Schizophrenia Research Forum - A Catalyst for Creative Thinking

Persistence of Memory in the Frontal Cortex: Timely Roles for Glutamate and Dopamine

4 February 2009. Unlike long-term potentiation, which can take minutes or more to activate and is relatively persistent, or long-term memory, which also requires the formation of new proteins and connections between neurons, delayed afterpolarization (dADP) provides a new mechanism whereby single neurons sustain a depolarization in the absence of sustained input. An impressive body of work published in the February 1, 2009 issue of Nature Neuroscience by Donald Cooper of University of Texas Southwestern Medical Center, Dallas, Kyriaki Sidiropoulou of the Chicago Medical School, and colleagues describes dADP as a mechanism for high-turnover memory in single neurons. The authors suggest that alterations of this short-term persisting membrane depolarization might contribute to working memory or attention deficits in disorders like schizophrenia.

Using patch-clamp recording from slice cultures of rat limbic cortex, a model for the human prefrontal cortex, they showed that metabotropic glutamate receptor (mGluR) agonism with the compounds ACPD or DHPG induced a long-lasting increase in layer V pyramidal neuron depolarization in response to brief depolarizing stimuli. mGluR agonism did not change the few action potentials generated by the depolarizing stimulation, but generated far more subsequent action potentials, and a far higher membrane depolarization, than normal. These cortical changes were calcium dependent and sodium independent, were not prevented by AMPA or NMDA receptor antagonism, were mostly blocked by the mGluR5 receptor antagonist, MPEP, and were absent in mGluR5 knockout mice. Neither the selective mGluR1 antagonist, LY367385, nor the use of mGluR1 knockout mice, prevented this effect, ruling out the mGluR1 receptor. Group 2 mGlu receptors also did not play a role, as their antagonism with LY431495 did not prevent dADP in response to ACPD.

The greatest dopamine input in the frontal cortex is to layer V pyramidal neurons, which also contain 20-fold more D1 receptors than D2 receptors. D1 agonism with SKF81297 blocked dADP induced by mGluR5 activation, whereas D2 agonism with quinpirole did not. The ability of the D1 antagonist SCH23390, or the protein kinase A (PKA) blocker, H89, to block the SKF81297 effect, and the ability of the PKA activator, forskolin, to mimic the D1 agonist effect, demonstrated that a D1-mediated induction of PKA is one pathway through which dADP activation by glutamate can be attenuated.

While the enabling role of mGluR5 agonism and attenuating effects of D1 agonism on dADP were elegantly revealed in these comprehensive studies, some of the functional implications of these interactions are puzzling. For example, how are increases in dopamine release from ventral tegmental area (VTA) neurons during reward to be reconciled with D1 receptor-mediated decreases in dADP they observed? Such a pairing would seem to lessen behavioral responses to reward or decrease memory of the behavior-reward association. The well-known, inverted U-shaped relationship between D1 activation and prefrontal cortex memory function (see SRF related news story and SRF news story) seems to be opposite for the inhibitory role of D1 receptors in the dADP response to glutamate, unless in vivo dopamine tone is at a maximum. This tone may not be modeled in their in vitro slice preparation, however, since it severs VTA inputs. It was also unclear how the D1-mediated decrease in dADP would "increase the signal-to-noise ratio by selecting only the strongest synaptic inputs for persistent activation." Wouldn't a stronger dopamine release during stimulants or reward actually diminish glutamate-stimulated dADP?

Sidiropoulou and colleagues also showed that chronic cocaine treatment of rats decreased the ability of D1 agonism with SKF81297 to block dADP induction by DHPG. Remarkably, this in vivo compensation persisted for at least two weeks, when it was assessed ex vivo in their brain slice preparation. The result may provide a mechanism for how chronic stimulant use in humans contributes to poor decisions, impulsivity, and inattentiveness, particularly during withdrawal. Whether inverted U-shaped or not, the relationships between the D1 or mGluR5 mechanisms and the dADP response need to be characterized in cocaine-sensitized versus vehicle-treated rats. This may help determine if stimulant-based adaptations in dADP contribute to psychostimulant alterations in attention, executive function, addiction, and relapse.—C. Anthony Altar.

Dopamine modulates an mGluR5-mediated depolarization underlying prefrontal persistent activity. Sidiropoulou K, Lu FM, Fowler MA, Xiao R, Phillips C, Ozkan ED, Zhu MX, White FJ, Cooper DC. Nat Neurosci. 2009 Feb;12(2):190-9. Epub 2009 Jan 25. Abstract

Comments on Related News

Related News: The New "Inverted U”—Cellular Basis for Dopamine Response Pinpointed

Comment by:  Andreas Meyer-Lindenberg
Submitted 8 February 2007
Posted 8 February 2007

This fascinating paper contributes to our mechanistic understanding of a fundamental nonlinearity governing the response of prefrontal neurons during working memory to dopaminergic stimulation: the “inverted U” response curve (Goldman-Rakic et al., 2000), which proposes that an optimum range of dopaminergic stimulation exists, and that either too little or too much dopamine impairs tuning, or the relationship between task-relevant (“signal”) and task-irrelevant (“noise”) firing of these neurons. On the level of behavior, this is predicted to result in impaired working memory performance outside the optimum middle range, and this has been confirmed in a variety of species. This is a topic of high relevance for schizophrenia where prefrontal dysfunction and related cognitive deficits, and dopaminergic dysregulation, have long been in the center of research interest (Weinberger et al., 2001), and may be linked (Meyer-Lindenberg et al., 2002). In particular, evidence for abnormally decreased dopamine levels in prefrontal cortex would predict that patients with schizophrenia are positioned to the left of the optimum. This line of thought has recently received impetus from genetic studies on COMT, the major enzyme catabolizing dopamine in prefrontal cortex (Tunbridge et al., 2004). Neuroimaging studies have shown that genetic variants with high COMT activity are positioned to the left, those with lower activity nearer the optimum of the inverted U curve, and that this position predicts nonlinear response to amphetamine stimulation (Mattay et al., 2003), as well as interactions between dopamine synthesis and prefrontal response (Meyer-Lindenberg et al., 2005). Variants with sub- (Egan et al., 2001; Nicodemus et al., 2007) or superoptimal (Gothelf et al., 2005) stimulation were associated with schizophrenia risk. Task-related and task-unrelated prefrontal function reacted in opposite ways to genetic variation in dopamine synthesis, suggesting a tuning mechanism (Meyer-Lindenberg et al., 2005). Recently, interacting genetic variants in COMT have also been found to affect prefrontal cortex function in an inverted U fashion (Meyer-Lindenberg et al., 2006).

A seminal contribution to the cellular mechanisms of the inverted U curve is the paper by Williams (one of the authors of the current study) and Goldman-Rakic in Nature 1995 (Williams and Goldman-Rakic, 1995). In this work, dopamine D1 receptor antagonists were used and shown to increase prefrontal cell activity in low levels, whereas high levels inhibited firing. This implicated a mechanism related to D1 receptors and suggested that the neurons studied were to the right of the optimum on the inverted U curve, that is, their dopamine stimulation was excessive. The present study, from Amy Arnsten’s lab at Yale, further defines the cellular mechanisms underlying the inverted U curve in recordings from PFC neurons of awake behaving monkeys exposed to various levels of stimulation by a dopamine 1 receptor agonist. A spatial working memory paradigm was used, enabling the determination of the degree to which the neurons were tuned by comparing the firing rate to stimuli in the preferred spatial stimulus direction (“signal”) to the firing rate to nonpreferred stimuli (“noise”). The authors recorded both from neurons that were highly tuned (supposedly receiving optimum stimulation) and neurons that were less tuned. As would be predicted from the model, highly tuned neurons did not improve, or worsened, during stimulation, while weakly tuned neurons became more focused in their activity profile. It is not quite clear to me why the previous paper (Williams and Goldman-Rakic, 1995) found neurons that were predominantly to the right of the optimum, while this work identified neurons using a similar paradigm that were either to the left or near the optimum. Perhaps it is because Williams and Goldman-Rakic (Williams and Goldman-Rakic, 1995) screened neurons for a response to the D1 antagonist first. In both studies, extracellular dopamine was not actually measured, meaning that the state of basal stimulation can only be inferred indirectly from the response to the iontophoresed agonist or antagonist. Importantly, the effect of D1 stimulation was always suppressive; effects on tuning were due to the fact that the reduction in response to the signal and the noise were different in extent, such that for weakly tuned neurons and low levels of D1 stimulation, the noise firing was more suppressed than that of the signal, resulting in increased signal to noise. In a second set of pharmacological experiments, which included validation in a rat working memory model, the authors show that these effects are cAMP, but not PKC-dependent, suggesting a preferential cellular mechanism through Gs-proteins, which might be useful for exploration of more specific drug targets.

This work has interesting implications for our understanding of prefrontal function in schizophrenia. Since dopamine stimulation was found to be almost exclusively suppressive, cortical dopamine depletion in schizophrenia would be predicted to lead to relatively increased, but inefficient (untuned) cortical cognitive response, as has indeed been observed (Callicott et al., 2000). However, it is an open question precisely how cortical physiology assessed by imaging relates to these cellular events. The data by Arnsten suggest that each patch of prefrontal cortex will contain a population of neurons at various states of tuning that will respond differently to drug-induced or cognitively related changes in extracellular dopamine, with some improving, some decreasing their tuning. Depending on whether imaging signals and tasks are more sensitive to overall firing rate, or to specific signal-to-noise properties, the resulting blood flow change might be quite different. Perhaps this contributes to some of the puzzling discrepancies between hypo- and hyperactivation both being observed in comparable tasks and regions of prefrontal cortex in schizophrenia.


1. Goldman-Rakic PS, Muly EC 3rd, Williams GV. D(1) receptors in prefrontal cells and circuits. Brain Res Brain Res Rev. 2000 Mar;31(2-3):295-301. Review. No abstract available. Abstract

2. Weinberger DR, Egan MF, Bertolino A, Callicott JH, Mattay VS, Lipska BK, Berman KF, Goldberg TE. Prefrontal neurons and the genetics of schizophrenia. Biol Psychiatry. 2001 Dec 1;50(11):825-44. Review. Abstract

3. Meyer-Lindenberg A, Miletich RS, Kohn PD, Esposito G, Carson RE, Quarantelli M, Weinberger DR, Berman KF. Reduced prefrontal activity predicts exaggerated striatal dopaminergic function in schizophrenia. Nat Neurosci. 2002 Mar;5(3):267-71. Abstract

4. Tunbridge EM, Bannerman DM, Sharp T, Harrison PJ. Catechol-o-methyltransferase inhibition improves set-shifting performance and elevates stimulated dopamine release in the rat prefrontal cortex. J Neurosci. 2004 Jun 9;24(23):5331-5. Abstract

5. Mattay VS, Goldberg TE, Fera F, Hariri AR, Tessitore A, Egan MF, Kolachana B, Callicott JH, Weinberger DR. Catechol O-methyltransferase val158-met genotype and individual variation in the brain response to amphetamine. Proc Natl Acad Sci U S A. 2003 May 13;100(10):6186-91. Epub 2003 Apr 25. Abstract

6. Meyer-Lindenberg A, Kohn PD, Kolachana B, Kippenhan S, McInerney-Leo A, Nussbaum R, Weinberger DR, Berman KF. Midbrain dopamine and prefrontal function in humans: interaction and modulation by COMT genotype. Nat Neurosci. 2005 May;8(5):594-6. Epub 2005 Apr 10. Abstract

7. Egan MF, Goldberg TE, Kolachana BS, Callicott JH, Mazzanti CM, Straub RE, Goldman D, Weinberger DR. Effect of COMT Val108/158 Met genotype on frontal lobe function and risk for schizophrenia. Proc Natl Acad Sci U S A. 2001 Jun 5;98(12):6917-22. Epub 2001 May 29. Abstract

8. Nicodemus KK, Kolachana BS, Vakkalanka R, Straub RE, Giegling I, Egan MF, Rujescu D, Weinberger DR. Evidence for statistical epistasis between catechol-O-methyltransferase (COMT) and polymorphisms in RGS4, G72 (DAOA), GRM3, and DISC1: influence on risk of schizophrenia. Hum Genet. 2007 Feb;120(6):889-906. Epub 2006 Sep 28. Abstract

9. Gothelf D, Eliez S, Thompson T, Hinard C, Penniman L, Feinstein C, Kwon H, Jin S, Jo B, Antonarakis SE, Morris MA, Reiss AL. COMT genotype predicts longitudinal cognitive decline and psychosis in 22q11.2 deletion syndrome. Nat Neurosci. 2005 Nov;8(11):1500-2. Epub 2005 Oct 23. Abstract

10. Meyer-Lindenberg A, Nichols T, Callicott JH, Ding J, Kolachana B, Buckholtz J, Mattay VS, Egan M, Weinberger DR. Impact of complex genetic variation in COMT on human brain function. Mol Psychiatry. 2006 Sep;11(9):867-77, 797. Epub 2006 Jun 20. Abstract

11. Williams GV, Goldman-Rakic PS. Modulation of memory fields by dopamine D1 receptors in prefrontal cortex. Nature. 1995 Aug 17;376(6541):572-5. Abstract

12. Callicott JH, Bertolino A, Mattay VS, Langheim FJ, Duyn J, Coppola R, Goldberg TE, Weinberger DR. Physiological dysfunction of the dorsolateral prefrontal cortex in schizophrenia revisited. Cereb Cortex. 2000 Nov;10(11):1078-92. Abstract

View all comments by Andreas Meyer-Lindenberg

Related News: The New "Inverted U”—Cellular Basis for Dopamine Response Pinpointed

Comment by:  Terry Goldberg
Submitted 6 April 2007
Posted 6 April 2007

In this landmark study, Arnsten and colleagues used a full dopamine agonist in awake behaving monkeys to make key points about the inverted U response at the cellular level and how this maps to the behavioral level. There were a number of surprises. The first was that stimulation of the D1 receptor had consistently suppressive effects on neuronal firing during delays in a working memory task. The second was that when responses were optimized, suppressive effects differentially affected non-preferred directional neurons, rather than preferred direction neurons. Thus, it appeared that noise was reduced rather than signal amplified. Too much D1 stimulation resulted in suppression of both classes of neurons.

The implications of this work are important because it suggests that there is a neurobiological algorithm at work that can reliably produce this unexpected physiological pattern (perhaps as the authors suggest on the basis of baseline activity). It remains to be elucidated whether the D1 receptor effects are mediated by glutamatergic neurons or GABA interneurons, or both. There is another layer of complexity to the story. As Arnsten and colleagues note, possible excitatory influences of D1 stimulation may not have been observed because endogenous dopamine had already triggered this process. It is unclear if D2 receptors in the cortex have a role in shaping or terminating this activity.

Last, it is tempting to speculate about the implications of these findings for other types of tasks that engage prefrontal cortex in humans. What does tuning mean in the context of tasks like the N Back which demands updating, the ID/ED test from the CANTAB, which involves suppression of salient distractors at early set shifting stages, or a task which demands heavy doses of cognitive control like the flanker task, all of which have been shown to be sensitive to manipulations of the dopamine system (Goldberg et al., 2003; Jazbec et al., 2007; Diaz-Asper et al., in press; Blasi et al., 2005)?

View all comments by Terry Goldberg

Related News: MTHFR, COMT Genes Work Together to Bring Down Cortical Activation in Schizophrenia

Comment by:  Jennifer Barnett (Disclosure)
Submitted 19 December 2008
Posted 19 December 2008

The recent studies of Prata and colleagues and Roffman and colleagues shed considerable further light on the ongoing mysteries of the catechol-O-methyltransferase Val158Met polymorphism and its effects on the proposed “inverted-U” shape of cortical dopamine function. Both study teams should be congratulated on these high-quality studies using what are, for neuroimaging experiments, impressive numbers of both patients and controls.

Our understanding of the effects of the COMT Val/Met polymorphism in humans remains incomplete despite no shortage of elegant studies and intriguing results. In their landmark 2001 paper, Egan and colleagues reported that Val carriers showed poorer cognitive function, a higher risk for schizophrenia, and reduced prefrontal efficiency when compared with Met carriers. These associations, along with a multitude of other psychological and psychiatric phenotypes, have since been tested in labs across the world. Meta-analyses of the available data have concluded that there is little influence of the Val/Met polymorphism on risk for schizophrenia (Allen et al., 2008; Fan et al., 2005; Munafo et al., 2005) or cognitive function (Barnett et al., 2008). Perhaps because of the increased cost and difficulty of collecting imaging data compared with cognitive or disease status, rather fewer studies have been published testing the hypothesis that Val/Met affects prefrontal cortical efficiency, but those few (e.g., Ho et al., 2005) do appear consistent with the original report .

Prata et al. (2008) studied the effects of Val/Met on cortical activation during a verbal fluency task and report an interesting, if somewhat unintuitive result: that there are opposite effects of genotype on task performance and cortical activation in patients with schizophrenia, compared with those seen in healthy controls. In patients, Val alleles were associated with poorer task performance, while in controls, there was no significant difference between genotype groups. The trend, however, was for better task performance among Val-carrying controls, and the group x genotype interaction term was significant. These results were interestingly reflected in regional activation patterns, where in the right peri-Sylvian region Val alleles were associated with increased activation in patients, and decreased activation in controls. Further analyses suggested that these group x genotype interactions may partly reflect genetically driven differences in functional connectivity. Explanations for these opposite effects in patients and controls are consistent with an inverted-U shape of dopaminergic function where patients lie on the left-hand side of the U (suboptimal dopamine) and controls lie somewhat to the right of the center, such that increased cortical dopamine (as experienced by Met carriers) is slightly disadvantageous. Interestingly, we found the same pattern when comparing the effect of Val/Met genotype on N-back performance in patients and controls (Barnett et al., 2008); it is good to see these non-linear behavioral results supported by structural and functional imaging data.

The Val/Met polymorphism is certainly not the only determinant of COMT function, and we now know that other SNPs within the gene greatly affect the amount of COMT expressed (Nackley et al., 2006). Moreover, in affecting cortical dopamine and norepinephrine, COMT does not operate alone. Roffman and colleagues’ study (Roffman et al., 2008) very nicely demonstrates how much we have still to learn about potential gene-gene interaction (epistatic) effects. They studied brain activation during a working memory task and analyzed the combined effects of Val/Met and a functional polymorphism in MTHFR, a gene with plausible biological interactions with COMT. In this study, COMT genotype alone did not predict variation in activation in dorsolateral prefrontal cortex. There was a three-way interaction, however, between COMT and MTHFR genotypes and diagnostic group, such that MTHFR genotype appeared to modulate prefrontal activation most in Val/Val patients (who would be expected to have the lowest prefrontal dopamine), and among Met/Met controls (who would be expected to have the highest prefrontal dopamine, potentially putting them beyond the optimal level in the inverted-U model).

Despite considerable interest in gene-gene and gene-environment interactions among schizophrenia researchers, replications of such interactions have been relatively few and far between. While it is notoriously difficult to demonstrate biological interaction from statistical data alone, Roffman’s study provides us with hope that a really good hypothesis may still give us reason to try and do so.


Allen NC, Bagade S, McQueen MB, Ioannidis JP, Kavvoura FK, Khoury MJ, Tanzi RE, Bertram L. Systematic meta-analyses and field synopsis of genetic association studies in schizophrenia: the SzGene database. Nat Genet. 2008 Jul 1;40(7):827-34. Abstract

Barnett JH, Scoriels L, Munafò MR. Meta-analysis of the cognitive effects of the catechol-O-methyltransferase gene Val158/108Met polymorphism. Biol Psychiatry. 2008 Jul 15;64(2):137-44. Abstract

Fan JB, Zhang CS, Gu NF, Li XW, Sun WW, Wang HY, Feng GY, St Clair D, He L. Catechol-O-methyltransferase gene Val/Met functional polymorphism and risk of schizophrenia: a large-scale association study plus meta-analysis. Biol Psychiatry. 2005 Jan 15;57(2):139-44. Abstract

Ho BC, Wassink TH, O'Leary DS, Sheffield VC, Andreasen NC. Catechol-O-methyl transferase Val158Met gene polymorphism in schizophrenia: working memory, frontal lobe MRI morphology and frontal cerebral blood flow. Mol Psychiatry. 2005 Mar 1;10(3):229, 287-98. Abstract

Munafò MR, Bowes L, Clark TG, Flint J. Lack of association of the COMT (Val158/108 Met) gene and schizophrenia: a meta-analysis of case-control studies. Mol Psychiatry. 2005 Aug 1;10(8):765-70. Abstract

Nackley AG, Shabalina SA, Tchivileva IE, Satterfield K, Korchynskyi O, Makarov SS, Maixner W, Diatchenko L. Human catechol-O-methyltransferase haplotypes modulate protein expression by altering mRNA secondary structure. Science. 2006 Dec 22;314(5807):1930-3. Abstract

Prata DP, Mechelli A, Fu CH, Picchioni M, Kane F, Kalidindi S, McDonald C, Howes O, Kravariti E, Demjaha A, Toulopoulou T, Diforti M, Murray RM, Collier DA, McGuire PK. Opposite Effects of Catechol-O-Methyltransferase Val158Met on Cortical Function in Healthy Subjects and Patients with Schizophrenia. Biol Psychiatry. 2008 Dec 1; Abstract

Roffman JL, Gollub RL, Calhoun VD, Wassink TH, Weiss AP, Ho BC, White T, Clark VP, Fries J, Andreasen NC, Goff DC, Manoach DS. MTHFR 677C --> T genotype disrupts prefrontal function in schizophrenia through an interaction with COMT 158Val --> Met. Proc Natl Acad Sci U S A. 2008 Nov 11;105(45):17573-8. Abstract

View all comments by Jennifer Barnett

Related News: MTHFR, COMT Genes Work Together to Bring Down Cortical Activation in Schizophrenia

Comment by:  S.H. Lin
Submitted 15 January 2009
Posted 19 January 2009
  I recommend the Primary Papers

The “inverted-U” shape of cortical dopamine function with regard to the COMT Val158Met polymorphism is an interesting issue worthy of discussion. The COMT enzyme may modulate the balance of tonic and phasic dopamine function depending on the area-specific neurochemical environment (Bilder et al., 2004). There is thought to be a complex nonlinear relationship between dopamine availability and brain function (Williams et al., 2007).

Our study (Liao et al., 2008) examined the relationships of three COMT SNPs—rs737865 in intro 1, rs4680 in exon 4 (Val158Met), and downstream rs165599—to schizophrenia and its related deficits in neurocognitive function in families of patients with schizophrenia in Taiwan. The study results indicated that the Val allele was associated with better performance on the WCST (i.e., greater Categories Achieved and Conceptual Level Response and fewer Perseverative Errors) or CPT (i.e., greater d'), which might be explained by an “inverted U” shaped relationship between dopamine levels and prefrontal cortex function (Cools and Robbins 2004; Mattay et al., 2003). This model reveals that an optimal functioning occurs within a narrow range of dopamine level, and both excessive and insufficient dopamine levels impair working memory performance. Our results indicate that the genetic variants in COMT might be involved in modulation of neurocognitive functions, hence conferring increased risk to schizophrenia.


Bilder, R.M., Volavka, J., Lachman, H.M. & Grace, A.A. (2004) The catechol-O-methyltransferase polymorphism: relations to the tonic-phasic dopamine hypothesis and neuropsychiatric pheno-types. Neuropsychopharmacology 29, 1943–1961. Abstract

Cools, R. and Robbins, T.W. (2004) Chemistry of the adaptive mind. Philos Transact A Math Phys Eng Sci 362, 2871–2888. Abstract

Liao S.Y., Lin S.H., Liu C.M., Hsieh M.H., Hwang T.J., Liu S.K., Guo S.C., Hwu, H.G., Chen W.J. (2008): Genetic variants in COMT and neurocognitive impairment in families of patients with schizophrenia. Genes, Brain and Behavior. Abstract

Mattay, V.S., Goldberg, T.E., Fera, F., Hariri, A.R., Tessitore, A., Egan, M.F., Kolachana, B., Callicott, J.H. and Weinberger, D.R. (2003) Catechol O-methyltransferase val158-met genotype and individual variation in the brain response to amphetamine. Proc Natl Acad Sci USA 100, 6186–6191. Abstract

Williams, H.J., Owen, M.J. and O‘Donovan, M.C. (2007) Is COMT a susceptibility gene for schizophrenia? Schizophr Bull 33, 635–641. Abstract

View all comments by S.H. Lin