Schizophrenia Research Forum - A Catalyst for Creative Thinking

BDNF In the Nucleus Accumbens—Too Much of a Good Thing?

15 February 2006. The brain-derived neurotrophic factor (BDNF) has drawn much attention from neuroscientists since Yves-Alain Barde and colleagues isolated it from pig's brain in 1982 (Barde et al., 1982). It enhances the survival and migration of neurons from the earliest stages of neural development and continues helping those cells form and maintain networks into adulthood (see, e.g., Egan et al., 2003). But according to an article in the February 10 issue of Science, in the more complex domain of behavior, too much BDNF can sometimes be a bad thing. Eric Nestler, Olivier Berton, and colleagues at the University of Texas Southwestern Medical Center in Dallas, with collaborators at several other institutions, report a new mouse model of social withdrawal that has an excess of BDNF in the mesolimbic dopamine system and can be reversed with antidepressant compounds. Beyond their relevance to depression, the findings may be broadly applicable to other psychiatric disorders that have elements of social withdrawal, including schizophrenia.

A range of studies, from postmortem neurochemical studies to work in experimental models, has linked dysregulation of BDNF to depression, although the guiding hypothesis until recently has been that a deficit of the growth factor contributes to the disorder. Indeed, studies in animal models of depression have suggested that boosting brain BDNF reverses behavioral deficits in these models, perhaps by stimulating neurogenesis (for review, see Hashimoto et al., 2004). However, Nestler and colleagues have previously reported an instance that goes against that grain. (Eisch et al., 2003). They reproduced the beneficial effects of antidepressants on an animal model of stress by blocking BDNF activity in the mesolimbic dopamine system—the projection from dopaminergic neurons in the ventral tegmental area (VTA) of the midbrain to the nucleus accumbens (N Ac) in the basal forebrain.

In their current article, first author Berton and colleagues describe the creation of a new model of social withdrawal, a common feature of depression. Mice who were repeated "losers" in interactions with other, aggressive mice developed an aversion to contact with any other mice, even ones that didn't look like the aggressors and were non-threatening. This effect was reversed with either of the antidepressants imipramine (Tofranil) or fluoxetine (Prozac), administered chronically for four weeks, but not acutely, after the stressful conditioning. "It's been hard for researchers to find a condition in animals that responds to chronic administration of antidepressants. This is one of the few tests in which animals respond to chronic antidepressants, rather than acute antidepressants, and that's a very important part of this study because antidepressants only work in humans after long-term administration," said Nestler, quoted in a press release from UT Southwestern.

The researchers next investigated whether the mesolimbic dopamine system was involved in the social withdrawal. They report that levels of cFos, an immediate early gene used as a marker of neuronal activity, increased significantly in both the dopamine neurons of VTA and their targets in the N Ac, and this effect could be detected even 4 weeks after the end of the 10-day conditioning period. Given their previous findings regarding BDNF's role in the mesolimbic dopamine system in a stress model, the authors investigated whether BDNF also plays a role in the social withdrawal model. They found that BDNF levels increased significantly in the N Ac of socially withdrawn mice following the social stress, and remained elevated for 4 weeks.

The scientists next looked at the effect of removing BDNF from VTA cells. Cells located in the N Ac make very little BDNF, and most BDNF in the nucleus appears to be secreted by the terminals of the VTA axons that enter the area. This extracellular BDNF in the N Ac has effects at dopaminergic synapses on both the pre- and postsynaptic membranes via its TrkB receptors. The researchers used a viral vector to knock down the BDNF produced in VTA neurons before exposing mice to the social defeat paradigm. About 75 percent of the dopaminergic neurons were successfully transfected, with virtually complete loss of BDNF mRNA in those cells. This manipulation eliminated the social withdrawal—despite repeated losses in interactions with aggressive mice, the "losers" still approached strange mice readily, supporting a critical role for BDNF in the conditioning of this social aversion.

An important methodologic concern is that the experimental manipulation could have perturbed the cells more broadly than just reducing BDNF, or may even have killed a subset of the dopaminergic neurons. However, the authors find that there was no reduction in tyrosine hydroxylase, the rate-limiting enzyme in dopamine synthesis (used as a marker for dopamine), and there was no apparent loss of cells. This last consideration is important, since studies in the other region of dopaminergic neurons, the substantia nigra, have found that suppressing BDNF kills, rather than protects dopaminergic cells there. Berton and colleagues note that VTA cells are less vulnerable than their substantia nigra cousins (Hung and Lee, 1996), which are selectively, and mysteriously, decimated in Parkinson disease.

Hypothesizing that the lasting behavioral conditioning reflects changes in the genetic targets of BDNF's signaling cascade in N Ac neurons, Nestler's team next used DNA microarrays to examine gene transcription in the N Ac. They report that 309 genes were up-regulated just after the social defeat conditioning, with 127 still elevated 4 weeks later, whereas 17 were immediately down-regulated, of which nine remained elevated 4 weeks later. These transcriptional alterations were eliminated by the virally mediated knockdown of BDNF in VTA neurons. Antidepressants were able to eliminate most of the transcription changes that persisted after 4 weeks.

Among the alterations in gene expression, the authors noted particularly players in the BDNF signaling cascade, such as phosphatidylinositol 3-kinase (PI3K) and mitogen activated protein kinase (MAPK). "These microarray data thereby suggest that chronic treatment with antidepressant restores social approach behaviors partly by interfering with the activity of neurotrophic cascades that mediate experience-induced neuroadaptations in the mesolimbic dopamine pathway," write the authors.

The effect of these findings on drug treatment options for depression remains to be seen, given the different roles that BDNF appears to play in different brain areas (for a new review on strategies for treatment of depression see Berton and Nestler, 2006). Likewise, it’s unclear at present how these finding might relate to schizophrenia. BDNF has been assayed in this disorder by various means, with findings of reduced expression in postmortem brain (see, e.g., Knable et al., 2004). Genetic association studies have failed to find consistent links between the BDNF gene and schizophrenia (most recently a negative study from Chen et al., 2006), though two recent studies report associations when the phenomenon under study is simplified to either psychosis (Rosa et al., 2006) or schizophrenia with a lifetime history of depressive symptoms (Schumacher et al., 2005). This last group also found an association of the BDNF gene with major depression, illustrating the revived interest in the notion of common etiologic factors, including genetic factors, for schizophrenia spectrum and affective disorders, especially the disorders that share psychosis as a symptom (for review, see Maier et al., 2005; Craddock et al., 2006).—Hakon Heimer.

Berton O, McClung CA, DiLeone RJ, Krishnan V, Renthal W, Russo SJ, Graham D, Tsankova NM, Bolanos CA, Rios M, Monteggia LM, Self DW, Nestler EJ. Science. 10 Feb 2006;311(5762):864-8. Abstract

Comments on News and Primary Papers
Comment by:  NN Kudryavtseva
Submitted 23 February 2006
Posted 23 February 2006

Berton and colleagues show very impressive data of molecular studies demonstrating numerous changes of gene expression in brain under repeated social defeats. However, the behavioral or pharmacological data that the authors use to support the development of depression in socially defeated mice may be interpreted otherwise.

The authors used decreases in the level of social communication (they called it avoidance-approach behavior) in defeated losers as parameters of depression. We repeatedly noted in our experiments on the social model of depression induced by social confrontations in mice of the C57BL/6J strain (Kudryavtseva et al., 1991) that even one or two social defeats lead to a decrease of communication in mice. Thus, avoidance behavior cannot be used as a specific parameter of depression; rather, it may represent anxiety. However, our experiments demonstrated that longer experience of defeats over 20-30 days (but not 10 days, as used by Berton et al.) in male mice produces development of a depression-like state (anxious depression): similarities of symptoms, etiological factors (social unavoidable emotional stress, permanent anxiety), sensitivity to chronic antidepressants and anxiolytics (imipramine, tianeptine, citalopram, fluoxetine, buspirone, etc.), as well as brain neurochemistry changes (serotonergic and dopaminergic systems) (Kudryavtseva et al., 1991; for reviews see Kudryavtseva, Avgustinovich, 1998; Avgustinovich et al., 2004).

In our molecular studies, we also demonstrated changes of gene expression in the brains of male mice after daily agonistic interactions. Three experimental groups were compared: the losers with repeated experience of social defeats; winners with repeated aggression accompanied by social victories; and controls (very important—the same strain). In has been shown that MAOA and SERT mRNA levels in the raphe nuclei of the losers were higher than in the controls and winners. TH and DAT gene expression in the ventral tegmental area was higher and κ opioid receptor gene expression was lower in the winners in comparison with the losers and controls (see Filipenko et al., 2001; 2002; Goloshchapov et al., 2005; reviewed in Kudryavtseva et al., 2004). Thus, there are different specific changes in gene expression in different brain areas in male mice with opposite social behaviors—winners and losers.

As for BDNF, there is an emerging body of data suggesting that different mood disorders are associated with changed BDNF. I think that changes of BDNF gene expression in the losers may be nonspecific for depression state. Expression of the BDNF gene in the winners should be investigated to confirm or reject this idea.

Again, Berton et al. (2006) have demonstrated very impressive data. Taking into consideration these data and our molecular studies, it may be suggested that the sensory contact paradigm (sensory contact model) may be used for the study of association between agonistic behavior and gene expression. We called this scientific direction “From behavior to gene” (reviewed in Kudryavtseva et al., 2004), as an addition to the traditional “From gene to behavior.”


Kudryavtseva NN, Bakshtanovskaya IV, Koryakina LA. Social model of depression in mice of C57BL/6J strain. Pharmacol Biochem Behav. 1991 Feb;38(2):315-20. Abstract

Kudryavtseva NN, Avgustinovich DF. (1998) Behavioral and physiological markers of experimental depression induced by social conflicts (DISC). Aggress Behav. 24:271-286.

Filipenko ML, Alekseyenko OV, Beilina AG, Kamynina TP, Kudryavtseva NN. Increase of tyrosine hydroxylase and dopamine transporter mRNA levels in ventral tegmental area of male mice under influence of repeated aggression experience. Brain Res Mol Brain Res. 2001 Nov 30;96(1-2):77-81. Abstract

Filipenko ML, Beilina AG, Alekseyenko OV, Dolgov VV, Kudryavtseva NN. Repeated experience of social defeats increases serotonin transporter and monoamine oxidase A mRNA levels in raphe nuclei of male mice. Neurosci Lett. 2002 Mar 15;321(1-2):25-8. Abstract

Kudryavtseva et al. (2004) Changes in the expression of monoaminergic genes under the influence of repeated experience of agonistic interactions: From behavior to gene. Genetika, 40(6):732-748.

Avgustinovich DF, Alekseenko OV, Bakshtanovskaia IV, Koriakina LA, Lipina TV, Tenditnik MV, Bondar' NP, Kovalenko IL, Kudriavtseva NN. [Dynamic changes of brain serotonergic and dopaminergic activities during development of anxious depression: experimental study] Usp Fiziol Nauk. 2004 Oct-Dec;35(4):19-40. Review. Russian. Abstract

Goloshchapov AV, Filipenko ML, Bondar NP, Kudryavtseva NN, Van Ree JM. Decrease of kappa-opioid receptor mRNA level in ventral tegmental area of male mice after repeated experience of aggression. Brain Res Mol Brain Res. 2005 Apr 27;135(1-2):290-2. Epub 2005 Jan 6. Abstract

View all comments by NN Kudryavtseva

Comments on Related News

Related News: Rett-icent Neurological Disorder Reveals Some Secrets

Comment by:  Mary Reid
Submitted 9 February 2006
Posted 10 February 2006

The suggestion by Rudolph Jaenisch and colleagues that decreased BDNF may in fact explain the brain pathology seen in Rett syndrome is most interesting.

I would like to propose that it is possible that Mecp2 mutations resulting in increased leptin levels may explain this. Melzner and colleagues (1) report that methyl-CpG binding proteins participate in transcriptional repression and regulation of the human leptin gene. Brunetti et al. (2) find that leptin inhibits depolarization-induced norepinephrine and dopamine release. Ivy and colleagues (3) find that noradrenergic and serotonergic blockade inhibits BDNF mRNA activation following exercise and antidepressant. Is it possible that increased leptin as a consequence of an Mecp2 mutation has resulted in reduced noradrenergic stimulation of BDNF transcription? The study by the Elefteriou group (4) reporting that elevating serum levels reduces bone mass, which is a characteristic of Rett syndrome, would seem to support the above hypothesis.


1. Melzner I, Scott V, Dorsch K, Fischer P, Wabitsch M, Bruderlein S, Hasel C, Moller P. Leptin gene expression in human preadipocytes is switched on by maturation-induced demethylation of distinct CpGs in its proximal promoter. J Biol Chem. 2002 Nov 22;277(47):45420-7. Epub 2002 Sep 3.

2. Brunetti L, Michelotto B, Orlando G, Vacca M. Leptin inhibits norepinephrine and dopamine release from rat hypothalamic neuronal endings. Eur J Pharmacol. 1999 May 21;372(3):237-40.

3. Ivy AS, Rodriguez FG, Garcia C, Chen MJ, Russo-Neustadt AA. Noradrenergic and serotonergic blockade inhibits BDNF mRNA activation following exercise and antidepressant. Pharmacol Biochem Behav. 2003 Apr;75(1):81-8.

4. Elefteriou F, Takeda S, Ebihara K, Magre J, Patano N, Kim CA, Ogawa Y, Liu X, Ware SM, Craigen WJ, Robert JJ, Vinson C, Nakao K, Capeau J, Karsenty G. Serum leptin level is a regulator of bone mass. Proc Natl Acad Sci U S A. 2004 Mar 2;101(9):3258-63. Epub 2004 Feb 20.

View all comments by Mary Reid

Related News: Rett-icent Neurological Disorder Reveals Some Secrets

Comment by:  Mary Reid
Submitted 13 March 2007
Posted 14 March 2007

I had proposed that mutations in MECP2 may result in increased leptin levels in Rett syndrome. I see that Blardi and colleagues report that patients with classic Rett syndrome and preserved-speech variant had leptin values significantly higher than controls. Ashwood et al. also report elevated leptin levels in those with early-onset autism.

View all comments by Mary Reid

Related News: Prozac and Plasticity—Antidepressant’s Action an Eye Opener

Comment by:  Keri Martinowich
Submitted 30 April 2008
Posted 1 May 2008

The recent paper by Vetencourt et al., showing that fluoxetine can restore neuronal plasticity in the adult visual system, has quite obvious and exciting potential applications for the treatment of amblyopia. However, beyond this potential unexpected use for fluoxetine in the clinical treatment of eye disorders, lie implications and new insight into antidepressant mechanisms of actions in mood disorders. It has been speculated for some time that the need for chronic treatment with antidepressants to achieve a therapeutic effect is dependent on changes in neuronal and synaptic plasticity. Time and again, regulation of BDNF has emerged as a candidate underlying various depressive and/or anxiety-like phenotypes, as well as being a possible mediator of the effect of antidepressant/mood-stabilizer drugs. It is now well accepted that beyond its role as a trophic factor during development, BDNF plays a key role in regulating neuronal plasticity in the adult central nervous system.

Clinicians, patients and clinical trials alike attest that antidepressants have strong effects, but the understanding of how exactly they exert their function has remained a mystery despite intense speculation and study. Mechanistic understanding of antidepressant function may provide valuable clues about the functional effects of these drugs, which will likely facilitate our ability to develop treatments with faster therapeutic onset and fewer or less severe side effects by allowing us to precisely target the necessary sites of action.

Reports like the current one are refreshing to the field, reminding us that sometimes, the answers to our questions might be best answered by looking in a different direction rather than focusing and refocusing on the current studies. The sites of action of antidepressant drugs and mood stabilizing therapies are likely to be not just in one place, but in several or even many, interacting circuits within the limbic system of the brain. Teasing apart these circuits and visualizing the effects that antidepressants have on these functional circuits is of course, technically challenging. The comparative simplicity and better understanding of the visual system, makes it an attractive system for the present type of molecular studies. Clearly, biology has made use of redundancy with systems that work well. This biological redundancy may afford us the use of more manageable systems that allow us to observe physiological effects of antidepressants and mood-stabilizing drugs that can be extrapolated to areas of the brain important for regulation of mood and behavior.

View all comments by Keri Martinowich