Enhancement by lithium of cAMP-induced CRE/CREB-directed gene transcription conferred by TORC on the CREB basic leucine zipper domain

Abstract

The molecular mechanism of the action of lithium salts in the treatment of bipolar disorder is not well understood. As their therapeutic action requires chronic treatment, adaptive neuronal processes are suggested to be involved. The molecular basis of this are changes in gene expression regulated by transcription factors such as CREB (cAMP-response-element-binding protein). CREB contains a transactivation domain, in which Ser119 is phosphorylated upon activation, and a bZip (basic leucine zipper domain). The bZip is involved in CREB dimerization and DNA-binding, but also contributes to CREB transactivation by recruiting the coactivator TORC (transducer of regulated CREB). In the present study, the effect of lithium on CRE (cAMP response element)/CREB-directed gene transcription was investigated. Electrically excitable cells were transfected with CRE/CREB-driven luciferase reporter genes. LiCl (6 mM or higher) induced an up to 4.7-fold increase in 8-bromo-cAMP-stimulated CRE/CREB-directed transcription. This increase was not due to enhanced Ser119 phosphorylation or DNA-binding of CREB. Also, the known targets inositol monophosphatase and GSK3beta (glycogen-synthase-kinase 3beta) were not involved as specific GSK3beta inhibitors and inositol replenishment did not mimic and abolish respectively the effect of lithium. However, lithium no longer enhanced CREB activity when the CREB-bZip was deleted or the TORC-binding site inside the CREB-bZip was specifically mutated (CREB-R300A). Otherwise, TORC overexpression conferred lithium responsiveness on CREB-bZip or the CRE-containing truncated rat somatostatin promoter. This indicates that lithium enhances cAMP-induced CRE/CREB-directed transcription, conferred by TORC on the CREB-bZip. We thus support the hypothesis that lithium salts modulate CRE/CREB-dependent gene transcription and suggest the CREB coactivator TORC as a new molecular target of lithium.

Figure 1. Effect of lithium on forskolin-stimulated CRE/CREB-directed gene transcription

HIT-T15 cells were transfected with the indicated reporter gene construct 4×SomCRET81Luc. At 7 h prior to harvest, cells were treated with 20 mM LiCl and 6 h prior to harvest with 10 μM forskolin. At 48 h after transfection, luciferase activity was determined. Values are expressed as a percentage of the controls that received solvent only. Values are means±S.E.M. for four independent experiments each performed in duplicate. FSK, forskolin; TK, thymidine kinase promoter. ***P<0.001, two-way ANOVA, Bonferroni post-hoc test.

Figure 2. Effect of lithium on cAMP- and membrane depolarization-induced CRE/CREB-directed gene transcription

HIT-T15 cells were transfected with the reporter gene construct 4×SomCRET81Luc. At 7 h prior to harvest, cells were treated with 20 mM LiCl and 6 h prior to harvest with 2 mM 8BrcAMP or 45 mM KCl. At 48 h after transfection, luciferase activity was determined. Values are expressed as a percentage of control (no 8BrcAMP, no KCl). Values are means±S.E.M. for three independent experiments each performed in duplicate. ***P<0.001, two-way ANOVA, Bonferroni post-hoc test.

Figure 3. Concentration-response curve for the effect on cAMP-induced CRE/CREB-directed gene transcription

HIT-T15 cells were transfected with the reporter gene construct 4×SomCRET81Luc. At 7 h prior to harvest, cells were treated with increasing concentrations of LiCl (20 μM–20 mM) and 6 h prior to harvest with 1 mM 8BrcAMP. At 48 h after transfection, the luciferase activity of the cells was determined. Values for 8BrcAMP-treated cells that received no LiCl were set to 100%. Values are means±S.E.M. for three independent experiments each performed in duplicate. *P<0.05, **P<0.01, Student's t test.

Figure 4. Effect of (a) GSK3β inhibitors and (b) inositol on cAMP-induced CRE/CREB-directed gene transcription

HIT-T15 cells were transfected with the reporter gene construct 4xSomCRET81Luc. (a) At 7 h prior to harvest, cells were treated with 20 mM LiCl or the GSK3β inhibitors SB215763 (30 μM) or SB415286 (50 μM) and 6 h prior to harvest with 1 mM 8BrcAMP. (b) At 8 h prior to harvest 1 mM myoinositol was added, 7 h prior to harvest cells were treated with 20 mM LiCl and 6 h prior to harvest with 1 mM 8BrcAMP. At 48 h after transfection, luciferase activity was determined. Values are expressed as a percentage of the controls that received solvent only. Values are means±S.E.M. for three independent experiments each performed in duplicate. ***P<0.001, two-way ANOVA, Bonferroni post-hoc test.

Figure 5. Effect of lithium on Ser¹¹⁹ phosphorylation of CREB

HIT-T15 cells were treated 3 h prior to harvest with 20 mM LiCl and 2 h prior to harvest with 2 mM 8BrcAMP. Cell lysates were subjected to Western blot analysis and immunostained with specific antibodies to CREB and phospho-CREB. (a) Immunoblot from a typical experiment. (b) Densitrometric quantification of the bands. Specific bands of P-CREB (phospho-CREB) and CREB from the same lysate were quantified and the ratio between P-CREB and CREB was calculated. Values are means±S.E.M. for three independent experiments.

Figure 6. Effect of lithium on the binding of CREB to the somatostatin CRE as revealed by EMSA

HIT-T15 cells were treated with 20 mM LiCl 2 h prior to harvest and with 2 mM 8BrcAMP 1 h prior to harvest. Nuclear extracts were incubated with labelled oligodeoxynucleotides containing the somatostatin CRE (SCE). Unlabelled wild-type (wt) or mutant (mut) oligonucleotides as competitors were added at a 450-fold molar excess. Protein complexes were separated on non-denaturing SDS/PAGE and probes visualized by autoradiography. (a) Autoradiography in a typical experiment. The arrow indicates the mutation-sensitive complexes. (b) Densitometric quantification of specific bands. Values are means±S.E.M. of six independent experiments. Optical density=absorbance.

Figure 7. Effect of lithium on cAMP-induced transcriptional activity of CREB wild-type and CREB mutant lacking the bZip, using the heterologous GAL4-system

HIT-T15 cells were transfected with the GAL4-driven reporter gene construct G5E1B-Luc and GAL4-fusion proteins containing full-length CREB or CREB1–261. At 7 h prior to harvest, cells were treated with 20 mM LiCl and 6 h prior to harvest with 2 mM 8BrcAMP. At 48 h after transfection, luciferase activity was determined. Values are expressed as a percentage of the respective untreated control. Values are means±S.E.M. for three independent experiments each performed in duplicate. ***P<0.001, two-way ANOVA, Bonferroni post-hoc test.

Figure 8. Effect of lithium on TORC-mediated transcription

HIT-T15 cells were transfected with either the GAL4-driven reporter gene construct G5E1B-Luc and GAL4-fusion proteins containing full length CREB mutant R300A or CREBbZip (a) or the truncated rat somatostatin promoter-driven reporter gene −65SMSLuc (b). Expression vector for TORC1 was co-transfected where indicated. At 7 h prior to harvest, cells were treated with 20 mM LiCl and 6 h prior to harvest with 2 mM 8BrcAMP. At 48 h after transfection, luciferase activity was determined. Values are expressed as a percentage of the respective untreated control. Values are means±S.E.M. for three independent experiments each performed in duplicate. ***P<0.001, two-way ANOVA, Bonferroni post-hoc test.