REF. 9). Iproniazid, a drug registered for the treatment of tuberculosis, was found to elevate the mood of patients that received it,and subsequent studies in patients who were Is mood chemistry? depressed but did not have tuberculosisshowed its effect as an antidepressant9. Sim-ultaneously and independently, imipramine, Eero Castrén an experimental antihistamine with a tricyclicstructure, was found to have antidepressant Abstract The chemical hypothesis of searching for the genes that might be asso- effects. These discoveries revolutionized the depression suggests that mood disorders ciated with these familial disorders, and recognition and treatment of mood disorders.
are caused by a chemical imbalance in the researchers hope to uncover a molecule that is In retrospect, it seems unbelievable that brain, which can be corrected by malfunctioning in people with depression8.
imipramine was introduced to the market antidepressant drugs. However, recent These research strategies seem to be based on only several years after its antidepressant evidence indicates that problems in an extension of the monoamine hypothesis, effects were discovered, mainly because the information processing within neural the chemical hypothesis of depression (FIG. 2), company producing it was unsure that the networks, rather than changes in chemical which proposes that mood disorders are number of patients who would benefit from balance, might underlie depression, and that caused by structural or functional changes in antidepressant treatment was sufficiently antidepressant drugs induce plastic changes particular molecules in the brain, and that high9. This now sounds incredible given the in neuronal connectivity, which gradually lead antidepressants function by counteracting current estimate that major depression is to improvements in neuronal information these molecular changes. Over the last few the single most expensive disorder faced processing and recovery of mood.
decades, the view that depression is produced by Western societies and that, overall, anti- by a chemical imbalance in the brain has depressants are among the best selling drugs.
The first antidepressants were discovered by become widely accepted among scientists, Soon after this discovery, imipramine and chance almost 50 years ago, when drugs that clinicians and the public.
iproniazid were found to increase the extra- had been developed for other disorders were However, during the past decade, several cellular concentrations of two important found to elevate the mood of psychiatric observations indicated that there might be neurotransmitters — serotonin and nor- patients. Soon after this, drugs with anti- an alternative hypothesis to the chemical adrenaline — in the brain, by blocking their depressant activity were shown to increase the view of depression. This network hypothesis re-uptake back to nerve endings or by extracellular concentrations of two important proposes that mood disorders reflect pro- inhibiting the main metabolizing enzyme, monoamine neurotransmitters in the brain — blems in information processing within par- monoamine oxidase, respectively. As drugs serotonin (5-hydroxytryptamine or 5-HT) and ticular neural networks in the brain and that that alleviate depression increase extracellular noradrenaline — by inhibiting their cata- antidepressant drugs and other treatments monoamine concentrations, it was proposed bolism or reuptake to nerve endings. These that alleviate depression function by gradu- that depression might be produced by a findings were the basis for the monoamine ally improving information processing serotonin or noradrenaline deficiency at hypothesis of depression, which proposes that within these networks (FIG. 3). This review functionally important receptor sites in the mood disorders are caused by a deficiency in discusses the evidence supporting and con- brain1–4 (FIG. 1), a proposal that is now known serotonin or noradrenaline at functionally tradicting the network hypothesis and the as the monoamine hypothesis of depression.
important receptor sites in the brain1–4 (FIG.1).
implications of the network view on drug Initially, the idea that a complex psychiatric It soon became evident that the development and the treatment of mood disorder such as depression could be pro- monoamine hypothesis in its original form duced by biochemical changes was met with could not explain all of the effects of anti- widespread scepticism among psychiatrists depressants5. Therefore, the focus of research The chemical hypothesis
and laymen. Nevertheless, during the last few was directed towards the receptors and intra- We will soon be celebrating the fiftieth decades this hypothesis has strongly influ- cellular signal transduction molecules that are anniversary of the discovery of antidepres- enced views about the pathophysiology of regulated by antidepressant treatment6,7.
sants, although the exact date and place of mood disorders, among not only pharma- Furthermore, because mood disorders often the discovery is a matter of dispute (for the cologists, but also clinicians, other scientists run in families, genetic studies have been history of the discovery of antidepressants, see and the public4.
2005 Nature Publishing Group
principal role of the nervous system is not tohandle chemicals but to store and processinformation. Arvid Carlsson, one of the mainarchitects of the concept of chemical neuro-transmission in the brain, stated in his Nobellecture, ‘However, it must be recognized thatthe brain is not a chemical factory but anextremely complicated survival machine'18.
Although chemical neurotransmitters are crucial for the transfer of information bet- Figure 1 Monoamine hypothesis of mood disorders. a In the normal brain, monoamine
ween neurons, information in the brain is not neurotransmitters (yellow) are released and bind to receptors on the postsynaptic neuron. Transmission is stored in a chemical form but is thought to be terminated by re-uptake of the transmitter. b In depression, the decreased concentration of monoamine at
processed by the complex interactions of neu- synaptic sites produces a mood disorder. c Blockade of the re-uptake sites (grey) increases the
rons in neural networks19,20. These networks concentration of monoamine neurotransmitters available at receptor sites and restores mood.
develop through interactions with the envi-ronment, and the neuronal structure of, andneurotransmission in these networks are con- The monoamine hypothesis focused the growth factors and their receptors, and stantly being refined through activity-depen- interest of the pharmaceutical industry on intracellular signalling molecules4,6,7,15 (FIG. 2).
dent synaptic plasticity to optimally process monoamine metabolism for decades.
As a result of this development, the mono- and store relevant information21 (BOX 1). So, Imipramine, which inhibits the re-uptake of amine hypothesis has evolved to what could disorders of the nervous system, including both serotonin and noradrenaline (and vari- be called a chemical or molecular hypothesis depression, might represent disturbances in ous other receptors and enzymes), has now of depression. This hypothesis presumes that the activity-dependent information process- been largely replaced by a host of molecules mood disorders are produced by long-term ing of the brain, rather than in the chemical that inhibit the uptake of either serotonin or changes in the production or activity of mol- balance of signalling molecules.
noradrenaline more selectively, and ipron- ecules in the brain and that antidepressants It should be noted that the chemical and iazid, which inhibits monoamine oxidase, function by counteracting these molecular network hypotheses are not mutually exclu- the main metabolizing enzyme for mono- changes. Motivated by this hypothesis, sive, but are complementary. As the synthesis amines, has given way to subtype-selective researchers are using large-scale DNA micro- and release of several important signalling monoamine oxidase inhibitors. Although array searches to look for genes that are up- or molecules are regulated by neuronal activity, this focused drug development effort has downregulated in depression or by anti- changes in the activity of neural networks clearly been successful from the point of view depressant treatments, in the hope that the produce changes in the concentration of these of safety, it has been less successful in terms molecules that are encoded by these genes signalling molecules. Therefore, although of efficacy. Modern antidepressants are no might be used as targets in the development the initial effects of antidepressants are obvi- more effective than the first generation of of new antidepressant drugs4,16,17.
ously chemical and are, in most cases, directed drugs that were discovered several decades towards the metabolism of monoamines, the ago, and electroconvulsive shock treatment ensuing adaptive changes in the concentra- remains the most effective treatment for tions of those signalling molecules are tightly It was recognized early on linked to the structure of the neural network, It was recognized early on that several that several observations and might be a consequence of the altered observations conflict with a simple link information processing rather than its cause.
conflict with a simple link between monoamine concentrations in the According to this view, antidepressants initiate brain and depression11. For example, deple- between monoamine a ‘self-repair' process, whereby plasticity in tion of dietary tryptophan, which signifi- neural networks and chemical neurotrans- concentrations in the brain cantly decreases the concentration of sero- mission indivisibly cooperate and gradually tonin in the brain, produces either no effects and depression.
bring about mood elevation.
or only a mild dysphoria in healthy volunteersand does not influence the mood of untreated Evidence for the network hypothesis
patients with depression12,13. More impor- The network hypothesis
The evidence that supports the network tantly, although the effects of antidepressants But is this view correct? Is mood chemistry? hypothesis of depression and antidepressant on monoamine metabolism can be seen soon Observations that have been made during the action is limited and mostly indirect. Part of after administration, it typically takes several last few years indicate that there might be an the problem is the lack of appropriate weeks of continued treatment for the clinical alternative to the chemical view of depression experimental models of depression. In par- antidepressant response to appear11. The dis- and the action of antidepressants5. This new ticular, there are no relevant and widely covery that long-term antidepressant treat- hypothesis, the network hypothesis, proposes accepted in vitro models of what might be ment produces adaptive changes in that problems in activity-dependent neuronal going on in the brain during depression.
monoamine receptors and in their coupling communication might underlie depression, Furthermore, methods for direct measure- to intracellular signal transduction14 caused and that antidepressants might work by ment of changes in neural networks in vivo the research focus to shift towards the effects improving information processing in the are only now being developed and have not that long-term antidepressant treatments affected neural networks (FIG. 3). A key aspect yet been applied to neuropharmacological have on the concentrations of neuropeptides, of the network view is the recognition that the 2 4 2 MARCH 2005 VOLUME 6
2005 Nature Publishing Group
Imaging studies in patients with depres- sion have revealed reduced grey matter volume in the prefrontal cortex32–34 and the hippocampus35–39. Morphological changes in the hippocampus are associated with, andmight be preceded by, functional deficits, such as memory impairment35. As the neu- ronal processes and synapses take up most of the space in the grey matter, reduced volumemight mean reduced neuronal complexity and connectivity. To at least some degree, these morphological alterations seem to be reversible by antidepressant therapies34,40.
The results of a recent study show that reduced hippocampal volume is particularly common in patients with depression who suffered a childhood trauma41, which indi- Cytocrome c cates that severe stress during a critical devel- opmental period might have lasting effects on the morphology of the brain. These data support the hypothesis that mood disordersare associated with compromised informa- tion processing in crucial neural networks, Failure of
and that the action of antidepressants might result in morphological and physiologicalreorganization of specific neuronal connec-tions in the brain. Furthermore, imaging andgenetic studies are beginning to elucidate Genetic and
which neural structures are involved in differ-ent mental health disorders and which circuits Figure 2 The chemical hypothesis of depression. The intracellular pathways that are affected by
might be important targets for successful mood disorders and antidepressants. AKT, protein kinase B; BAD, BLC-associated death promoter;BCL2, B-cell leukaemia/lymphoma 2; BCL-X, BCL2-like protein 1; BDNF, brain-derived neurotrophic factor; CREB, cyclic AMP responsive element binding protein; ERK, mitogen activated protein Perhaps the most important evidence for kinase 1; GR, glucocorticoid receptor; GSK3, glycogen synthase kinase 3; MEK, ERK kinase; the network hypothesis is the recent obser- VPA, valproate; NA, noradrenaline; P, phosphate; RAF, RAF proto-oncogene; ROS, reactive oxygen vation that antidepressants increase the species; Ras GTP, Ras GTPase-activating protein; RSK2, ribosomal protien S6 kinase polypeptide 3; production of new neurons in the rodent TRKB, neurotrophic tyrosine kinase receptor type B; 5-HT, 5-hydroxytryptamine (serotonin).
hippocampus42. Importantly, the increased Adapted, with permission, from REF. 6 (2001) Macmillan Magazines Ltd. neurogenesis that is brought about bychronic antidepressant treatment correlateswith the behavioural effects produced by Monoamines, particularly serotonin, It is well known that the plasticity of neural antidepressants43. Newly generated neurons have a significant role during brain develop- connectivity in the brain is greater and more differentiate over time, and are only mature ment23. Genetic elimination of the 5-HT1A extensive during critical periods of postnatal enough to participate in information receptor produces anxiety-type behaviour in development than in adults, and that func- processing several weeks after their birth44.
adult mice, but only when the receptor was tional structures that are formed during The fact that this time course correlates with absent during early postnatal development critical periods remain relatively stable in the delayed onset of the clinical effects of — its absence in adulthood produces no adulthood31 (BOX 1). These data highlight the antidepressants has created a lot of excite- behavioural effects24. Furthermore, muta- effects that serotonin and antidepressant ment among neuropharmacologists. In the tions in monoamine oxidase A produce treatments have on the plasticity of neural hippocampi of rodents that have received behavioural alterations in both men and networks, and link the effects of antidepres- antidepressant treatment, the elimination of mice25,26, and disrupt the developmental sants with the environmental manipulations neurons through apoptotic cell death organization of thalamocortical inputs and that are known to modulate neural network increases simultaneously with increased cortical modules in the brains of rodents23,27.
formation during development31. Moreover, neurogenesis, which indicates that antide- A similar disruption in thalamocortical they indicate that the effects of the drugs pressants might increase neuronal turnover organization has also been produced with might be more robust in the brain during rather than neurogenesis per se45. This effect the administration of antidepressants and early postnatal development and qualita- might be functionally analogous to the monoamine oxidase inhibitors during early tively different when compared with the overproduction of neurons that occurs postnatal development27,28. Furthermore, effects seen in the adult brain, which could during the development of the peripheral antidepressant treatment during early post- have significant implications for the pre- nervous system (BOX 1), and indicates that natal life can produce permanent behav- scription of antidepressants to children and antidepressants might facilitate optimiza- ioural disturbances in adult animals29,30.
pregnant mothers.
tion of neuronal connectivity by increasing NATURE REVIEWS NEUROSCIENCE VOLUME 6 MARCH 2005 2 4 3
2005 Nature Publishing Group
a Healthy
b Depressed
of BDNF in the hippocampus and cortex52–54,and the injection of BDNF into the brain oroverexpression of its receptor in transgenicmice produces similar behavioural responsesto those typically observed after treatmentwith antidepressant drugs55,56. Consistentwith the importance of BDNF in the effectsof antidepressants, transgenic mice withreduced BDNF expression or signalling fail to show these characteristic behaviouralresponses after the administration of anti-depressants54, which indicates that normalBDNF signalling might be both necessary andsufficient for a normal antidepressant effect. Itshould be emphasized that the crucial point isnot the increased molecular concentrationsof BDNF as such, but the importance ofthis neurotrophin as a mediator of activity-dependent neuronal plasticity (BOX 1).
In conclusion, the data that are summa- rized above provide some evidence that anti- c During treatment
d Recovered
depressants, through their acute effects onmonoamine metabolism, activate processesof plasticity, which are thought to graduallylead to improved information processing inthe neural networks that are involved inmood regulation. These processes, whichinclude neurogenesis and selective neuralelimination, arborization and retraction ofaxons and dendrites, and synaptic formationand pruning, are expected to take time todevelop and mature, which is consistent withthe delayed appearance of the clinical effectsof antidepressants.
However, observations have also been made that seem to be incompatible with thenetwork hypothesis. Although depletion of tryptophan — the rate-limiting factor ofserotonin synthesis — does not influence themood of healthy volunteers and untreatedpatients with depression12,13,57, it does pro- Figure 3 The network hypothesis of depression. a In the healthy brain, information is processed
duce a rapid relapse of depressive symptoms in partially overlapping neural networks. b In depression, information processing in some networks
in about 50% of remitted patients who are does not function properly. c Antidepressant treatment enhances connectivity in neural networks.
d Activity-dependent pruning of synapses selects out and stabilizes the active synapses and networks.
being, or have recently been treated withserotonin selective antidepressants12,13.
Furthermore, there is a circadian variation the choice of neurons available for selection connectivity without any net change in in mood, and sleep deprivation rapidly through activity-dependent mechanisms. At synaptic number20. Unfortunately, it is difficult improves the mood of patients with depres- a more subtle level, antidepressant drugs to quantify synaptic turnover in vivo.
sion, albeit temporarily58. These relatively can enhance the sprouting of axons46 and One possible mechanism through which rapid effects on mood are difficult to recon- dendrites47, and support the morphological antidepressants might enhance the plasticity cile with the view of their production by a maturation of the newborn neurons47.
of neuronal connections in the hippocampus gradual change in the structure of mood- These data indicate that, in addition to their and cerebral cortex is the activation of neuro- elevating neural networks. It is obvious that established function in elevating neuronal trophin signalling48,49. Brain-derived neuro- more experimental work will be necessary to turnover in the dentate gyrus, antidepres- trophic factor (BDNF), which is produced test the new model, but the rapid develop- sants might also stimulate the turnover of and released by neurons in an activity- ment of neurophysiological methods and axonal branches and synaptic contacts, dependent manner50, has been proposed to the imaging of neural networks will help us thereby providing more material for activ- be a crucial factor in the selection and stabi- to gain further insights into the relationship ity-dependent selection. It should be noted lization of active synaptic contacts51 (BOX 1).
between the effects of antidepressants and that increased synaptic turnover might lead Both antidepressants and electroconvulsive neural plasticity, which might become a to significant reorganization of neuronal shocks increase the expression and signalling fruitful area of further research.
2 4 4 MARCH 2005 VOLUME 6
2005 Nature Publishing Group
be expected to be more beneficial than either Box 1 Activity-dependent refinement of neural networks
treatment alone, and there is evidence that During the development of the peripheral nervous system, neurons are produced in excess
this might be the case59. As the network and compete to innervate the target tissues61. Neurotrophic factors, which are secreted by
hypothesis emphasizes the importance of the target tissues (for example, muscles) in limited quantities, and which are required by
environmental information in the process of innervating neurons for their survival, select from the competing neurons those that best
activity-dependent selection of neurons and innervate the target, and the defeated neurons are eliminated by apoptosis. This competitive
synapses (BOX 1), it predicts that full recovery process matches the number of neurons and targets and ensures optimal innervation of the
would not even be possible with drug treat- target cells61–63. In the CNS, neurons target other neurons and competitive selection takes
ment alone, but that external stimuli, such as place at several levels: among neurons (as in the peripheral nervous system), among axon
social communication, would be required to branches and among synaptic contacts64. In the brain, the release of target-derived trophic
provide environmental input for the selec- factors is activity dependent: the innervating neuron must sufficiently stimulate the target
tion of the appropriate network connections neuron in order to induce the production and release of the trophic factor50,51. Neurons can
(FIG. 3). The role of environmental stimula- also cooperate to increase the release of trophic factors: simultaneously active neurons that
tion might also be related to the fact that innervate the same target neuron can induce the release of trophic factors through a much
major depression is typically cyclical and lower level of activity than is required for a single innervating neuron51. This activity-
often self-limiting, and many patients dependent cooperative selection of simultaneously active neurons and the elimination of
improve with time even in the absence of inactive and incoherent contacts is considered to be crucial in the development of large,
coherently active neural networks: ‘the neurons that fire together, wire together'
21. Through
active treatment.
these mechanisms, neuronal structure and neurotransmission are optimized to best store
Finally, the effects of antidepressant drugs and process relevant information. During critical periods in early postnatal development,
on network plasticity in the brain might much larger changes in the connectivity and organization of neural networks take place
explain why they are effective for many than is possible in the adult brain31. Nevertheless, even the adult brain still shows significant
neuropsychiatric disorders, including anxiety, plasticity. The crucial event in activity-dependent plasticity is not the formation of neuronal
obsessive-compulsive disorder, eating dis- contacts (which might occur stochastically and in excess) but the activity-dependent
orders, chronic pain and tinnitus. It has been selection and stabilization of those synapses that mediate useful signals, together with the
proposed that some of these disorders, par- selective pruning and elimination of those that produce random noise20. Therefore,
ticularly chronic pain, might be produced by neurogenesis and synaptogenesis cannot simply be considered beneficial, and neuronal
aberrant neuronal connectivity60.
death and synapse elimination harmful; what matters is the optimization of the
The hypothesis that mood represents a signal–to–noise ratio in the network. Neurotrophins are important in this process through
functional state of neural networks might their selective release from active connections; indiscriminate increases in the levels of
sound incompatible with the efforts of ratio- neurotrophic factors or their signalling (as, for example, would be produced by a
nal drug development. However, the data neurotrophic factor receptor agonist) is not expected to be beneficial to the network, as both
reviewed above indicate that the antidepres- active and inactive connections would be similarly supported.
sant drugs that have been used successfullyfor several decades might function by initi-ating such plastic processes, apparently plasticity. Therefore, psychological and indirectly, by influencing monoamine The view of mood disorders as problems of pharmacological therapies, electroconvulsive metabolism. It is possible that a similar information processing in the brain has shock treatment and placebo effects might all process could also be initiated through several important implications. As devel- lead to improved information processing and other pharmacological mechanisms, which opmental neurobiologists have been investi- mood recovery through mechanisms that ini- might become the targets of new anti- gating activity-dependent plastic processes tiate similar processes of plasticity (FIG. 4). In depressants that could help patients who are for decades, collaboration between neuro- this scenario, a combination of drug treat- resistant to current drugs and only respond pharmacologists, developmental neuro- ment and psychological rehabilitation would to electroconvulsive shock treatment.
scientists and behavioural geneticists shouldbe encouraged. Recent studies clearly showthat genetic manipulation of neural circuitsand assessment of the consequences through in vivo recording techniques and behav-ioural assays might provide an incrediblepotential for begining to define the relation-ship between circuit properties, behaviouraldeficits and potential therapeutics23,24.
During development and in adults, train- ing and rehabilitation produce functional andanatomical changes in neural networks,which are reflected in the gradual improve- ment of the rehearsed action. Analogously, psychotherapy, cognitive behavioural therapy Figure 4 A combinatorial approach for treating depression based on the network hypothesis.
and other forms of psychological rehabilitation Depression might reflect disturbed information processing in neural networks (left panel). Antidepressant could also have therapeutic effects on mood drugs, electroconvulsive shock and psychotherapy can all induce activity-dependent plasticity, which disorders through use-dependent neuronal gradually leads to the recovery of connectivity in the affected neural networks (right panel).
2005 Nature Publishing Group
18. Carlsson, A. A half-century of neurotransmitter research: 45. Sairanen, M., Lucas, G., Ernfors, P., Castrén, M. & Our view of mood disorders and the impact on neurology and psychiatry. Nobel lecture.
Castrén, E. BDNF and antidepressant drugs have action of antidepressants is beginning to different but coordinated effects on neuronal turnover, proliferation and survival in the adult dentate gyrus. change from a chemical view towards a net- 19. Buzsaki, G. Large-scale recording of neuronal J. Neurosci. 25, 1089–1094 (2005).
work hypothesis, in which problems in infor- ensembles. Nature Neurosci. 7, 446–451 (2004).
46. Vaidya, V. A., Siuciak, J. A., Du, F. & Duman, R. S.
20. Hua, J. Y. & Smith, S. J. Neural activity and the dynamics Hippocampal mossy fiber sprouting induced by chronic mation processing and neuronal connectivity of central nervous system development. Nature electroconvulsive seizures. Neuroscience 89, 157–166
in the brain are the central questions.
Neurosci. 7, 327–332 (2004).
21. Katz, L. C. & Shatz, C. J. Synaptic activity and the 47. Fujioka, T., Fujioka, A. & Duman, R. S. Activation of cAMP However, several important issues have yet to construction of cortical circuits. Science 274, 1133–1138
signaling facilitates the morphological maturation of be addressed. How are neural networks newborn neurons in adult hippocampus. J. Neurosci. 24,
22. Varela, F., Lachaux, J. P., Rodriguez, E. & Martinerie, J.
319–328 (2004).
formed and sustained during development The brainweb: phase synchronization and large-scale 48. Altar, C. A. Neurotrophins and depression. Trends and in adults? Which neuronal connections integration. Nature Rev. Neurosci. 2, 229–239 (2001).
Pharmacol. Sci. 20, 59–61 (1999).
23. Gaspar, P., Cases, O. & Maroteaux, L. The 49. Castrén, E. Neurotrophic effects of antidepressant drugs.
are involved, and are the same networks developmental role of serotonin: news from mouse Curr. Opin. Pharmacol. 4, 58–64 (2004).
affected in all patients? How do antidepres- molecular genetics. Nature Rev. Neurosci. 4, 1002–1012
50. Thoenen, H. Neurotrophins and neuronal plasticity.
Science 270, 593–598 (1995).
sants initiate and support the process of plas- et al. Serotonin receptor acts during 51. Poo, M. M. Neurotrophins as synaptic modulators.
ticity? Does the delayed appearance of clini- development to establish normal anxiety-like behaviour in Nature Rev. Neurosci. 2, 24–32 (2001).
the adult. Nature 416, 396–400 (2002).
52. Nibuya, M., Morinobu, S. & Duman, R. S. Regulation of cal antidepressant effects reflect slow 25. Brunner, H. G., Nelen, M., Breakefield, X. O., Ropers, H. H.
BDNF and trkB mRNA in rat brain by chronic maturation of neuronal connections, and & van Oost, B. A. Abnormal behavior associated with a electroconvulsive seizure and antidepressant drug point mutation in the structural gene for monoamine treatments. J. Neurosci. 15, 7539–7547 (1995).
could these processes be accelerated? In any oxidase A. Science 262, 578–580 (1993).
53. Russo-Neustadt, A. A., Beard, R. C., Huang, Y. M. & case, it is increasingly evident that the activ- et al. Aggressive behavior and altered amounts Cotman, C. W. Physical activity and antidepressant of brain serotonin and norepinephrine in mice lacking treatment potentiate the expression of specific brain- ity-dependent plasticity and connectivity of MAOA. Science 268, 1763–1766 (1995).
derived neurotrophic factor transcripts in the rat neural networks must be considered when et al. Lack of barrels in the somatosensory hippocampus. Neuroscience 101, 305–312
cortex of monoamine oxidase A-deficient mice: role of a designing future strategies of antidepressant serotonin excess during the critical period. Neuron 16,
54. Saarelainen, et al. Activation of the TrkB neurotrophin treatment, whether they are pharmacologi- 297–307 (1996).
receptor is induced by antidepressant drugs and is 28. Xu, Y., Sari, Y. & Zhou, F. C. Selective serotonin reuptake required for antidepressant-induced behavioral effects. cal, psychological or combinatorial.
inhibitor disrupts organization of thalamocortical J. Neurosci. 23, 349–357 (2003).
somatosensory barrels during development. Dev. Brain 55. Shirayama, Y., Chen, A. C., Nakagawa, S., Russell, D. S.
Eero Castrén is at the Neuroscience Center,
Res. 150, 151–161 (2004).
& Duman, R. S. Brain-derived neurotrophic factor University of Helsinki, Finland.
29. Ansorge, M. S., Zhou, M., Lira, A., Hen, R. & Gingrich, J. A.
produces antidepressant effects in behavioral Early-life blockade of the 5-HT transporter alters models of depression. J. Neurosci. 22, 3251–3261
emotional behavior in adult mice. Science 306, 879–881
56. Siuciak, J. A., Lewis, D. R., Wiegand, S. J. & Lindsay, R. M.
30. Feng, P., Ma, Y. & Vogel, G. W. The critical window of Antidepressant-like effect of brain-derived neurotrophic Bunney, W. E. Jr & Davis, J. M. Norepinephrine in brain development from susceptive to insusceptive.
factor (BDNF). Pharmacol. Biochem. Behav. 56, 131–137
depressive reactions. A review. Arch. Gen. Psychiatry 13,
Effects of clomipramine neonatal treatment on sexual 483–494 (1965).
behavior. Brain Res. Dev. Brain Res. 129, 107–110
57. Van der Does, A. J. W. The effects of tryptophan Coppen, A. The biochemistry of affective disorders. Br. depletion on mood and psychiatric symptoms. J. Affect. J. Psychiatry 113, 1237–1264 (1967).
31. Berardi, N., Pizzorusso, T. & Maffei, L. Critical periods Disord. 64, 107–119 (2001).
Schildkraut, J. J. The catecholamine hypothesis of during sensory development. Curr. Opin. Neurobiol. 10,
58. Wirz-Justice, A. & Van den Hoofdakker, R. H. Sleep affective disorders: a review of supporting evidence. Am. 138–145 (2000).
deprivation in depression: what do we know, where do J. Psychiatry 122, 509–522 (1965).
32. Bremner, J. D. et al. Reduced volume of orbitofrontal we go? Biol. Psychiatry 46, 445–453 (1999).
Wong, M. L. & Licinio, J. From monoamines to genomic cortex in major depression. Biol. Psychiatry 51, 273–279
Adolescents With Depression Study targets: a paradigm shift for drug discovery in Team. Fluoxetine, cognitive-behavioral therapy, and depression. Nature Rev. Drug Disc. 3, 136–150 (2004).
33. Botteron, K. N., Raichle, M. E., Drevets, W. C., Heath, A. C.
their combination for adolescents with depression: Nestler, E. J. et al. Neurobiology of depression. Neuron & Todd, R. D. Volumetric reduction in left subgenual Treatment for Adolescents With Depression Study 34, 13–25 (2002).
prefrontal cortex in early onset depression. Biol. (TADS) randomized controlled trial. JAMA 292, 807–820
Manji, H. K., Drevets, W. C. & Charney, D. S. The cellular Psychiatry 51, 342–344 (2002).
neurobiology of depression. Nature Med. 7, 541–547
34. Drevets, W. C. Neuroimaging and neuropathological 60. Wolpaw, J. R. & Tennissen, A. M. Activity-dependent studies of depression: implications for the spinal cord plasticity in health and disease. Annu. Rev. Coyle, J. T. & Duman, R. S. Finding the intracellular cognitive–emotional features of mood disorders. Curr. Neurosci. 24, 807–843 (2001).
signaling pathways affected by mood disorder Opin. Neurobiol. 11, 240–249 (2001).
61. Levi-Montalcini, R. The nerve growth factor: thirty-five treatments. Neuron 38, 157–160 (2003).
35. MacQueen, G. M. et al. Course of illness, hippocampal years later. EMBO J. 6, 1145–1154 (1987).
Kennedy, J. L., Farrer, L. A., Andreasen, N. C., Mayeux, R.
function, and hippocampal volume in major depression.
62. Barde, Y.-A. Trophic factors and neuronal survival.
& George-Hyslop, P. The genetics of adult-onset Proc. Natl Acad. Sci. USA 100, 1387–1392 (2003).
Neuron 2, 1525–1534 (1989).
neuropsychiatric disease: complexities and conundra? 36. Sheline, Y. I., Gado, M. H. & Kraemer, H. C. Untreated 63. Huang, E. J. & Reichardt, L. F. Neurotrophins: roles in Science 302, 822–826 (2003).
depression and hippocampal volume loss. Am. neuronal development and function. Annu. Rev. The Antidepressant Era (Harvard Univ. Press, J. Psychiatry 160, 1516–1518 (2003).
Neurosci. 24, 677–736 (2001).
Cambridge, Massachusetts, 1997).
37. Sheline, Y. I. Neuroimaging studies of mood disorder 64. Cohen-Cory, S. The developing synapse: construction 10. Duman, R. S. & Vaidya, V. A. Molecular and cellular effects on the brain. Biol. Psychiatry 54, 338–352 (2003).
and modulation of synaptic structures and circuits.
actions of chronic electroconvulsive seizures. J. ECT 14,
et al. Hippocampal changes in patients with a Science 298, 770–776 (2002).
181–193 (1998).
first episode of major depression. Am. J. Psychiatry 159,
11. Nestler, E. J. Antidepressant treatments in the 21st 1112–1118 (2002).
century. Biol. Psychiatry 44, 526–533 (1998).
et al. Quantitative MRI of the hippocampus I would like to thank H. Rauvala, M. Saarma, M. Castrén and 12. Delgado, P. L. How antidepressants help depression: and amygdala in severe depression. Psychol. Med. 30,
R. Galuske for their comments to the manuscript, and the Sigrid mechanisms of action and clinical response. J. Clin. 117–125 (2000).
Jusélius Foundation, Sohlberg Foundation and the Academy of Psychiatry 65 Suppl. 4, 25–30 (2004).
40. Drevets, W. C., Bogers, W. & Raichle, M. E. Functional Finland and for support.
13. Booij, L., Van der Does, A. J. & Riedel, W. J. Monoamine anatomical correlates of antidepressant drug treatment depletion in psychiatric and healthy populations: review.
assessed using PET measures of regional glucose Competing interests statement Mol. Psychiatry 8, 951–973 (2003).
metabolism. Eur. Neuropsychopharmacol. 12, 527–544
The author declares no competing financial interests.
14. Sulser, F., Vetulani, J. & Mobley, P. L. Mode of action of antidepressant drugs. Biochem. Pharmacol. 27,
41. Vythilingam, et al. Childhood trauma associated with 257–261 (1978).
smaller hippocampal volume in women with major Online links
15. Duman, R. S., Heninger, G. R. & Nestler, E. J. depression. Am. J. Psychiatry 159, 2072–2080 (2002).
A molecular and cellular theory of depression. Arch. Gen. 42. Malberg, J. E., Eisch, A. J., Nestler, E. J. & Duman, R. S.
Psychiatry 54, 597–606 (1997).
Chronic antidepressant treatment increases The following terms in this article are linked online to:
16. Knuuttila, J. E., Toronen, P. & Castrén, E. Effects of neurogenesis in adult rat hippocampus. J. Neurosci. 20,
Mouse Genome Informatics:
antidepressant drug imipramine on gene expression in rat 9104–9110 (2000).
prefrontal cortex. Neurochem. Res. 29, 1235–1244
et al. Requirement of hippocampal neurogenesis for the behavioral effects of 17. Newton, S. S. et al. Gene profile of electroconvulsive antidepressants. Science 301, 805–809 (2003).
seizures: induction of neurotrophic and angiogenic 44. van Praag, H. et al. Functional neurogenesis in the adult factors. J. Neurosci. 23, 10841–10851 (2003).
hippocampus. Nature 415, 1030–1034 (2002).
Access to this interactive links box is free online.
2 4 6 MARCH 2005 VOLUME 6
2005 Nature Publishing Group


and Treatment of in Primary Care This educational activity is supported through pooled-funded educational grants from Pfizer Inc. and Abbott Laboratories to the Physicians' Institute for Excellence in Medicine to develop and offer grants to accredited CME providers for improvement in the diagnosis

MisiÓn de la universidad autÓnoma de yucatÁn