Distinct lithium-induced gene expression effects in lymphoblastoid cell lines from patients with bipolar disorder

Abstract

Lithium is the most commonly prescribed medication for the treatment of bipolar disorder (BD), yet the mechanisms underlying its beneficial effects are still unclear. We aimed to compare the effects of lithium treatment in lymphoblastoid cell lines (LCLs) from BD patients and controls. LCLs were generated from sixty-two BD patients (based on DSM-IV) and seventeen healthy controls matched for age, sex, and ethnicity. Patients were recruited from outpatient clinics from February 2012 to October 2014. LCLs were treated with 1mM lithium for 7 days followed by microarray gene expression assay and validation by real-time quantitative PCR. Baseline differences between groups, as well as differences between vehicle- and lithium-treated cells within each group were analyzed. The biological significance of differentially expressed genes was examined by pathway enrichment analysis. No significant differences in baseline gene expression (adjusted p-value < 0.05) were detected between groups. Lithium treatment of LCLs from controls did not lead to any significant differences. However, lithium altered the expression of 236 genes in LCLs from patients; those genes were enriched for signaling pathways related to apoptosis. Among those genes, the alterations in the expression of PIK3CG, SERP1 and UPP1 were validated by real-time PCR. A significant correlation was also found between circadian functioning and CEBPG and FGF2 expression levels. In summary, our results suggest that lithium treatment induces expression changes in genes associated with the apoptosis pathway in BD LCLs. The more pronounced effects of lithium in patients compared to controls suggest a disease-specific effect of this drug.

Keywords: Apoptosis; Bipolar disorder; Gene expression; Lithium; Lymphoblastoid cell lines; Morningness.

Figure 1

Cell viability after treatment of lymphoblastoid cells with lithium. Lithium at three different concentrations (0.1, 1 or 10 mM) did not alter cell viability assessed by 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MMT) reduction assay. Each dashed line represents the 100% viability of vehicle-treated cells. Each bar represents average of the percentage of vehicle ± standard error of the mean.

Figure 2

Lithium-induced differentially expressed genes enriched for ‘regulation of programmed cell death’ (GO ID: 0043067). Only the sixteen genes (out of 26) that show direct or indirect relationships (according to the IPA Knowledge Base) are depicted. Green represents down-regulation of the gene, while red represents up-regulation. The network was made with the software Ingenuity® Pathway Analysis, version 31813283 (IPA, QIAGEN).

Figure 3

Validation of microarray expression data. A–C) Real-time quantitative PCR validation for the differences in the expression of PIK3CG, SEPR1 and UPP1 after lithium (Li) treatment in lymphoblastoid cells from patients with bipolar disorder (BD I). RQ stands for the ‘relative quantification’ calculated by the delta delta Ct method. D–F) Microarray results showing the differences in PIK3CG, SERP1 and UPP1 expression after lithium treatment.