Catecholamines
 

The catecholamines (CA) are a group of 3 transmitters which lie on a common pathway of synthesis. They were among the first transmitters identified, and they were the first to be definitively localized in the central nervous system with the development of
the histofluorescence method of Falck and Hillarp. In fact the visualization of the catecholamines revolutionized behavioural neuroscience and neuroanatomy by showing it was possible to identify the neurochemical characteristics of connections between
structures.

The catecholamines adrenaline and noradrenaline were first identified as chemical messengers in the sympathetic division of the autonomic nervous system. Noradrenaline is the transmitter at sympathetic post-ganglionic fibres innervating smooth muscle, while
adrenaline and noradrenaline are released by the adrenal gland. Dopamine was first identified as a transmitter in the CNS but was later found in the sympathetic ganglia.
 

The cell bodies of noradrenaline and adrenaline containing neurons are all in the brain stem but they innervate almost all the forebrain through 2 pathways. The dorsal noradrenaline pathway originates in small nuclei (about 3000 cells in the rat) in the dorsal pons called the locus coeruleus. This is so named because in humans it contains melanin and it looks blue. The fibres from the locus coeruleus (A6) arc forwards and down to joint the MFB and run forward through the hypothalamus after which they run dorsally and innervate the cortex. Their projections are remarkable because a single NA neuron may meander through many regions of the cortex. There are also some fibres projecting caudally. The ventral pathway originates in several groups of noradrenaline cells in the pons and medulla which run forward to joint the MFB and innervate the brain stem and hypothalamus. There are also a small number of neurons that release adrenaline.

(Figures of NA and DA pathways)

The dopamine pathways are even more restricted in distribution. The major groups lie in the substantia nigra (which looks black in humans) and the adjacent ventral tegmental area. They innervate the dorsal and ventral striatum and related structures via the MFB.

The striatum is innervated by the cells of the substantia nigra in a systematic mapping. Medial cells project to medial striatum and lateral cells to lateral striatum. The dorsal and ventral mapping is inverted - ventral SN to dorsal striatum. The cells of the VTA project to the ventral striatum, olfactory tubercle and, in the rat, restricted areas of the cortex including the medial and lateral frontal cortex and cingulate gyrus. These pathways are known as the nigro-striatal pathway, meso-limbic and meso-cortical pathways respectively.

A small branch of the mesocortical pathway projects to the lateral habenula of the thalamus.

There are several short pathways. There are cells scattered through the dorsal part of the hypothalamus and perivenricular grey. They terminate locally and send collaterals to spinal dorsal horn and the intermediolateral cellsc columns where the autonomic pre-ganglionic neuros are

The tuberoinfundibular neurons (A12) release dopamine into the pituitary portal system where it acts to inhibit prolactin release.o not make r
 

Synthesis

The common precursor is the amino acid tyrosine which comes from the diet. Tyrosine is converted to L-DOPA by the enzyme tyrosine hydroxylase (TyrOHase) which can be blocked by drugs such as alpha-methyl-p-tyrosine. L-DOPA is converted to dopamine by DOPA decarboxylase (inhibited by benserazide or alpha-methyl-DOPA). In DA neurons the pathway stops here. Both these enzymes are cytosolic and dopamine is taken up into vesicles by an uptake mechanism in the vesicular membrane. In NA neurons the vesicles contain the enzyme dopamine-beta-hydroxylase (DBH), which converts dopamine to noradrenaline. DBH can be inhibited by drugs like FLA-57. DBH has been very important in the evidence for exocytotic release because as it is too large to cross the cell membrane but as it is inside vesicles it is released alongside NA and can be found in the extracellular space.

In adrenaline containing neurons there is one additional enzyme - phenylethanolamine-n-methyl-transferase (PNMT for short) which converts NA to A.
 

The rate limiting step in the pathway is tyrosine hyrdoxylase which about 75% saturated under resting conditions. so it is slightly sensitive to the availability of tyrosine which is taken up by a mechanisms common to many amaino acids. In normal circumsatnces manipulating tyrosine availability has little effect but when the firing rate is high, supplementation leads to a moderate increase in turnover.

More important there is complex regulation of tyrosine hydroxylase activity.

1. end product inhibition - NA and Da bind to tyrOHase and inhibitit it. Reactivated by phosphorylation by cAMP which liberates the CA.

2.Phosphorylation by protein kinases (CAM-KII, ERK extracellular signal-regulated protein kinases, PKA). The enzyme has three sites on its reguilatory domain ( serine 19, 31, 40) where phosphorylation activates the enzyme. In adrenal chromaffin cells the
phosphorylation is induced by cacium influx, or by recptors for Acetylcholine and VIP that are coupled through PIP2 and cAMP second messenger systems.

3. There is also long-term up regulation of tyrosine hyroxylase. This increase in the production of enzyme takes hours to occur rather than minutes like the other mechansisms. It is blocked by inhibition of protein sythesis and has been shown to be mediated by enhanced transcription of the gene. Apparently this does not occur in all CA cells groups. Induction after reserpine treatment is strong in locus coeruleus, modest in VTA but does not occur in the SNc. Induction also occurs for DA betaOHase and PNMT,
proably by simlar mechanisms.
 

Storage, release and regulation.

A number of important features of chemical neurotransmission were first recognized in the catecholamines including regulation of synthesis by the product, the presence of more than one pool of transmitter in the neuron, regulation of release by receptors on the
terminal membrane (autoreceptors). As so often has happened, these discoveries were
the result of analysis of drug effects.
 

In the vesicles CA are found along with ATP and acidic proteins. There appear to be two pools of CA, one is a pool of free CA which is readily and preferentially released.

The other pool is bound in a complex with ATP and the acidic proteins and is normally not released. After synthesis in the cytoplasm, DA is taken up into vesicles by the vesicular monoamine transporter (different from the re-uptake pump). Since the trnsport is agains a steep concentration gradient energy is required which is supplied by protons. The vesicular memebrane also contains an ATPase which pumps protons from the cytoplasm to the interior of the vesicle. This produces a pH of 5.5 inside compared to 7 outside as well as a potential difference (+ve inside). The electrochemical gradient drives DCA inside as H+ goes out. It is very efficient and can maintain a 135K to 1 gradient of CA. The same transporter molecule is found in NA, DA and 5HT neurons.
 

The drug reserpine binds to the vesicular uptake transporter and displaces CA from its storage sites. The displaced CAs leak into the cell cytoplasm where they are destroyed by the degradative enzyme monoamine oxidase. The result is profound depletion of CA, and failure of CAergic transmission. Reserpine is the active ingredient of a herbal tranquillizer from Rauwolfia serpentina discovered by ancient Indian physicians. Another group of catecholaminergic drugs is the sympathomimetics which incudes amphetamine and methylphenidate (RITALIN). Amphetamine is a relative of ephedrine (used in cold preps) which comes from a medicinal plant identified in ancient China. Methylphenidate is a synthetic substitute. These drugs increase the release of CAs but by different mechanisms

- amphetamine displaces CA from the small newly synthesized pool while methylphenidate, displaces the larger bound pool.

The mechanism of CA releasing by amphetamine is still a matter of debate.

The release has a number of properties

1. Release occurs from the cytoplasm (newly synthesized0

2. Release is not blocked by tetorodotoxin i.e not CA dependent - in this respect it resembles spontaneous (i.e unstimulated release)

3. Release is dependent on the CA transporter - blocked by inhibitors of the re-uptake pump Amphetamine is also a weak uptake inhibitor. The current view is that the primary mechanisms of release is exocytosis, and this is consistent with the fact that DA-beta- OHase is released along with NA. It is thought that release also occurs by reversal of the memebrane transporter for DA and NA
 

Inactivation

Once CAs are released the primary mechanism of inactivation is reuptake by a specific saturable, energy dependent transporter in the pre-synaptic membrane (re-uptake pump).

The transporter

The re-uptake pump or CA transporters belong to a family of memebrane proteins that include transporters for all monamines, glutamate and GABA. They are glycoproteins and possess 12 transmembrane domaines. NA and DA cells express only the gene for the particular transporter but the transporters are not very specific. The energy for transport is provided by the Na+ gradient.
 

CA that escapes re-uptake can be deaminated by Catechol-o-methyl-transferase (COMT). There are inhibitors of COMP but they are not very important psychopharmacologically because reuptake accounts for the greater part of CA anactivation. The reuptake pumps are slightly different for different catecholamines and can be inhibited by a number of drugs, some which are relatively specific for one catecholamine. There are many of these drugs because they were discovered as anti-depressants and drug companies have synthesized many variants. Some like protriptyline or desipramine are quite selective for noradrenaline, and have no effect on dopamine uptake while nomifensine is selective for dopamaine. Others like cocaine block noradrenaline and dopamine uptake. The release produced by amphetamine is not calcium dependent indicating that it does not facilitate exocytosis. It is now thought that amphetamine may work by making the CA uptake pump run backwards.

Another group of drugs that can increase CA transmission is the MAOIs which are also antidepressants (e.g pargyline, isoniazid). There are two forms ofMAO, MAO/A which prefers NA and 5HT, inhibited by clorgyline; MAO/B relatively non-specific and accepts DA, inactivated by deprenyl). These drugs reduce the production of de-aminated metabolites and can be dangerous in conjunction with drugs like reserpine, the uptake inhibitors or sympathomimetics. They have been less used in treatment of depression because of fears of adverse reactions caused by tyramine, which is a sympathomimetic found in a number of foods such as mature cheeses and red wine and is normally broken down by MAO.

After inhibition of MAO, CA levels in the neuron do not increase very much because tyrosine hydroxylase activity is regulated by two several mechanisms. Firstly high levels of CAs in the neuron compete at the site that binds the (pteridine) cofactor and reduces the rate of synthesis by what is called end product inhibition. Secondly, the production of tyrosine hydroxlase itself is not fixed but can increase when there is prolonged demand.

The enzyme can also be activated by being phosphorylated by several protein kinases associated with second messenger systems and/or Calcium influx.

Catecholamine receptors

The concept of subtypes of receptors was first proposed in regard to catecholamine receptors, when Ahlquist (1948) suggested that adrenergic receptors were of two varieties (alpha and beta) on the basis of comparisons of the relative potencies of a series of agonists.

Adrenergic receptors

There are now thought to be two groups of adrenergic receptors and two families of dopamine receptors, all of which seem to operate via second messengers. The alpha and beta receptors of Ahlquist have now been divided into subtypes. Beta receptors are linked to stimulation of adenylyl cyclase. B1 are dense in heart and cerebral cortex, B2 in lung and cerebellum, B3 are associated with fat cells and non-shivering thermogenesis. Beta receptors are stimulated by adrenaline and isoproterenol and blocked by propanalol. Since relatively little is known about the role of Beta receptors in the CNS we will not discuss their pharmacology further.

Alpha receptors are divided into A1 (found postsynaptically in blood vessels) and A2 which are located on presynaptic nerve terminals. A2 receptors inhibit adenylyl cyclase and their major action is to inhibit the release of noradrenaline. They are the archetypical autoreceptor and the concept of autoreceptor was developed to explain the effects of a number of drugs acting on this receptor. The classic A2 agonist is clonidine (used in treatment of hypertension) and there are a number of closely related drugs used to reduce nasal congestion(e.g. oxymetazoline)

A specific antagonist at this receptor is yohimbine.
 

Dopaminergic receptors. The dopamine receptors have been divided into families on the basis of information about their protein structure but the pharmacological implications of this division are still being worked out.

The D1 receptor family (D1 and D5) stimulates adenylyl cyclase. The D1 receptors are present on intrinsic neurons in the striatum. Little is known about the D5 receptor. A specific agonist at D1 receptors is SKF-38393 and a specific antagonist is SCH-23390.

The D2 receptors are the more studied and they have been identified by molecular biological techniques. There appear to be two varieties of D2 (A and B) and the D3 and D4 receptors also belong to this family. The D2 family of receptors are linked to inhibition of adenylyl cyclase. They are stimulated by drugs like apomorphine (non-selective) and quinpirole (selective). They are blocked by a large number of antipsychotic drugs.

Haloperidol, pimozide and raclopride are examples of selective antagonists. Alpha- flupenthixol is an often used non-selective antagonist at D1 and D2 receptors. The pharmacology of the other subtypes is still uncertain but the atypical neurleptic clozapine
is a antagonist at D4 receptors.
 

CA and Behavior
 

There is abundant evidence that CA systems play an extremely important role in behaviour. In fact with total removal of the CA systems, voluntary behaviour is completely suppressed. The question that is still not completely answered, is, what are the roles of
the various CA pathways in the catastrophic loss of behaviour produced by CA depletion?
 

The importance of CA in behavior was first exposed by the effects of the drug reserpine which produces a severe loss of brain CA after a single dose. Carlsson et al 1957 observed that reserpine leads to behavioral depression which involves
akinesia - lack of movement hunched posture unresponsiveness to stimulation sleeplessness as shown by EEG
 

Normal behavior is restored by the CA precursor L-dopa but not by 5HTP (5HT precursor), showing that CA are involved. Further proof of this conclusion was provided by Rech et al 1966 who showed that similar depression of behavior could be produced by AMPT which blocks synthesis of CA, but not 5HT.

Other treatments which increase the output of CA also increase activity.

Amphetamine increases locomotion and exploratory behaviour over a wide range of doses, and in all species tested. The effects of amphetamine are blocked by drugs that block the synthesis of CA, especially AMPT (Weissman, Koe and Tenen 1965).
 

Stereotypy
 

Hyperactivity is not the only behavior induced by raised CA activity. If an animal is treated with relatively large doses of amphetamine or apomorphine a peculiar syndrome results. The animal repeats some small behaviour pattern over and over again. The behaviour observed depends on the species. Randrup & Monkvaad (1972).
 

Rats - sniffing on the floor, gnawing, biting floor, movements of head and forelegs

Cats - move head side to side

Dogs - run in circles or back and forth along a fixed route

Man - repetitive tasks, tidying up, sorting handbag, washing car
 

These behaviors are called stereotyped because of their fixed, repetitive nature.
 

The behaviors are almost certainly due to raised DA activity because they are induced by DA agonists such as apomorphine and blocked by DA antagonists.
 

Locus of Actions

Given that all these effects can be observed the question arises as to whether they depend on some general DA mechanism or represent the action of different DA systems.

Kelly, Seviour & Iversen (1975) made small lesions with the specific neurotoxin 6- OHDA, in either the NAS or CN. 6-OHDA is taken up by the specific uptake mechanism of catecholamine cells and decomposed to release hydrogen peroxide which kills the CA cell but does not damage other cells which lack the CA uptake mechanism. Kelly et al's results showed an interesting differentiation between the nucleus accumbens (NAS or ventral striatum) and the striatum or caudate-putamen (CP).

Location Behaviour Behaviour of Lesion Blocked Facilitated accumbens amphetamine locomotion Apomorphine locomotion
striatum amphetamine stereotypy Apomorphine stereotypy
 

You will recall that amphetamine is an indirect CA stimulant (DA releaser) and, therefore, a lesion of CA cells would be expected to abolish the effect of amphetamine. This is what happened, except that different effects were abolished when the lesions were
in the NAS or the CP. Locomotor stimulation was blocked by lesions of the NAS, while stereotypy was blocked by the CN lesions. The opposite effect was seen when the rats were treated with the direct agonist, apomorphine. NAS lesions potentiated locomotion while CN lesions potentiated steoroypy. These latter effects are due to denervation supersensitivity. When tissue is denervated there is frequently a proliferation of receptors so that the transmitter becomes more potent in its effect. In this case apomorphine was observed to produce effects at doses much lower than those required in the intact animal.

Moreover, apomorphine produced those effects of amphetamine that were blocked by the 6-OHDA lesions.
 

These early experiments illustrate the profound influence of dopamine systems on behaviour but they give little clue as to the role played by DA in the organization of behaviour. In order to devolp more understanding of current views on the role of dopamine
we need to look in more detail at the neural systems of the basal ganglia (see fig. Below).

The basal ganglia comprise the striatum (caudate-putamen in humans) and the globus pallidus, which are are large masses of grey matter at the core of the cerebral hemispheres. The striatum receives excitatory input from the sensorimotor and prefrontal cortex, and from the thalamus. It receives a modulatory dopamine input from the SNc. Its output (GABA plus peptides) is directed to the GP and to the SNr which projects to the thalamus, brainstem and tectum. The GP receives additional input from the subthalamic
nucleus and SNc and projects back to the thalamus, subthalamus and SNr. The system is usually viewed as a hierarchical series of feedback loops, but this makes its operation very difficult to understand unless one can distinguish the output of the system from
feedback pathways. There is considerable evidence that the effective output is the cells of the SNr and their projections to the superior colliculus, peduculopontine region, and the thalmus. The evidence is, in summary, that manipulations of the nigrotectal/pontine pathway duplicates the effects of major disruption of other parts of the circuit and preempts the effects of manipulations of other parts of the circuit.

(figure 20.7 Feldman, Meyer and Quenzer)

This is illustrated in experiments examining the circing elicited by unilateral damage to the basal ganglia. In the early 70's Urban Ungersedt observed that after unilateral damage to the nigrostriatal DA pathway animals would bend to one side and walk in circles if treated with DA agonists. The circles were ipsiversive (towards the lesion) with amphetamine and contraversive with apomorphine. To understand the reasoning you need to remember to remember the effects of amphetamine and apomorphine. On the DA denervated side amphetamine would be ineffective, since it releases endogenous DA, while apomorphine would be extra effective because DA receptors were supersensitive. It follows that the animals appear to be orienting themselves towards the side opposite the most active dopamine system (note that this is the side of the body controlled by the hemisphere with the more active DA system).

Injection of a GABA (inhibitory) agonist into the SNr produces rapid contraversive circling while an antagonist produced ipsiversive circling. Cutting off the inputs from the forebrain does not prevent the effect of GABA agonist in the SNr. Since an inhibitory substance in the SNr releases behaviour controlled by the same hemisphere we can deduce that the SNr normally inhibits the tectal-pontine motor mechanisms and that DA's nett effect is to inhibit the SNr. This idea was confirmed by elegant experiments in the monkey (Hikosaka and Wurtz) in which it was shown that eye movements controlled by the superior colliculus
were associated with pauses in the tonic activity of SNr neurons. The data from the circling model are also generally consistent with clinical observations of the effects of lesions of the basal ganglia The SNr is normally excited by the input from the subthalamic nucleus.

The effect of DA release in the striatum is stimulate the direct inhibitory pathway from the striatum to the SNr and through the indirect pathway via the globus pallidus to inhibit the subthalamic nucleus. Inhibition of the subthalamic nucleus removes the excitatory drive on SNr neurons.

Parkinson's Disease
 

The two classic basal ganglia diseases are Parkinson's disease and Huntington's chorea.

Parkinson's disease is named after the physician who first described it. The clinical symptoms are bradykinesia (slowness of movement), rigidity and tremor. In the early stages of the disorder the individual has difficulty in modulating force and fluidity of
movement, so that before the patient becomes obviously impaired in movement he may show a reduced mobility of the facial muscles, or cease to swing his arms in time with walking. As the disease progresses the loss of mobility becomes increasingly severe until the patient is eventually bedridden.

The most debilitating symptoms of the disease - the rigidity and bradykinesia - appear to be a consequence of degeneration of the catecholamine neurons of the brain, most particularly the DA cells of the substantia nigra. Histologically the cells of the SN pars
compacta and locus coeruleus disappear while the striatum looks normal. Biochemical measurements show that there is severe depletion of dopamine and noradrenaline and the enzymes that synthesize them. The degeneration of DA neurons presumably takes many years, but at the onset of symptoms the depletion of DA is already near 80%. Other transmitters are also depleted but to a lesser degree. The similarity between the reserpine syndrome and PD suggested that the depletion of dopamine might be the cause of the PD symptoms. There are two lines of pharmacological evidence that the loss of DA is the cause of PD symptoms.
 

1. Akinesia and rigidity are markedly relieved by treatment with the CA precursor L-Dopa and also by the direct DA agonists.
 

2. Parkinsonian symptoms are induced when non-PD persons are treated with the DA blocking anti-psychotic drugs such as chlorpromazine or haloperidol.
 

PD is a disease of old age, it is rarely found in persons under 50 and the prevalence increases with age after 50. In 1982 neurologists at a Ca. hospital were surprised to receive several young patients with what appeared to be classic PD. The only thing they had in common was that they were all drug addicts and had used what they believed to be "synthetic heroin" a few days before the onset of the neurological symptoms. The neurologists succeeded in finding a toxic constituent and identifying it as 1-methyl-4- phenyl-1,2,3,6-tetrahydropyridine or MPTP. These patients responded to L-Dopa though, as in older patients, the response diminished with time and side effects increased.

Monkeys proved to be sensitive to this neurotoxin but rodents showed temporary depletion without cell damage. Research on this problem has yielded some interesting clues to the possible cause of PD. It turns out that MPTP is not the damaging agent. Rather MPTP is converted to a charged metabolite called 1-methyl-4-phenylpyridium ion (MPP+) which is taken up by the CA neurons. The enzyme involved is MAO-B which is blocked by deprenyl, selective MAO-B inhibitor.
 

(Fig 20.12 Feldeman, Meyer and Quenzer)
 

The other common form of the enzyme MAO-A is blocked by clorgyline. The neurotoxic effect of MPTP is blocked by deprenyl but not by clorgyline.
 

An additional complication is that not all CA cells are equally afffected. The SNC is destroyed while the VTA and LC suffer relatively little damage. This observation confirms that the PD symptoms are due primarily to loss of the nigrostriatal system. However it raises the question of why the damage is selective in this way. It is suggested that the answer lies in the presence of large amounts of neuromelanin in the SNC. This substance is a dark pigment which gives the SN its name. It is a derivative of tyrosine and can be formed from dopamine. In 1986 it was shown that neuromelanin binds with high affinity to MPP+ and may trap it in the SN cells. Monkeys and humans have high levels of neuromelanin in the substantia nigra while rodents have little or no neuromelanin in these nuclei. Thus primates may be particularly susceptible to this toxin.
 

The MPTP story has led to a the suggestion that an MPTP-like substance may be the cause of natural Parkinsonism. Twin studies have shown that there is poor concordance for PD in identical twins. Barbeau, in Quebec, has presented evidence that the incidence of PD is higher in areas that have high pesticide use or have petrochemical plants. Both these lines of evidence point to an environmental influence in PD. The idea has received some support from clinical studies of the use of deprenyl in early stage PD. The researchers measured the progress of PD in several groups of patients that were randomly assigned to deprenyl or no deprenyl treatment crossed with alpha-tocopherol or no alpha- tocopherol (antioxidant). They determined the progress of the disease by the timing of the neurologist's decision to begin L-DOPA treatment.

(Fig 20.20 Feldman, Meyer and Quenzer)

Fig 20.20 below shows the results of the study ignoring the effects of alpha-tocopherol. The effect of deprenyl was a dramatic reduction in the probability of reaching the end point by a given date i.e. the rate of progress of the disease seemed to be slowed down. This evidence remains controversial. Some commentators believe that deprenyl slows the disease progression while others claim that it simply acts as an antiparkinson drug like L- DOPA. Good histopathological data on brains of idividuals with long-term deprenyl treatment has not be reported yet.

It has also been suggested that deprenyl may be beneficial but that it may not work via MAOB inhibition.
 

Sources
 

1. D'Amato, R.J., Lipman, Z.P. and Snyder, S.H. Selectivity of Parkinsonian neurotoxin MPTP: Toxic metabolite MPP+ binds to neuromelanin. Science 231:987-989, 1986.
 

2. Di Chiara, G., Porceddu, M.L., Morelli, M., Mulas, M.L. and Gessa, G.L. Substantia nigra as an output station for striatal dopaminergic responses: role of a GABA-mediated inhibition of pars reticulata neurons. Naunyn-Schmied.Arch.Pharmacol. 306:153-159, 1979.
 

3. Feldman, Meyer and Quenzer, Principles of neuropsychopharmacology. Sinauer, 1996 4. Gerfen, C.R. The neostriatal mosaic: multiple levels of compartmental organization. Trends in Neurosciences 15:133-138, 1992.
 

5. Langston, J.W., Irwin, I. and Langston, E.B. Pargyline prevents MPTP-induced Parkinsonism in primates. Science 225:1480-1482, 1984.
 

Dopamine and reinforcement

Three and a half decades ago Olds and Milner (1954) discovered that rats would work for electrical stimulation of their own brains. From the fact that animals will work for stimulation at some brain areas but not others they postulated that the brain contains specialized neural circuitry for the mediation of reward, and that this circuitry can be anatomically mapped using brain stimulation techniques.

Today the 4 main paradigms used to explore the neurochemical basis of reinforcement in animals are:- 1) ICSS, 2) Self-administration, 3) Administration of natural reinforcers such as food and water, 4) Conditioned place preference (CPP). The evidence that DA is involved in reinforcement is superficially quite straightforward but there some difficult problems in interpreting it. Simply put the evidence is

1. There is considerable anatomical correspondence between the ICSS system and the distribution of DA axons and terminals.

2. Depletion of DA, but not NA, blocks ICSS

3. Drugs which release DA such as amphetamine or cocaine facilitate ICSS.

4. A number of drugs which release DA or stimulate DA receptors are reinforcers in self- administration paradigms e.g. amphetamine, cocaine, apomorphine.

5. DA antagonists attenuate the reinforcing effect of food and water.

6. All of the DA agonists examined have been shown to produce CPP.

The problems associated with statements 2 - 6 is that , to varying degrees, all of the behavioral paradigms used to assess reward are partially confounded by the fact that DA may also mediate locomotion and the formation of associations (learning). In other words, it is often the case that it is difficult to determine whether a DA agonist or antagonist is increasing or reducing behavioral expression of reward by directly inhibiting reward itself, or instead, inhibiting performance capability (locomotion), or the association formation (learning), required for the expression of reward.

1. ICSS can be elicited from many regions of the brain that contain DA cells or terminals and highly effective sites are found along the course of the MFB in which run the axons of the nigrostriatal and mesolimbic pathways. However the association of DA and ICSS is rather weak evidence because there are many sites where ICSS is found and DA neurons are not. The one clear cut correspondence is the cortex innervated by the mesocortical DA system, this is the only cortical area that does support ICSS.
 

2. Blocking DA transmission unequivocally interferes with ICSS and other operant behaviours. One of the first studies was that of Breese et al. who showed that ICSS was blocked by 6-OHDA lesions that depleted brain DA whereas lesions that reduced NA by 98% had little effect. Similarly ICSS is severely suppressed by DA receptor blocking drugs such as chlorpromazine, spiroperidol, haloperidol and pimozide. However the same treatments block operant responding for food and water reward and disrupt avoidance behaviour, though they have less effect on escape.

These observations leave open the possibility that the effect of DA blockade on ICSS may be partially attributable to performance, and/or association formation, deficits.

3. What can be concluded from the data above is that DA is somehow important for the occurrence of ICSS and other motivated behaviours but they do not show that DA is involved in the process of reinforcement itself. In fact it turns out to be very difficult to evise a satisfactory test of the DA hypothesis. In essence the DA hypothesis states that the experience of reward or reinforcement involves a signal being passed through a DA napse at some stage of the process. In other words, release of DA somewhere in the brain is a critical event in reward. One prediction of this theory is that release of DA, or equivalent DA receptor activity, should be rewarding. This prediction is fairly well confirmed. When the presentation of a DA agonist is made contingent on a response, e.g. lever pressing, the rate of that response is increased. The effectiveness of DA agonists as reinforcers has been confirmed for amphetamine and cocaine (Pickens and Harris, 1968) and for apomorphine (Baxter et al 1974). As might be expected these effects are abolished by large doses of DA blocking agents. More recently it has been shown that the mesolimbic dopamine projection is involved in the rewarding effect of amphetamine and cocaine. 6-OHDA lesions of the NAS blocked self-administration of amphetamine and cocaine (Lyness et al.)

4. The performance of self-administration is not the same as other operant behaviour because of the peculiar nature of IV drug administration. The level of the self-administered drug in the blood is regulated so that the rate of response is inversely related to the unit dose of the drug. If the drug is diluted the rate of SA goes up. This provides a unique opportunity to test the effect of DA antagonists on self-administration. If the level of DA activity is being regulated, the addition of a competitive DA antagonists should cause the rate of SA to rise. This excludes the confounding variable of antagonist-induced performance deficits. Yokel and Wise (1975) showed that this was indeed the case. Very recently Lepore and Franklin (1992) have shown that a similar result is found if animals are trained to administer electrical brain stimulation is a manner that mimics drug self- administration. In this experiment animals lever press to increase the stimulation pulse frequency of a continuous train of pulses that is automatically decreasing in frequency (Note: it is known that the reinforcing effect increases as frequency increases). Thus by responding at regular intervals an animal can maintain the frequency in any range it desires. In this test the response rate is inversely related to the "dose" (i.e. the size of the increase in stimulation frequency used as a reinforcement), and to the elimination 1/2 time
(i.e. the more slowly the drug is eliminated, the more slowly the animal responds). When animals performing this task are given a DA antagonist they increase their response rate and raise the maintained frequency. Since the drug produces increases in response rate the effect cannot be explained by some kind of performance deficit. In contrast the DA antagonists depress responding for conventional self-stimulation in which the reinforcement is a brief train of stimulation pulses.

5. Wise et al (1978) concluded that DA blockade attenuates the rewarding impact of a stimulus which normally sustains operant responding for food and water reinforcement. With food-rewarded responding, Wise et al found that they could train animals sufficiently so that they showed no deficits on the first day of DA antagonist treatment suggesting that this dose of the antagonist did not produce performance deficits. However, there was no evidence of a reward deficit on the first day either. Clear cut evidence that the drug attenuated the rewarding value of food appeared on subsequent days of testing. Over successive days of testing the amount of responding progressively decreased, so that the drug-treated animals, like non-rewarded animals, learned to respond less and less with experience in the test condition. It was as if the DA antagonist was depriving the food of the rewarding value of its taste.

Further studies, using home cage controls, indicated that it was not the drug (the pharmacological experience) that was producing the progressive decrease in responding, but it was rather a consequence of the drug experience in the context of normal reward in
the operant task (the psychopharmacological experience) which is important. If animals have a series of 4 injections of DA antagonist in the test situation, they develop a progressive unwillingness to respond, however, if they receive the same set of injections in a different environmental context (the home cage), they respond, when given the pportunity, as if they had not had any prior drug experience. It seems that the animals learn not to expect reinforcement from food under DA antagonist conditions.

6. All of the DA agonists tested (including the D1 selective agonist SKF 38393 if it is directly injected in the ventral striatum) have been reported to produce CPP. DA antagonists do not produce significant preferences or aversions. The CPP's produced by
most of the DA agonists (cocaine, methylphenidate, bupropion, nomifensine) are attenuated by DA antagonists.

The tentative conclusion can be drawn that DA release is a sufficient condition for reinforcement i.e. release of DA in the brain meets the essential requirements for it to be called a reinforcer. The question still to be answered is whether DA release is a necessary condition for reward. The problem here is that we can only detect the presence of reward by changes in behavioural output (recall the definition) but drugs and procedures used to alter DA function have direct effects on behavioural output. The problem is to find changes in behavioural response that can only be ascribed to changes in reward and not changes in performance, or learning, capability. One approach has been to look at the pattern of behaviour when an animal is performing under the influence of a DA blocking drug. When the reinforcer is omitted extinction occurs and this behavioural change has a characteristic form. If DA block really blocks the rewarding effect of brain stimulation we would expect extinction of the response to occur. Fouriezos and Wise showed that indeed extinction-like declines in responding did occur and were followed by spontaneous recovery. However it can be argued that the motor effects of DA block cannot be distinguished from extinction. The PD patient appears to be excessively fatiguable and a common observation is that PD patients begin an action which gradually dies away. Thus an extinction-like decline in responding could be produced by PD.
 

Another approach is to look for properties of reinforcers that can be tested in the drug free state. One of these is the partial reinforcement effect. When an animal is given a proportion of reward-free trials during training, resistance to extinction is increased. Ettenberg used this phenomenon to show that DA blockers interfere with reward. Rats were trained to run a straight alley for food or water one trial per day. One group got no special treatment (continuous reinforcement group). Two other groups were treated so that on alternate days the food or water reward was omitted (partial reinforcement group) or rats were treated with haloperidol and given the food or water which they consumed normally (DA blocker group). When they were later tested in extinction the haloperidol- treated rats were more resistant to extinction than the continuous reinforcement group i.e. a partial reinforcement effect was observed, though it was not as strong as the true partial reinforcement effect.

Part 2

Given that dopamine is involved in reinforcement, which dopamine system is involved? Evidence suggests that it is the VTA projection to the accumbens that is involved in reinforcement. Animals will learn to press a lever for direct injections of dopamine or amphetamine into the nucleus accumbens. Curiously, they will not lever press for direct injections of cocaine. However cocaine is a local anaesthetic and it is thought that it may block transmission of the dopamine sensitive cells and thus prevent the reinforcing signal from being transmitted out of the nucleus accumbens.

If the accumbens receptors are important for the reinforcing effect of brain stimulation and drugs were would expect that blocking these receptors would block reinforcement. Several studies have shown that microinjections of dopamine blocking drugs into the accumbens attenuate the reinforcing effects of brain stimulation, cocaine and amphetamine. A particularly interesting experiment is one which compared the effect of cocaine and apomorphine after lesions of the dopamine neurons. If you inject the
neurotoxin, 6-hydroxydopamine, into the nucleus accumbens it destroys the dopamine terminals in the accumbens without affecting other dopamine terminals in the brain. Since cocaine acts by releasing dopamine from the terminals it should be ineffective as a reinforcer after 6-hydroxydopamine. This is in fact what happens when the accumbens terminals are destroyed even though cocaine continues to release dopamine in other parts of the brain. On the other hand apomorphine acts directly on the receptors.

When dopamine terminals are destroyed the post synaptic neurons produce more receptors, apparently to compensate for the lack of normal postsynaptic activity. This makes the cells supersensitive to dopamine, i.e. they respond to much lower concentrations of dopamine than normal. It would be predicted that if the reinforcement system was supersensitive to dopamine the reinforcing effect of direct agonists like apomorphine should be exaggerated and Roberts (1989) showed that this was indeed the case.

Another interesting technique to study the psychological effects of drugs is the drug discrimination technique. The animal is trained to lever press for food reward and has 2 levers that it can press. Presses on one lever are reinforced when the animal is given a drug (e.g cocaine) and presses on the other lever are reinforced when the animal is given saline. The animal learns to use the sensations evoked by the drug as cues to which lever will be reinforced. On the test day the animal is given another drug (e.g.
amphetamine) and the test is which lever does the animal press. If the animal presses the "drug" lever we infer that the new drug feels more like cocaine than saline, or vice versa.

Wood and Emmett-Oglesby (1989) used this procedure but instead on given a different drug in the test they injected cocaine into one of the 3 major dopamine terminal regions, the prefrontal cortex, striatum or nucleus accumbens. The rats pressed the cocaine lever only when cocaine was injected into the accumbens.
 

The theory that brain stimulation and drugs are reinforcing because they activate the neural system that is involved in the reinforcing effects of natural rewards implies that this system should be active when the animal is receiving reinforcement. More specifically it implies that the dopamine system should be active when reinforcement occurs. Recently it has become possible to test this theory directly by measuring the release of dopamine in the brain when reinforcers are presented. Dopamine release can be measured by microdialysis and volammetry. In microdialysis a minute tube made of semipermeable membrane is implanted in the brain so that substances like dopamine can diffuse into a fluid passing through the tube. This fluid is collected and analyzed by liquid chromatography. In voltammetry a specially treated electrode, that makes it possible to measure the minute current produced by the oxidation of dopamine at the electrode surface, is implanted in the brain.

Phillips et al 1989 and Fiorino et al, 1993 showed, using voltammetry and in vivo dialysis, that when a rat pressed a lever for reinforcing brain stimulation, the stimulation increased the release of dopamine in the accumbens.

(Figure from Phillips et al 1989)

Natural reinforcers such as water or food also induce dopamine release. It is also possibleto show that dopamine release depends on the reinforcing nature of the natural stimulus rather than on its chemical composition. The sweet substance saccharine normally
increases dopamine release and is an effective reinforcer. However if the taste ofsaccharine is paired with an aversive stimulus such as lithium chloride, which makes an animal sick, saccharine can become an aversive stimulus. This treatment also alters the
response to saccharine so that it inhibits dopamine release instead of increasing it. Morerecent studies using faster voltammetric methods (Richardson & Gratton, 1996) show that here is a surge of dopamine release when a reinforcer is first presented but after a few presentations dopamine release begins to occur just before the response that producesthe reinforcer, and dopamine output declines when the reinforcer is presented. These data suggest that dopamine release is correlated with the appearance of a novel stimulus and with the execution of a conditioned response, rather than with reinforcing events.
 

What is stimulated by brain stimulation?

The evidence that dopamine is involved in reinforcement suggests that brain stimulation must directly activate dopamine neurons and researchers struggled for many years to try to demonstrate this. The problem was to isolate the neurons that are responsible for reinforcement in order to identify them. This can be done by making use of the properties of axonal conduction. you will recall that the axon potential is a wave of depolarization that is carried along the axon. For brief period after depolarisation the axon
is not stimulable while polarisation is re-established. This is the refractory period. The speed of conduction and the length of the refractory period are related to the diameter of the fibre. Large fibres conduct faster and have a shorter refractory period. Since
dopamine neurons are small unmyelinated fibre they conduct very slowly. If we could establish the conduction velocity of the reinforcing neurons we could see if it was compatible with the speed of dopamine neurons.

This difficult task has been accomplished by John Yeoman's group at U of T. and Peter Shizgal's group at Concordia.Yeomans' method depends on the idea that the more stimulation pulses get through to the terminals the bigger the reinforcement. If we give a self-stimulating animal two stimulation pulses rather than 1, the reinforcing effect of the stimulation should be increased. However if we move the second pulse closer and closer to the first pulse eventually the second pulse will fall in the refractory period of the first pulse and will not be transmitted. The reinforcing effect will therefore diminish when the inter-pulse interval falls within the refractory period of the neuron. In fact we use trains of paired pulses rather than single pairs but the logic is the same. By this method Yeomans found that the refractory period for MFB neurons involved in reinforcement was under 1.2 msec which is quite short, implying that the neurons are fairly large myelinated fibres. Using a different method, Shizgal has also used a "collision" method to reach similar conclusions. In this method two action potentials are started at slightly different times in different parts of the axon and they travel in both directions away from the points of excitation. If the timing is right the forward conducting action potential from one electrode collides with the backward conducting pulse from the other electrode and they are anihilated. With longer intervals between the pulses both forward conducting pulses get through to the nerve terminals. If the distance between
the electrodes is known the conduction velocity can be calculated. Shizgal estimated a conduction velocities of 2-8 m/sec, which is also quite fast and agrees with Yeomans' estimate of the type of fibre. Thus the neurons stimulated by the self-stimulation electrode cannot be dopamine neurons.

Given that the primary neurons are not dopaminergic, what are they? One useful piece of information would be which way do they run in the MFB. Bielajew and Shizgal (1986) have provided evidence that the cell bodies are in the forebrain and the axons travel
caudally to the midbrain in the region of the VTA. They placed 2 electrodes in the MFB, one at the rostral and one at the caudal end. Through one electrode they passed a depolarising current and through the other a hyperpolarising current.

(Carlson fig 16.19)

The animal could turn on the current by pressing a lever. When the depolarising current was caudal the animals pressed faster than when the depolarising current was rostral. That is, when the action potential had to pass through a hyperpolarised region of the axon, the reinforcing effect of stimulation was attenuated. It can be inferred from this that the neurons conduct from rostral to caudal along the MFB. Though it has not been shown directly it is thought that the primary neurons must connect with the dopamine neurons and activate them indirectly.
 

One important implication of the research on reinforcement has been to increase our understanding of drug addiction. It has been shown that the majority of drugs that have high abuse potential are direct or indirect stimulants of the dopamine system or its
connections. Thus amphetamine, cocaine and nicotine all increase the release of dopamine. Opioids also stimulate dopamine release and, in addition, may act directly on the cells of the nucleus accumbens. We now believe that drugs are addictive because
they activate the reinforcement system and are thus directly reinforcing.
 
 

Bibliography

1 Feldman, R.S., Meyer, J.S. and Quenzer, L.F. Principles of neuropsychopharmacology, Sinauer, Sunderland, Mass. 1997,

 Reprint : In File,

2 Lepore, M. and Franklin, K.B.J. Modelling drug kinetics with brain stimulation: dopamine antagonists increase self-stimulation, Pharmacol. Biochem. Behav. 41 (1992) 489-496.

Reprint : In File,
 

3 Wise, R.A. The role of reward pathways in the development of drug dependence, Pharmacol. Ther. 35 (1987) 227-263.

Reprint : In File,
 

4 Yokel, R.A. and Wise, R.A. Attenuation of intravenous amphetamine reinforcement by central dopamine blockade in rats. Psychopharmacology, 48 (1976) 311-318.

Reprint : Not in File,