Académique Documents
Professionnel Documents
Culture Documents
0165-6147/$ see front matter 2008 Elsevier Ltd. All rights reserved. doi:10.1016/j.tips.2008.05.004 Available online 26 June 2008
433
Opinion
The development of neuroendocrine challenge techniques enabled assessments of brain 5-HT function to be
carried out in unmedicated depressed patients. These
investigations yielded more reliable evidence of abnormalities in brain 5-HT function, albeit in hypothalamic
regions [12]. The most consistent changes were found by
using 5-HT-reuptake inhibitors; acute drug challenge is
reliably associated with diminished prolactin release in
unmedicated depressed patients [12,13]. This suggests
that depressed patients have impaired 5-HT-mediated
neuroendocrine responses. This could be due to diminished
synaptic 5-HT levels or reduced sensitivity of postsynaptic
5-HT receptors in hypothalamic regions. It is not clear,
however, whether this abnormality is specific for depression, as opposed to, for example, anxiety disorders. In
addition, the prolactin response to intravenous administration of the SSRI citalopram remains blunted even when
patients have recovered and withdrawn from drug treatment [13] (Figure 1). This suggests that impaired 5-HTmediated prolactin release might be associated with
vulnerability to depression rather than the acute symptomatic state.
More modern investigations have employed brain-imaging techniques, in particular ligand imaging in conjunction with positron emission tomography (PET) and single
photon emission tomography (SPET) [14]. This enables
more direct investigation of 5-HT receptors in the living
human brain. These data have provided consistent and
convincing evidence that the binding density of 5-H1A
receptors shows a widespread decrease in depressed
patients [15,16]. However, as with the prolactin response
Figure 1. Mean (SEM) prolactin response, measured as area under the curve
(AUC), to citalopram (10 mg intravenously) in healthy controls (labelled as
controls) (n = 16), unmedicated depressed subjects (labelled as acute) (n = 13)
and unmedicated recovered depressed subjects (labelled as recov) (n = 16). Data
are taken, with permission, from Ref. [13].
434
Reliability of finding
High
Modest
Poor
High
Weak, might be stronger in suicidal subgroup
High (also diminished in recovered patients)
High (also diminished in recovered patients)
Poor
Poor
to citalopram, the binding appears to remain low in recovered depressed patients, suggesting that it is not a marker
of the acute depressive state [17]. In addition 5-HT1A-receptor binding also is lowered in patients with panic
disorder [18]; however, these patients usually have high
rates of co-morbid depression.
Other 5-HT receptors, for example, the 5-HT2A receptor
and the 5-HT transporter, also have been studied widely in
imaging investigations in depressed patients. However,
the data from this work are contradictory [19,20], showing
that even the use of state-of-the-art imaging to measure 5HT receptors in unmedicated patients, classified with
reliable diagnostic systems, is insufficient to overcome
clinical heterogeneity or technical variability between
different laboratories. In schizophrenia the ability to use
PET and SPET to measure in vivo dopamine release has
produced important advances in understanding pathophysiology [21]. It has been difficult to apply the same methodology to measure 5-HT release in vivo, but a recent study
shows that it might be possible to use 5-HT1A ligand PET to
measure the release of 5-HT onto 5-HT1A autoreceptors in
the raphe nuclei [22]. This could be an important methodological development.
The data reviewed so far show unequivocally that unmedicated depressed patients have abnormalities in aspects
of brain 5-HT activity. Although this is a useful start, there
are many important caveats before we can link these
changes with depression. For example, with a few exceptions, we do not know how specific any of these changes are
for depression as opposed to other psychiatric disorders, for
example, anxiety states and schizophrenia. Also many of
the reported abnormalities apparently are present even
when patients are recovered from depression. Although the
latter observation suggests that the changes are not
merely epiphenomena of the acute depressed state, it does
raise questions about how these 5-HT abnormalities are
involved in the experience of being depressed. They might
be markers of vulnerability, indicating that people with
abnormal brain 5-HT function are more likely to become
depressed, perhaps when exposed to psychosocial stress. It
also could be, however, that these changes are scars of
previous acute depressive episodes and their treatment.
Studies in high-risk groups before the onset of clinical
illness will be needed to resolve this issue.
An interesting genetic vulnerability factor is allelic
variation in the promoter region of the gene encoding
the 5-HT transporter (5HTT), whereby carriers of the short
allele of 5HTT (the low-activity allele) have a greater risk
of developing depression in response to stressful life events
[23]. Intriguingly, subjects with a short allele of 5HTT also
Opinion
show increased neural responses to fearful facial expression in functional imaging studies [24], suggesting that the
short allele might facilitate processing of negative
emotional information, perhaps thereby worsening the
psychological impact of adverse life experiences. However,
this account has been criticized as implausible because of
the small effect size of polymorphic variation on the phenotypic expression of complex traits [25]. In addition
carriers of the short allele of the 5HTT do not seem much
more likely than carriers of the long allele to experience
depressive illness, which would be expected if they were
more susceptible to stress [26].
Tryptophan depletion, 5-HT and depression
Another approach to elucidating the role of 5-HT in depression is to assess the effect of lowering brain 5-HT function
and examining its clinical and neuropsychological consequences. The technique employed, tryptophan depletion
(TRD), takes advantage of the fact that is possible to lower
plasma and brain tryptophan by administering an amino
acid mixture free of the 5-HT precursor, tryptophan.
Because the conversion of tryptophan to 5-hydroxytryptophan by tryptophan hydroxylase is the rate-limiting step in
5-HT synthesis, TRD is able to produce a transient lowering of brain 5-HT activity [27].
In healthy volunteers with no vulnerability factors for
the development of depression, TRD does not produce
consistent changes in mood [28]. This suggests that lowering brain 5-HT is not in itself sufficient to cause clinical
depressive symptomatology. However, in recovered
depressed patients withdrawn from medication, TRD
causes, in a high proportion of participants, a transient
but striking return of depressive symptomatology [28,29].
This suggests that, in those with risk factors for depression, lowering brain 5-HT function can, indeed, cause
clinical depressive symptomatology. Interestingly the
same is found when catecholamines are depleted with amethyl-para-tyrosine [30], showing that recovered depressed patients are psychologically vulnerable to both 5-HT
and catecholamine deficiency.
Why should recovered depressed patients show a
depressive response to TRD when healthy controls do
not? It is worth noting at this point that people at high
genetic risk of depression who have not suffered a personal
episode of illness do not experience clinical symptomatology after TRD, although they might be more likely than
controls to experience transient dysphoria, measurable on
adjective check lists [28]. This suggests that most of the
overt clinical relapse experienced by recovered depressed
patients after TRD is related to the experience of having
been depressed before. This could be a consequence of
previous treatment, for example, with SSRIs, but in our
own TRD study [29] we also saw clinical relapses after TRD
in patients who had recovered spontaneously or after
psychotherapy. Therefore, it seems more likely that
depression itself might cause persistent neurobiological
changes that produce a depressogenic response when
affected individuals are exposed to low brain 5-HT activity.
What could these neurobiological changes be? We
already have noted that recovered depressed patients have
evidence of abnormal brain 5-HT activity, so perhaps the
Vol.29 No.9
Opinion
Conflict-of-interest statement
The author has acted as a paid advisor to the following
companies involved in the development of antidepressant
drugs: Eli Lilly, GlaxoSmithKline, Lundbeck, Servier and
Wyeth.
Acknowledgements
The studies of the author were supported by the Medical Research
Council.
References
1 Coppen, A. (1967) The biochemistry of affective disorders. Br. J.
Psychiatry 113, 12371264
2 Cowen, P.J. (1990) A role for 5-HT in the action of antidepressant
drugs. Pharmacol. Ther. 46, 4351
3 (2005) Eli Lilly: correction and apology. BMJ 330, 211
4 Kirsch, I. et al. (2008) Initial severity and antidepressant benefits: a
meta-analysis of data submitted to the food and drug administration.
PLoS Med. 5, e45 DOI: 10.1371/journal.pmed.0050045 (http://
medicine.plosjournals.org)
5 Whittington, C.J. et al. (2004) Selective serotonin reuptake inhibitors
in childhood depression: systematic review of published versus
unpublished data. Lancet 363, 13411345
6 Lacasse, J.R. and Leo, J. (2005) Serotonin and depression: A disconnect
between the advertisements and the scientific literature. PLoS
Med. 2, e392 DOI: 10.1371/journal.pmed.0020392 (http://medicine.
plosjournals.org)
7 Healy, D. (2002) The Creation of Psychopharmacology, Harvard
University Press
8 Healy, D. et al. (1985) Peripheral adrenoceptors and serotonin
receptors in depression: changes associated with response to
treatment with trazodone or amitriptyline. J. Affect. Disord. 9, 285296
9 Yeo, I.B. (1888) An address on the therapeutics of the uric acid
diathesis. BMJ 1, 1619
10 Coppen, A. et al. (1978) Platelet 5 hydroxytryptamine accumulation in
depressive illness. Clin. Chim. Acta 87, 165168
11 Cowen, P.J. et al. (1989) Decreased plasma tryptophan levels in major
depression. J. Affect. Disord. 16, 2731
12 Cowen, P.J. (2001) Neuroendocrine markers of depression and
antidepressant drug action. In Antidepressants (Leonard, B.E. et al.,
eds), pp. 95107, Birkhauser Verlag
13 Bhagwagar, Z. et al. (2002) State and trait abnormalities in serotonin
in major depression. Br. J. Psychiatry 180, 2428
14 Grasby, P.M. (2002) Imaging the neurochemical brain in health and
disease. Clin. Med. 2, 6773
15 Sargent, P.A. et al. (2000) Brain serotonin1A receptor binding
measured by positron emission tomography with [11C]WAY-100635:
effects of depression and antidepressant treatment. Arch. Gen.
Psychiatry 57, 174180
16 Drevets, W.C. et al. (2007) Serotonin-1A receptor imaging in recurrent
depression: replication and literature review. Nucl. Med. Biol. 34, 865
877
17 Bhagwagar, Z. et al. (2004) Persistent reduction in brain serotonin1A
receptor binding in recovered depressed men measured by positron
436
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36