ICOSR 2009—Dopamine Signaling Meets Modern Genetics
As part of our ongoing coverage of the 2009 International Congress on Schizophrenia Research (ICOSR), 28 March to 1 April 2009, in San Diego, California, we bring you session summaries from some of the Young Investigator travel award winners. We are grateful for this summary by Elizabeth Tunbridge of the University of Oxford, U.K.
28 April 2009. Chair Alessandro Bertolino helped start the 30 March sessions by opening a symposium entitled “Dopamine signaling in the era of molecular and genetic pathways.” He began by emphasizing that no one doubts the involvement of dopaminergic dysfunction in schizophrenia, although it remains to be determined whether this dysfunction is a primary abnormality or secondary to other pathological processes. Therefore, the aim of the symposium was to describe recent evidence linking genetic determinants of dopamine signaling to phenotypes relevant to the pathophysiology of schizophrenia.
Bertolino described some of his recent studies investigating links between genetic variants in the dopamine D2 receptor (DRD2) gene and brain activation during working memory. He demonstrated that a polymorphism within the DRD2 gene, which is in linkage disequilibrium with a schizophrenia risk polymorphism, is associated with reduced expression of the D2 short (D2S) variant in postmortem human brain. Since the D2S variant is thought to act as a presynaptic autoreceptor, reductions in its expression would be anticipated to result in excessive presynaptic dopamine release. Consistent with this hypothesis, the polymorphism is associated with reduced D2 ligand binding in human striatum in vivo. Furthermore, in healthy controls the variant is linked with relatively poorer working memory performance and excessive (“inefficient”) task-related brain activation. Intriguingly, he presented recent data which show opposite effects of this DRD2 polymorphism on brain activation in healthy controls and patients with schizophrenia: the allele that is associated with greater activation in healthy controls predicts lower activation in the patient group. However, rather than reflecting greater efficiency of processing (as appears to be the case in the healthy volunteers, who perform well on the working memory task), in the patient group the reduced activation associated with the risk allele appears to result from insufficient engagement of the prefrontal-striatal circuitry, since the patient group has both reduced brain activation and poorer task performance compared with healthy controls. Finally, he concluded by showing interactive effects of DRD2 and the dopamine transporter (DAT) gene on prefrontal-striatal activation during working memory. Thus, DRD2 + DAT combinations that predict intermediate levels of dopamine release resulted in most efficient activation, while combinations indicative of both very high and very low dopamine release predicted relatively inefficient patterns of activation. These results are consistent with the inverted-U relationship known to exist between dopamine levels and prefrontal performance, whereby intermediate levels of dopamine result in optimal prefrontal function while either too much, or too little, dopamine impairs performance.
Moving away from polymorphisms in genes directly involved in dopaminergic signaling, Danny Weinberger described data linking genetic variation in the voltage-sensitive potassium channel gene KCNH2 with dopaminergic phenotypes. He showed that polymorphisms within the KCNH2 gene are associated with schizophrenia in several different cohorts and demonstrated that risk-associated variants predict expression of a brain-specific KCNH2 splice variant. When overexpressed in cultured cortical neurons, this splice variant shows abnormal repolarization currents compared with the full-length form of KCNH2. Therefore, the overexpression of the splice variant would be predicted to prevent the adaptation of cortical neurons to repetitive stimulation, thereby resulting in cortical dysregulation due to an impaired ability to regulate pyramidal excitability. Consistent with this hypothesis, KCNH2 risk-associated variants map onto suboptimal brain function. Even in healthy volunteers, risk-associated variants predict reduced hippocampal volume as well as inefficient hippocampal engagement during a test of episodic memory. These variants are also associated with reduced IQ and processing speed, and map onto general intellectual function by factor analysis, as well as predicting inefficient prefrontal activation during working memory. Notably, these associations between risk-associated variants within KCNH2 and brain function are reminiscent of patterns of brain activity associated with the high activity Val158 allele of catechol-O-methyltransferase (COMT), an enzyme which breaks down prefrontal dopamine. Therefore, his group investigated whether KCNH2 and COMT show interactive or additive effects on brain activation. The interaction between KCNH2 and COMT was found to be additive, suggesting that they impact on prefrontal function via independent mechanisms. Interestingly, he noted that KCNH2 has a very high affinity for dopamine D2 antagonists, such as the antipsychotic drugs: the novel KCNH2 variant may itself represent a novel antipsychotic target.
Next, Hao-Yang Tan explored the cognitive neuroscience of dopamine imaging genetics, describing the individual and interactive effects of genes on network activation during working memory performance. He first presented data demonstrating that the COMT Val158Met polymorphism modulates activation of network activity underlying the manipulation and active maintenance components of working memory. This was less prominent during the simple retrieval of stored information. He also demonstrated modulation of working memory-related brain activation by the metabotropic glutamate receptor 3 (GRM3) gene, as well as by AKT1, which encodes an effector kinase coupled to dopamine D2 receptors. Expression of AKT1 is reduced in schizophrenia, and reduced expression is predicted by a polymorphism that increases risk for developing schizophrenia. This same risk-associated polymorphism also predicts inefficient activation of the prefrontal cortex during working memory performance. Strikingly, there were epistatic interactions of both GRM3 and AKT1 with COMT, with most inefficient processing occurring in individuals with two risk alleles. He concluded by demonstrating different patterns of connectivity between prefronto-parietal-striatal brain regions in the working memory network during different task components, and presented preliminary data suggesting that there may be interactive effects of COMT and AKT1 on these patterns of connectivity.
Finally, Emiliana Borrelli moved away from human neuroimaging studies to demonstrate how she is using transgenic mice to unravel the mechanisms by which the dopamine D2 receptor affects brain function. She began by presenting data from the D2 knockout mouse, which shows hypolocomotion and impaired coordination. These mice also show abnormal cocaine-induced dopamine release, possibly due to the loss of D2-mediated autoreceptor functions as well as the normal presynaptic dimerization of D2 and the dopamine transporter. However, D1-mediated signaling remains intact in D2 knockout animals: administration of a D1 agonist increases locomotor activity and induces expression of the immediate early gene cFos, which is thought to reflect neuronal activation, similar to its effects in wild-type mice. She highlighted the complexity of D2 receptor biology by demonstrating that several cocaine-induced effects are lost in the D2 knockout, likely as the result of the loss of normal D2-mediated regulation of striatal GABA and acetylcholine release, indicating that D2 plays a critical role as heteroreceptor in striatal neurons. In order to understand the mechanisms underlying some of the more complex aspects of D2’s biology, she has developed mice specifically lacking either the short (D2S) or long (D2L) form of this receptor. There are notable differences between these mice, suggesting that D2S and D2L may mediate separate actions of dopamine, perhaps coupled to different downstream signaling cascades. She concluded that D2S may mediate presynaptic actions, whilst D2L is located and functions predominantly post-synaptically.
Taken together, these findings show how researchers are beginning to move away from linking individual genes and single brain regions, to understanding how multiple genes interact to modulate entire networks. This symposium also highlighted the importance of understanding how splice variants can contribute to the complex relationship between individual genes and brain function, demonstrating that alternative variants can act via complementary, or even opposing, mechanisms.—Elizabeth Tunbridge.