The mood stabiliser lithium suppresses PIP3 signalling in Dictyostelium and human cells

Abstract

Bipolar mood disorder (manic depression) is a major psychiatric disorder whose molecular origins are unknown. Mood stabilisers offer patients both acute and prophylactic treatment, and experimentally, they provide a means to probe the underlying biology of the disorder. Lithium and other mood stabilisers deplete intracellular inositol and it has been proposed that bipolar mood disorder arises from aberrant inositol (1,4,5)-trisphosphate [IP(3), also known as Ins(1,4,5)P(3)] signalling. However, there is no definitive evidence to support this or any other proposed target; a problem exacerbated by a lack of good cellular models. Phosphatidylinositol (3,4,5)-trisphosphate [PIP(3), also known as PtdIns(3,4,5)P(3)] is a prominent intracellular signal molecule within the central nervous system (CNS) that regulates neuronal survival, connectivity and synaptic function. By using the genetically tractable organism Dictyostelium, we show that lithium suppresses PIP(3)-mediated signalling. These effects extend to the human neutrophil cell line HL60. Mechanistically, we show that lithium attenuates phosphoinositide synthesis and that its effects can be reversed by overexpression of inositol monophosphatase (IMPase), consistent with the inositol-depletion hypothesis. These results demonstrate a lithium target that is compatible with our current knowledge of the genetic predisposition for bipolar disorder. They also suggest that lithium therapy might be beneficial for other diseases caused by elevated PIP(3) signalling.

Fig. 1

Effect of lithium on Dictyostelium cells undergoing chemotaxis. (A) Wild-type cells treated with 10 mM LiCl or NaCl (control) were stimulated with 10 μM of 2′-deoxy-cAMP and cAMP production was then measured over 5 minutes. (B) Analysis of chemotaxis for wild-type and mutant cell strains after lithium or control treatment. Representative tracks are shown for each experiment and mean±s.e.m. is shown for the parameters: chemotactic index (CI), directionality (D) (a measure of cell turning where 1 is a straight line) and speed. (C) Wild-type cells were treated with 10 mM LiCl or 10 mM NaCl (control) and the amount of F-actin was measured through TRITC-phalloidin binding. The graph shows the fluorescence at time intervals after cAMP stimulation, displayed as relative values compared with prestimulation fluorescence. All data points are the mean±s.e.m. of three independent experiments.

Fig. 2

Lithium suppresses PIP₃ signalling. (A) Translocation of the PIP₃ marker PH_CRAC-GFP to the plasma membrane upon uniform stimulation of control and lithium-treated cells with 1 μM cAMP. The graph shows quantification of PH_CRAC-GFP by measuring the cytosolic fluorescence intensity. The values plotted are the means of three independent experiments±s.e.m. (B) Lithium inhibition of PKB. Aggregation-competent cells were treated for 1 hour with 10 mM LiCl or NaCl (control) and then stimulated with 1 μM cAMP. Samples were analysed by western blotting with an antibody that recognises phosphorylation at both threonine 278 on PkbA and threonine 309 on PkgB (Kamimura et al., 2008) (lc is a non-specific band that acts as a loading control). To confirm specificity, the same antibody was used on wild-type cells, pkbA-null mutant cells and pi3k1-5-null mutant cells that cannot induce PIP₃- dependent phosphorylation. (C) Quantification of PIP₃ synthesis. Cells were stimulated with cAMP under the same conditions as in (B). Phosphorylation of PkbA was measured using an anti-phosphothreonine antibody, as described by Lim et al. (Lim et al., 2001). A direct measurement of the quantity of PIP₃ was made by using a PIP₃ mass ELISA. The results are displayed as percentage increase over the unstimulated level (mean±s.e.m.). (D) Wild-type cells treated with 10 mM LiCl or NaCl (control) were stimulated with 1 μM cAMP and cGMP synthesis was then measured over 30 seconds.

Fig. 3

Lithium effects on PIP synthesis. (A) The total mass of PIP₂ in Dictyostelium cells was measured by alkaline hydrolysis of membrane lipids followed by measuring the released IP₃ (Chilvers et al., 1991). Dictyostelium cells contain between 270–290 picomoles of PIP₂ and lithium-treated cells showed no detectable difference compared with control (NaCl) cells. (B) Measurement of PIP and PIP2 following treatment with 10 mM LiCl. Wild-type cells show a significant (P<0.01, paired t-test) decrease in both PIP and PIP₂ following lithium treatment compared with NaCl control treatment. This decrease is reversed by increased expression of the ImpA1 protein. (C) Lithium treatment decreases the amount of PTEN-GFP associated with the plasma membrane of cells undergoing chemotaxis. Images show typical cells. (D) Lithium treatment increases the basal level of PIP₃ and p-PkbA in unstimulated cells. (E) Wild-type cells expressing PTEN-GFP were pulsed for 5 hours and treated with 10 mM NaCl or LiCl for 1 hour. Translocation following uniform stimulation with 1 μM cAMP was quantified by measuring the change in cytosolic fluorescence intensity. Values plotted are the means of three independent experiments and error bars indicate the s.e.m. Note that the fluorescence intensity measurements are the relative changes compared with the unstimulated level that has been normalized for each cell. This means that although lithium-treated cells had less PTEN on their membrane in the unstimulated state, the kinetics of loss from the membrane remained the same as that seen for the NaCl control-treated cells. (F) A cDNA encoding ImpA1 was expressed in wild-type Dictyostelium cells from the actin15 gene promoter. Samples were analysed for p-PkbA content as in Fig. 2B.

Fig. 4

Lithium suppresses PIP₃ signalling in HL60 cells. (A) HL60 cells expressing PH_Akt-GFP were differentiated to neutrophils and then globally stimulated with fMLP. PH_Akt-GFP translocates from the cytosol to the plasma membrane within 20 seconds. At 60 seconds, cells have returned to a polarised state with localised at the leading edge. (B) Translocation of the PH_Akt-GFP is measured by the same method used in Dictyostelium. Graph shows quantification by measuring the cytosolic fluorescence intensity. Values plotted are the means of three independent experiments ± s.e.m.