Neurobiological Changes Induced By Chronic Administration Of Antidepressants

The actions of all the compounds described so far seem to underpin the monoamine hypothesis. Yet an outstanding problem in treating depression is that the therapeutic response is both slow and progressive: a significant improvement usually takes at least 2-3 weeks and sometimes much longer. Obviously, if we are to explain the therapeutic effects of antidepressants, we must search for long-term neurochemical changes that occur after their prolonged administration.


The first indication that some neurochemical changes developed only after prolonged treatment with antidepressants came from landmark experiments carried out by Vetulani and Sulser in the mid-1970s (Vetulani et al. 1976). They found that repeated, but not a single, administration to rats of any of the antidepressants which were available at that time (i.e. MAOIs, TCAs, iprindole and even simulated electroconvulsive therapy) attenuated the increase in cAMP in the cerebral cortex induced by ¿-adrenoceptor agonists. They suggested that antidepressants desensitised ¿-adrenoceptors by uncoupling the receptor from what is now recognised as the Gs-protein so that it can no longer synthesise the ¿-adrenoceptor second messenger, cAMP. Shortly afterwards, it was found that this desensitisation was usually paralleled by downregulation of ¿1- (but not ¿2-) adrenoceptors. This action is even shared by repeated electroconvulsive shock (Stanford and Nutt 1982) but not by drugs that are ineffective in relieving depression (e.g. neuroleptics).

A logical conclusion from this work was that depression is caused by hyperresponsive ¿-adrenoceptors. At first, this might seem to undermine Schildkraut's suggestion that depression is caused by a deficit in noradrenergic transmission. However, proliferation of receptors is the normal response to a deficit in transmitter release and so the opposite change, downregulation of ¿-adrenoceptors by antidepressants, would follow an increase in the concentration of synaptic noradrenaline. This would be consistent with both their proposed mechanism of action and the monoamine theory for depression.

Nonetheless, there are many reasons to be confident that ¿-adrenoceptor desensitisation does not explain the therapeutic effects of antidepressants. First, with the development of more selective ligands for use in radioligand binding studies, it became evident that ¿-adrenoceptor downregulation can occur after only 2-3 days of drug treatment (Heal et al. 1989). Second, maprotiline, most of the SSRIs, and even some of the newer TCAs have no effect on ¿-adrenoceptor binding or function. Third, and the greatest problem of all, citalopram increases the ¿-adrenoceptor-mediated cAMP response without changing receptor density. Evidently, we must look elsewhere to find an explanation for the neurobiology of depression and its treatment.


There is a good deal of evidence that the therapeutic effects of antidepressants could involve adaptive changes in 5-HT1A receptors. Postsynaptic 5-HT1A receptor responses became implicated because the hyperpolarisation of hippocampal CA3 pyramidal neurons that follows ionophoretic administration of 5-HT was found to be increased after chronic treatment with most (but not all) antidepressants (Chaput, de Montigny and Blier 1991). Others suggested that antidepressants attenuate postsynaptic 5-HT1A responses because the hypothermia, evoked by their activation, is diminished by antidepressants (Martin et al. 1992).

More recently, a series of studies using microdialysis in vivo has suggested that long-latency changes in presynaptic 5-HTiA receptors could underlie the therapeutic lag in antidepressant treatment. In these experiments, a single dose of either a SSRI (e.g. fluoxetine or paroxetine), or a MAOI (e.g. tranylcypromine) increased the concentration of extracellular 5-HT in the dorsal Raphe nucleus but not in the brain areas to which these neurons project (e.g. the frontal cortex or striatum; see Hervas et al. 1999). The suggested explanation for this regional difference was that the accumulation of extracellular 5-HT in the Raphe nuclei, caused by the SSRIs blocking its reuptake, activates somatodendritic 5-HTjA receptors and so inhibits the firing of serotonergic neurons. This results in reduced impulse flow to their terminals so that extracellular 5-HT does not increase there despite blockade of its reuptake (Fig. 20.6). Obviously, if iuwjV -

Bupropion Reeptor Profile

O 5-HT transporter

O 6-HT1B receptor


Figure 20.6 Schematic representation of the effects of 5-HT reuptake inhibitors on serotonergic neurons, (a) 5-HT is released at the somatodendritic level and by proximal segments of serotonergic axons within the Raphe nuclei and taken up by the 5-HT transporter. In these conditions there is little tonic activation of somatodendritic 5-HT1A autoreceptors. At nerve terminals 5-HT1B receptors control the 5-HT synthesis and release in a local manner. (b) The blockade of the 5-HT transporter at the level of the Raphe nuclei elevates the concentration of extraneuronal 5-HT to an extent that activates somatodendritic autoreceptors (5-HT1A). This leads to neuronal hyperpolarisation, reduction of the discharge rate and reduction of 5-HT release by forebrain terminals. (c) The exposure to an enhanced extracellular 5-HT concentration produced by continuous treatment with SSRIs desensitises Raphe 5-HT1A autoreceptors. The reduced 5-HT1A function enables serotonergic neurons to recover cell firing and terminal release. Under these conditions, the SSRI-induced blockade of the 5-HT transporter in forebrain nerve terminals results in extracellular 5-HT increases larger than those observed after a single treatment with SSRIs. (Figure and legend taken from Hervas et al. 1999 with permission)

this is correct, then blocking 5-HTiA receptors in the Raphé nuclei should prevent these changes. This was confirmed by the finding that SSRIs did increase the concentration of extracellular 5-HT in the cortex and failed to reduce neuronal firing rate if the 5-HTjA receptor antagonist, WAY 100635, was co-administered, either systemically or by infusion directly into the dorsal Raphe nucleus.

More importantly for this discussion is the finding that chronic administration of an antidepressant produces a similar increase in the concentration of extracellular 5-HT in the terminal field together with recovery of neuronal firing. Presumably this is because the prolonged elevation of extracellular 5-HT around the neurons in the Raphe causes progressive desensitisation of the somatodendritic 5-HTjA receptors. At this point, inhibition of their firing does not occur and so more 5-HT is released in the cortex (see Hervas et al. 1999).

If long-latency 5-HTjA receptor downregulation explains the antidepressant therapeutic lag, then 5-HTjA receptor antagonists might reduce the delay in treatment response. This prediction has been tested in the clinic using combined treatment with paroxetine and the mixed ¿-adrenoceptor/5-HTjA antagonist, pindolol and the majority of studies report a successful outcome (see Hervas et al. 1999). However, it remains uncertain whether this effect of pindolol is due to its actions at presynaptic 5-HTjA receptors. If, as suggested earlier, postsynaptic 5-HTjA receptors are involved in the therapeutic effects of antidepressants, then co-administration of a 5-HTjA receptor antagonist of this receptor might well diminish any antidepressant effect. Pindolol is said to avoid this problem by its selective antagonism of presynaptic, but not postsynaptic, 5-HTjA receptors, but this is controversial.

A related strategy would be to inactivate the 5-HTjB/jD autoreceptors which are found on serotonergic nerve terminals and so prevent feedback inhibition of 5-HT release in the terminal field. These drugs would not prevent the impact of indirect activation of 5-HTjA receptors, and the reduced neuronal firing, by SSRIs (described above), but they would augment 5-HT release in the terminal field once the presynaptic 5-HTjA receptors have desensitised. Selective 5-HTjB/jD antagonists have been developed only recently but will doubtless soon be tested in humans.


The extensive literature on long-latency changes in neurotransmitter receptors following chronic administration of antidepressants reflects the intense effort that has been invested in the search for the cause of their therapeutic lag. Indeed, apart from developing compounds that help patients who currently do not respond to any existing treatment, the most pressing problem in this field is to reduce the delay in treatment response. Yet, despite the numerous investigations of the effects of antidepressants on a wide range of transmitter receptors, few consistent findings have emerged. Results tend to vary not only from laboratory to laboratory and between different brain regions but they also vary with the species and compound tested. The most promising changes are summarised in Table 20.7 but, so far, these do not fit into a scheme that explains either depression or its reversal by antidepressants.

Obviously one limitation of all this work is that the drug effects have been tested in 'normal' animals. So far, the neurochemical changes induced by long-term drug treatment have not been tested in combination with procedures such as learned helplessness, but it cannot be assumed that they will be the same as those in normal (non-depressed) subjects.

Continue reading here: The Hpa Axis And Depression

Was this article helpful?

0 0