How a single protein was found to link schizophrenia and depression to drugs of abuse and addiction.
By Per Svenningsson and Paul Greengard 

In the early 1980s, a search for brain proteins that might be involved in communication between nerve cells turned up a 32-kilodalton protein-kinase substrate with some remarkable properties. Our group had been working under the assumption that the signaling pathways that regulate synaptic transmission might resemble the signaling pathways working elsewhere in the body, particularly the endocrine signaling involved in the breakdown of glycogen in muscle and liver.

Inspired by the work of Earl Sutherland and Edwin Krebs, who showed that cyclic AMP acting through a cyclic AMP-dependent protein kinase (PKA) facilitated this breakdown, we were tantalized by the possibility that analogous systems might be operating in the brain to produce physiological responses. John Kebabian, James Nathanson, Jy-Fan Kuo, Howard Schulman, and Mary Kennedy began looking...


Around this time we wondered if there might also be brain-region-specific or even neuron-specific phosphoproteins. The striatum was a good starting place because there are few cell types. Exposing striatal tissue to dopamine or cAMP revealed a whole family of phosphoproteins, one of which was easily extracted by acid. So in 1983, Ivar Walaas and Dana Aswad came to characterize what would become known as dopamine- and cAMP-regulated phosphoprotein, Mr 32kDa (DARPP-32).1

Fast forward 25 years, and DARPP-32 has emerged as a master regulator in neuronal responses.2 Found as a major target for dopamine-activated PKA in medium spiny neurons in striatum, DARPP-32 plays a critical role in regulating kinase/phosphatase signaling cascades in striatonigral neurons, which contain high levels of D1 receptors, as well as in striatopallidal neurons, which predominantly express D2 receptors. Aberrations in the cascades that pass through this protein contribute to the etiology of several common neuropsychiatric disorders, including schizophrenia, depression, and addiction to drugs of abuse. Although present in far lower concentrations, it also regulates functions of neuronal subpopulations in cortex, hippocampus, hypothalamus, and cerebellum. Its scarcity, but functional importance, in these neuronal tissues speaks to the potency of the protein once in its activated form.

In 1984 we found that phosphorylation by PKA at Thr34 converts DARPP-32 into a powerful high-affinity inhibitor of the multifunctional serine/threonine protein phosphatase, PP-1.3 PP-1 which in turn, regulates the phosphorylation state and activity of many physiological effectors, including neurotransmitter receptors, voltage-gated ion channels, and transcription factors.2 This positioned DARPP-32 as a powerful player in regulating neuronal physiology, and it made us question whether other neurotransmitters in the striatum might also affect its phosphorylation. It turned out that every neurotransmitter known to affect the physiology of the cells in the striatum also affects the phosphorylation state of DARPP-32, but in different ways. Some increase phosphorylation, some decrease it; conversely, some increase dephosphorylation, and some decrease it.

The dynamics of DARPP-32's circuitry are context dependent. For example, recent studies show that dopamine regulates DARPP-32 phosphorylation in both directions (see figure).4 Dopamine, via D1 receptor-mediated activation of PKA, stimulates phosphorylation of Thr34. Conversely, dopamine, via the D2 receptors, inhibits PKA and Thr34 phosphorylation. The downregulation of Thr34 phosphorylation by D2 receptor agonists also involves activation of a phospholipase C signaling cascade that includes PP-2B, a protein responsible for dephosphorylating DARPP-32 at Thr34.

Glutamatergic activation of PP-2B, via NMDA and AMPA receptor activation, is the major mechanism involved in dephosphorylation at Thr34 in medium spiny neurons.5 Activation of mGlu5 receptors stimulates the phosphorylation of DARPP-32 at Thr34 by potentiating adenosine A2A receptor/cAMP/PKA signaling.5 Adenosine A2A receptors are selectively expressed in the striatopallidal neurons, and stimulation of these receptors counteracts inhibitory effects of D2 receptors on Thr34 phosphorylation.

The pathophysiology of depression remains largely unknown.

During the period from the late 1980s through the 1990s, it was found that DARPP-32 is phosphorylated at multiple sites in addition to Thr34. Phosphorylation at Thr75, by Cdk5, converts the protein into an inhibitor of PKA.6 Phosphorylation at Ser97 by CK2 increases the ability of PKA to phosphorylate Thr34. CK1-mediated phosphorylation of DARPP-32 at Ser130 limits dephosphorylation of Thr34 by PP-2B. Serotonin, via 5-HT4 or 5-HT6 receptor-mediated activation of PKA, increases phospho-Thr34, but decreases phospho-Thr75. By activating 5-HT2A/C receptors, serotonin increases Ser130 phosphorylation.

As an increasing number of biochemical cascades controlled by DARPP-32 phosphorylation were observed, it became important to determine whether these cascades were physiologically relevant. This was made possible by technological advances in electrophysiological recording, and perhaps to an even greater extent by the generation of mouse mutants. The generation of DARPP-32 knockout (KO) mice in 1998,7 and the more recent generation in 2003 of mice in which individual phosphorylation sites have been mutated,8 cemented the idea that DARPP-32 mediates the biochemical, electrophysiological, gene transcriptional, and behavioral effects of the neurotransmitters dopamine, serotonin, glutamate, GABA, and adenosine.

Dysfunctions of these neurotransmitters are involved in the pathobiology of several common psychiatric diseases. However, the aberrations in intracellular signaling that arise from dysfunctions in the neurotransmitter systems are still far from understood. During recent years, a major focus of our laboratory has been to determine the extent of DARPP-32's involvement in molecular and cellular disturbances in various disease states.


The dopaminergic system stimulates reward-related behaviors. By virtue of its regulation by dopamine, as well as by other neurotransmitters linked to the actions of drugs of abuse, such as serotonin and glutamate, DARPP-32 was positioned to play an important role in either mediating or modulating the short- and long-term actions of drugs of abuse. In the beginning of the 1990s, we began investigating how drugs of abuse might affect the phosphorylation state of DARPP-32.

Cocaine and amphetamine - Cocaine inhibits dopamine reuptake, whereas amphetamine promotes dopamine release from nerve terminals through a weak-base-mediated reverse transport mechanism. Acute treatment of mice with cocaine or amphetamine increases phosphorylation of DARPP32 at Thr34 and decreases phosphorylation at Thr75. In contrast, chronic treatment with cocaine upregulates the expression of both cdk5 and p35, a cofactor for cdk5, in striatum. This leads to increased phosphorylation of Thr75 consequently decreasing phosphorylation at Thr34.9

The ERK pathway is important for long-term synaptic plasticity at both transcriptional and posttranscriptional levels in reward-related brain circuitries and is activated by many widely abused drugs. This pathway is controlled by protein phosphorylation. Jean-Antoine Girault decided in 2005 to study the involvement of DARPP-32 in regulating the phosphorylation events in the ERK pathway. Interestingly, ERK activation by cocaine and d-amphetamine was absent in DARPP-32 knockout mice and in mice with a point mutation of Thr34. Treating wild type animals with cocaine or amphetamine also increases the phosphorylation state of CREB and Fos/Jun genes, including DFosB, which may coordinate alterations in gene expression, leading to long-term changes in neuronal function underlying addiction. Notably, Eric Nestler found that cocaine-mediated activation of CREB, c-Fos, and DFosB is strongly attenuated in DARPP-32 knockout mice. Thus, there is biochemical data suggesting a role for DARPP-32 in the actions of psychostimulants.

In a series of experiments, using different tests, we have also provided evidence that DARPP-32 is involved in various behavioral effects of psychostimulants. In 1998, we showed that locomotor responsiveness to a single injection of cocaine is attenuated in DARPP-32 knockout mice compared to wild-type mice. Similarly, the acquisition of cocaine-induced place-preference is also significantly attenuated in DARPP-32 knockout mice as well as in Thr34Ala-DARPP-32 phosphomutant mice. Sensitization to cocaine administration, a well-known behavioral effect of repeated cocaine administration, can be either enhanced or diminished in DARPP-32 knockout mice. The discrepancy remains to be clarified, but may be related to the differences in the relative levels of phospho-Thr75 and phospho-Thr34.

A recent study using phosphomutant mice has shown that acquisition of cocaine self-administration required significantly more time in Thr34Ala-DARPP-32 mice and that both Thr34Ala- and Ser130Ala-mice self-administered more cocaine than their respective wild-type controls. Notably, levels of phospho-Thr34 and phospho-Ser130 were reduced after self-administration of cocaine in wild-type mice. Likewise, the ability of D-amphetamine to disrupt prepulse inhibition of the startle response and induce repetitive movements was significantly counteracted in both Thr34Ala- and Ser130Ala-DARPP-32 mice.8 Thus, the phosphorylation states of Thr34 and Ser130 play important roles in modulating the effects of cocaine and amphetamine.

LSD and PCP - D-amphetamine, LSD, and PCP represent three distinct classes of psychotomimetic agents. In 2003 we showed that all three of these compounds upregulate Thr34- and Ser130 phosphorylation.8 We also found that the disruption of prepulse inhibition of the startle response and the induction of repetitive movements by amphetamine, LSD, and PCP are counteracted in the DARPP-32 knockout mice as well as in Thr34Ala-DARPP-32 and Ser130Ala-DARPP-32 mice. These studies demonstrated that all these classes of psychotomimetics, through distinct upstream signaling pathways, activate the same final common pathway and help to explain the fact that all these drugs have similar psychotomimetic actions.


The pathophysiology of schizophrenia remains poorly understood, but there appears to be a hyperfunction of the dopamine system at least in some brain regions. Also, emerging genetic evidence suggests that there is a hypofunction of NMDA receptor-mediated neurotransmission in a subclass of schizophrenic patients. These lines of evidence along with several others were suggestive that disturbances in DARPP-32 function may occur in schizophrenia.

In 2002 we found that DARPP-32 levels in prefrontal cortex are decreased in some patients with schizophrenia,10 indicating that perturbations in DARPP-32-regulated signal transduction cascades may occur. Just as experimental evidence from animal studies in 2003 implicated DARPP-32 in mediating the actions of psychotostimulants, administration of the typical antipsychotics haloperidol, raclopride, or eticlopride increases Thr34 phosphorylation. This likely happens in cell populations distinct from those affected by psychotomimetics. Much higher concentrations of the typical neuroleptic agent raclopride are required to induce catalepsy in DARPP-32 knockout mice than in their wild-type littermates. Further, several additional behavioral effects (reduction in sniffing, locomotion, rearing, grooming, and chewing), exerted through D2 receptor antagonism, are also reduced in DARPP-32 knockout mice.


Like schizophrenia, the pathophysiology of depression remains largely unknown, but several lines of evidence have suggested impairments of synaptic plasticity, synapse morphology, and cell proliferation. Such evidence emerged from neuroanatomical, biochemical and electrophysiological studies. The immediate action of many currently used antidepressants is to increase the synaptic availability of serotonin and/or noradrenaline. It is, however, unlikely that this early action, per se, is antidepressant, but rather that the increased levels of these neurotransmitters initiate cellular adaptations that ultimately lead to an antidepressant action. Thus there is a temporal delay in the onset of action of all antidepressants. Alterations in intracellular signaling probably underlie some of these adaptations, as agents that modulate intracellular signaling, such as rolipram (a phosphodiesterase type 4 inhibitor) and lithium (a GSK-3 inhibitor), are effective in the treatment of mood disorders. In 2002, we performed a series of biochemical experiments demonstrating that serotonin can regulate DARPP-32 phosphorylation, indicating that intracellular signaling cascades regulated by serotonin/DARPP-32 pathways may be implicated in the pathophysiology of bipolar and unipolar depression.

In fact, as early as the 1990s, Nestler had shown that chronic treatment with the mood-stabilizer, lithium, increases DARPP-32 levels in the rat

DARPP-32 has enabled our work to branch out in a multitude of directions.

frontal cortex. In 2004, using a convergent approach that integrates brain gene-expression data from mice treated with a stimulant, methamphetamine, and a mood stabilizer, valproate, with linkage loci data from human genetic studies and changes in postmortem brains from patients, DARPP-32 was reported to be a major candidate gene for bipolar depression.

Since we had found that serotonin regulates DARPP-32 phosphorylation, we examined the effects of the selective serotonin reuptake inhibitor, fluoxetine (Prozac), on DARPP-32 phosphorylation.11 Either acute or chronic administration of fluoxetine increases phosphorylation of Thr34 and decreases phosphorylation at Thr75. Acute, but not chronic, administration of fluoxetine also increases phosphorylation at Ser130. The evidence suggested that several signaling cascades account for the effects on these three phosphorylation sites, and the changes work synergistically, each contributing to the inhibition of PP-1. Chronic treatment with fluoxetine or imipramine also increased the total levels of DARPP-32 mRNA and protein in prefrontal cortex and hippocampus.

Phosphorylation at Thr34 of DARPP-32 increases the phosphorylation state and efficacy of several ionotropic receptors, including AMPA receptors. Allosteric potentiators of AMPA receptor function have robust antidepressant-like actions in tests of "learned helplessness," such as the forced-swim test and the tail-suspension test. Fluoxetine decreased immobility in wild type mice in the tail suspension test. Interestingly, the effect of fluoxetine in this test was significantly attenuated in DARPP-32 knockout mice. Thus, the data strongly suggested to us that a signaling pathway involving DARPP-32 induces an increase in AMPA receptor phosphorylation and conductance and may play an important role in mediating the antidepressant actions of fluoxetine. This work was key to our realization of the importance of DARPP-32 in areas of the brain other than the striatum.


The evidence that DARPP-32 is involved in the serotonin/serotonin-receptor signaling pathway that mediates the antidepressant effects of fluoxetine led us to a search for proteins that might modulate the actions of serotonin receptors. Earlier this year, using protein interaction cloning, we uncovered a serotonin receptor-interacting protein, p11, the level of which may correlate with the liability to depression.12

Using protein interaction cloning, 5-HT1B receptors were found to interact with p11. P11 is a member of the S100 EF-hand protein family, but it is unique in this family for having lost its calcium-binding properties. Previous work showed that p11 dimers form a tetrameric complex with annexin II to enhance its expression at the cell surface. Several papers have reported that p11 increases the levels of ion channels, including NaV1.8, TASK-1, TRP5/6, and ASIC-1 at the cell surface. Likewise, p11 increases localization of 5-HT1B receptors at the cell surface.

Given the importance of serotonin and 5-HT1B receptors in depression-, stress-, and anxiety-related states, we then performed experiments to study the involvement of p11 in depression-like states. Interestingly, p11 expression is increased in mice treated with antidepressant drugs or electroconvulsive therapy. Conversely, decreased levels of p11 are found in a genetic animal model of depression and in postmortem brain tissue from depressed patients. Overexpression of p11 increases 5-HT1B receptor function in cells and recapitulates certain actions of antidepressant drugs in mouse behavioral paradigms. Conversely, p11-knockout mice exhibit a depression-like phenotype, and have reduced responsiveness to stimulation of 5-HT1B receptors and to the stimulatory action of imipramine in the tail suspension test of antidepressant efficacy. The possible relationship between DARPP-32 and p11, both of which appear to play important roles in susceptibilities to depression, remains to be determined.


A large body of work has revealed that DARPP-32, when phosphorylated at Thr34, acts as an amplifier of PKA-mediated signaling through its ability to potently inhibit PP-1. This amplifying property of DARPP-32 is critical for dopaminergic signaling, but it is also used by multiple other neurotransmitters, neuromodulators, and neuropeptides in several brain regions, placing it at the center of a vast number of physiological inputs and responses. Accumulating evidence suggests that additional phospho-sites of DARPP-32, including Thr75, Ser97, and Ser130, are functionally important, and the recent generation of mice in which each of these phosphorylation sites have been individually mutated should help dissect their precise roles.

Future studies will also be aimed at delineating the neuronal subpopulations in which phosphorylation of DARPP-32 occurs. For example, the fact that neuroleptics, as well as amphetamine and cocaine, upregulate Thr34-DARPP-32 phosphorylation in striatum suggests that the upregulation of DARPP-32 by neuroleptics and psychostimulants may occur in different subpopulations of striatal neurons. Thus, it seems possible that the psychostimulants preferentially act in the D1-receptor-enriched striatonigral neurons, whereas the neuroleptics, at least the typical ones, act in the D2-receptor-enriched striatopallidal neurons. To test this hypothesis, we are developing mice that express tagged DARPP-32 in striatonigral or striatopallidal neurons, respectively.

DARPP-32, because of its role as an informational integrator, has enabled our work to branch out in a multitude of directions investigating the cell populations, protein kinases, protein phosphatases, and phosphoprotein substrates that mediate the actions of drugs of abuse, neuroleptics, and antidepressants.

Per Svenningsson is an assistant professor in the Department of Physiology and Pharmacology at the Karolinska Institute, Stockholm, Sweden, and at The Rockefeller University. Paul Greengard won the 2000 Nobel Prize in Physiology or Medicine and is Vincent Astor Professor at The Rockefeller University.


DARPP-32 at the center of it all:
View animation

More neuroscience coverage:

Meeting of the Minds
36 recommendations Europeans came up with when they gathered to talk about brain science.

Singing in the Bird Brain
The songbirds studied by Fernando Nottebohm aren't the only ones singing praises.

Patch Clamping Unhooked
How to record a single neuron in a moving animal

Ten Steps to Better Patch Clamping
An expert on the technique shares his secrets.

Paul Greengard and Per Svenningson discuss DARPP-32
Listen as senior editor Brendan Maher interviews Paul Greengard and Per Svenningson about the implications of their work on DARPP-32 and the role of p11 in Prozac's mechanism of action

Owing to space limitations only selected references are given. For a complete reference list, please contact the authors.

1. S.I. Walaas et al., "A dopamine- and cyclic AMP-regulated phosphoprotein enriched in dopamine-innervated brain regions," Nature, 301:69-71, 1983.
2. P. Greengard, "The neurobiology of slow synaptic transmission," Science, 294:1024-30, 2001.
3. H.C. Hemmings Jr., et al., "DARPP-32, a dopamine-regulated neuronal phosphoprotein, is a potent inhibitor of protein phosphatase-1," Nature, 310:503-5, 1984.
4. A. Nishi et al., "Bidirectional regulation of DARPP-32 phosphorylation by dopamine," J Neurosci, 17:8147-55, 1997.
5. P. Svenningsson et al., "DARPP-32: an integrator of neurotransmission," Ann Rev Pharmacol Toxicol, 44:269-96, 2004.
6. J.A. Bibb et al., "Phosphorylation of DARPP-32 by Cdk5 modulates dopamine signaling in neurons," Nature, 402:669-71, 1999.
7. A.A. Fienberg et al., "DARPP-32: regulator of the efficacy of dopaminergic neurotransmission," Science, 281:838-42, 1998.
8. P. Svenningsson et al., "Diverse psychotomimetics act through a common signaling pathway," Science, 302:1412-5, 2003.
9. J.A. Bibb et al., "Effects of chronic exposure to cocaine are regulated by the neuronal protein Cdk5," Nature, 410:376-80, 2001.
10. K.A. Albert et al., "Evidence for decreased DARPP-32 in the prefrontal cortex of patients with schizophrenia," Arch Gen Psychiatry, 59:705-12, 2002.
11. P. Svenningsson et al., "Involvement of striatal and extrastriatal DARPP-32 in biochemical and behavioral effects of fluoxetine (Prozac)," Proc Natl Acad Sci, 99:3182-7, 2002.
12. P. Svenningsson et al., "Alterations in 5-HT1B receptor function by p11 in depression-like states," Science, 311:77-80, 2006.

Interested in reading more?

Magaizne Cover

Become a Member of

Receive full access to digital editions of The Scientist, as well as TS Digest, feature stories, more than 35 years of archives, and much more!
Already a member?