of the multiple schedule

and were conducted Monday through Friday.

Acute cocaine administration and dose-response function determination. Once

lever pressing and licking rates were stable, as determined by visual inspection of the

data (approximately 85 sessions), an injection of saline was administered (IP) prior to

an experimental session. After observing no disruption in response rates (lever

pressing and licking), subjects received an IP injection of saline or cocaine (1.0, 3.0,








88
5.6, 10.0, 17.0, or 30.0 mg/kg) on Tuesdays and Fridays of each week. The series of

injections began with saline and the subsequent doses of cocaine were presented in a

descending order. The effects of each dose were assessed at least twice (range: 2 4).

Repeated saline administration. After the acute dose-response function was

determined and before cocaine was administered before every session (Monday

through Friday), all subjects received injections of saline (IP) prior to the start of the

experimental session. This phase lasted for 5 consecutive days (when lever pressing

and licking rates were observed to be stable).

Repeated cocaine administration and dose-response function determination. After

observing no disruption in response rates following repeated saline administration,

cocaine was given prior to each session. The dose of cocaine chosen for each subject

was individually determined so that doses would be functionally equivalent. The

selection criterion was a dose that decreased licking (during FI or FT) but that did not

completely eliminate it. Due to the exceptionally low baseline licking rates for subject

78-1, a dose was chosen that produced a lick rate increase. Subject 78-1 received 1.0

mg/kg, Subject 78-2 received 5.6 mg/kg, and Subject 78-5 received 17.0 mg/kg, while

the other Subjects (78-3, 78-4, and 78-6) were given 10.0 mg/kg cocaine before every

session. Each subject was exposed to the drug dose in the experimental context for 46

sessions. After this time, the dose-response function was redetermined by presenting

saline or cocaine (1.0, 3.0, 5.6, 10.0, 17.0, or 30.0 mg/kg) twice per week as

substitutes for the repeated dose. The effects of each dose, with the exception of the

dose that was given daily, was assessed at least twice (range: 2 4).








89
Data collection. The data collected in Experiment 2 were the same as in

Experiment 1. Data were collected in each component (n=4) and presented as two

overall session averages (for FI and FT schedules). Which schedule (FI or FT) began

each session was also recorded and evaluated for systematic effects on the subsequent

data obtained.

Results

Effects of acute cocaine administration on lever-press rates. Figure 12 shows lever

presses per minute for subjects' responding maintained by an FI schedule in a MULT

FI 60-s FT 60-s schedule context. The data from both FI components are combined

because there was little variability between responding in the two components. Mean

lever-press rates across subjects ranged from 5.88 to 33.60 presses per minute with all

but one subject's rates noted to be less than 15 per minute. These rates were lower

than those seen for responding during the FI 60-s schedule alone (Experiment 1).

Occasional administration (acute dosing) of cocaine (open symbols) generally

resulted in no change in pressing rates at the low to medium doses (1.0 10.0 mg/kg).

Cocaine only produced rate increases for two rats, 78-1 at 3.0 through 17.0 mg/kg and

78-4 at 10.0 and 17.0 mg/kg. Theses two subjects had the lowest overall lever-press

rates during control (non-drug) conditions. At the largest dose (30.0 mg/kg) lever

pressing was greatly attenuated or suppressed in most subjects.

Figures 13 and 14 present lever-press rates as a percent of rates following control

sessions and sessions that were preceded saline administration, respectively. The rate

increases after administration of the mid-range to large doses of cocaine during the








90
acute dose-response function (open symbols) are evident for Subjects 78-1 and 78-4.

Lever-press rates for the other subjects did not differ greatly from rates in control or

post-saline-injection sessions, except at the largest dose of cocaine administered (30.0

mg/kg).

The mean numbers of food pellets delivered during both FI components for all

subjects are presented in Table 2. Under control conditions, all subjects earned all or

almost all of the food pellets. The two subjects with the lowest overall lever-press

rates also were the two subjects that earned fewer reinforcers (78-1 and 78-4) than the

remaining subjects during non-drug conditions. When the effect of cocaine

administration was to increase these subjects' lever-pressing rates, there was often an

increase in the number of food pellets presented. In general, decreases in the number

of food pellets delivered for all subjects were only observed after a large dose of

cocaine (17.0 or 30.0 mg/kg) was injected.


Effects of repeated cocaine administration on lever-press rates. Following repeated

administration of the chronic dose, which varied for each rat and is indicated on the

figures with an asterisk, a relatively large change in baseline (saline) responding is

indicated for only one subject, 78-5 (see Figure 12, filled symbols). Lever-press rates

decreased for this subject, which had the highest initial baseline rate and was given the

largest chronic dose of cocaine (17.0 mg/kg). For half of the subjects, repeated

exposure to a chronic dose of cocaine resulted in either no change from the initial

dose-response function (78-2) or a decrease in pressing rates (78-1 and 78-5). For two








91
of the other three subjects (78-3 and 78-4), responding increased slightly when the

lowest doses (1.0 and 3.0 mg/kg) were presented and decreased dose-dependently for

doses equal to or greater than the chronic dose. Lever-press rates for subject 78-6

remained similar to those recorded during the acute dose-response function after

administration of small or medium doses of cocaine and were increased after

administration of 10.0 mg/kg and 17.0 mg/kg.


The proportional changes in responding after saline and cocaine administration are

presented in Figures 13 and 14 (filled symbols). After repeated exposure to cocaine,

the rate-increasing effects of cocaine when compared to control rates of responding,

were attenuated for Subjects 78-1 and 78-4. However, when the slightly decreased

pressing rate following saline administration is taken into account for Subject 78-1, an

increase after injections of 5.6, 10.0, and 17.0 mg/kg are shown for lever-pressing

rates.

The total number of food pellets delivered did not change greatly following

repeated exposure to cocaine, with the exception of the decrease recorded for Subjects

78-4 and 78-5 after administration of the largest doses (see Table 2). The number of

obtained food pellets increased, relative to the acute phase, by modest amounts for

some subjects at some doses of cocaine. These increases did not always correspond

with an increase in lever-pressing rates (e.g., Subjects 78-2 and 78-3 after injections of

17.0 mg/kg cocaine).








92
Effects of acute cocaine administration on lever-press distributions and post-

reinforcement pause. Figure 15 (a-f) shows that under control (non-drug) conditions

(open symbols), subjects emitted more lever presses during the later portions of the

interval. In general, the number of lever presses made in the earlier segments of the

interval increased as the dose of cocaine increased. For most subjects, the

distributions at the largest doses, particularly 17.0 mg/kg, were even across the

interval. Increased lever pressing during the early portions of the interval is also noted

in the dose-dependent decreases in the post-reinforcement pauses presented in Figure

16 (open symbols).


Effects of repeated cocaine administration on lever-press distributions and post-

reinforcement pauses. After repeated exposure to chronic cocaine, redetermination of

the dose-response function (Figure 15 a-f, filled symbols) revealed that for the subjects

with increased response rates (78-3 at 1.0 10.0 mg/kg, 78-4 at 3.0 mg/kg, and 78-6 at

10.0 17.0 mg/kg), more lever presses were being emitted during the later portions of

the interval. Often this change later in the interval was accompanied by decreased

responding in the earlier segments of the interval. Thus, the positively accelerated

curve either remained intact or approached control (non-drug) distributions. When

decreased overall lever-press rates were observed, it was usually due to an overall

decrease in the number of lever presses throughout the interval at the largest doses and

a decrease in pressing toward the end of the interval at the low to medium doses (e.g.,

78-5). Again, the positively accelerated function was maintained. Post-reinforcement



























Figure 12. Average lever presses per minute as a function of dose of cocaine. Open
circles represent control (non-drug) rates (C) and rates after acute administration of
various doses of cocaine and saline (S). Filled circles represent rates following repeated
exposure to cocaine with the repeated dose for each subject indicated by an asterisk. Bars
around the mean indicate standard error.