Gordon J. Mogenson

FROM MOTIVATION TO ACTION: FUNCTIONAL INTERFACE BETWEEN THE LIMBIC SYSTEM AND THE MOTOR SYSTEM

Limbic forebrain structures and the hypothalamus are essential in the initiation of food-seeking, escape from predators and other behaviors essential for adaptation and survival. Neural integrative activities subserving these behaviors initiate motor responses but the neural interface between limbic and motor systems has received relatively little attention. This neglect has been in part because of the emphasis on the motor control of the movements and on the contributions of the cerebral cortex, cerebellum, spinal cord and other components of the motor system, but more importantly, because of a lack of relevant anatomical evidence of connections. Anatomical findings obtained in recent years now make it possible to investigate the neural interface between limbic and motor systems—neural mechanisms by which “motivation” gets translated into “action”. It has been proposed that the nucleus accumbens is a key component of this neural interface since it receives inputs from limbic forebrain structures, either directly or indirectly via the ventral tegmental area of Tsai, and sends signals to the motor system via the globus pallidus. The nucleus accumbens has been implicated in locomotion, a fundamental component of attack, feeding and other behaviors utilized in adaptation and survival. It has also been implicated in oral motor responses, utilized in feeding, drinking, vocalization and other adaptive responses. The role of the nucleus accumbens and its functional relationship with the ventral tegmental area and globus pallidus has been investigated using neuropharmacological-behavioral techniques to initiate and disrupt locomotor and ingestive responses and using electrophysiological recording techniques. The results of these investigations are interpreted in relation to a proposed model of the limbic-motor interface and further experiments are suggested. This model for the initiation of actions by limbic forebrain structures (e.g. the “emotice brain”) is considered in relation to what is known about the initiation of actions by cognitive processes involving previous experience and learning, which include research is how the “emotive brain” and the “cognitive brain” operate together in response initiation.

Evidence for a projection from the lateral preoptic area and substantia innominata to the 'mesencephalic locomotor region' in the rat.

A series of anatomical and electrophysiological experiments have been carried out to examine the organization of a direct projection from the substantia innominata and the lateral preoptic area of the hypothalamus, referred to collectively as the subpallidal region, to the pedunculopontine nucleus and adjacent parts of the dorsal midbrain in the adult rat. In the first series of experiments, the retrogradely transported fluorescent tracer SITS, which does not appear to be taken up by fibers-of-passage, was injected into the pedunculopontine nucleus, and the distribution of labeled neurons was plotted in the substantia innominata and lateral preoptic area, as well as in adjacent regions including the medial preoptic area, the bed nucleus of the stria terminalis and parvocellular parts of the paraventricular nucleus. Then, the anterogradely transported lectin PHA-L, which also does not appear to be taken up in effective amounts by fibers-of-passage, was injected into parts of the substantia innominata and lateral preoptic area that project directly to the pedunculopontine nucleus. These experiments demonstrated that fibers from both regions descend through the medial forebrain bundle and give rise in the pedunculopontine nucleus to a terminal field that contains many structures with the appearance of terminal boutons. They also indicated that individual fibers from the subpallidal region innervate both the pedunculopontine nucleus and adjacent parts of the central gray, and that the pathway innervates areas along the length of the medial forebrain bundle on its way to the dorsal midbrain. In a third series of experiments the retrogradely transported fluorescent tracer True Blue was injected into upper thoracic levels of the spinal cord, and it was found that the region of the pedunculopontine nucleus that receives the densest input from the subpallidal region contained many retrogradely labeled neurons on both sides of the brain. And finally, a series of electrophysiological experiments demonstrated that single-pulse stimulation of the substantia innominata and the lateral preoptic area altered the firing rate of a majority of the neurons in and around the pedunculopontine nucleus, and that excitatory and inhibitory responses occurred about equally. These results clearly suggest that the subpallidal region projects directly to the pedunculopontine nucleus and adjacent regions including the central gray and the superior colliculus.(ABSTRACT TRUNCATED AT 400 WORDS)

Electrophysiological responses of neurones in the nucleus accumbens to hippocampal stimulation and the attenuation of the excitatory responses by the mesolimbic dopaminergic system

Extracellular single unit recordings were obtained from neurones in the nucleus accumbens of urethane anaesthetized rats. Single pulse stimulation (300-800 microA, 0.15 ms, 0.5-1.5 Hz) of the ventral subiculum of the hippocampus strongly excited silent and spontaneously active (3-6 spikes/s) medial accumbens neurones. The majority of neurones excited by hippocampal stimulation were quiescent and identified only by the elicited action potentials. Neurones on the dorso-medial border of the nucleus accumbens and adjacent lateral septum, with a faster spontaneous discharge rate (8-12 spikes/s), were inhibited by hippocampal stimulation. In the ventral border of the accumbens and the olfactory tubercle, hippocampal stimulation also inhibited the fast-firing (greater than 20 spikes/s) neurones. When trains of 10 conditioning pulses (300-800 microA, 0.15 ms, 10 Hz) were delivered to the ventral tegmental area (VTA) 100 ms before each single-pulse stimulation of the hippocampus, the excitatory responses of the silent and spontaneously active accumbens neurones were attenuated. The possibility of this relatively prolonged attenuation effect being dopamine-mediated was supported by several lines of evidence. Dopamine, applied iontophoretically, reduced markedly the excitatory response of accumbens neurones to hippocampal stimulation. Iontophoretically applied dopamine mimicked the attenuating effect produced by VTA conditioning stimulation in the same neurone. The attenuating effects of VTA conditioning stimulation on the activation of accumbens neurones by hippocampal stimulation was reduced by: (1) administration of 6-hydroxydopamine to the VTA 2 days and 7-9 days prior to the recording session, (2) the intraperitoneal injection of haloperidol 1 h before the recording session, and (3) the iontophoretic application of trifluoperazine to accumbens neurones. These observations support the hypothesis that the attenuating effects of the mesolimbic dopamine system on limbic inputs to the nucleus accumbens may have a role in limbic-motor integration.

Response of nucleus accumbens neurons to amygdala stimulation and its modification by dopamine.

Extracellular single unit recordings were obtained from the nucleus accumbens of urethane anesthetized rats. It was found that electrical stimulation of the basal lateral and basal medial nuclei of the amygdala produced strong excitatory responses in neurons of the nucleus accumbens, in particular the medial region. Latencies of activation were relatively short with a mean of 10.7 ms. Dopamine applied iontophoretically had a marked attenuating effect on the excitatory response of nucleus accumbens neurons to amygdala stimulation. The spontaneous activity of all neurons recorded from the nucleus accumbens was also suppressed by dopamine, but the excitatory response was more sensitive to dopamine inhibition than the spontaneous activity. Neurons in the nucleus accumbens showed a variety of responses to single-pulse electrical stimulation of the ventral tegmental area (VTA). Some units in the nucleus accumbens received convergent inputs from both the amygdala and the VTA. Stimulation of the VTA also attenuated the response of nucleus accumbens neurons to excitatory inputs from the amygdala. A train of 10 pulses (0.15 ms, 200--600 microA) at 10 Hz delivered to the VTA at 100 ms before stimulation of the amygdala caused attenuation of the original excitatory response. The attenuating effect could be observed irrespective of whether individual single-pulse stimulation of the VTA elicited a response in that particular accumbens neuron or not. 6-Hydroxydopamine injected into the VTA 2 days prior to the recording experiment, or haloperidol injected intraperitoneally 1 h before the recording session, abolished this attenuating effect. However, responses to single-pulse stimulations of the VTA were not abolished. The results suggest that the attenuation of the excitatory response to amygdala stimulation was due to the release of dopamine from mesolimbic dopaminergic neurons. Responses to single-pulse stimulations of the VTA were probably due to activation of non-dopaminergic neurons projecting from the same area. It is suggested as a working hypothesis that this inhibitory effect of dopamine may be an important function of the mesolimbic dopamine pathway in modulating the extent to which limbic structures can exert an influence on the motor system through the accumbens.

Injections of dopaminergic, cholinergic, serotoninergic and gabaergic drugs into the nucleus accumbens: effects on locomotor activity in the rat

Abstract Locomotor activity was elicited by injecting dopamine into the nucleus accumbens of the chronically cannulated rat. Dopamine was also injected together with cholinergic, serotoninergic and GABAergic agonists and antagonists to investigate the possible contribution to locomotor activity of these putative neurotransmitters. Carbachol elicited a transient enhancement of dopamine-stimulated activity. Atropine attenuated carbachol-stimulated enhancement but did not attenuate dopamine-stimulated activity. Serotonin attenuated dopamine-stimulated activity. The peripheral serotonin antagonist, methysergide, also attenuated locomotion. The administration of GABA elicited a bimodal response in locomotion, the lower dose eliciting a small increase in locomotion and the larger dose eliciting a reduction. The GABA antagonist, picrotoxin, elicited only increases in locomotion at all doses. These results suggest that there is not a cholinergic interneuron on the dopamine-stimulated pathway subserving locomotion but that both cholinergic and serotoninergic projections may modulate locomotor activity. Further they suggest that such modulation must be at least one inhibitory interneuron away from the dopamine synapse on the locomotor pathway. In addition, these results suggest a direct influence of GABAergic interneurons on the pathway subserving locomotor activity within the nucleus accumbens.

A study of the contribution of hippocampal—accumbens—subpallidal projections to locomotor activity

Locomotor activity recorded in an automated open-field apparatus was increased substantially by unilateral injections of carbachol, a cholinergic agonist, into the dentate gyrus of the hippocampus. Hyperactivity elicited in this way was reduced significantly when glutamate antagonists were injected into the ipsilateral nucleus accumbens. Injecting gamma-aminobutyric acid into the ipsilateral subpallidal region also reduced the hyperactivity from injections of carbachol into the dentate gyrus. When these compounds were injected into the contralateral accumbens and subpallidal region, respectively, there was little or no reduction in the carbachol-elicited locomotor activity. These observations suggest that neural pathways fom hippocampus to accumbens to subpallidal region may contribute to locomotor activity.

Evidence that an accumbens to subpallidal GABAergic projection contributes to locomotor activity

Neural projections from nucleus accumbens to subpallidal region, which contains a major GABAergic component, have been demonstrated with anatomical and electrophysiological techniques. The possible contribution of this GABA projection to the initiation of locomotor activity was investigated using neuropharmacological techniques. Injecting picrotoxin, a GABA antagonist, into the ventral globus pallidus increased locomotor activity measured in an open-field test, confirming findings. Locomotor activity was also increased when picrotoxin was injected into the lateral preoptic area, the sublenticular part of the substantia innominata and bed nucleus of the stria terminalis. In another series of experiments locomotor activity initiated by injecting dopamine into the nucleus accumbens was attenuated by pretreating the lateral preoptic area, the substantia innominata and ventral globus pallidus with GABA. These observations provide evidence that GABAergic projections from accumbens to subpallidal region contribute to locomotor activity and raise the possibility that they have a role in exploratory locomotion and in certain goal-directed behaviors.

Nucleus accumbens to globus pallidus GABA projection: electrophysiological and iontophoretic investigations.

Extracellular recordings were obtained from neurons in the nucleus accumbens and globus pallidus of urethane anesthetized rats. Eight neurons in the nucleus accumbens were activated antidromically following stimulation of the globus pallidus. Calculated conduction velocities were 0.4-1.5 m/sec, indicative of small unmyelinated fibers. A total of 74 of 153 neurons in the globus pallidus responded to stimulation of the nucleus accumbens. Of these neurons 4 (2.7%) were excited only, 46 (30.1%) were inhibited only and 24 (15.7%) had sequential effects to excitation and inhibition. Iontophoretic application of picrotoxin was found to attenuate or abolish the poststimulus inhibition in one-half of the neurons examined. The 74 neurons which responded to stimulation of the nucleus accumbens had slower firing frequencies and generally more random firing patterns than neurons which did not respond to stimulation. Fifty-three per cent of all globus pallidus neurons examined had increased spontaneous firing frequencies following the iontophoretic administration of picrotoxin alone. This is indicative of the removal of a tonic GABA input onto these neurons. Most neurons examined had decreased spontaneous firing frequencies following the iontophoretic application of GABA which could be blocked by the iontophoretic application of picrotoxin. The results from antidromic activation, slow conduction velocity, sensitivity to GABA and picrotoxin, and picrotoxin attenuation of the poststimulus inhibitory effect provide evidence of a direct GABAergic projection from the nucleus accumbens to the globus pallidus in the rat.

Electrophysiological studies of neurons in the ventral tegmental area of tsai

Extracellular recordings were obtained from single neurons in the ventral tegmental area of rats anesthetized with urethane. It was found that the area appeared to contain two groups of neurons with distinctly different spike durations, firing rates and firing patterns. One group (group A) had properties similar to those of nigral dopaminergic neurons: slow random firing rates, unusually long spike durations and slow conduction velocities. The discharge rate of the majority of these neurons was reduced by iontophoretically applied dopamine. It was concluded that neurons of this group were probably A10 dopaminergic neurons. The other group (group B) had relatively faster and rhythmical firing rates, short spike durations and faster conduction velocities and were considered to be non-dopaminergic. Forty-nine units in the ventral tegmental area were antidromically activated by electrical stimulation of the nucleus accumbens. Units antidromically activated included neurons of group A and group B, suggesting that the nucleus accumbens received dural projections of dopaminergic and non-dopaminergic fibres from the ventral tegmental area. The discharge rate of 141 out of 142 neurons tested in the ventral tegmental area (group A: 66/66, group B: 75/76) was found to be reduced by GABA. The inhibition was blocked by the simultaneous application of picrotoxin. Picrotoxin alone activated 47.7% of 155 units tested. These results provide further evidence of a GABAergic input to dopaminergic and non-dopaminergic neurons projecting to the limbic forebrain structures.

Evidence that projections from substantia innominata to zona incerta and mesencephalic locomotor region contribute to locomotor activity

A series of anatomical, electrophysiological and behavioral experiments was carried out in the rat to investigate the possible functional significance of a recently demonstrated neural pathway from the substantia innominata of the subpallidal forebrain to the mesencephalic locomotor region. Following injections of the anterogradely transported lectin PHA into the substantia innominata labeled fibers with terminal boutons were observed in the zona incerta, dorsal to the medial part of the subthalamic nucleus, and some appeared to continue on to the pedunculopontine nucleus. Electrophysiological recordings of action potentials were made from neurons in the substantia innominata and some of these neurons were activated antidromically by single-pulse stimulation of the zona incerta and/or by single-pulse stimulation of the pedunculopontine nucleus as well. Neurons in the zona incerta responded orthodromically to stimulation of the substantia innominata. Locomotor activity was initiated by injecting picrotoxin, a GABA antagonist, unilaterally into the substantia innominata through chronic cannulae, as reported previously. This picrotoxin-initiated locomotor activity was reduced significantly when procaine (a neuronal blocker) was injected into the ipsilateral zona incerta. Injecting procaine into the contralateral zona incerta had little or no effect on the picrotoxin-initiated locomotor activity. Taken together these observations suggest the tentative working hypothesis that projections from the substantia innominata to the zona incerta as well as the pedunculopontine nucleus may contribute to the locomotor component of adaptive behaviors resulting from limbic forebrain integrative activities, an hypothesis that can now be investigated further.