Skip to content
Skip to navigation menu


Prof Adrian Harwood  -  PhD

Lithium has the simplest structure of all therapeutic agents and its bioactive properties on bipolar depression have been known for over 100 years. Surprisingly, the basis of its clinical use is still unclear. Furthermore, lithium is used at relatively high doses and has a number of potentially hazardous side-effects. An understanding of the molecular mechanisms of lithium action may shed light on the biological basis of mood disorders and lead to improved therapeutics.

Lithium is also well known to embryologists because of its potent teratogenic effects. We are using genetic, biochemical and cell biological techniques to investigate the lithium sensitive signal transduction pathways required for development of the simple eukaryote, Dictyostelium discoideum. This organism has much of the complexity of metazoa, but the molecular genetic advantages of a micro-organism. Dictyostelium has both a cytoskeleton and a network of signal transduction pathways that, in most aspects, resemble those of mammalian cells. For instance, they contain tyrosine kinases, SH2 domain proteins, G-proteins and small GTPases. Dictyostelium has a 34 Mbp, haploid genome and contains approximately 13,000 genes. The genetic source material has been greatly expanded by completion of its genome sequence (see

Two well-characterised lithium targets are known: inositol monophosphatase (IMPase) and glycogen synthase kinase-3 (GSK-3). Inhibition of IMPase blocks phosphoinostide (PI) signalling, whereas inhibition of GSK-3 affects the cellular response to insulin and Wnt proteins. GSK-3 is also of interest as it regulates the regulation of nuclear transcription factors, such as the Dictyostelium STATa protein, NF-ATc and c-jun, and is involved in the progression of Alzheimer's disease.

We have established that DictyosteliumGSK-3 is required for correct pattern formation during development causing a switch from spore to stalk cell fate. We have shown that this is due to its regulation by extracellular cAMP via the cAR3 receptor. We have identified a number of GSK-3 targets in Dictyostelium and these include both novel genes and a homologue of beta-catenin, a protein that acts in the Wnt signalling pathway in metazoa. Earlier in development, GSK-3 is required for cells to become chemotaxis competent, and loss of GSK-3 prevents cells from migrating towards the chemotractants, cAMP and folate.

The Dictyostelium fruiting body possess a constriction around the stalk tube (arrow). Cells transformed with the F-actin binding protein ABPD-GFP show a ring of actin filaments encircling the stalk tube at this point. Each cell in this ring is connected via an adherens junction.

Movie 1: The Dictyostelium fruiting body form adherens junctions to connect a ring of cells positioned towards the top of the stalk tube (arrow). Cells transformed with the F-actin binding protein ABPD-GFP show a ring of F-actin filaments encircling the stalk tube at this point (see Grimson et al (2000) Nature 408:727-731).

In contrast, we have established that lithium inhibition of PI signaling suppresses PIP3 signalling leading to specific affects on chemotaxis. This indicates that PIP3 signalling, an important cellular signaling pathway, is sensitive to lithium mediated inositol-depletion. In neurons, lithium leads to altered growth cone morphology and behaviour, potentially altering neuronal interactions. We have found that the alterative mood stabilizers, valproic acid (VPA) and carbamazepine suppress PI signalling and alter cell behaviour in a similar way to lithium, indicating that they target a common cellular process.

A neuronal growth cone treated with lithium. Lithium induces both an alteration in microtubule dynamics (stained green with anti-acetyl tubulin antibodies) and an expansion of the growth cone (stained blue with calcein).

Figure 2: A neuronal growth cone treated with lithium. Lithium induces both an alteration in microtubule dynamics (stained green with anti-acetyl tubulin antibodies) and an expansion of the growth cone (stained blue with calcein), (See Williams et al (2002) Nature 417: 292-295).

Movie 2: Dictyostelium cells and neuronal growth cones treated lithium exhibit altered moprphology and chemotaxis (see King et al (2009) DMM 2 306-312 and Shimshoni et al (2009) Neuropharmacology 56 831-837)

To pursue the modulatory mechanisms that affect lithium sensitivity, we have isolated a collection of lithium resistant Dictyostelium mutants. One of these mutants lacks the Dictyostelium prolyl oligopeptidase (DpoA) and escapes the effects of lithium by having elevated concentrations of inositiol phosphates. Remarkably, altered prolyl oligopeptidase activity has been associated with clinical depression and other mental illnesses. We have found that a second mood stabilizer, valproic acid (VPA) also affects InsP signalling in Dictyostelium and loss of dpoA confers cross-resistance to the effects of VPA. We have shown that DpoA mediates a previously unknown, but conserved, signaling pathway to alter PI signaling.

We are currently using both Dictyostelium and cellular neurobiology to investigate the interaction of GSK-3 and phosphoinositide signalling with mood stabilizing drugs. Our Lithium resistant mutants offer a unique opportunity to investigate the molecular mechanisms of drug action and this could have direct implications for the study of mood disorders.

Current Grants

Wellcome Trust Project Grant: Investigation of lithium-induced macro-autophagy in Dictyostelium.

Wellcome Trust Project Grant: Investigation of calcium signalling in the regulation of PTEN and lithium sensitivity during Dictyostelium chemotaxis.

DTI Technology Programme Grant: Nanothether Biochemistry (with Trevor Dale, Paola Borri and Phil Davies [School of Chemistry]).


Trevor Dale Cardiff University

Alan Kimmel, National Institutes of Health, USA

Galila Agam, 
Ben-Gurion University of the Negev, Israel

Meir Bialer,
 Hebrew University of Jerusalem, Israel 

Brian Dean, Mental Health Research Institute, Melbourne, Australia.