Circadian clock genes and cell survival in bipolar disorder: Insights into lithium responsiveness and molecular mechanisms in patient-derived neural progenitor cells and mouse neurons.

Bipolar disorder (BD) is characterized by disrupted circadian rhythms and neuronal loss. Lithium amplifies circadian rhythms and is neuroprotective, indicating mechanistic overlap across cellular systems. We examined the role of neuroprotection in determining lithium response and how circadian clock genes regulate apoptosis. In stem-cell-derived neuronal progenitor cells (NPCs) from BD patients, and in immortalized mouse hippocampal neurons, lithium was neuroprotective against staurosporine (STS)-induced apoptosis. However, neuroprotection did not distinguish lithium-responders (Li-Rs) from nonresponders (Li-NRs). Disruption of selected clock genes-period circadian regulator 1 (PER1), basic helix-loop-helix ARNT like 1 (ARNTL, also known as BMAL1), and nuclear receptor subfamily 1 group D member 1 (NR1D1; also known as REV-ERBα)-with small interfering RNA (siRNA) had distinct effects on apoptosis in mouse neurons and human NPCs. In BD patient NPCs, reduced expression of PER1 and BMAL1 increased apoptosis more in Li-NR NPCs vs. Li-R NPCs. Reduced expression of REV-ERBα caused more extensive apoptosis in BD vs. control NPCs, but did not distinguish Li-Rs vs. Li-NRs. The REV-ERB agonist GSK4112 was neuroprotective in mouse neurons and control NPCs, but did not reduce apoptosis in BD NPCs. Simultaneous treatment of NPCs with STS and GSK4112 revealed BD- and Li-R-associated differences in apoptosis-related gene expression. In summary, the pharmacological effects of lithium in promoting neuroprotection are not sufficient to distinguish between Li-Rs and Li-NRs. However, genetic and pharmacological manipulations of the circadian clock reveal distinct roles in apoptosis in BD patient NPCs, some of which do distinguish Li-Rs and Li-NRs. We conclude that genetic disruption of cellular circadian clocks may predispose neurons to apoptosis preferentially in BD and/or Li-NR. These findings imply that differences in BD between Li-Rs/Li-NRs primarily arise from defining biological characteristics of patient subgroups, not from distinct pharmacodynamic factors following lithium treatment.