Figure 1.

Ratios of BCL2/ BAD (panel A), BCL2/BAK1 (panel B), IRS2/BAD (panel C) and IRS2/BAK1 (panel D) in lithium responders and non-responders over the 8 weeks of the study. In each case, the ratio of anti- to pro-apoptotic genes increased in lithium responders over the first month of treatment and then returned to baseline, while the opposite pattern was observed in lithium non-responders.