The neural substrates for the rewarding and dopamine-releasing effects of medial forebrain bundle stimulation have partially discrepant frequency responses.

Midbrain dopamine neurons have long been implicated in the rewarding effect produced by electrical brain stimulation of the medial forebrain bundle (MFB). These neurons are excited trans-synaptically, but their precise role in intracranial self-stimulation (ICSS) has yet to be determined. This study assessed the hypothesis that midbrain dopamine neurons are in series with the directly stimulated substrate for self-stimulation of the MFB and either perform spatio-temporal integration of synaptic input from directly activated MFB fibers or relay the results of such integration to efferent stages of the reward circuitry. Psychometric current-frequency trade-off functions were derived from ICSS performance, and chemometric trade-off functions were derived from stimulation-induced dopamine transients in the nucleus accumbens (NAc) shell, measured by means of fast-scan cyclic voltammetry. Whereas the psychometric functions decline monotonically over a broad range of pulse frequencies and level off only at high frequencies, the chemometric functions obtained with the same rats and electrodes are either U-shaped or level off at lower pulse frequencies. This discrepancy was observed when the dopamine transients were recorded in either anesthetized or awake subjects. The lack of correspondence between the psychometric and chemometric functions is inconsistent with the hypothesis that dopamine neurons projecting to the NAc shell constitute an entire series stage of the neural circuit subserving self-stimulation of the MFB.

Psychophysical inference of frequency-following fidelity in the neural substrate for brain stimulation reward.

The rewarding effect of electrical brain stimulation has been studied extensively for 60 years, yet the identity of the underlying neural circuitry remains unknown. Previous experiments have characterized the directly stimulated ("first-stage") neurons implicated in self-stimulation of the medial forebrain bundle. Their properties are consistent with those of fine, myelinated axons, at least some of which project rostro-caudally. These properties do not match those of dopaminergic neurons. The present psychophysical experiment estimates an additional first-stage characteristic: maximum firing frequency. We test a frequency-following model that maps the experimenter-set pulse frequency into the frequency of firing induced in the directly stimulated neurons. As pulse frequency is increased, firing frequency initially increases at the same rate, then becomes probabilistic, and finally levels off. The frequency-following function is based on the counter model which holds that the rewarding effect of a pulse train is determined by the aggregate spike rate triggered in first-stage neurons during a given interval. In 7 self-stimulating rats, we measured current- vs. pulse-frequency trade-off functions. The trade-off data were well described by the frequency-following model, and its upper asymptote was approached at a median value of 360 Hz (IQR = 46 Hz). This value implies a highly excitable, non-dopaminergic population of first-stage neurons. Incorporating the frequency-following function and parameters in Shizgal's 3-dimensional reward-mountain model improves its accuracy and predictive power.

A new view of the effect of dopamine receptor antagonism on operant performance for rewarding brain stimulation in the rat.

RATIONALE: Previous studies of neuroleptic challenges to intracranial self-stimulation (ICSS) employed two-dimensional (2D) measurements (curve shifts). Results so obtained are ambiguous with regard to the stage of neural processing at which the drug produces its performance-altering effect. We substituted a three-dimensional (3D) method that measures reward-seeking as a function of both the strength and cost of reward. This method reveals whether changes in reward seeking are due to drug action prior to the output of the circuitry that performs spatiotemporal integration of the stimulation-induced neural activity. OBJECTIVES: The aim of this study was to obtain new information about the stage of neural processing at which pimozide acts to alter pursuit of brain stimulation reward (BSR). METHODS: Following treatment with pimozide (0.1 mg/kg) or its vehicle, the proportion of trial time allocated to working for BSR was measured as a function of pulse frequency and opportunity cost. A surface defined by Shizgal's reward-mountain model was fitted to the drug and vehicle data. RESULTS: Pimozide lowered the cost required to decrease performance for a maximal BSR to half its maximal level but did not alter the pulse-frequency required to produce a reward of half-maximal intensity. CONCLUSIONS: Like indirect dopamine agonists, pimozide does not alter the sensitivity of brain reward circuity but changes reward-system gain, subjective effort costs, and/or the value of activities that compete with ICSS. The 3D method is more sensitive and informative than the 2D methods employed previously.

Functional imaging of neural responses to expectancy and experience of monetary gains and losses.

Neural responses accompanying anticipation and experience of monetary gains and losses were monitored by functional magnetic resonance imaging. Trials comprised an initial "prospect" (expectancy) phase, when a set of three monetary amounts was displayed, and a subsequent "outcome" phase, when one of these amounts was awarded. Hemodynamic responses in the sublenticular extended amygdala (SLEA) and orbital gyrus tracked the expected values of the prospects, and responses to the highest value set of outcomes increased monotonically with monetary value in the nucleus accumbens, SLEA, and hypothalamus. Responses to prospects and outcomes were generally, but not always, seen in the same regions. The overlap of the observed activations with those seen previously in response to tactile stimuli, gustatory stimuli, and euphoria-inducing drugs is consistent with a contribution of common circuitry to the processing of diverse rewards.

Brain reward circuitry and the regulation of energy balance.

Reward signals contribute to the regulation of energy balance by influencing switching between feeding and competing behaviors. Properties of natural rewards are mimicked by electrical stimulation of certain brain regions. The rewarding effect produced by stimulating the perifornical region of the hypothalamus is modulated by body weight and is attenuated both by leptin and insulin. Research is reviewed concerning the dependence of the rewarding effect of perifornical stimulation on long-term energy stores and the effects of two neuropeptides implicated in the regulation of energy balance, neuropeptide Y and corticotropin-releasing hormone. It is proposed that the potentiating effect of weight loss on perifornical self-stimulation is not tied to an increased propensity to eat or to an enhancement of food reward per se, but resembles the influence of long-term energy stores on non-ingestive behaviors that defend body weight, such as hoarding.

Fos expression following self-stimulation of the medial prefrontal cortex.

Self-stimulation of the medial prefrontal cortex and medial forebrain bundle appears to be mediated by different directly activated fibers. However, reward signals from the medial prefrontal cortex do summate with signals from the medial forebrain bundle, suggesting some overlap in the underlying neural circuitry. We have previously used Fos immunohistochemistry to visualize neurons activated by rewarding stimulation of the medial forebrain bundle. In this study, we assessed Fos immunolabeling after self-stimulation of the medial prefrontal cortex. Among the structures showing a greater density of labeled neurons in the stimulated hemisphere were the prelimbic and cingulate cortex, nucleus accumbens, lateral preoptic area, substantia innominata, lateral hypothalamus, anterior ventral tegmental area, and pontine nuclei. Surprisingly, little or no labeling was seen in the mediodorsal thalamic nucleus or the locus coeruleus. Double immunohistochemistry for tyrosine hydroxylase and Fos showed that within the ventral tegmental area, a substantial proportion of dopaminergic neurons did not express Fos. Despite previous suggestions to the contrary, comparison of the present findings with those of our previous Fos studies reveals a number of structures activated by rewarding stimulation of both the medial prefrontal cortex and the medial forebrain bundle. Some subset of activated cells in the common regions showing Fos-like immunoreactivity may contribute to the rewarding effect produced by stimulating either site.

Modulation of brain reward circuitry by leptin.

Leptin, a hormone secreted by fat cells, suppresses food intake and promotes weight loss. To assess the action of this hormone on brain reward circuitry, changes in the rewarding effect of lateral hypothalamic stimulation were measured after leptin administration. At five stimulation sites near the fornix, the effectiveness of the rewarding electrical stimulation was enhanced by chronic food restriction and attenuated by intracerebroventricular infusion of leptin. In contrast, the rewarding effect of stimulating neighboring sites was insensitive to chronic food restriction and was enhanced by leptin in three of four cases. These opposing effects of leptin may mirror complementary changes in the rewarding effects of feeding and of competing behaviors.

Effects of NMDA lesions of the medial basal forebrain on LH and VTA self-stimulation.

Rewarding stimulation of the medial forebrain bundle (MFB) increases Fos-like immunoreactivity in many brain areas, including an ipsilateral, basal forebrain region extending from the medial preoptic area (MPO) to the lateral preoptic area, and substantia innominata. Excitotoxic lesions of the lateral portion of this region have been found to produce large sustained or transient increases in the number of pulses required to maintain half-maximal lever-pressing (required number of pulses) for MFB stimulation. In the present study, changes in self-stimulation of the lateral hypothalamus and ventral tegmental area were assessed following excitotoxic lesions of more medial structures, including the MPO and bed nucleus of the stria terminalis. Increases in the required number of pulses (up to 0.16 log10 units) were seen in only 2 of 10 subjects. In two other rats, the reward effectiveness of the stimulation was moderately increased after the lesion as manifested in decreases of up to 0.14 log10 units in the required number. No appreciable change from baseline was seen in the remaining six subjects. The simplest interpretation of these results is that neurons with cell bodies in the medial portion of the basal forebrain may make a smaller contribution to the rewarding effect of MFB stimulation than neurons in the lateral portion.

Early onset of demyelination after N-methyl-D-aspartate lesions of the lateral hypothalamus.

The function of neurons residing in a particular brain area is often assessed by injecting glutamatergic excitotoxins into that area and determining the consequences for the behavior of interest. However, injections of excitotoxins into the central nervous system not only kill local neurons but also demyelinate fibers of passage. Previous studies suggest that the myelin damage is triggered by a delayed inflammatory response to cell death mediated by monocytes of peripheral origin. If so, demyelination should commence only after recruitment of monocytes, their passage through the blood-brain barrier, and their metamorphosis into macrophages. This process is estimated to require at least 48 h. Using a hematoxylin (Weil) stain and immunohistochemistry for myelin basic protein, we looked for signs of demyelination at various times after injections of N-methyl-D-aspartate into the lateral hypothalamus. Demyelination was seen within 24 h after the lesion, sooner than predicted by the monocytic infiltration hypothesis. This finding has implications for interpreting effects of excitotoxic lesions and for developing means of improving their specificity.

Fos-like immunoreactivity in the caudal diencephalon and brainstem following lateral hypothalamic self-stimulation.

Fos immunohistochemistry was used to stain neurons in the caudal diencephalon, midbrain and hindbrain driven by rewarding stimulation of the lateral hypothalamus (LH). Increases in Fos-like immunoreactivity were most pronounced ipsilateral to the site of stimulation and tended to be confined within discrete structures such as the posterior LH, arcuate nucleus, ventral tegmental area (VTA), central gray, dorsal raphé, pedunculopontine area (PPTg), parabrachial nucleus, and locus coeruleus. At least two of these structures, the VTA and PPTg, have been implicated in medial forebrain bundle self-stimulation.

Fos-like immunoreactivity in forebrain regions following self-stimulation of the lateral hypothalamus and the ventral tegmental area.

According to the descending-path hypothesis, the direct excitation of descending fibers linking the lateral hypothalamus (LH) and ventral tegmental area (VTA) contributes to the rewarding effect produced by electrical stimulation of the medial forebrain bundle (MFB). To visualize forebrain neurons activated by stimulation of both the LH and VTA, Fos-like immunoreactivity (FLIR) in forebrain regions was assessed following self-stimulation of these two sites in male rats. Among the regions where FLIR was greater in the stimulated hemisphere following either LH or VTA stimulation were the anterior LH, the substantia innominata, and the bed nucleus of the stria terminalis, and olfactory tubercle. These findings are analyzed with reference to the effects of forebrain lesions on self-stimulation of the MFB. Advantages and limitations of using FLIR to identify neurons activated by rewarding stimulation are discussed.

Neural basis of utility estimation.

The allocation of behavior among competing activities and goal objects depends on the payoffs they provide. Payoff is evaluated among multiple dimensions, including intensity, rate, delay, and kind. Recent findings suggest that by triggering a stream of action potentials in myelinated, medial forebrain bundle axons, rewarding electrical brain stimulation delivers a meaningful intensity signal to the process that computes payoff.

Increased ipsilateral expression of Fos following lateral hypothalamic self-stimulation.

Immunohistochemical labeling of Fos protein was used to visualize neurons activated by rewarding stimulation of the lateral hypothalamic level of the medial forebrain bundle (MFB). Following training and stabilization of performance, seven rats were allowed to self-stimulate for 1 h prior to anesthesia and perfusion. Brains were then processed for immunohistochemistry. Two control subjects were trained and tested in an identical manner except that the stimulator was disconnected during the final 1 h test. Among the structures showing a greater density of labeled neurons on the stimulated side of the brains of the experimental subjects were the septum, lateral preoptic area (LPO), medial preoptic area, bed nucleus of the stria terminalis, substantia innominata (SI), and the lateral hypothalamus (LH). Several of these structures, the LPO, SI, and LH, have been implicated in MFB self-stimulation by the results of psychophysical, electrophysiological, and lesion studies.

Behavioral measures of conduction velocity and refractory period for reward-relevant axons in the anterior LH and VTA.

Psychophysical techniques were used to estimate the conduction velocities and refractory periods of reward-relevant fibers in the anterior lateral hypothalamus (ALH) and ventral tegmental area (VTA). Six male rats of the Long-Evans strain served as subjects. Stimulation consisted of trains of pulse pairs delivered to the ALH and VTA; each stimulation site received one pulse from each pair. Collision of antidromic and orthodromic action potentials along the same reward-relevant axons was inferred from a decrease in the effectiveness of the stimulation at short intra-pair intervals. Significant collision-like effects were obtained for 5 of the 6 subjects, supporting the notion that reward-relevant neurons directly link the ALH and VTA. Estimates of the slowest conduction velocities ranged from 0.4 to 5.5 m/s. Refractory period estimates, obtained by delivering the trains of pulse pairs to a single stimulation site, ranged from 0.5 to 5.4 ms. The overlap between these psychophysical estimates and the electrophysiological estimates obtained in the accompanying paper is consistent with the notion that reward-relevant MFB fibers arise in one or several rostral MFB nuclei.

Effects of excitotoxic lesions of the basal forebrain on MFB self-stimulation.

Electrolytic lesions of the anterior medial forebrain bundle (MFB) have been shown to attenuate the rewarding impact of stimulating more caudal MFB sites. In the present study, excitotoxic lesions were employed to determine the relative contribution of somata or fibers of passage contributing to that effect. Changes in reward efficacy were inferred, at three currents, from lateral displacements of the curve relating the rate of responding to the number of stimulation pulses per train. After baseline data were collected from stimulation sites in the lateral hypothalamus (LH) and the ventral tegmental area (VTA), 70 nmol of N-methyl-D-aspartic acid was injected via cannulae aimed at basal forebrain sites. Three subjects were injected with vehicle and served as controls. In 5 out of 15 cases, lesions encompassing the lateral preoptic area, anterior LH, and substantia innominata resulted in long-lasting, large increases (0.2-0.47 log10 units) in the number of pulses required to maintain half-maximal rates of self-stimulation for low currents delivered via the LH electrode; smaller increases (0.08-0.33 log10 units) were noted at moderate and high currents. Seven rats with similar or more dorsally located damage showed moderate or transient increases in the number of pulses required to maintain half-maximal rates of LH or VTA self-stimulation. Vehicle injections did not affect behaviour. Varying degrees of demyelination were seen, mostly removed from the electrode tip, and in locations that varied substantially across subjects manifesting similar changes in self-stimulation. These results support the notion that somata in the basal forebrain give rise to some of the directly activated fibers subserving self-stimulation of the MFB.

Physiological measures of conduction velocity and refractory period for putative reward-relevant MFB axons arising in the rostral MFB.

Extracellular recordings were obtained, in urethane-anesthetized rats, from 44 neurons in the rostral bed nuclei of the medial forebrain bundle (MFB). These cells were antidromically activated by stimulation of MFB sites that typically support self-stimulation. Recording sites included the magnocellular preoptic nucleus, substantia innominata, ventral pallidum, olfactory tubercle, and horizontal limb of the diagonal band. Refractory period estimates ranged from 0.35 to 1.20 ms (mean +/- SD = 0.72 +/- 0.30 ms, n = 15) for stimulation sites in the lateral hypothalamic and ventral tegmental areas when using currents of twice threshold and procedures designed to estimate excitability at or near the site of stimulation. Interelectrode conduction velocity estimates ranged from 1.48 to 20.0 m/s (mean +/- SD = 9.26 +/- 7.22 m/s, n = 11) and were obtained by dividing the interelectrode distance by the difference in the response latency from the two MFB stimulation sites. The refractory period and conduction velocity estimates for these neurons overlap the psychophysically derived estimates for MFB reward neurons. These data are consistent with the hypothesis that neurons arising in the rostral bed nucleus of the MFB compose at least part of the directly activated substrate for MFB self-stimulation.

Attenuation of medical forebrain bundle reward by anterior lateral hypothalamic lesions.

Psychophysical data consistent with rostro-caudal conduction along reward-relevant neurons linking the lateral hypothalamus (LH) and ventral tegmental area (VTA) have lead to the hypothesis that some of the directly activated neurons responsible for medial forebrain bundle (MFB) self-stimulation arise anterior to the level of the LH. This hypothesis has been challenged on the grounds that lesions to the anterior LH (ALH) often fail to degrade the rewarding value of stimulating more posterior MFB sites. The present study was aimed at investigating the effect of lesion location and stimulation current on the efficacy of ALH lesions in an effort to account for the inconsistencies in the earlier data. Self-stimulation thresholds were obtained for LH and VTA sites by estimating the number of pulses per stimulation train required for half-maximal responding at each of 3 currents. Electrolytic lesions (anodal, 1.0 mA for 10 s) were then made to the ALH at varying medial-lateral coordinates. In 7 of the 14 rats with MFB stimulation sites, lesions to the ALH produced increases in threshold which often declined over the next several days to weeks; in 5 cases thresholds remained elevated by 0.1 to 0.25 log10 units above baseline up to end of testing. In all but one case, the effective lesions were centered in the lateral ALH. Increases in threshold were more likely to be detected when stimulating at low currents; at low currents fewer neurons are recruited and the lesion can have a greater proportional effect on threshold. These data support the hypothesis that cell bodies, terminals, or fibers of passage in the ALH contribute to the rewarding effect of stimulating more posterior MFB sites.

Self-stimulation of the MFB following parabrachial lesions.

It has been reported previously that the parabrachial region supports robust self-stimulation. In the present study, we determined whether lesions of the parabrachial nucleus (PBN) influence the rewarding effect of medial forebrain bundle (MFB) stimulation. In 10 rats, stimulation electrodes were aimed at the lateral hypothalamus and/or ventral tegmental area and a lesioning electrode aimed at the PBN. Rate-frequency curves were collected at each of three stimulation currents at each electrode, before and after lesioning. Four rats showed virtually no change in the frequency required to sustain half-maximal performance following lesioning, and two showed some postlesion decreases. Only two rats showed substantial postlesion increases in required frequency; the lesions in these subjects damaged the region ventral to the superior cerebellar peduncle, just caudal to the decussation of the peduncle, but spared the PBN. Thus, the reward effectiveness of MFB stimulation does not appear to be altered substantially following PBN lesions but may decrease following damage to the neighboring pedunculopontine region.

Evidence implicating both slow- and fast-conducting fibers in the rewarding effect of medial forebrain bundle stimulation.

A behavioral version of the collision test was used to determine whether reward-relevant neurons directly link self-stimulation sites in the lateral hypothalamic (LH) and ventral tegmental (VTA) areas. Five male rats served as subjects. Trains of conditioning (C) and test (T) pulses were delivered to the two stimulation sites, each site receiving one of the pulses from each pair. The C-T interval was varied from 0.2-17.3 ms, and the effectiveness of the paired pulse stimulation was estimated by comparing the rate-number curve obtained at each C-T interval to rate-number curves obtained with trains of evenly spaced single pulses delivered via one electrode. For 4 of the subjects, stimulation effectiveness increased with the C-T interval, and the form of this increase was similar regardless of which electrode delivered the C-pulses. These increases in effectiveness are consistent with recovery from collision block in reward-relevant fibers stimulated at both sites. The domain of the rising portion of the effectiveness versus C-T interval curve spanned 2.2-7.7 ms. Such a gradual rise suggests that the directly stimulated substrate is composed of fibers with a wide range of conduction velocities and/or refractory periods. The discrepancy between these gradually rising collision curves and the steeply rising curves obtained in previous collision studies may have been due to inadequate sampling of the rate-number function in the earlier studies.

Thermogenesis in brown adipose tissue is activated by electrical stimulation of the rat dorsal raphe nucleus.

Brainstem mechanisms involved in the central control of thermogenesis in brown adipose tissue (BAT) were investigated, in urethane-anaesthetized rats, by observing changes in the temperature of interscapular BAT following electrical stimulation of sites in the dorsal raphe nucleus (DRN). Large increases in temperature could be produced by single 30 s stimulation trains. The magnitude of the temperature increase grew as a function of both current and frequency. Non-linear temporal summation was observed when near-threshold trains were delivered in close succession. These results are consistent with the view that ascending projections from the DRN relay thermal information to the basal forebrain and hypothalamus.