Dopaminergic function

Studies on platelets, lymphocytes, changes in cerebrospinal fluid metabolites of brain monoamines and post-mortem studies suggest that a major abnormality in both noradrenergic and serotonergic function occurs in depression, and that such changes could be causally related to the disease process.

Less attention has been paid to the possible involvement of dopamine in this disorder. However, anhedonia is a characteristic feature of major depression, and a defect in dopaminergic function is thought to be causally involved in this symptom (Reference WillnerWillner, 1983). The concentration of the main dopamine metabolite, homovanillic acid (HVA), is decreased in the CSF of patients with depression, particularly those with psychomotor retardation. Patients with depression who attempted suicide were also found to have a decreased urinary excretion of HVA and the second major dopamine metabolite, dihydroxyphenylacetic acid (DOPAC). The dopamine deficit may be a reflection of the degree of retardation rather than of the psychological state, as similar changes in CSF HVA concentrations have been reported to occur in patients with Parkinson's disease. Changes in the basal ganglia may overshadow any changes in the mesolimbic dopaminergic system, as the contribution of this area is relatively minor.

Bupropion (not marketed in Europe but available in North America) amineptine and nomifensin (withdrawn because of the rare occurrence of haemolytic anaemia) owed their antidepressant efficacy to their ability to increase central dopaminergic function. There are also open-label studies to suggest that the novel dopamine receptor agonist roxindole and the selective dopamine uptake inhibitor pramipexole may have antidepressant action.

Why is there a delay in onset of the antidepressant response?

Delay in onset of the therapeutic effect of antidepressants, which clearly is unrelated to the acute actions of these drugs on monoamine reuptake transporters or monoamine oxidase, is probably related to the adaptive changes that occur in the pre- and post-synaptic receptors that govern neurotransmitter release.

Antidepressant action usually has a gradual onset of over 2–3 weeks before the optimal effect is obtained (World Health Organization Mental Health Collaborating Centres, 1989). Much of the early improvement is probably associated with a reduction in anxiety and improvement in sleep caused by the sedative action of the drugs. Delay in onset of the therapeutic response cannot be easily explained by the pharmacokinetic profile of the drugs, as peak plasma (and presumably brain) concentrations are usually reached in 7–10 days. Furthermore, the 2- to 3-week delay is also seen in many patients given electroconvulsive therapy (ECT). Box 4 summarises some of the changes in neurotransmitter receptors that occur in the cortex of rat brain following ECT or the chronic administration of antidepressants. Adaptational changes occur in adrenoceptors, serotonin, dopamine and type B γ-aminobutyric acid (GABA-B) receptors (Reference LeonardLeonard, 1994a ). GABA-B receptors play a role in enhancing noradrenaline release in the cortex and in this respect differ fundamentally from the inhibitory GABA-A receptors that facilitate central GABAergic transmission. A decrease in the activity of GABA-B receptors may therefore contribute to the reduced central noradrenergic tone in depression.

Changes in cholinergic function

In the bulbectomised rat model of depression, a decrease in cortical muscarinic receptors occurs that returns to control values following treatment with antidepressants (Reference Earley, Glennon and LallyEarley et al, 1994).

The anticholinergic activity of the tricyclic antidepressants is usually associated with their unacceptable peripheral side-effects and most second-generation antidepressants lack such side-effects. Possible support for a cholinergic hypothesis of depression is that the cholinesterase inhibitor pyridostigmine, when administered to drug-free patients with depression, causes an enhanced activation of the anterior pituitary glands, as shown by the release of growth hormone (Reference O'Keane, O'Flynn and LuceyO'Keane et al, 1992). This suggests that the muscarinic receptors are supersensitive in patients with depression. Janowsky et al (1986) postulated that depression arises as the result of an imbalance between the central noradrenergic and cholinergic systems; in depression, the activity of the former system is decreased and, conversely, in mania it is increased. As most antidepressants enhance noradrenergic function, it is hypothesised that the reduction in cholinergic activity is a consequence of the increase in noradrenergic activity.

In summary, irrespective of the specificity of the antidepressnts following their acute administration, it can be speculated that a common feature of all of these drugs is to correct the abnormality in neurotransmitter receptor function. Such an effect of chronic antidepressant treatment may parallel the time of onset of the therapeutic response.

Link between the serotonergic and noradrenergic systems

Chronic administration of antidepressants enhances the inhibitory response of forebrain neurons to microiontophoretically applied 5-HT. This enhanced response is blocked by lesions of the noradrenergic projections to the cortex. This dual effect could help to explain enhanced serotonergic function that arises after chronic administration of antidepressants or ECT. Conversely, impairment of serotonergic function by means of selective neurotoxins (e.g. 5,7-dihydroxytryptamine) or a 5-HT synthesis inhibitor (e.g. parachlorophenylalanine) largely prevents the decrease in functional activity of cortical β-adrenoceptors that usually arises following chronic antidepressant treatment. 5-HT_1b receptors are located on serotonergic nerve terminals and act as autoreceptors; stimulation by serotonin decreases the further release of this amine. Chronic administration of selective SSRIs slowly desensitises the inhibitory 5-HT_1b autoreceptors and thereby enhances serotonin release.

Likewise, the 5-HT_1a somatodendritic receptors inhibit the release of serotonin and it is posulated that the enhanced release of the transmitter following chronic administration of SSRIs is a consequence of adaptive down-regulation of the inhibitory 5-HT_1a receptors (Reference Blier, de Montigny and ChaputBlier et al, 1990). The validity of this hypothesis is supported by the pharmacological effect of 5-HT_1a antagonists. Thus, the beta-blocker and 5-HT_1a antagonist pindolol, in combination with fluoxetine or paroxetine, may enhance the therapeutic efficacy of the SSRI and, in some studies, reduce the time to onset of the peak therapeutic effect (Reference Tome, Isaac and HarteTome et al, 1997). However, several investigators have not been able to replicate such findings (see Reference McAskill, Mir and TaylorMcAskill et al, 1998).

Serotonin can also regulate dopamine turnover. A positive correlation exists in patients with depression between the HVA and 5-HIAA concentrations in the CSF (Reference AgrenAgren, 1980). Stimulation of the 5-HT cell bodies in the median raphe causes reduced firing of the substantia nigra, where dopamine is the main neurotransmitter. Serotonin plays an important role in modulating dopaminergic function in many regions of the brain, including the mesolimbic system. The effects of some antidepressants that show an apparent selectivity for the serotonergic system could be ascribed to secondary changes in dopaminergic pathways.

Role of the glutamatergic system in the action of antidepressants

The glutamatergic system is the main excitatory neurotransmitter pathway in the brain. Tricyclic antidepresssants (TCAs) inhibit the binding of dizolcipine to the ion channel of the main glutamate receptor, the NMDA-receptor in the brain (Reference Reynolds and MillerReynolds & Miller, 1988). Newer antidepressants have a qualitatively similar effect (Reference Kitayama, Yaga and KayaharaKitayama et al, 1997). Whether this is due to direct action on glutamate receptor sites or indirect action via the glycine receptor site is uncertain. Glycine and drugs modulating the glycine site show antidepressant-like activity in animal models of depression. Thus, antidepressants may act as functional NMDA receptor antagonists.

Intracellular changes that occur following chronic antidepressant treatment

The coupling of the NDMA-receptor on the cell surface to the intracellular second messenger is brought about by a member of the G-protein family. Beta-adrenoceptors are linked to adenylate cyclase and, depending on the subtype of receptors, 5-HT is linked to either adenylate cyclase (5-HT_1a, 5-HT_1b) or phospholipase (5-HT_2a, 5-HT_2c). Activation of phospholipase results in an intracellular increase in the secondary messengers diacylglycerol and inositol triphosphate (IP3), the IP3 then mobilising intraneuronal calcium.

The net result of the activation of the secondary messenger systems is to increase the activity of the various intracellular protein kinases (termed third messengers) that phosphorylate cell proteins to produce a physiological response. Racagni et al (1991) have investigated the effect of chronic antidepressant treatment on the phosphorylation of proteins associated with the cytoskeletal structure of the nerve cell. Antidepressants could affect the function of the cytoskeleton by changing the component of the associated protein phosphorylation system. Thus, both typical (e.g. desipramine) and atypical antidepressants (e.g. (+)oxaprotiline, a specific noradrenaline reuptake inhibitor, and fluoxetine, an SSRI) increased the synthesis of a microtubule fraction. These changes occurred only after chronic antidepressant treatments. Antidepressants might thereby change neuronal signal transduction processes distal to the receptor.

Glucocorticoid receptors: adaptive changes following antidepressant treatment

Glucocorticoid receptors have been identified in the nuclei of catecholamine and 5-HT-containing cell bodies in the brain. Glucocorticoid receptors activate deoxyribonucleic acid-binding proteins, which can modify the transcription of genes. Chronic administration of imipramine increases glucocorticoid receptors in rat brain, particularly the noradrenergic and serotonergic cell body regions.

Lymphocyte glucocorticoid receptors are subsensitive in patients with depression. The failure of the negative feedback mechanism that regulates the secretion of adrenal glucocorticoids further suggests that the central glucocorticoid receptors are subsensitive. Hypersecretion of cortisol is a characteristic feature of many patients with major depression (Reference DinanDinan, 1994). Central neurotransmission occurring in depression may, in part, result from the effects of chronic glucocorticoids.

Glucorticoid synthesis inhibitors (e.g. metyrapone) may be able to reduce the abnormality in neurotransmitter function by decreasing the cortisol concentration.

Typical antidepressants increase the density of glucocorticoid receptors in rats. Such an effect could increase the negative feedback mechanism and thereby reduce the release of cortisol. In support of this hypothesis, there is preliminary clinical evidence that metyrapone (and the steroid synthesis inhibitor ketoconazole) may have antidepressant effects. Recently, several lipophilic antagonists of corticorophin-releasing factor (CRF) type 1 receptor, which appears to be hyperactive in the brain of patients with depression, have been shown to be active in animal models of depression. Clearly, this is a potentially important area for antidepressant development (Reference Mansbach, Brooks and ChenMansbach et al, 1997).

Glucocorticoid receptors are present in high density in the amygdala, and neuroimaging studies have shown that the amygdala is the only structure in which the regional blood flow and glucose metabolism consistently correlate positively with the severity of depression (Reference Lesser, Mena and BooneLesser et al, 1994). This hypermetabolism appears to reflect an underlying pathological process as it also occurs in asymptomatic patients and in the close relatives of patients.

Effects of antidepressants on endocrine–immune functions

Stress is frequently a trigger factor for depression in vulnerable patients. Corticorophin-releasing factor is elevated in the CSF of untreated patients with depression, which presumably leads to the hypercortisolaemia that usually accompanies the condition. One of the consequences of elevated plasma glucocorticoids is a suppression of some aspects of cellular immunity. Many aspects of both cellular (for example, natural killer cell activity, T-cell replication) and non-cellular (for example, raised acute phase proteins) immunity are abnormal in untreated patients with depression (see Reference Song and LeonardSong & Leonard, 1995). This may contribute to the susceptibility of patients with depression to physical ill health.

A link between CRF, the cytokines that orchestrate many aspects of cellular immunity and the prostaglandins of the E series has been the subject of considerable research in recent years. Prostaglandin E2 (PGE2) concentrations are raised in the plasma of untreated patients with depression and are normalised following effective treatment with TCAs. Raised PGE2 concentrations reflect increased proinflammatory cytokines (particularly interleukins 1 and 6 and tumour necrosis factor,), which occur as a consequence of increased macrophage activity in the blood and brain. In the brain, the microglia function as macrophages and produce such cytokines locally. Thus, the increased synthesis of PGE2 may contribute to the reduction in amine release in the brain that appears to underly the pathology of depression. It has recently been postulated that antidepressants normalise central neurotransmission by reducing brain concentrations of both the cytokines and PGE2 by inhibiting central and peripheral macrophage activity together with cyclooxygenase type 2 activity in the brain. Cyclooxygenase is the key enzyme in the synthesis of the prostaglandins. The usefulness of TCAs in severe rheumatoid arthritis may reflect the inhibitory action of such drugs on cyclooxygenase activity. These effects of TCAs, together with those in glucocorticoid receptor function, may thus normalise defective central neurotransmission.

Antidepressants and changes in neuronal structure

Another possible mechanism whereby antidepressants may change the physical relationship between neurons in the the brain is by inhibiting neurite outgrowth from nerve cells. Amitriptyline inhibits neurite outgrowth from chick embryonic cerebral explants in vivo (Reference Wong, Bruch and FarbmanWong et al, 1991). A common mode of action of all antidepressants could be to modify the actual structure of nerve cells and possibly eliminate inappropriate synaptic contacts that are responsible for behavioural and psychological changes associated with depression.

It is also possible that chronic antidepressant treatment may affect pathways that involve receptor interactions with protein tyrosine kinases, by increasing specific growth factor synthesis or by regulating the activity of proinflammatory cytokines (Reference Duman, Henninger and NestlerDuman et al, 1997). These pathways control many aspects of neuronal function that ultimately underlie the ability of the brain to adapt and respond to environmental stimuli. The infusion of one of these transcription factors (brain-derived neutrophic factor) into the mid-brain of rats results in antidepressant-like activity (Reference Siuciak, Lewis and WiegandSiuciak et al, 1996), an action associated with an increase in the synthesis of tryptophan hydroxylase, the rate limiting enzyme in the synthesis of serotonin.