Psychopharmacology

Psychopharmacology 56, 211- 215 (1978)

9 by Springer-Verlag 1978

Some Effects of Chlordiazepoxide and d-Amphetamine on Response Force During Punished Responding in Rats STEPHEN C. FOWLER and A. W. PRICE Department of Psychology, University of Mississippi, University, Mississippi 38677, U.S.A.

Abstract. Rats were reinforced with water on a continuous reinforcement schedule and were also punished with electric shock for every fifth response applied to a silent, isometric, force-sensing manipulandum. Oral doses of chlordiazepoxide (3.0, 9.0, 27.0 mg/kg) increased both conventional rate and force of punished responding. In contrast, d-amphetamine (0.8, 1.6, 3.2 mg/kg, by gavage) further decreased conventional rate and force of response, but this latter drug increased the rate of recorded responses that were lower than the 15-g force criterion for response consequences. The results for chlordiazepoxide are viewed in terms of its anxiolytic properties, while the d-amphetamine data appear to support a theory of amphetamine effects based on the concept of stereotyped behaviors.

Key words." Chlordiazepoxide - d-Amphetamine Response force havior - Rat

Punishment -

Stereotyped be-

Chlordiazepoxide increases the rate of operant responding that has been suppressed by responsecontingent electric shock (Geller et al., 1962; Cook and Davidson, 1973; McMillan and Leander, 1976). In contrast, amphetamine appears to decrease further the rate of such punished responding (Hendry and Van Toller, 1964; Cook and Davidson, 1973; McMillan and Leander, 1976). While there are no reports of the effects of these two drugs on response force during a punishment procedure, one study has demonstrated that response-contingent shock reduces response force as well as response rate (Filion et al., 1969). This study uses the response-force variable as an additional means of characterizing the effects of chlordiazepoxide and d-amphetamine on punished operant responding.

Since force of response reflects the motor features or topographical variants of behavior, this dependent variable-which is not observable with rate measures alone (Falk, 1969; Fowler, 1974b; Fowler et al., 1977)-may supply important information about drug effects. For example, Fowler et al. (1977) showed that chlordiazepoxide can substantially increase response force at a dose that produces a marked reduction in rate of appetitive responding. A second reason for studying response force concerns the measurement of incipient or subcriterion responses that fail to achieve the force criterion for programmed response consequences. In the conventional operant conditioning apparatus with a microswitch lever, these subcriterion responses are not recorded at all because they are of insufficient force to close the microswitch. The relevance of this distinction between criterion and subcriterion responses (Notterman and Mintz, 1965) for punishment procedures is vividly illustrated by Hill and Tedeschi's (1971) description of a rat in a conflict situation: "When not under the influence of a drug rats make many abortive approaches to a lever response..., frequently touching the lever but withdrawing the paw at the last moment. Under the influence of anxiolytics, however, responses appear to be emitted unhesitatingly and deliberately, despite the fact that the reactions to the shock appear to be as pronounced as ever" (p. 277). If a sufficiently sensitive force transducer is used in place of the lever, the frequency and amplitude of the subcriterion responses can be measured while keeping the force required for programmed consequences at the typical 1 5 - 2 5 g . Moreover, this technique provides a method for quantifying, in terms of peak force of response, the 'deliberate' responding produced by anxiolytics. Since changes in the frequency distribution of response force can result in changes in the observed criterion or conventional

0033-3158/78/0056/0211/$ 1.00

212

rate of response, a rate measure that includes subcriterion responses may not necessarily parallel the criterion rate. In view of the foregoing considerations, response force and rate for responses both above and below a 15-g force criterion served as dependent measures of effects of chlordiazepoxide and d-amphetamine during a punishment procedure that delivered water reinforcers on a continuous reinforcement basis and concurrently presented electric foot shock on a fixed-ratio 5 schedule (Hendry and Van Toller, 1964).

MATERIALS AND METHODS Subjects. The animals were six male Sprague-Dawley rats (Holtzman Co.) averaging about 225 g in body weight. These rats were maintained under conditions of water deprivation that permitted 20-min access to water approximately 30 rain after each daily session. Food was continuously available in the individual home cages.

Apparatus, The conditioning chamber measured 23-cm long, 20-cm wide, and 19-cm high and was fitted with a grid floor composed of 6.5-ram diameter stainless steel rods parallel to the front of the chamber. A rectangular opening in the front panel of the chamber was 3.0-cm wide and 2.5-cm high and permitted access to the manipulandum positioned outside the chamber. The lower edge of the manipulandum aperture was 5.5 cm above the grid floor. A Sanborn force transducer (Model FTA-100) served as the silent, practically isometric manipulandum (downward excursion of 0.4 mm at a force of 200 g). This transducer is a linear-variable-difference transformer capable of measuring both static and dynamic forces. The portion of the transducer available to the animal was a horizontal disk 18 mm in diameter, the surface of which was 6.0 cm above the level of the grid floor. The transducer was positioned so that the center of the disk was 3.0 cm from the outside of the chamber wall. Protruding into the chamber from the lower right front wall was a stainless steel drinking cup, which was serviced by a solenoid valve that permitted delivery of calibrated volumes of water. Scrambled electric shock was supplied to the floor rods by a BRS shocker-scrambler (Model 2903). The chamber, manipulandum, shocker-scrambler, and solenoid valve were enclosed in a soundattenuating compartment. Programming of contingencies and recording of data were accomplished with a laboratory computer (PDP8/e) and associated peripherals. The apparatus was programmed to record peak force and rates of responding. Details of these techniques may be found elsewhere (Fowler, 1974a; Fowler and Leberer, 1977). Under computer control, an analog-to-digital converter sampled the analog voltage from the transducer every 0.01 s. From these measurements the peak forces of individual responses above a 4-g threshold (cf. Notterman and Mintz, 1965) were obtained on-line. The peak force of a response is simply the maximum force amplitude attained by a response, which itself is defined by the force rising and then dropping below the 4-g threshold. Peak force was recorded with a precision of • 0,5 g, and session times, on which average rate of response measures were based, were taken with a precision o f • 1.0s.

Procedure. The rats were manually shaped by the method of successive approximations to reach through the aperture in the chamber wall to exert downward vertical force on the manipu]andum. This procedure was undertaken to develop relatively uniform response topography and to preclude operandum biting. Throughout the

Psychopharrnacology 56 (1978) experiment, the response consequence, whether shock or water (0.05 ml), was delivered upon response termination. This is an important procedural departure from the method used by Hendry and Van Toiler (1964), who delivered shock simultaneously with lever depression so that it was possible for unconditioned responses to shock to influence the operant response. In the present situation the recording of such unconditioned reactions to shock was unlikely because the rat's paw was being withdrawn when the shock occurred. Subsequent to shaping, each subject received 28 daily sessions of bar-pressing without punishment, in which responses of 15 g or more (the criterion force) produced water on continuous reinforcement. These were followed by 36 sessions in which the criterion responses (15 g or more) also produced foot shock (the aluminum front panel was also part of the shock circuit) on variableratio 5 (VR5). Sessions were terminated after 50 water reinforcers or 20 min, whichever occurred first. Conditioning sessions were carried out at the same time each day (11 a.m. + 1 h). Shock intensity was adjusted individually for each animal to achieve about 50 ~ reduction in rate of criterion responses. After VR5, eight sessions of fixed-ratio 5 (FR5) shock were conducted. Using this latter schedule, the effects ofchlordiazepoxide(Hoffmann-La Roche: 3.0, 9.0, 27.0 mg/kg) and d-amphetamine (Smith, Kline and French: 0.8, 1.6, 3.2 mg/kg) were assessed. Drugs were administered by gavage 45 min before the rat was placed in the chamber and were given every third day. In all cases drugs were diluted with sufficient 0.9 ~ saline to yield a dose volume of 1.0 ml. The day before each drug session, 1.0 ml of physiological saline was intubated. On nonintubation days, which followed each drug day and preceded each saline day, animals were merely exposed to the conditioning procedures. The drugs and dosages (expressed as the salt) were given in randomized order. One determination of drug effect was made for all doses, except for 1.6 mg/ kg d-amphetamine, which was given twice. Drug effects were characterized by three dependent variables: mean peak force of all responses (responses having forces > 4 g), rate of criterion responses (responses having forces > 15 g), and rate of subcriterion responses (responses having forces > 4 g but < 15 g). Drug effects were expressed as a proportion of the saline control value obtained on the day immediately preceding the drug day. The 95 ~ confidence intervals shown in Figure 1 are based on the t distribution and on the assumption that the proportions are adequately modeled by a log-normal distribution.

RESULTS Prior to the introduction of electric foot shock the group mean for rate of criterion responses was 18.3 responses/min, and for average peak force the group mean was 23.5 g. After 42 sessions of shock administration the criterion response rates for saline sessions ranged from a low of 2.8 responses/rain to a high of 10.2 responses/min. Similarly, group means for average peak force for shock-saline days ranged from 11.7 g to a maximum of 14.7 g. These data illustrate the degree and stability of suppression achieved for the rate of criterion responses and average peak force variables recorded for the period during which the drug effects were assessed.

Chlordiazepoxide. Panel A of Figure 1 presents the dose-response data for the three dependent variables

S. C, Fowler and A. W. Price: Chlordiazepoxide, d-Amphetamine, and Response Force

measured during the punishment procedure. As expected, chlordiazepoxide increased criterion rate of punished responding for the three doses used. Rate of subcriterion responses increased at 3.0 and 9.0 mg/ [-I FORCE OF R'S24G [ ] RATE OF R'S 215G []

RATE OF R'S < 15G

~ 5s z o

~

o/ ili ELO 20~ Z

~

IO

2 2o_0

.

~

.

.

.

.

.

i_

270 3.0 9,0 CHLORDIAZEPOXIDE (MG/KG)

B

~S Oo

8o

1.0 . . . . . . . . . . . . .

Bib_

13-0 0

kg, but fell well below control values at the highest dose. The decline in rate of subcriterion responses at the 27.0 mg/kg dose is a reflection of the substantial increase in average peak force seen at the same dose (see Fig. 1). The highest dose of chlordiazepoxide produced such forceful responding that very few responses fell below the ~5-g criterion. This kind of responding, subsequent to chlordiazepoxide administration, is depicted in Figure 2 by the top two stripchart recordings. Response waveforms emitted under the influence of chlordiazepoxide are generally above 15 g and are quite 'ballistic' (i.e., the force rises and falls rapidly) relative to control performance.

d-Amphetamine. As shown in Panel B of Figure 1,

.

z~ o o 1o:7

213

Q8 1,6 3.2_ d-AMPHETAMINE (MG/KG)

Fig, 1. Data for average peak force (open bars), rate of criterion responses (horizontal cross-hatch), and rate of subcriterion responses (oblique cross-hatch) as a function of dose of chlordiazepoxide (Panel A) and d-amphetamine (Panel B). The 95 ~ confidence intervals are shown at top of each histogram bar. All data are group mean proportions based on six rats, except for data at 3.2 mg/kg dose of d-amphetamine, which are based on five rats

d-amphetamine decreased both the average peak force and the rate of criterion responding at the highest dose used. An unexpected finding is the increase in rate of subcriterion responses produced by each of the doses of d-amphetamine, illustrated for one subject in the lower three tracings of Figure 2. Compared to the responses recorded for saline, those emitted after d-amphetamine administration were primarily subcriterion responses. The increase in the rate of emission of subcriterion responses suggests that the subjects were very active, but that the activity had been pushed below the force criterion by virtue of the drug-related reduction in response force, which is especially evident at the 3.2 mg/kg dose. DISCUSSION The fact that chlordiazepoxide increased rate of responding that was suppressed by response-contingent

RAT 5 SALINE IS

s RAT 5 CHLORDIAZEPOXIDE 270

RAT I SALINE

RAT AMPHETAMINE5.2

RAT I AMPHETAMINE 5.2

Strip-chart recordings for representative subjects illustrating the effects of chlordiazepoxide (top two tracings) and d-amphetamine (lower three tracings) on individual response waveforms. The breaks in base-lines indicate pauses of several seconds when recorder was off Fig. 2.

214

shock generally agrees with the literature (Geller et al., 1962; Cook and Davidson, 1973; McMillan and Leander, 1976). Further, the results for this drug extend previous observations by showing that peak force of punished responding is also elevated by chlordiazepoxide. Two experiments using appetitive procedures have examined the effects of chlordiazepoxide on response force (Fowler, 1974b; Fowler et al., 1977). These studies demonstrated that chlordiazepoxide, over a range of doses similar to those used here, resulted in relatively higher response force (compared to nondrugged controls) during extinction and FR, whereas response force during CRF was unaffected by the drug. Previous work with rats has shown that FR schedules and extinction engender comparatively high force emission despite the lack of any experimental requirement for such effortful behavior (Mintz, 1962; Notterman and Mintz, 1965). If one assumes that the high forces occasioned by extinction and FR reflect an aversive aspect of nonreinforcement, then the drug-related increases observed here and in the aforementioned reports may be seen as instances of chlordiazepoxide's tendency to enhance behavioral output in situations where responses have aversive consequences. Methodologically, the present techniques of response measurement provide a means for objectively describing some of the quantitative and qualitative changes in responding produced by chlordiazepoxide. For example, Hill and Tedeschi's (1971) qualitative behavioral description of 'deliberate' bar-pressing induced by anxiolytics is clearly supported by the response force data in Figure 1 (Panel A) and by the response characteristics shown in Figure 2 for chlordiazepoxide. d-Amphetamine further reduced average peak force of response that had already been suppressed by response contingent shock. In accord with previous reports (Hill and Tedeschi, 1971 ; Cook and Davidson, 1973), a reduction in criterion or conventional rate of responding was also observed. However, d-amphetamine increased rate of subcriterion responding. This later result was foreshadowed by the work of Morrison and Stephenson (1973), who reported that in appropriate doses d-amphetamine increased the tendency for rats to touch but not to depress the lever after punishment was instituted. The current data for subcriterion responding provide additional evidence for this lever-touching behavior. Lyon and Robbins (1975) argue that a decrease in observed rate of response in a conventional operant test situation (with microswitch lever) does not necessarily indicate a slowdown or reduction in behavior. In contrast, they theorize that rats are made progressively more active by increasing doses of d-amphetamine (up to

Psychopharmacology 56 (1978)

10.0 mg/kg i.p.); however, the number of classes of emitted behavior narrows in direct proportion to the dose, so that 10 mg/kg produces only a few stereotyped behaviors occurring at a high rate. In the usual operant situation, such typical stereotyped behaviors as sniffing and lateral head movements are simply incompatible with topographies necessary for the operant response. Within this theoretical framework, the increase in rate of low-force responses is a reflection of the transition from effective manipulandumoriented behavior to more stereotyped 'off-lever' behaviors (cf. Fig. 4 in Lyon and Robbins, 1975). Regardless of the theoretical interpretations offered for the specific drug effects, the methodological implications of the present results appear straightforward. As Notterman and Mintz (1965) and Mintz et al. (1976) have cogently argued, the rate of responding observed in a given operant situation is determined at least in part by the force requirement selected for defining a response. In a different context, Thompson (1976) suggested that research in the physiology of learning, of which operant behavioral pharmacology may be considered a subset, would do well to use a behavioral response for which the exact amplitudetime course is measurable. Recording the force of operant responses having specific topographies appears to be one way of satisfying Thompson's criterion. The present results confirm the utility of such amplitude measures in complementing behavioral pharmacology techniques based on the rate variable. Acknowledgements. This research was supported by NIMH Grant R03MH27177. The authors thank Smith, Kline, and French, which supplied the d-amphetamine, and Hoffmann-La Roche, which provided the chlordiazepoxide.

REFERENCES Cook, L., Davidson, A. B.: Effects of behaviorally active drugs in a conflict-punishment procedure in rats. In: The benzodiazepines, S. Garatinni, E. Mussini, and L. O. Randall, eds., pp. 327-345. New York: Raven 1973 Falk, J. L. : Drug effects on discriminative motor control. Physiol. Behav. 4, 421-427 (1969) Filion, R. D. L., Fowler, S. C., Notterman, J. M. : Some effects of simultaneous force-proportional positive and negative reinforcement. J. Exp. Psychol. 82, 267-271 (1969) Fowler, S. C. : A minicomputer system for recording the dynamic properties of individual operant responses. Behav. Res. Meth. Instrum. 6, 288- 292 (1974a) Fowler, S. C. : Some effects ofchlordiazepoxide and chlorpromazine on response force in extinction. Pharmacol. Biochem. Behav. 2, 155-160 (1974b) Fowler, S. C., Filewich, R. J., Leberer, M. R. : Drug effects upon force and duration of response during fixed-ratio performance in rats. Pharmacol. Biochem. Behav. 6, 421-426 (1977) Fowler, S. C., Leberer, M. R. : Hardware techniques for analog processing using the State Systems PDP8 I/O interface. Behav. Res. Meth. Instrum. 9, 210-214 (1977)

S. C. Fowler and A. W. Price: Chlordiazepoxide, d-Amphetamine, and Response Force Geller, I., Kulak, J. R., Seifter, J. : The effects of chlordiazepoxide and chlorpromazine on a punishment discrimination. Psychopharmacologia (Bed.) 3, 374-385 (1962) Hendry, D. P., Van-Toiler, C. : Fixed-ratio punishment with continuous reinforcement. J. Exp. Anal. Behav. 7, 293-300 (1964) Hill, R. T., Tedeschi, D. H. : Animal testing and screening procedures in evaluating psychotropic drugs. In: An introduction to psychopharmacology, R. H. Rech and K. E. Moore, eds., pp. 237-288. New York: Raven 1971 Lyon, M., Robbins, T.: The action of central nervous system stimulant drugs: a general theory concerning amphetamine effects. In: Current developments in psychopharmacology, vol. 2, W. Essman and L. Valzelli, eds., pp. 79-163. New York: Spectrum 1975 McMillan, D. E., Leander, J. D.: Effects of drugs on schedulecontrolled behavior. In: Behavioral Pharmacology, S. D. Glick

215

and J. Goldfarb, eds., pp. 85-139. St. Louis: C. V. Mosby 1976 Mintz, D. E. : Force of response during ratio reinforcement. Science 138, 5 1 6 - 517 (1962) Mintz, D. E., Samuels, R. M., Barber, N. G.: Force and rate relations in responding during variable-interval reinforcement. J. Exp. Anal. Behav. 26, 387-393 (1976) Morrison, C. F., Stephenson, J. A.: Effects of stimulants on observed behavior of rats on six operant schedules. Neuropharmacology 12, 297 -- 310 (1973) Notterman, J. M., Mintz, D. E. : Dynamics of response. New York: Wiley 1965 Thompson, R. F.: The search for the engram. Am. Psychol. 31, 209 - 227 (1976)

Received June 21, 1977