Report

Diverse Psychotomimetics Act Through a Common Signaling Pathway

See allHide authors and affiliations

Science  21 Nov 2003:
Vol. 302, Issue 5649, pp. 1412-1415
DOI: 10.1126/science.1089681

Abstract

Three distinct classes of drugs: dopaminergic agonists (such as D-amphetamine), serotonergic agonists (such as LSD), and glutamatergic antagonists (such as PCP) all induce psychotomimetic states in experimental animals that closely resemble schizophrenia symptoms in humans. Here we implicate a common signaling pathway in mediating these effects. In this pathway, dopamine- and an adenosine 3′,5′-monophosphate (cAMP)–regulated phospho-protein of 32 kilodaltons (DARPP-32) is phosphorylated or dephosphorylated at three sites, in a pattern predicted to cause a synergistic inhibition of protein phosphatase–1 and concomitant regulation of its downstream effector proteins glycogen synthesis kinase–3 (GSK-3), cAMP response element–binding protein (CREB), and c-Fos. In mice with a genetic deletion of DARPP-32 or with point mutations in phosphorylation sites of DARPP-32, the effects of D-amphetamine, LSD, and PCP on two behavioral parameters—sensorimotor gating and repetitive movements—were strongly attenuated.

Schizophrenia has been associated with dysfunctions of dopaminergic, serotonergic, and glutamatergic neurotransmission (16). Indeed, dopaminergic agonists (e.g., D-amphetamine), serotonergic agonists [e.g., D-lysergic acid diethylamide (LSD)], and glutamatergic antagonists [e.g., phencyclidine (PCP)] are psychotomimetic in some healthy humans and worsen the symptomatology in schizophrenics (79). Schizophrenia is characterized by persistent negative symptoms, such as cognitive and emotional impairments, alternating with episodes of disabling positive symptomatology, such as hallucinations and delusions, implicating a highly dynamic pathophysiology.

Protein phosphorylation is a molecular switch that reversibly alters the function of many neuronal proteins, and its dysregulation could therefore contribute to psychotogenesis. DARPP-32 is a key regulator of kinase-phosphatase signaling cascades modulated by dopaminergic, serotonergic, and glutamatergic neurotransmission (10).

The effects of D-amphetamine, LSD, and PCP on sensorimotor gating (i.e., information processing and stimulus filtering) and repetitive movements (i.e., perseverative behavior and stereotypy) were compared in wild-type and DARPP-32 knockout (KO) mice. Sensorimotor gating is reduced in nonmedicated schizophrenic individuals and after intake of psychotomimetics (11). Prepulse inhibition (PPI) of the startle reflex is, both in humans and in rodents, the most appropriate test for sensorimotor gating currently available (11). Similarly, repetitive behaviors are often seen in schizophrenics (12), and drugs that elicit psychosis in humans also increase repetitive movements in rodents (2, 13). D-Amphetamine, LSD, and PCP disrupted PPI and caused an increase in repetitive movements in wild-type mice, but not in DARPP-32 KO mice (Fig. 1).

Fig. 1.

Sensorimotor gating and repetitive movements in response to treatment with distinct classes of psychotomimetic drugs in wild-type and DARPP-32 KO mice. Effects of (A and D) d-amphetamine, (B and E) LSD, and (C and F) PCP on (A to C) prepulse inhibition of the startle response (PPI 85 dB) and (D to F) repetitive movements in wild-type (WT) and DARPP-32 KO mice. In studies of prepulse inhibition, mice were injected intraperitoneally with d-amphetamine, LSD, and PCP at 7.5, 0.3, and 10 mg/kg, respectively, and in studies of repetitive movements mice at 2.5, 0.1, and 3 mg/kg. Black bars represent means (±SEM) for wild-type mice and white bars, for DARPP-32 KO mice. ***P < 0.001, **P < 0.01, *P < 0.05 compared with vehicle-treated mice; ### P < 0.001, ## P < 0.01, # P < 0.05 compared with wild-type mice given the same treatment by two-way analysis of variance (ANOVA) followed by Duncan's test.

Four distinct phosphorylation sites determine the function of DARPP-32 (10). Phosphorylation at Thr34 converts DARPP-32 into a potent inhibitor of protein phosphatase–1 (PP1) (14). Phosphorylated Ser97 increases the ability of protein kinase A (PKA) to phosphorylate DARPP-32 at Thr34 (15). Phosphorylated Ser130 prevents the dephosphorylation at Thr34 by protein phosphatase 2B (PP2B) (16). Phosphorylated Thr75 converts DARPP-32 into an inhibitor of PKA (17).

We studied the effects of D-amphetamine, LSD, and PCP on the phosphorylation state of Thr34–, Thr75–, Ser97–, and Ser130–DARPP-32 in normal mice and on other potential downstream targets of DARPP-32. Tissue samples were collected from frontal cortex, striatum, hippocampus, and cerebellum, because these structures contain DARPP-32 (10), are involved in sensorimotor gating (11), and are dysfunctional in schizophrenic individuals (1). D-Amphetamine, LSD, and PCP increased the levels of phosphorylated Thr34– and Ser130–DARPP-32 in frontal cortex and striatum (Fig. 2). Thr75–DARPP-32 was decreased by both D-amphetamine and LSD, but unaltered by PCP, in these regions. None of these psychotomimetics significantly regulated Ser97–DARPP-32 (18). The regulation of Ser133-CREB and Ser9–GSK-3β by D-amphetamine, LSD, and PCP was similar to that seen for Thr34– and Ser130–DARPP-32 (Fig. 2). There were no significant effects on Ser9-synapsin, Tyr705–STAT-3 (signal transducer and activator of transcription–3), or Thr286-CaMKII (calcium/calmodulin-dependent protein kinase II) in cortex or striatum (18), which indicated that these drugs were not nonspecifically altering protein phosphorylation. Analysis of tissue samples from hippocampus and cerebellum failed to reveal a consistent pattern of the effects of D-amphetamine, LSD, and PCP on Thr34–, Thr75–, and Ser130–DARPP-32 or on Ser133-CREB and Ser9–GSK-3β (18), which indicated regional specificity of the observed effects. Thus, D-amphetamine, LSD, and PCP share the ability to regulate the phosphorylation state of multiple sites of DARPP-32, Ser133-CREB, and Ser9–GSK-3β in frontal cortex and striatum.

Fig. 2.

Regulation of phosphorylation of DARPP-32, CREB, and GSK-3β in vivo by diverse psychotomimetic substances. Mice were injected intraperitoneally with vehicle, d-amphetamine (2.5 mg/kg), LSD (100 μg/kg), or PCP (3 mg/kg). After 15 min, mice were killed by focused microwave irradiation. Phosphorylation of Thr34–, Thr75–, and Ser130–DARPP-32, Ser133-CREB and Ser9–GSK-3β were studied in (A) frontal cortex and (B) striatum. Data were normalized to total levels for each of these proteins and represent means ± SEM for 8 to 10 mice per group. Qualitatively similar data were obtained with 7.5 mg/kg of d-amphetamine, 300 μg/kg of LSD, and 10 mg/kg of PCP (18). ***P < 0.001, **P < 0.01, and *P < 0.05 compared with vehicle-treated mice by one-way ANOVA followed by Dunnett′s test.

Frontal cortex and striatum are major recipients of dopaminergic inputs. We performed three series of experiments to evaluate whether altered dopaminergic neurotransmission contributed to the effects of D-amphetamine, LSD, and PCP on Thr34–DARPP-32 phosphorylation. In one series, the psychotomimetics were studied in mice pretreated with the dopaminergic toxin, 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine (MPTP). In such dopamine-depleted mice, the effect of D-amphetamine on Thr34–DARPP-32 phosphorylation was significantly attenuated (40% of wild type), but the effects of LSD or PCP (103 and 90% of wild type, respectively) were not. In a second series of experiments, the effects of the psychotomimetics were studied in horizontal striatal slices. In these slices, D-amphetamine, LSD, and PCP increased Thr34– and Ser130–DARPP-32 phosphorylation and decreased Thr75–DARPP-32 phosphorylation (table S1) as in whole animals (Fig. 2). Cocaine, but not procaine, mimicked the action of D-amphetamine on Thr34– and Thr75–DARPP-32. Serotonin mimicked the action of LSD. Ketamine and MK-801 [psychotomimetic N-methyl-D-aspartate (NMDA) receptor antagonists], but not 6-cyano-7-nitroquinoxaline-2,3-dione (CNQX, an AMPA receptor antagonist), mimicked the action of PCP. In a third series of experiments, the effects of the psychotomimetics were compared in slices made from wild-type and dopamine D1 receptor KO mice. The effect of D-amphetamine on Thr34–DARPP-32 phosphorylation was significantly attenuated (35% of wild type) in D1 KO mice, whereas the effects of LSD or PCP (90 and 88% of wild type, respectively) did not significantly differ between wild-type and D1 KO mice. Taken together, these results indicate that increased dopaminergic neurotransmission underlies the effect of D-amphetamine, but not of LSD or PCP, on Thr34–DARPP-32 phosphorylation. These data are consistent with a model in which D-amphetamine, LSD, and PCP predominantly modulate DARPP-32 phosphorylation through distinct actions on the dopaminergic, serotonergic, and glutamatergic systems, respectively.

To clarify the role of Thr34–, Thr75–, and Ser130–DARPP-32 in the actions of psychotomimetics, we used a gene knock-in method (Fig. 3A) to generate three lines of mutant mice in which Thr34–, Thr75–, or Ser130–DARPP-32 was replaced with alanine to prevent phosphorylation (T34A, T75A, and S130A mutants). We confirmed these mutations by immunoblot analysis (Fig. 3A). Using these phosphomutant mice, we examined the effects of psychotomimetics on the phosphorylation state of Ser133-CREB and Ser9–GSK-3β, both of which are regulated by both PP1 and PKA. No basal differences were observed in the phosphorylation state of Ser133-CREB or Ser9–GSK-3β between wild-type and phosphomutant mice (18). The psychotomimetic-induced increases of phospho-Ser133-CREB and phospho-Ser9-GSK-3β were significantly reduced in frontal cortex (Fig. 3B, 1 and 3) and striatum (Fig. 3B, 2 and 4), respectively, from T34A mutants, but unaltered in T75A mutants. Psychotomimetic-induced phospho-Ser133-CREB in frontal cortex and striatum from S130A mutants was also significantly reduced (Fig. 3B, 1 and 2). Furthermore, the effects of the psychotomimetics on Ser9-GSK-3β were reduced in striatum from these mutants. These studies demonstrate that DARPP-32 signaling pathways play an integral role in the effects of all three psychotomimetics on phosphorylation of GSK-3β and CREB.

Fig. 3.

Regulation of phosphorylation of Ser133-CREB and Ser9–GSK-3β and of c-fos mRNA expression by distinct classes of psychotomimetic drugs in wild-type and phosphomutant mice. (A) Introduction of point mutations in the DARPP-32 gene by yeast homologous recombination. (Top) Schematic drawing showing the strategy using a modified yeast-based system to introduce point mutations in a genomic clone of DARPP-32 (for details, see supporting online material). (Bottom left) The phosphomutant mice can be distinguished from their wild-type littermates by polymerase chain reaction (PCR). (Bottom right) Confirmation of the genotype of the phosphomutant mice by immunoblotting with phospho-specific antibodies against Thr34–, Thr75–, and Ser130–DARPP-32, respectively. (B) Effects of vehicle, d-amphetamine (2.5 mg/kg), LSD (100 μg/kg), and PCP (3 mg/kg) on the phosphorylation state of (1 and 2) Ser133-CREB, and (3 and 4) Ser9–GSK-3β and on (5 and 6) c-fos mRNAin (1, 3, and 5) frontal cortex and (2, 4, and 6) striatum from wild-type, T34A, T75A, and S130A phosphomutant mice. All data were normalized to vehicle-injected mice of the same genotype and represent means ± SEM for 8 to 10 mice per group. **P < 0.01 and *P < 0.05 compared with vehicle-treated mice; #P < 0.05 compared with wild-type mice given the same treatment by one-way ANOVA followed by Dunnett′s test.

To further characterize the involvement of DARPP-32 in mediating the actions of the psychotomimetics, c-fos mRNA was studied in wild-type and phosphomutant mice by in situ hybridization. c-fos, an immediate early gene stimulated by Ser9-GSK-3β [by an indirect disinhibitory mechanism (19)] and by Ser133-CREB, [by direct positive regulation (20)], is a measure of postsynaptic biochemical activation. All three psychotomimetics increased the levels of c-fos mRNA in cortical areas and in subregions of striatum from wild-type mice (Figs. 3B, 5 and 6, and fig. S1). The ability of all the psychotomimetics to increase c-fos mRNA in cingulate cortex and in the paraventricular region of striatum, a subregion that receives strong innervation from prefrontal and cingulate cortices, was reduced in T34A and S130A mutants (Fig. 3B, 5 and 6). Double-labeling in situ hybridization experiments showed that D-amphetamine, LSD, and PCP increased c-fos mRNA in neurons both with and without DARPP-32 throughout cortex (18). In striatum, the psychotomimetics induced c-fos expression in DARPP-32–positive neurons (18). These studies demonstrate that DARPP-32 signaling pathways also play an integral role in the effects of all three psychotomimetics on transcription of c-fos mRNA.

We next compared the behavioral effects of the psychotomimetics in the phosphomutant mice. The effects of D-amphetamine, LSD, and PCP on PPI and repetitive movements were all significantly attenuated with the T34A point mutation (Fig. 4). Furthermore, the effects of D-amphetamine and LSD on both behavioral parameters, as well as the effects of PCP on repetitive movements, were attenuated in the S130A phosphomutant mice (Fig. 4). Also, the effects of PCP were attenuated in T75A phosphomutant mice (Fig. 4).

Fig. 4.

Sensorimotor gating and repetitive movements in response to treatment with distinct classes of psychotomimetic drugs in wild-type and phosphomutant mice. Effects of (A and D) d-amphetamine, (B and E) LSD, and (C and F) PCP on (Ato C) prepulse inhibition of the startle response (PPI 85 dB) and (D to F) repetitive movements in wild-type, and T34A, T75A, and S130A mutants. In studies of prepulse inhibition, mice were injected intraperitoneally with d-amphetamine, LSD, and PCP at 7.5, 0.3, and 10 mg/kg, respectively, and in studies of repetitive movements at 2.5, 0.1, and 3 mg/kg. Data represent means ± SEM for 6 to 10 mice per group. ***P < 0.001, **P < 0.01 compared with vehicle-treated mice; ##P < 0.01, #P < 0.05 compared with wild-type mice given the same treatment by two-way ANOVA followed by Duncan′s test.

Our data point to the unexpected finding that Ser130–DARPP-32 plays a major role in the actions of psychotomimetic drugs. In vitro studies have shown that casein kinase 1–mediated phosphorylation of Ser130–DARPP-32 prevents the dephosphorylation by PP2B of phospho-Thr34–DARPP-32 (16). Consistent with this, in vivo, the psychotomimetics induced significantly more phospho-Thr34–DARPP-32 in wild-type mice than in S130A mutants in striatum (341 ± 23.2 versus 231 ± 16.6, 159 ± 8.8 versus 127 ± 10.7, and 185 ± 13.0 versus 141 ± 8.5 for D-amphetamine, LSD, and PCP, respectively), supporting a modulatory role for Ser130–DARPP-32 in the regulation of Thr34–DARPP-32 phosphorylation. The data indicate a hierarchical interdependence between Ser130– and Thr34–DARPP-32.

Our data demonstrate that the effects of D-amphetamine, LSD, and PCP on sensorimotor gating and repetitive movements depend on DARPP-32. These psychotomimetics regulate, through distinct mechanisms, phosphorylation of DARPP-32 as well as downstream effectors in the cortex and the striatum. Analysis of psychotomimetic-induced changes in phosphorylation of several sites on DARPP-32, complemented by behavioral analysis of the effects of these same drugs, indicates that phosphorylation at multiple residues of DARPP-32 (and by implication the kinases and phosphatases involved in the regulation of these phosphorylation sites) is imperative for biochemical and behavioral actions of psychotomimetic drugs.

A general pattern emerges from this study. Three pathways, involving regulation of the state of phosphorylation of Thr34–, Thr75–, and Ser130–DARPP-32, inhibit PP1, which leads to an increased state of phosphorylation of various PP1 substrates. Future studies may identify the precise PP1 substrates involved in the behavioral effects of psychotomimetic drugs.

Supporting Online Material

www.sciencemag.org/cgi/content/full/302/5649/1412/DC1

Materials and Methods

Figs. S1 and S2

Table S1

References and Notes

References and Notes

View Abstract

Stay Connected to Science

Navigate This Article