Bipolar disorder or manic-depression is a severe and common psychiatric illness that is often treated with mood-stabilizing drugs. Several molecular targets and signalling pathways have been implicated in the pathophysiology of the illness and in the mechanism of action of these drugs. However, the precise targets that are responsible for the therapeutic action and side-effects of the drugs are not known. In this PhD thesis we have analysed some of the potential drug targets, in particular extracellular signal-regulated kinase mitogen-activated protein kinase (ERK/MAPK) and glycogen synthase kinase-3 (GSK.3). which mood stabilizers are believed to regulate. We showed, using cell and biochemical assays, that the effects of lithium, valproate or carbamazepine on ERIC MAPK are not always observed in a given cell type. In addition, only lithium inhibits GSK3 directly and modulation of this kinase with potent and selective inhibitors does not mimic the effects of mood stabilizers on cortical and sensory neuron growth cone morphology. A major part of my thesis focused on the enzyme prolyl oligopeptidase (PO), which is thought to modulate the phosphatidylinositol pathway, as shown from studies in Dictyostelium discoideum and human astroglioma cells. PO inhibitors alone have no effect on sensory neuron growth cone morphology, but reverse the growth cone changes induced by mood stabilizers, mimicking addition of/m-o-inositol to the culture media. We studied PO null-mutant mice and analysed the effects of mood stabilizers on the morphology of growth cones from these mice. Unexpectedly, the PO null-mutant phenotype itself resembled wild-type neurons treated with a mood stabilizer and each drug had no further effect on growth cone morphology. These results show that PO is a critical component of signalling pathways involved in mood-stabilizing drug action on growth cones. Using viral delivery of native or catalytically-dead PO, we showed that each restored the wild-type phenotype. In order to better understand PO function, a yeast-tvvo-hybrid screen (Y2H) was also performed at Gla.xoSmithK.line to determine protein interactors for PO, and we analysed one of these interactors, namely GAP43. We showed that both the native and the catalytically-dead PO co-precipitate with the neuronal protein GAP43. These results show that there are additional biological effects of PO independent of its catalytic domain.

No description available.

The mood-stabilising drugs lithium and valproic acid (VPA) are used in the treatment of bipolar disorder, but the molecular mechanisms underlying their therapeutic effects are not well understood at present. Both drugs have been suggested to attenuate inositol-based signalling: Lithium by depleting the intracellular pool of inositol via uncompetitive inhibition of Inositol monophosphatase (IMPase) and Inositol polyphosphate-1-phosphatase (IPPase), and VPA by inhibition of inositol de novo synthesis. The therapeutic time courses for lithium and VPA treatment suggest that the drugs exert their effects through changing gene expression. Therefore, the aim of the present study was to test the hypothesis that lithium sensitivity in the model system Dictyostelium discoideum (D.discoideum) is caused by changes in gene expression. Real-time PCR and motility assays were used to investigate whether (1) changes in gene expression can be observed in lithium-resistant mutants, (2) overexpression of IMPase and lno1 leads to lithium resistance, and (3) lithium and analogues of the bipolar drug VPA affect gene expression. I found that (1) loss of prolyl oligopeptidase (DpoA) in the lithium resistant LisA mutant or loss of a chromatin-remodelling factor (mutation remains to be confirmed) in the lithium resistant LisG mutant increased lno1 and IPP1 expression. (2) Overexpression of lno1 or IMPase led to resistance against lithium, and also VPA and the VPA-analogue VGD. (3) Lithium and VPA- analogues caused distinct changes in gene expression: Lithium treatment increased the expression of enzymes involved in inositol phosphate signalling, with the exception of IMPase VGD decreased the expression of IMPase, IPP1 and dpoA. In addition, the LisG mutant, which showed increased IPP1 expression, was found to be cross-resistant to VGD. The present study shows that lithium sensitivity in D.discoideum correlates with changes in gene expression and suggests that increase in lno1 and IPP1 expression may confer lithium resistance in D.discoideum.

There is thought to be a link between characteristic reward-oriented processes in bipolar disorder and dysregulation in the behavioural activation system (BAS). This study aimed to assess differences between bipolar and healthy control individuals in their response to reward, failure and a reasoning task sensitive to data-gathering biases. Participants were 25 adults with bipolar I disorder and 25 healthy controls. Measures of sensitivity to reward and failure were collected during the first task (Go task), which included visual analogue ratings of mood and success expectancy, reaction time of button-pressing, and the difficulty level set by the participant. There were no significant differences between groups following reward feedback or failure feedback on the Go task. Results from the second task showed that bipolar individuals needed less data than controls before making a decision on the emotionally-neutral, difficult version of the task. The results are discussed in relation to current trends in bipolar research.

It has been six decades since the discovery of lithium for the treatment of bipolar disorder. There is, as yet, no conclusive evidence as to how lithium produces this therapeutic effect, since it is known to interact with multiple cellular targets. One of the most credible targets is the enzyme, inositol monophosphatase (lMPase), which plays a crucial role in cell signalling. My aim was to find a novel IMPase inhibitor and evaluate it as a possible lithium-like mood stabiliser by using enzyme, cell and whole animal experiments. To achieve this, I created recombinant human and mouse IMPase enzymes and then used these for screening and crystallisation. I used two different approaches for the small-molecule screening: substrate-based virtual screening and drug repositioning using a library of compounds with clinically proven safety. I identified ebselen as a novel IMPase inhibitor suitable for drug repositioning. I determined thatebselen inhibited IMPase noncompetitively, likely through a covalent modification on a cysteine. In cell cultures, ebselen was found to inhibit not just IMPase but other steps that resulted in accumulation of higher inositol phosphates. When injected intraperitoneally into mice, ebselen crossed the blood- brain barrier and exhibited inhibition of IMPase ex vivo. Moreover, in mice, ebselen simulates some, but not all, of the behavioural effects of lithium. I have determined that ebselen inhibits IMPase and acts as a partial lithium mimetic. Given that ebselen is safe in man, it warrants clinical testing for the treatment of bipolar disorder.