An Addiction Science Network Resource


Reprinted from R.U. Esposito, L.J. Porrino, and T.F. Seeger (1987), Brain stimulation reward: Measurement and mapping by psychophysical techniques and quantitative 2-[14C]deoxyglucose autoradiography. In M.A. Bozarth (Ed.), Methods of assessing the reinforcing properties the reinforcing properties of abused drugs (pp. 421-445). New York: Springer-Verlag.
 
Brain Reward System
Back to book TOC
(Search book contents) (Search entire ASNet)


Chapter 20

Brain Stimulation Reward:
Measurement and Mapping by Psychophysical Techniques
and Quantitative 2-[14C]Deoxyglucose Autoradiography
 

Ralph U. Esposito,1 Linda J. Porrino,2 and Thomas F. Seeger3

1Laboratory of Psychology and Psychopathology
2Laboratory of Cerebral Metabolism
3Biological Psychiatry Branch
National Institute of Mental Health
U.S. Public Health Service
U.S. Public Health Service
Department of Health and Human Services
Bethesda, Maryland 20205


Abstract
Brain stimulation reward is a useful model for the study of the neural mechanisms of reinforcement. A number of methods used to assess the reinforcing qualities of brain stimulation and the effects of drugs of abuse on this behavior are reviewed with emphasis on the relative advantages and disadvantages of each. In addition, the use of the quantitative 2-[14C]-deoxyglucose method as a novel means to map the neural substrates of drug-induced effects on brain stimulation reward is described.

 

Introduction

Since its discovery the phenomenon of rewarding brain stimulation (Olds & Milner, 1954) has received intensive study as a means to understand brain mechanisms involved in motivated behavior. Recent advances in neuroanatomy, electrophysiology, and neuropharmacology have indicated an important relationship between certain ascending catecholaminergic pathways and this behavior (Crow, 1976; Fibiger, 1978; German & Bowden, 1974). Regarding the reinforcing properties of drugs of abuse it is noteworthy that the two most widely abused and euphorigenic classes eworthy that the two most widely abused and euphorigenic classes of drugs, the opiates and psychostimulants, exert specific facilitative effects on self-stimulation to these same projection systems (Esposito & Kornetsky, 1978; Wise & Bozarth, 1981). Further studies, therefore, concerning the effects of these agents on brain stimulation reward, in conjunction with other techniques, should lead to important insights concerning the precise site and mechanisms of the reinforcing action of these agents. Conversely, and perhaps of more general significance, these same types of studies will also shed light on the neural substrates of goal-oriented or motivated behavior.

This chapter will briefly review methods used to assess the behavioral effects of drugs on brain stimulation reward and suggest adequate criteria for the design of future studies. Secondly, we will present recent developments concerning the use of quantitative 2-deoxyglucose autoradiography as a novel and unique means to identify the substrate of these drug-induced effects.

Methods for the Determination of Drug Effects
on Brain Stimulation Reward

Response Rate and Brain Stimulation Reward

Rate of response is by far the most commonly used measure of the reinforcing value of rewarding brain stimulation. Typically this type of experiment involves a rat lever pressing (in a stant involves a rat lever pressing (in a standard operant chamber) for a fixed intensity of brain stimulation on a continuous reinforcement schedule. The inadequacies of response rate as a measure of the reinforcing value of brain stimulation have been thoroughly discussed in detail elsewhere (Liebman, 1983; Valenstein, 1964), and only the major problems will be reviewed here.

Valenstein and associates have demonstrated on both empirical and logical grounds the inadequacy of response-rate measures. Thus, in rats given a choice between two levers, each activating a different electrode, the rate of responding was not found to correlate significantly with lever preference in a choice situation (Hodos & Valenstein, 1962) nor with measures of resistance to competition from other reinforcers such as food or the avoidance of foot shock (Valenstein & Beer, 1962). With respect to the choice situation in particular, it has been reliably noted that rats will choose "low-rate" septal stimulation over "higher-rate" lateral hypothalamic sites. This demonstration itself indicates that assertions regarding the rewarding value of stimulation to specific brain sites cannot be based on the simplistic equation of increasing response rate with increasing rewarding value. It is important to keep in mind that we know very little about animals’ preferred rates of responding for many brain sites and, furthermore, even less about the interactive effects oven less about the interactive effects of various drugs on these response patterns.

In addition to the above, more general interpretative problems are encountered when one attempts to measure the effects of drugs on the reinforcing value of brain stimulation. Drug induced changes on rate of responding for brain stimulation are often automatically interpreted as a specific change in the reinforcing value of the stimulation when, in fact, these changes may be due to any number of nonspecific drug effects on attention, arousal level, or various perceptual and/or sensory-motor functions related to the task performance per se. Drug-induced changes may also be related to other important nonspecific factors such as differences in predrug baseline rates of responding. Even if animals are equated in terms of number of responses/minute, important individual differences may still exist. For instance, some animals may actually be straining to meet a particular response rate criterion while others do so with ease. Differences in response rate after drug treatment are difficult to interpret when this type of pretreatment variable is taken into consideration. Thus, taking the drug morphine as an example, one can find studies, using rate of lever pressing as the dependent measure, showing that morphine can either increase, decrease, or have no effect on the rate of self-stimulation, depending upon dose, time of measurement, baseline rate, and pattermeasurement, baseline rate, and pattern of responding, plus a host of other variables such as general sedative or stimulatory effects of the drug (for a review, see Esposito & Kornetsky, 1978; cf. Reid, this volume). From these studies one can draw any preferred conclusion concerning the effects of morphine on rewarding brain stimulation.

Finally, an important methodological problem with rate in drug studies involving brain stimulation reward has been pointed out recently by Liebman (1983). It concerns the fact that a particular dose of a drug may have rather selective effects on high or low stimulation intensities (see Figure 1). It is clear that these important differences are obscured entirely when one examines the drug’s effects at a single intensity, as is the case in many, if not most, studies employing response rate as the dependent variable. Even if a clear shift in a rate-intensity function is noted, its interpretation is still subject to all of the aforementioned ambiguities relating to drug-induced changes on brain stimulation reward. For these reasons some investigators have adopted psychophysical methods in order to obtain more valid measures of the reinforcing effects of brain stimulation.
 

 
Response rate as a function of stimulus intensity
Figure 1: Response rate plotted as a function of stimulus intensity. A shows a clear drug-induced shift in a rate-intensity function. B shows a selective attenuation of response rate at higher stimulus intensities, while C shows a selective drug-induced attenuation of rate of responding at lower stimulus intensities. Reprinted with permission from Liebman, 1983. Copyright 1983 by ANKHO International, Inc.
 

Psychophysical Techniques and Brain Stimulation Reward

Contemporary psychophysical methods are essentially based on the systematic approaches to the quantitative analysis of sensory processing set by certain nineteenth century neurophysiologists, notably von Helmholtz and Weber. The actual founding of psychophysics, however, can be traced to Gustav Theodor Fechner, the physician, mathematician, physicist, philosopher, and poet, whose first book on the subject, modestly entitled Zend-Avesta, ober wher die Dinge des Himmels und des Jenseits (1851; or On the Nature of the Heavens and the Nonmaterial World), laid down the essential ideas detailed in the subsequent Elemente der Psychophysik, (1860) which marked the official beginning of the field of psychophysics. The general aim of the early psychophysicists was to find the quantitative relationships between physical events and their "mental" representation. Although rooted in sensory processes, Fechner and his scientific followers wanted the general principles of psychophysics to be applied to other areas so as to eventually achieve a complete, systematic and quantitative knowledge of the relationship of the physical world to mental processes. Few would agree that the field of psychophysics has achieved this goal, but nonetheless psychophysical methodology finally separated psychology as a field distinct from philosophy and has had a profound positive influence on experimental psychology with wide application in many areas, particularly in the area of sensory psychology.

To avoid the problems inherent in rate-dependent studies of brain stimulation reward, some investigators have adopted the use of classical psychophysical procedures. These investigators attempt to obtain a more valid measure of the reinforcing value of brain stimulation by determining the absolute threshold,4 generally expressed in terms of current intensity or frequency, necessary to support self-stimulation behavior to any particular brain site. This absolute reinforcing threshold is then used as a baseline against which the effect of various experimental manipulations such as lesions or drug administration can be expressed quantitatively in psychophysical units. This insures the specifi experimental manipulations such as lesions or drug administration can be expressed quantitatively in psychophysical units. This insures the specificity of effects and also permits quantification of the magnitude of any observed changes. Another distinct advantage is that the low stimulus intensities typically employed in these threshold determinations should more accurately reflect drug-induced changes in the excitability of the underlying neural activity with minimal tissue damage. Thus, in general, psychophysical studies of drug effects on brain stimulation reward determine the reinforcing threshold by obtaining a discrete measure of responding over a range of stimulus values, which by definition will maintain responding 50% of the time. An example of such a determination of reinforcing threshold before and after drug administration can be seen in Figure 2. As mentioned, several research groups have employed a variety of psychophysical techniques to obtain accurate estimates of the reinforcing value of brain stimulation and the specific effects of psychoactive drugs thereon. The remainder of the section will critically review some of the most representative of these endeavors, noting the relative advantages and disadvantages of each. These methods were chosen on the basis of their utility and representativeness, while no attempt to review the vast literature concerning drug effects on brain stimulation is intended.
 

4The term absolute threshold indicp>4The term absolute threshold indicates the lowest point on a particular stimulus continuum (i.e., electrical intensity or frequency) at which behavior will be maintained 50% of the time. This measurement is not to be confused with the differential threshold which refers to the lowest magnitude of physical change (increment or decrement relative to a standard) on a stimulus continuum required for a subject to detect a difference relative to a standard 50% of the time. In this chapter the term reinforcement threshold will refer to the absolute threshold. It is important to note that the historical use of the term absolute is misleading because due to the subject’s variability of response the threshold is not an absolute value but is a statistical approximation. 

The Two-Lever Reset Method

The method was first introduced by Stein and Ray (1960) who measured reinforcing thresholds for brain stimulation to the posterior hypothalamus and the ventral tegmentum. Typically in this procedure a rat is placed in a chamber with two levers available for responding. A response on one of the levers results in a positively reinforcing brain stimulation with a fixed decrease of intensity in a step-wise manner. Responding at the second lever resets the level of stimulation available at the first lever to its original level. The reinforcement threshold is defined as the averageorcement threshold is defined as the average intensity at which reset responses are made (see Figure 3).

Stein (1962) has employed the two-lever procedure to demonstrate the threshold lowering effect of amphetamine, and the potentiation of this effect by imipramine. A dissociation of these effects from simple rate measures was reported by Stein, who noted that amphetamine-induced increases in rate of responding on the "rewarding" lever were most evident after the threshold lowering effect had returned to baseline. Nazzaro, Seeger, and Gardner (1981) have used this procedure in slightly modified form to demonstrate the threshold-lowering effect of morphine on ventral tegmental self-stimulation, as well as the additive effects of the co-administration of amphetamine and morphine (Seeger & Carlson, 1981). The two-lever procedure represents a significant advancement over the simple rate-dependent free operant method as it generates reliable and stable thresholds, but it is not without disadvantages, both theoretical and practical.
 

 
Effects of morphine on BSR
Figure 2: The effects of 4 mg/kg of morphine, administered subcutaneously to a rat, on the reinforcing threshold for brain stimulation reward. The ordinate shows the percent brain stimulation reward. The ordinate shows the percent of trials in which the subject responded at each current intensity plotted on the abscissa. Current intensity was systematically varied in alternating ascending and descending series of 10 mA "step-size" increments and decrements, respectively. The data were collected immediately pre- and 10 minutes postinjection. Reinforcement thresholds are specified as the median reinforcing intensity at which the subject responded. Data adapted from Esposito and Kornetsky, 1977. 
 

The major theoretical problem with the method is related to the regular descending serial order presentations of the stimuli. The use of descending series may lead to response habituation and thus yield inaccurate or spurious threshold measurements (Engen, 1971). The serial order stimulus presentations may also enable the animal to anticipate the regular decrease in reinforcement value and learn to make reset responses at intensities above the actual reinforcing level, thereby maintaining the stimulation at a preferred suprathreshold level (Valenstein, 1964). Modification of the experimental environment and procedure can partly obviate these limitations. For example, more valid threshold measures may be obtained by using a large number of current steps, such that the magnitude of the current decrement is minimal, or by physically seprrent decrement is minimal, or by physically separating the two levers enough to insure that the animal spends a considerable amount of time away from the stimulus lever while resetting the current level. Empirical evaluation of these limitations, however, must await parametric investigations which systematically vary both the initial (i.e., maximal) stimulus intensity as well as the magnitude of the decremental step size. So far, work in this area is limited to one report which indicates that variation of the maximal current level does not necessarily result in significant alterations in the reinforcing threshold (Schaeffer & Holtzman, 1979; see also Fouriezos & Nawiesniak, this volume).
 

 
Illustration of the two-lever reset method
Figure 3: Schematic representation of the basic two-lever reset method determining intracranial self-stimulation thresholds in the rat. Level of stimulation current (amplitude) is plotted on the ordinate while experiment time is recorded along the abscissa. Responding on the stimulus in a fixed decremental "step-wise" manner. Response at the reset lever resets the level of stimulation available on the first lever to its original intensity. A measure of the average intensity at which reset responses are made is taken as the estimatech reset responses are made is taken as the estimate of the reinforcement threshold. Reprinted with permission from Stein, 1962. Copyright 1962 by Plenum Press.
 

In studies involving primates, the requirement of restraint in a chair precludes the physical separation of the levers. In this situation the animals quickly learn to alternate responses in order to keep the stimulus intensity at highly rewarding levels. Gardner (1971) has reported a modification of the two-lever procedure that makes it suitable for studies with primate subjects. This modified technique essentially involves the use of one response manipulandum only and requires the subject to withhold responding for a specified time, as in a modified DRL free-operant schedule, in order to obtain a reset of the stimulus intensity to its original high level. Seeger and Gardner (1979) have used this modified procedure to demonstrate neuroleptic-induced supersensitivity by threshold-lowering effects in the nucleus accumbens. Despite these modifications it must be noted that theoretically the current level at which animals reset the intensity is not the lowest level which will support behavior (i.e., the reinforcement threshold), but rather some preferred level of current intensity. Threshold measures using other methods typically yield lower intensities than the two-lever method (Valenstein, 1964).

Another potential prohod (Valenstein, 1964).

Another potential problem with the two-lever reset method can arise with the use of drugs such as amphetamine, for example, that cause nonspecific effects such as disinhibition of responding, stereotypy, response perseveration, and perseverative switching (Evenden & Robbins, 1983). Each of these nonspecific effects could conceivably distort the validity of threshold determination. Behaviorally, however, the more severe nonspecific high dose drug effects are sufficiently detrimental to the operant procedure such that performance of the task tends to become completely erratic, precluding any valid measures. The more subtle confounding factors may be addressed if empirical means are used to assess and separate nonspecific from specific response alterations. For example, Seeger, Carlson, and Nazzaro (1981) attempted to parcellate specific from nonspecific responding by statistical criteria. They recorded the current level at which an animal chose to reset current level and thus generated a reset frequency distribution, taking the mean of this distribution as the reinforcement threshold, while the standard deviation was adopted as a measure of response integrity (i.e., an increasing standard deviation would indicate deterioration of responding). By means of such analysis they were able to detect the specific, low dose (10 mg/kg) threshold-lowering effect of pentobarbital without concurrent response deterioration, as was rent response deterioration, as was noted at higher doses (e.g., 20 mg/kg) which produced marked ataxia.

In summary, the two-lever reset technique possesses features which clearly improve on the limitations of purely rate-dependent measures. It produces stable, reliable, and practical measures of relative reinforcement, which can be maximized for resistance to nonspecific confounding effects on performance. Its greatest utility may be in assessing drug effects, both with regard to specificity of action on reward mechanisms and also in its ability to monitor simultaneously the temporal pattern of drug effects, which can provide information concerning the time of onset and duration of action particularly important in those studies involving agonist/antagonist interactions (cf. Fouriezos & Nawiesniak, this volume).

The Post-Reinforcement Pause Method

Huston and Mills (1979) have devised a procedure to determine reinforcement thresholds which capitalizes on certain aspects of schedule control. In their procedure rats lever pressed to self-stimulate on a fixed ratio (FR) schedule, while superimposed on this schedule was another continuous reinforcement (CRF) schedule of reinforcement. The intensity of the reinforcing stimulus on the FR schedule was always kept at a highly rewarding value, while the value (i.e., stimulus intensity) of the CRF component was varied, in ascending and descending steps ed, in ascending and descending steps of 5 mA according to the method of limits, at different FR components. When the CRF value was low, and presumably nonreinforcing, responding was controlled by the FR component and accordingly the cumulative record showed characteristic high response rates and post-reinforcement pauses. At higher levels of CRF reinforcement intensity, this component came to control responding, and the record showed typical patterns of continuous reinforcement. By varying the CRF intensity between these two extremes, Huston and Mills found an intensity at which post-reinforcement pauses (an interresponse interval 3 or more standard deviations above the mean interresponse time at a given current level) appeared in the record and defined this as the reinforcement threshold (see Figure 4). Cassens and Mills (1973) have used this procedure to demonstrate opposite effects of lithium and amphetamine on intracranial reinforcement thresholds in rats. Amphetamine was found to lower the threshold while lithium elevated the threshold. In a follow-up study Cassens, Actor, Kling, and Schildkraut (1981) demonstrated a marked increase in reinforcing thresholds 24 to 48 hours after cessation of chronic amphetamine administration, which peaked between 24 to 72 hours. The post-reinforcement pause method was also used by Kelly and Reid (1977) to demonstrate the threshold lowering effects of morphine. In their ering effects of morphine. In their study rats with lateral hypothalamic implants were given 10 mg/kg of morphine daily for 20 consecutive days and tested on the threshold procedure. Morphine-induced reductions in the thresholds evidenced no tolerance over the entire testing period, even though the animals showed clear signs of physical dependence.

The post-reinforcement pause method is ingenious inasmuch as it relies on classic schedule control patterns to determine a threshold whereas virtually all other measures use the reinforcing brain stimulation itself both to control and to maintain responding when determining the reinforcing thresholds. This reliance on pattern of responding, however, can also present problems, particularly when trying to assess the effects of drugs which may disrupt timing or selectively enhance responding at low rates. Another problem with the current use of CRF stimuli superimposed on the FR schedule is related to the fact that sub-reward-threshold stimuli delivered to brain sites, which support self-stimulation, possess drive or incentive-induction properties (Coons & Cruce, 1968; Huston, 1971). Thus, considering that drive-inducing stimuli can certainly affect responding, it becomes theoretically impossible to determine if, for example, the appearance of CRF responding is due to summating level of drive-induction or control by reward factors. As in the case with the two-lever method, these theoretth the two-lever method, these theoretical problems do not detract from the method’s utility in assessing drug effects, as it does seem to generate stable and reliable effects. Finally, in the early studies employing this method the definition of what actually constituted a post-reinforcement pause was somewhat arbitrary and subjective. For example, Cassens and Mills (1973) noted that a " . . .lower limit of 7 seconds was arbitrarily chosen because this interval was just visually discernible on a cumulative record with a time base of 5 mm/min" (Cassens & Mills, 1973, p. 284). The procedure has since been automated (Cassens, Shaw, Dudding, & Mills, 1975) so that post-reinforcement pauses (PRP) are based upon their relative frequency. A PRP is defined as an interresponse interval 3 or more standard deviations outside the mean interresponse interval at any particular current level of CRF stimulation. On this basis a PRP/FR ratio can be calculated at each current level and a least square regression analysis applied to the resulting proportions as a function of current level in order to determine the line of best fit for the data. The threshold can thus be specified in terms of current intensity at which the proportion of PRP/FR is 0.5. Further refinements in this method have been made by Phelps and Lewis (1982; Lewis & Phelps, this volume). They interfaced a minicomputer (PDP8-Digital) with a constant current stimulator so that a constant current stimulator so that the stimulator functions as a voltage follower-regulator with continuous feedback. Instantaneous samples of the resistance (impedance) to the electric current are made and result in voltage adjustments to maintain constant current intensity. Another modification involved a reduction of the usual FR requirement (e.g., FR-40 reduced to FR-15), plus reducing the number of total FR components. These latter adjustments reduce the mean time necessary for threshold determination without affecting the reliability of the measurement. Data on total session time, reinforcement threshold, and average impedance values are collected by the computer and analyzed by a statistical program which summarizes and calculates means and standard errors of the mean and correlates each measure of a subject’s performance at each current range tested.
 

 
Cumulative record from a concurrent CRF:FR schedule
Figure 4: Cumulative record for a rat on a concurrent CRF:FR schedule. This rat received FR reinforcements of 120 mA on a FR-40 schedule. Changes in the CRF intensity are indicated along the base of the curve and are expressed in microamperes (mA) base to peak. The post-reinforcement pauses (Pont>A) base to peak. The post-reinforcement pauses (PRPs) are long and consistent when the CRF stimulation is low, such as 30 or 35 mA, but disappear when the CRF intensity is raised to 60 or 65 mA. When the PRPs disappear, the behavior is identical to that of a subject on a simple CRF schedule of equal reinforcements. Reprinted with permission from Cassens and Mills, 1973. Copyright 1973 by Springer-Verlag.
 

Using this method of analysis Lewis (1981; Lewis & Phelps, this volume) has been able to demonstrate normal age-related changes in motivation and their rejuvenation with amphetamine as assessed by the self-stimulation model. He reported a decline in responding vigorously and reliably at 6 months of age. The decline in behavior was characterized by a decrease in response rate, increased impedance to the stimulating current, and failure to complete the reinforcement threshold procedure described above. Priming, increased current, food deprivation, and reshaping were all ineffective in restoring the behavior. A single dose of d-amphetamine (0.25 to 0.30 mg/kg), however, reinstated the self-stimulation, although thresholds and impedance value were higher than the 6 month values, and priming was required in some cases. It is of interest that rate of responding remained unchanged, arguing against a nonspecific arousal. Microed, arguing against a nonspecific arousal. Microanalysis of the electrode tips in both young and old animals showed no significant differences in size of necrosis around the electrode tips, indicating that the behavioral decline was not likely to be a function of tissue damage or electrode defects. The most parsimonious explanation of these findings is that the amphetamine activated central motivational systems in the aged rats, possibly by a restoration of functioning of age-related functionally declining catecholamine systems. The effect on motivational systems is supported by the finding that all the amphetamine-treated aged rats showed an improved performance in at least two subsequent consecutive sessions without further drug treatment. It is possible that the amphetamine treatment may have aided a motivationally relevant component of reacquisition of the rather demanding threshold procedure. In any case, this study provides age-related information about motivational systems and suggests that the brain self-stimulation thresholds may provide useful information relative to the functional integrity of catecholamine systems which are important for this behavior and show decline (particularly dopaminergic cells) with age (Ordy, Brizee, Kaack, & Hansche, 1978). The full automation of this somewhat complex technique should now permit its wider use.

The Reward-Summation Method

The reward-summation method, designed by Gallistel and associates, has been used along with various modifications and adjunctive manipulations as part of a systematic attempt to measure the fundamental parameters prerequisite to an understanding of the basic psychophysical relationships underlying motivation and reward through use of the self-stimulation model (Gallistel, 1983). In this respect this method has proven to be particularly useful in disassociating performance variables from specific effects on reward.

In the basic paradigm a rat is placed in a start box and receives a train of priming pulses and then runs down an alley to a goal box where a single lever-press response results in the contingent delivery of a train of rewarding brain stimulation pulses. After this discrete trial the lever is retracted, and the animal is placed back in the start box until the initiation of another trial. Over trials the number of rewarding pulses available on the response lever is varied systematically in ascending and descending series (i.e., classical method of limits) through a wide range of values. Running speed in the alley typically was noted to increase precipitously from low to asymptotic speeds at a particular value of reinforcement available on the response lever. This value represents the point on the reward-summation function which can be taken as a measure of the absolute reinforcing threshold. Specific effects on reinforreshold. Specific effects on reinforcement threshold are seen as a shift in this "locus of rise" point on the reward-summation function (i.e., running speed plotted against the number of reward pulses in the stimulus train), while performance effects are independently reflected in the asymptotic level of running speed. Performance effects have, in fact, been shown to be quite sensitive to nonspecific factors such as running uphill, administration of muscle relaxants, et cetera, while the reinforcement threshold remained unchanged as assessed by locus of rise (Edmonds & Gallistel, 1974). Utilizing this method Franklin (1978) was able to demonstrate the specific reward-attenuating effect of the relatively selective dopamine antagonist, pimozide, as evidenced by a shift in the locus of rise without a concomitant change in the asymptotic level of running speed (see Figure 5).

The work of Gallistel and associates represents an empirical attempt to develop a psychophysics of intracranial self-stimulation. For instance, Edmonds and Gallistel (1974) and Gallistel, Rolls, and Greene (1969) have provided some empirical data supporting the proposed distinction between the neural circuitry mediating the drive-inducing effects of brain stimulation and that circuitry which mediates the rewarding or reinforcing component of such stimulation on the basis of differential sensitivity to variations in interpulse intervals within the stimulus trailse intervals within the stimulus train. To our knowledge, however, there have been few published investigations of the reinforcing properties of drugs of abuse that have used this method, most likely due to the time consuming and performance-demanding nature of the procedure which would lead to practical difficulties in studies of drugs with short duration of action or manifest tolerance or sensitization with repeated administration. However, it would seem that with appropriate modifications useful information could be gained from such investigations. For example, Belluzzi and Stein (1977) could have hypothesized that certain drugs (i.e., psychostimulants) may exert their major facilitative action on the motivational or drive-induction (i.e., behavior activating) properties of intracranial self-stimulation, while other drugs (i.e., opiates) may preferentially affect the drive-reduction or reinforcing aspect of goal-oriented behavior such as self-stimulation. To the extent that psychophysical methods such as these can empirically separate drive induction effects from reward effects they should be extremely useful in future studies attempting a finer dissection of the effects of drugs of abuse and, further, in discerning what drugs of abuse may teach us about the process(es) and substrate(s) of reinforcement.
 

 
The effect of pimozide on reward summation
Figure 5: The effect of pimozide on the reward summation function, the curve relating performance (i.e., running speed in an alley) to the number of pulses in the train of reinforcing stimulation. At a dose of 0.2 mg/kg the range is shifted by 1.0 log units (a 10-fold change) to the right of its normal locus, with no decline in the asymptotic level of performance. This indicates that the drug reduced the reinforcing efficacy of the stimulation without impairing the innumerable other processes that translate a reinforcing effect into an observable behavior. Reprinted with permission from Franklin, 1978. Copyright 1978 by ANKHO International, Inc.
 

The Double-Staircase Method

This method, based on a modification of the original procedure as proposed by Cornsweet (1962), has been employed in only one study concerned with the effects of drugs on brain stimulation reward. This study (Marcus & Kornetsky, 1974) represents the sole use of the technique successfully applied to infrahuman subjects and, additionally, was the first psychophysical demonstration of an "acute-onset" (30 minutes postinjection) specific facilitative effect of morphine on rewarding brain stimulation.

For the determination rewarding brain stimulation.

For the determination of positive reinforcement thresholds on this procedure, a trial began with the delivery of a noncontingent 0.5 second pulse train. A response (discrete wheel turn) within 7.5 seconds of this stimulus was immediately followed by a contingent stimulus identical in all parameters to the noncontingent stimulus. The response also terminated the trial, so that only one response-contingent stimulus was available on each trial. Failure to respond had no scheduled consequences, and the trial terminated after 7.5 seconds. Thus, the noncontingent stimulus served two functions: (a) It served as a discriminative stimulus for the availability of response-contingent stimulation, and (b) it also served as a comparative stimulus in that it was a predictor of the stimulus amplitude (and hence reinforcement strength) of the contingent stimulus. Trials began on the average of once every 15 seconds and intertrial or "error" responding resulted in a 15 second "time-out" period (see Figure 6).

The amplitude of the stimuli presented on a given trial was selected according to the double-staircase procedure. Briefly, the stimulus intensities were determined by two independent "staircases" alternating with one another. One staircase determined the stimulus intensity presented on odd-numbered trials, while the other staircase determined the intensity on the even-numbered trials. The occurrence of a successful response onoccurrence of a successful response on trial n, indicating that the stimulus was reinforcing, resulted in a decrease by one step of the intensity on trial n + 2; failure to respond resulted in an increase by one step on trial n + 2. Similarly, the occurrence or nonoccurrence of a response on trial n + 1 determined the intensity on trial n + 3. Step size was selected on the basis of preliminary investigations so as to conform with certain practical and theoretical requirements (Dixon & Massey, 1957). Thresholds were calculated on the basis of the number of responses (or failures) at each intensity level (Dixon & Massey, 1957).

On this procedure morphine administered subcutaneously reliably lowered the reinforcing threshold in all subjects at a dose range of 2 to 8 mg/kg and lasted for a period ranging from 1 to 3 hours postinjection. In terms of psychophysical criteria for drug studies, this study is technically exemplary in several respects. Notably, the use of two interlocking staircases circumvents many commonly encountered problems in other procedures. The threshold determination is independent of rate and/or pattern of responding, and nonspecific responding is controlled for by the time-out component. The quasi-random order of stimulus presentation eliminates the confounding effects of serial order presentations with brain stimulation as well as errors of anticipation or response habituation (Boren & Malis, 1961; Vabituation (Boren & Malis, 1961; Valenstein, 1964). Since the stimuli to be presented are determined on the basis of the subject’s responding during the course of the experiment, there is by definition a high density of data points at intensities near the absolute threshold. This increases both the efficiency (e.g., allowing for a maximum information per trial ratio) and the validity of the measurement. The tracking or titration-like feature of the method permits more "on-line" monitoring of drug-induced changes (e.g., onset, magnitude, and duration of blockade) that are critical and other types of drug interactions such as additivity, synergism, or potentiation. The main problem with this technique relates to its practical limitations. Because the procedure does not allow a high density of suprathreshold stimuli, the thresholds tend to "drift" up over daily testing sessions, with some animals showing signs of extinction, thus presenting obvious problems for studies involving chronic drug administration (unpublished observations).
 

 
Illustration of the discrete-trial instrumental procedure
Figure 6: A diagrammatic and pictorial representation of the basic discrete-trial instrumental procedure utilized in the double-staircase and modental procedure utilized in the double-staircase and modified method of limits procedures (see text) for determining a reinforcing threshold with brain stimulation reward. I. Depicts the sequence of events that occur when a subject does not respond. II. Depicts the sequence of events when a subject emits a correct response (i.e., wheel turn within the appropriate time constraint). S1 represents the noncontingent stimulus which initiates the trial while S2 represents a stimulus that is delivered contingently upon a subject-emitted correct response. S2 is identical to S1 within each trial at a given current intensity throughout the variation of the amplitude parameter. Reprinted with permission from Kornetsky and Bain, 1983. Copyright 1983 by Elsevier/North Holland Biomedical Press.
 

The Modified Method of Limits

In this method animals were trained on the same basic discrete-trial procedure as described above for the double-staircase method (see above and Figure 6). Determination of reward thresholds involved variation of the stimulus intensities according to the classical method of limits. Stimuli were presented in alternating descending and ascending series with a step size of 10 mA. Ten trials were given in succession at each step size or interval. A descending series t each step size or interval. A descending series was initiated at a previously determined intensity which invariably yielded a contingent response in at least 9 out of 10 trials, and then 10 more successive trials were conducted at the next lowest interval and so on. Five or more responses at a particular intensity were arbitrarily scored as a plus for the interval. Descending series were conducted until minus scores were achieved in two successive intervals. An ascending series was then started at one step size below the lowest intensity in the descending series and continued until a level was reached in which there were at least 9 responses out of 10 trials, whereupon a descending series would be initiated at least one interval above the last intensity used in the ascending series. Threshold was determined by calculating the arithmetic mean (x) in microamperes of the midpoints between intervals in which the animal made greater than five responses (a plus score) and less than five responses (a minus score). Results could also be plotted as the psychophysical function of responses plotted against stimulus intensity as shown in Figure 7.

Esposito and Kornetsky have employed this method to assess the effects of opiates and other major agents of abuse. With respect to the opiates, morphine was found to produce an immediate (i.e., 15 minutes post subcutaneous injection of 2 to 8 mg/kg) lowering of the reward threshold at a number of catecholaminethreshold at a number of catecholamine-opiate-receptor rich sites within the brain (Esposito & Kornetsky, 1977; Esposito, McLean, & Kornetsky, 1979). In contrast, naloxone administered alone was without effect at these same sites (Kornetsky, Esposito, McLean, & Jacobson, 1979; Perry, Esposito, & Kornetsky, 1981). Of particular interest was that these threshold-lowering or reward-enhancing effects showed no evidence of tolerance (Esposito & Kornetsky, 1977). This latter finding has subsequently been replicated by a number of other investigators and has important basic and theoretical implications for our understanding of opiate reward mechanisms (for discussion, see Esposito & Kornetsky, 1978; Kornetsky et al., 1979). An important advantage of this method is sensitivity and, accordingly, it has been able to discriminate between the mixed opiate agonist/antagonists on the basis of their relative euphorigenic or abuse potential. Specifically, the abused agent, pentazocine, was found to lower reward thresholds, while other mixed agonist/antagonists (e.g., cyclazocine, nalorphine) that do not possess abuse potential, but are potent analgesics, did not (Kornetsky & Esposito, 1979). Flexibility is another advantage of this psychophysical procedure inasmuch as it has been able to demonstrate the threshold elevating or reward-attenuating effect of neuroleptics (Esposito, Faulkner, & Kornetsky, 1979; Esposito, Perramp; Kornetsky, 1979; Esposito, Perry, & Kornetsky, 1981) and document the threshold-lowering or reward-enhancing effects of morphine (see above) and the psychostimulants amphetamine and cocaine (Esposito, Motola, & Kornetsky, 1978; Esposito, Perry, & Kornetsky, 1980), independent of the many nonreward-specific actions these drugs possess. Recently, Hubner, Bain, and Kornetsky (1983) have demonstrated a synergistic action of morphine/amphetamine combinations on brain stimulation reward thresholds with this procedure consistent with the "street" users predilection for this drug combination. It is noteworthy that the opiate antagonist, naloxone, was able to block or greatly attenuate the reward-enhancing effectiveness of both amphetamine and cocaine at doses of naloxone which when administered independently do not alter reward thresholds (see Figure 8; Esposito et al., 1980; Kornetsky, Bain, & Reidl, 1981). These latter findings are in agreement with those of Seeger, Nazzaro, and Gardner, who used the two-lever method (see above) and suggest an important dopaminergic-endogenous opiate interaction(s) in the mediation of central reward processes (for detailed discussion, see Esposito, 1984).
 


 
Example of the data collection procedure for the modified method of limits
Figure 7: An example of the data collection procedure followed during the modified method of limits for determination of reinforcing thresholds. Ascending and descending series for PRE and POST drug sessions are indicated by the arrows at the top of each column or series, with the individual series threshold at the bottom of each column. The numbers within the columns represent the number of contingent responses at each intensity. The total number of responses at each intensity is indicated at the right-hand column after each session (PRE & POST). Response as a function of intensity can also be plotted as in the accompanying graph. The data show the threshold-lowering effect of a subcutaneous (6 mg/kg) dose of morphine in a rat. Reprinted with permission from Esposito and Kornetsky, 1977. Copyright 1977 by the American Association for the Advancement of Science.
 

Reinforcing-threshold measures derived from this method have also been compared to rate-intensity functions derived from the same placements in the same subject. With few exceptions (e.g., ventral tegmentum), there was little correlation between these two measures at a variety of commonly studied sites in self-stimulation studies (Payton, Kornetsky, & Rosene, 1983). This method has also been used to differentiate the threshold for brain so been used to differentiate the threshold for brain stimulation reward from the threshold for brain stimulation detection and, additionally, the dose-related differential effect of cocaine on these measures within the same subjects (Kornetsky & Esposito, 1981). The reward threshold was determined in the usual manner, whereas the measurement of the brain stimulation detection threshold simply required varying the initial stimulus (i.e., noncontingent) in a quasi-random manner through a predetermined range of sub-to-suprathreshold (nonreinforcing) intensities according to the procedure of the classical method of constant stimuli while keeping the second or response-contingent stimulus at a highly rewarding value in order to maintain responding (see Figure 9).
 

 
Effects of various drug combinations on BSR thresholds
Figure 8: Effects of various combinations of naloxone with cocaine, d-amphetamine, or morphine on brain stimulation reinforcement threshold. Z-scores are based on the mean and standard deviation of threshold changes (POST-PRE) after respective vehicle injections. Z-scores + 2 are considered significantly different. Data are based on the mean Z-score for six subjects with cocaine, four subjects with d-amphetamine, and a typical morphine subject. The naloxone bline, and a typical morphine subject. The naloxone blockade or attenuation of the respective drug effect was observed in all animals. Reprinted with permission from Kornetsky and Wheeling, 1982. Copyright 1982 by Elsevier/North Holland Biomedical Press.
 

Two groups of investigators have employed modified versions of this basic method, essentially varying current frequency rather than amplitude (Porrino & Coons, 1979; Schenk, Williams, Coupal, & Shizgal, 1980). For example, using such frequency variations Schenk et al. (1980) reported variable effects of morphine on self-stimulation in rats with placements in the lateral hypothalamus and dorsal raphé. Varying frequency rather than amplitude has the advantage of controlling for current spread at higher stimulus intensities; however, the use of different frequencies probably incurs the recruitment of different neural populations (Myers, 1971).
 

 
Effects of cocaine on detection and reinforcement thresholds
Figure 9: Dose related effects of cocaine on the reinforcing threshold for rewarding brain stimulation and the threshold for brain stimulation detection in the same subjects. The graph shows a clear dissociation betwection in the same subjects. The graph shows a clear dissociation between effects on a sensory dimension (i.e., detection) and a reinforcement dimension (see text for details). Data are expressed as Z-scores based on the respective mean and standard deviation of effects of vehicle (saline) injections. (N=4). Adapted with permission from Kornetsky and Esposito, 1981. Copyright 1981 by Elsevier/North Holland Biomedical Press.
 

The modified method of limits has several distinct practical and theoretical advantages. The procedure is easily acquired and, most importantly for drug studies, places minimal performance demands on the subjects. Reliable and stable baseline thresholds can be obtained within one week, and the procedure, unlike many others, works well at various brain sites. Responding is maintained throughout the experimental session with minimal stimulation at low intensities, and specific effects of drugs from diverse classes have been demonstrated.

In sum, there are a number of techniques which attempt to meet the basic criteria for an adequate assessment of drug effects on brain stimulation reward (see Table 1). These basic criteria should be used as guideposts in the design of future studies because it is essential to insure specificity of drug effects before attempting the study of the neural substrate of these effects.
 


 
Table 1
Criteria for Adequate Psychophysical Threshold Determination
1. 
2. 
3. 
4. 

5. 
6. 
7. 
8. 

Easy to train with minimal performance demands on subject 
Control for serial order stimulus presentations, response set, and contrast effects 
Dissociate effects on reward from effects on detection thresholds 
Maintain responding throughout the session with minimal stimulation at the lowest intensities 
Yield stable thresholds for chronic studies 
Have adequate flexibility to assess drugs from different classes 
Allow sensitivity to dose-response effects, plus agonist/antagonists 
Possess reliability and validity

 

Metabolic Mapping of the Neural Substrate
of Brain Stimulation Reward

In the first part of this chapter, we have focused on ways in which drug effects on brain stimulation reward can be assessed. In parallel, the problem of determining the neural substrates of these actions of psychoactive substances, whether studied alone or in conjunctthese actions of psychoactive substances, whether studied alone or in conjunction with brain stimulation reward, has been approached in quite similar fashion. Much of our present knowledge regarding the identity of the pathways that are specifically related to intracranial self-stimulation (ICSS) has been obtained from neuroanatomical tracing methods, electrophysiological recording techniques and studies of the effects of lesions, among others.

The recent development by Sokoloff and colleagues of the 2-[14C]deoxyglucose method affords a novel and unique opportunity to map functional neural pathways simultaneously in all anatomical components of the central nervous system. This method, therefore, allows the identification of complex neural circuits that are functionally active during behavioral or pharmacologicall or pharmacological manipulation. The technique is based on the close relationship between energy metabolism (In the brain glucose is virtually the exclusive substrate for metabolism.) and functional neural activity. By measuring glucose utilization in different regions of the brain, it is possible to estimate the level of functional activity within that structure. The technique makes use of a radioactive labeled analogue of glucose, 2-[14C]deoxyglucose, which like glucose is transported into neurons and phosphorylated by hexokinase, but unlike glucose deoxyglucose cannot be further metabolized and is therefore trapped within the cells. Thus, quantitative autoradiography can be used permitting not only the measurement of actual rates of glucose utilization in individual brain regions, but also a pictorial representation of the relative rates of glucose utilization throughout the entire brain. For detailed discussion of the theoretical bases of the radioactive deoxyglucose method and reviews of some of its applications, see Sokoloff (1981; 1982) and Sokoloff et al. (1977).

To date the 2-[14C]deoxyglucose method has proved highly useful in determining the sites of actions of pharmacological agents. The method has the particular advantage of allowing the visualization of alterations in metabolic activity throughout the entire brain, thereby making possible the identification of the complex neurathe identification of the complex neural circuitry which mediates the complete behavioral response to such pharmacological challenges. Studies with amphetamine, for instance, have shown that at high doses (i.e., those eliciting stereotypic responses) altered metabolic activity was evident throughout the dopaminergic nigrostriatal system and the rest of the extrapyramidal system (Orzi, Dow-Edwards, Jehle, Kennedy, & Sokoloff, 1983; Porrino, Lucignani, Dow-Edwards, & Sokoloff, 1984b; Wechsler, Savaki, & Sokoloff, 1979). In contrast, lower doses (i.e., those which have been shown to lower thresholds for ICSS and to facilitate exploratory behavior) produced changes in glucose utilization that were limited mainly to the dopaminergic mesolimbic system, in particular the nucleus accumbens (Porrino et al., 1984b).

In contrast to these pharmacological studies, early work with the 2-[14C]deoxyglucose technique to map the functional pathways activated during brain stimulation reward were not as promising. In these experiments which utilized modifications of the original methodology of Sokoloff and colleagues, rewarding brain stimulation to the medial forebrain bundle, medial prefrontal cortex, or the locus coeruleus produced no consistent pattern of changes in relative rates of glucose utilization when comparisons were made between stimulated and unstimulated sides of the brain (Yadin, Guarini, & Gallistel, 1983).

Work in this laboratory with the fully quantitative 2-[14C]deoxyglucose autoradiographic method at a different stimulation site, however, has yielded quite a different picture. In our studies we assessed changes in functional activity by comparing the actual rates of glucose utilization in various brain regions of rats self-stimulating to the ventral tegmental area (VTA) and passive control rats. We chose the VTA because there is extensive evidence for a significant role for dopamine in the mediation of ICSS, in particular the mesocorticolimbic system which receives its dopaminergic innervation predominantly from cells in the VTA (Clavier & Routtenberg, 1980; Phillips & Fibiger, 1978). In these experiments the standard protocol for the 2-[14C]deoxyglucose technique as developed in this laboratory by Sokoloff and colleagues was followed. On the day of the experiment each rat was lightly anesthetized with halothane/nitrous oxide and implanted with femoral arterial and venous catheters that exited the skin at the nape of the neck. Thus, the animals were free to move and respond in the chamber during the experimental session. After surgery the animals were allowed 3 to 4 hours to recover. Both self-stimulating rats were allowed to begin lever pressing for brain stimulation to the VTA (current parameters: biphasic rectangular wave pulses, 100 Hz, 250 mmA, 400 msec train). The procedure for the measurement of glucose utilization was initiated by the intravenous injection of a pulse of 2-[14C]deoxyglucose at 125 mCi/kg. Timed arterial samples were taken for the next 45 minutes. At the end of this time, animals were killed by the intravenous administration of an overdose of sodium pentobarbital. The brains were rapidly removed and frozen in isopentane cooled to -45 °C. The blood samples were immediately centrifuged and plasma concentrations of 2-[14C]deoxyglucose were determined by means of liquid scintillation counting and plasma glucose concentrations were assayed as well.

The brains were then cut in 20 mm sections in a cryostat maintained at -22 °C. The sections were picked up on glass coverslips and dried on a hot plate. Sections were autoradiographed along with a set of calibrated standards on Kodak OM-1 x-ray film for approximately 12 days. The autoradiographs were analyzed by quantitative densitometry. Optical density measurements for each structure were made in a minimum of four brain sections, both ipsilateral and contralateral to the electrode site. Tissue 14C concentrations were determined from the optical densities and a calibration curve obtained from the densitometric analysis of the standards. Local cerebral glucose utilization (LCGU) in erebral glucose utilization (LCGU) in each structure was then calculated from the 14C tissue concentrations, time courses of the plasma 2-[14C]deoxyglucose and glucose concentrations and the appropriate constants according to the operational equation of Sokoloff et al. (1977).

LCGU was calculated in the terminal fields of the VTA and in various sensory and motor areas. There was an intense increase in metabolic activity at the site of stimulation in the VTA extending laterally into the substantia nigra. Discrete fiber activation was evident through the medial forebrain bundle into the diagonal band of Broca rostrally and in the pontine reticular formation caudally. Within the terminal fields of the VTA a very consistent pattern of LCGU changes were found in the self-stimulating animals compared to the unstimulated controls. Bilateral increases in LCGU were found in the nucleus accumbens, lateral septum, mediodorsal nucleus of the thalamus and in the hippocampus. Ipsilateral to the stimulation site, increased LCGU was evident in the medial prefrontal cortex, bed nucleus of the stria terminalis, and within the basolateral and central nuclei of the amygdala, as well as in the locus coeruleus and the medial parabrachial nucleus (Esposito, Porrino, Seeger, Crane, Everist, & Pert, 1984).

These results show that self-stimulation to the VTA is associated with the discrete activation of projection fibers and sel activation of projection fibers and selective terminal fields of the VTA rather than a diffuse, nonspecific pattern of altered activity throughout the brain. They also show that the 2-[14C]deoxyglucose method can be quite useful in the analysis of the neural substrates of goal-oriented behavior. In a further experiment we have shown that the pattern of changes in LCGU found in self-stimulating animals is specific to this behavior and not merely the result of the electrical stimulation itself. When glucose utilization in self-stimulating animals was compared to that in animals which received noncontingent electrical stimulation delivered by the experimenter (at rates and current parameters comparable to those of the self-stimulating rats) to VTA placements shown to support self-stimulation, a very different pattern of changes was found (Porrino, Esposito, Seeger, Crane, Pert, & Sokoloff, 1984a). These differences are summarized in Table 2.
 

Table 2
Significant Changes in Metabolic Activity Following
Stimulation of the Ventral Tegmental Area
Bilateral Ipsilateral
SELF-STIMULATION
nucleus accumbens 
lateral septum td VALIGN=TOP>nucleus accumbens 
lateral septum 
bed nucleus of the stria terminalis 
hippocampus (CA3) 
mediodorsal thalamus 
locus coeruleus 
medial parabrachial nucleus 
dorsal raphé
medial prefrontal cortex 
central amygdala 
basolateral amygdala
EXPERIMENTER-ADMINISTERED STIMULATION
locus coeruleus 
dorsal raphé
lateral septum 
mediodorsal thalamus 
hippocampus (CA3)


Note: Projection fields of the ventral tegmental area in which Projection fields of the ventral tegmental area in which increased glucose utilization was found following intracranial self-stimulation or experimenter-administered stimulation. In bilateral fields, changes were ipsilateral and contralateral to the stimulation site. In ipsilateral fields, increases were confined to the side of stimulation.
 

We believe that these results can form the basis of future studies in which the effects of drugs such as the psychostimulants and the opiates on the substrates of rewarding brain stimulation can be assessed. Preliminary results of an experiment in which amphetamine (0.5 mg/kg) was administered to animals lever pressing for stimulation to the VTA at current levelslation to the VTA at current levels one third those used in our earlier work revealed a complex pattern of alterations in metabolic activity which was similar in many areas to those changes seen in animals receiving either drug or stimulation alone. In other areas, however, effects were seen which were unique to this combination of treatments, indicating that the facilitative actions of amphetamine on rewarding brain stimulation may be exerted in areas that are not predicted by the effects of rewarding brain stimulation alone or amphetamine’s actions on other behaviors or other measures of reward (Seeger, Porrino, Esposito, Crane, Pert, & Sokoloff, in preparation).

Conclusions

The 2-[14C]deoxyglucose method can be used to identify pathwaysused to identify pathways underlying goal-oriented behavior and the changes that result in such behavior following the administration of psychostimulants or opiates. This requires first the careful definition and assessment of the behavior to be studied, with methods such as those outlined in the first part of this chapter. The analysis of the substrates of behavioral changes requires the use of the fully quantitative 2-[14C]deoxyglucose autoradiographic technique, which takes into account various factors including plasma glucose levels and other physiological changes that can influence the amount of radioactive label present in the tissue. In this way it is possible not only to make valid comparisons between sites ipsilateral and contralaterall and contralateral to the brain stimulation, but, more importantly, to make valid comparisons between subjects. In summary, the fully quantitative 2-[14C]deoxyglucose method, used with appropriate psychophysical techniques, can form the basis for research directed towards an understanding of the complex patterns of neural activity mediating the behavioral process of reinforcement.

Acknowledgment

We thank Dr. Agu Pert and Dr. Louis Sokoloff for their critical commentary on earlier versions of this manuscript, J. D. Brown for his help in preparing the figures, and Brenda Sandler for her editorial assistance.

References

Belluzzi, J. D., & Stein, L. (1977). Enkephalin may mediate euphoria and drive reduction reward. Nature, 266, 556-558.

Boren, J. J., & Malis, J. L. (1961). Determining thresholds of aversive brain stimulation. American Journal of Physiology, 201, 429-433.

Cassens, G. P., Actor, C., Kling, M., & Schildkraut, J. (1981). Amphetamine withdrawal: Effects on the threshold of intracranial reinforcement. Psychopharmacology, 73, 318-322.

Cassens, G. P., & Mills, A. W. (1973). Lithium and amphetamine: Opposite effects on threshold of intracranial reinforcement. Psychopharmacologia, 30, 283-290.

Cassens, G., Shaw, C., Dudding, K. E., & Mills, A. W. (1975). On-line Mills, A. W. (1975). On-line brain stimulation: Measurement of threshold of reinforcement. Behavioral Research Methods and Instrumentation, 7, 145-150.

Clavier, R. M., & Routtenberg, A. (1980). In search of reinforcement pathways: A neuroanatomical odyssey. In A. Routtenberg (Ed.), Biology of reinforcement: Facets of brain stimulation reward (pp. 81-107). New York: Academic Press.

Coons, E. E., & Cruce, J. A. F. (1971). Lateral hypothalamus: Food current intensity in maintaining self-stimulation. Science, 159, 1117-1119.

Cornsweet, T. N. (1962). The staircase method in psychophysics. American Journal of Psychology, 75, 485-491.

Crow, T. J. (1976). Specific monoamine systems as reward pathways: Evidenceward pathways: Evidence for the hypothesis that activation of the ventral mesencephalic dopaminergic neurones and noradrenergic neurones of the locus coeruleus complex will support self-stimulation responding. In A. Wauquier & E. T. Rolls (Eds.), Brain-stimulation reward (pp. 211-238). New York: American Elsevier Publishing.

Dixon, W. J., & Massey, F. J. (1957). Introduction to statistical analysis. New York: McGraw Hill.

Edmonds, D. E., & Gallistel, C. R. (1974). Parametric analysis of brain stimulation reward in the rat. III. Effects of performance variables on the reward summation function. Journal of Physiological and Comparative Psychology, 87, 876-883.

Edmonds, D. E., & Gallistel, C. R. (1977). Reward vs. performance Reward vs. performance in self-stimulation: Electrode specific effects of alpha-methyl-p-tyrosine on reward in the rat. Journal of Comparative and Physiological Psychology, 91, 962-974.

Engen, T., Kling, J. W., & Riggs, L. A. (1971). Experimental psychology (3rd edition). New York: Holt, Rhinehart, and Winston.

Esposito, R. U. (1984). Cognitive-affective integration: Some recent trends from a neurobiological perspective. In H. Weingartner & E. Parker (Eds.), Memory consolidation; Toward a psychobiology of cognition (pp. 15-63). Hillsdale, NJ: Lawrence Erlbaum.

Esposito, R. U., Faulkner, W., & Kornetsky, C. (1979). Specific modulation of brain stimulation reward by haloperidol. Pharmacologyeridol. Pharmacology Biochemistry & Behavior, 10, 937-940.

Esposito, R. U., & Kornetsky, C. (1977). Morphine lowering of self-stimulation thresholds: Lack of tolerance with long-term administration. Science, 195, 189-191.

Esposito, R. U., & Kornetsky, C. (1978). Opioids and rewarding brain stimulation. Neuroscience & Biobehavioral Reviews, 2, 115-122.

Esposito, R. U., McLean, S., & Kornetsky, C. (1979). Effects of morphine on intracranial self-stimulation to various brain stem loci. Brain Research, 168, 425-429.

Esposito, R. U., Motola, A. H. D., & Kornetsky, C. (1978). Cocaine: Acute effects on reinforcement thresholds for self-stimulation behaviorself-stimulation behavior to the medial forebrain bundle. Pharmacology Biochemistry & Behavior, 8, 437-439.

Esposito, R. U., Perry, W., & Kornetsky, C. (1980). Effects of d-amphetamine and naloxone on brain stimulation reward. Psychopharmacology, 69, 187-191.

Esposito, R. U., Perry, W., & Kornetsky, C. (1981). Chlorpromazine and brain-stimulation reward: Potentiation of effects by naloxone. Pharmacology Biochemistry & Behavior, 15, 903-905.

Esposito, R. U., Porrino, L. J., Seeger, T. F., Crane, A. M., Everist, H. D., & Pert, A. (1984). Changes in local cerebral glucose utilization during rewarding brain stimulation. Proceedings of the National Academyof the National Academy of Sciences, 81, 635-639.

Evenden, J. L., & Robbins, T. W. (1983). Increased response switching, perseveration, and perseverative switching following d-amphetamine in the rat. Psychopharmacology, 80, 67-73.

Fechner, G. T. (1851). Zend-avesta, oder uber die dinge des himmels und des jenseits. Leipzig: Voss.

Fechner, G. T. (1860). Elemente der psychophysik. Leipzig: Breitkopf & Hartel.

Fibiger, H. C. (1978). Drugs and reinforcement mechanisms: A critical review of the catecholamine theory. Annual Review of Pharmacology and Toxicology, 18, 37-56.

Franklin, K. B. J. (1978). Catecholamines and self-stimulation: Reward and performance effects dissociated. Pharmacology Biochemistry &cology Biochemistry & Behavior, 9, 813-820.

Gallistel, C. R. (1983). Self-stimulation. In J. A. Deutch (Ed.), The physiological basis of memory consolidation (2nd ed., pp. 270-351). New York: Academic Press.

Gallistel, C. R., Rolls, E. T., & Greene, D. (1969). Neuron function inferred by behavioral and electrophysiological measurement of refractory period. Science, 166, 1028-1030.

Gardner, E. L. (1971). An improved technique for determining brain reward threshold in primates. Behavioral Research Methods and Instrumentation, 3, 273-274.

German, D. C., & Bowden, D. M. (1974). Catecholamine systems as the neural substrate for intracranial self-stimulation: A hypothesis. Brain Research, 73, 381-419.

Hodos, W., & Valenstein, E. S. (1962). An evaluation of response rate as a measure of rewarding intracranial stimulation. Journal of Comparative and Physiological Psychology, 55, 80-84.

Hubner, C. B., Bain, G. T., & Kornetsky, C. (1983). Morphine and amphetamine: Effect on brain stimulation reward. Society for Neuroscience Abstracts, 9, 893.

Huston, J. P. (1971). Relationship between motivating and rewarding stimulation of the lateral hypothalamus. Physiology & Behavior, 6, 711-716.

Huston, J. P., & Mills, A. (1971). Threshold of reinforcing brain stimulation. Commueshold of reinforcing brain stimulation. Communications in Behavioral Biology, 5, 331-340.

Kelly, K., & Reid, L. D. (1977). Addictive agents and intracranial stimulation: morphine and thresholds for positive intracranial reinforcement. Bulletin of the Psychonomic Society, 10, 298-300.

Kornetsky, C., & Bain, G. (1983). Effects of opiates on rewarding brain stimulation. In J. E. Smith & J. E. Lane (Eds.), The neurobiology of opiate reward processes (pp. 237-256). Amsterdam: Elsevier/North Holland Biomedical Press.

Kornetsky, C., Bain, G., & Reidl, E. M. (1981). Effects of cocaine and naloxone on brain stimulation reward. The Pharmacologist, 23, 192.

Kornetsky, C., & Esposito, R. U. (1979). Euphorigenic drugs: Effects on the reward pathways of the brain. Federation Proceedings, 38, 2473-2476.

Kornetsky, C., & Esposito, R. U. (1981). Reward and detection thresholds for brain stimulation: Dissociative effects of cocaine. Brain Research, 209, 496-500.

Kornetsky, C., Esposito, R. U., McLean, S., & Jacobson, J. O. (1979). Intracranial self-stimulation thresholds: A model for the hedonic effect of drugs of abuse. Archives of General Psychiatry, 36, 321-330.

Kornetsky, C., & Wheeling, H. S. (1982). Theoretical and methodological issues in the use of animal models of drug abuse. In M. Y. Spreglestein & ls of drug abuse. In M. Y. Spreglestein & A. Levy (Eds.), Behavioral models and analysis of drug action (pp. 21-38). Amsterdam: Elsevier/North Holland Biomedical Press.

Lewis, M. J. (1983). Age related decline in brain stimulation reward: Rejuvenation by amphetamine. Experimental Aging Research, 7, 225-234.

Liebman, J. M. (1983). Discriminating between reward and performance: A critical review of intracranial self-stimulation methodology. Neuroscience & Biobehavioral Reviews, 7, 45-72.

Marcus, R., & Kornetsky, C. (1974). Negative and positive intracranial reinforcement thresholds: Effects of morphine. Psychopharmacologia, 38, 1-13.

Myers, R. D. (Ed.) (1963). Methods in psychobiology (Vol. 1). New York: Academic Press.

Nazzaro, J. M., Seeger, T. F., & Gardner, E. L. (1981). Morphine differentially affects ventral tegmental and substantia nigra brain reward thresholds. Pharmacology Biochemistry & Behavior, 14, 325-331.

Ordy, J. M., Brizee, K. R., Kaack, B., & Hansche, J. (1978). Age differences in short-term memory and cell loss in cerebral cortex of the rat. Gerontology, 24, 276-285.

Orzi, F., Dow-Edwards, D., Jehle, J., Kennedy, C., & Sokoloff, L. (1983). Comparative effects of acute and chronic administration of amphetamine on local cerebral glucose utilization in the conscious rat. Journal of Cerebrion in the conscious rat. Journal of Cerebral Blood Flow and Metabolism, 3, 154-160.

Payton, M., Kornetsky, C., & Rosene, D. (1983). Brain stimulation reward: Thresholds versus response rates from various brainstem loci. Society for Neuroscience Abstracts, 9, 977.

Perry, N., Esposito, R. U., & Kornetsky, C. (1981). Effects of chronic naloxone treatment on brain-stimulation reward. Pharmacology Biochemistry & Behavior, 14, 247-249.

Phelps, R. W., & Lewis, M. S. (1982). A multifunctional on-line brain stimulation system. Behavioral Research Methods and Instruments, 14, 323-328.

Phillips, A. G., & Fibiger, H. C. (1978). The role of dopamine in maintaining intracranial self-stimulation in the ventral tegmentum, nucleus accumbens, and medial prefrontal cortex. Canadian Journal of Psychology, 32, 58-66.

Porrino, L. J., & Coons, E. E. (1979). Effects of GABA on stimulation-induced feeding and self-stimulation. Pharmacology Biochemistry & Behavior, 12, 125-130.

Porrino, L. J., Esposito, R. U., Seeger, T. F., Crane, A. M., Pert, A., & Sokoloff, L. (1984a). Metabolic mapping of brain during rewarding self-stimulation. Science, 244, 306-309.

Porrino, L. J., Lucignani, G., Dow-Edwards, D., & Sokoloff, L. (1984b). Correlation of dose-dependent effects of acute amphetamine administration on behfects of acute amphetamine administration on behavior and local cerebral metabolism in rats. Brain Research, 307, 311-320.

Schaeffer, G. J., & Holtzman, S. G. (1979). Free-operant and auto-titration brain self-stimulation procedures in the rat: A comparison of drug effects. Pharmacology Biochemistry & Behavior, 10, 127-135.

Schenk, S., Williams, T., Coupal, A., & Shizgal, P. (1980). A comparison between the effects of morphine on the rewarding and aversive properties of lateral hypothalamic and central gray stimulation. Physiological Psychology, 8, 372-378.

Seeger, T. F., & Carlson, K. R. (1981). Amphetamine and morphine: Additive effects on ICSS threshold. Society for Neuroscience Abstracts,roscience Abstracts, 7, 974.

Seeger, T. F., Carlson, K. R., & Nazzaro, J. M. (1981). Pentobarbital induces a naloxone-reversible decrease in mesolimbic self-stimulation threshold. Pharmacology Biochemistry & Behavior, 15, 583-586.

Seeger, T. F., & Gardner, E. L. (1979). Enhancement of self-stimulation behavior in rats and monkeys after chronic neuroleptic treatment: Evidence for mesolimbic supersensitivity. Brain Research, 175, 49-57.

Sokoloff, L. (1981). Localization of functional activity in the central nervous system by measurement of glucose utilization with radioactive deoxyglucose. Journal of Cerebral Blood Flow and Metabolism, 1, 7-36.

Sokoloff, L. (1982). The radioactive deoxyglucose method: theory, proceduree method: theory, procedure and applications for the assessment of local glucose utilization in the central nervous system. In B. W. Agranoff & M. H. Aprison (Eds.), Advances in neurochemistry (pp. 1-82). New York: Plenum Press.

Stein, L. (1962). Effects and interactions of imipramine, chlorpromazine, reserpine, and amphetamine on self-stimulation: Possible neurophysical basis of depression. In J. Wortis (Ed.), Recent advances in biological psychiatry (Vol. 4, pp. 297-311). New York: Plenum Press.

Stein, L., & Ray, O. S. (1960). Brain stimulation reward "thresholds" self-determined in rat. Psychopharmacologia, 1, 251-256.

Valenstein, E. S. (1964). Problems of measurement and interpretation and interpretation with reinforcing brain stimulation. Psychological Review, 71, 415.

Valenstein, E. S., & Beer, B. (1962). Reinforcing brain stimulation in competition with water reward and shock avoidance. Science, 137, 1052-1054.

Wechsler, L. R., Savaki, H. E., & Sokoloff, L. (1979). Effects of d- and l-amphetamine on local cerebral glucose utilization in the conscious rat. Journal of Neurochemistry, 32, 15-22.

Wise, R. A., & Bozarth, M. A. (1981). Brain substrates for reinforcement and drug self-administration. Progress in Neuro-Psychopharmacology, 5, 467-474.

Yadin, E., Guarini, V., & Gallistel, C. R. (1983). Unilaterally activated systems in rats self-stimulating in the medial forebrain bundle,he medial forebrain bundle, medial prefrontal cortex, or locus coeruleus. Brain Research, 266, 39-50.


©1987 Springer-Verlag (printed version)
©2000-2009 Addiction Science Network (web-enhanced version)



ASNet Home

ASNet Profile

Illicit Drug Index

Virtual Lab Tour

Research Reports

Drug Classification

A Primer on Drug Addiction

Experimental Methods

Treatment Resources

Biological Basis of Addiction

Before Prohibition: Early Psychoactive Medicines

Links to Other Websites

Click here to enter the Addiction Science Network Discussion Forum

Brain Reward System©1999-2009 Addiction Science Network
This page was last revised 05 April 2009 11:06 EDT.
Send comments to: feedback@AddictionScience.net
Report technical problems to: webmaster@AddictionScience.net