In this study, we compared changes in gene-expression in peripheral blood among a group of depressed subjects with BD over a period of eight weeks following the initiation of treatment with lithium. We identified 127 genes whose expression changed differentially in lithium responders and non-responders. Pathway analysis of the differentially expressed genes using GeneGO Metacore software showed that regulation of apoptosis was the most significantly affected pathway among these genes. Strikingly, among lithium responders, several anti-apoptotic genes including BCL2 and insulin receptor substrate 2 (IRS2) were up-regulated, while pro-apoptotic genes, including BCL2-antagonist/killer 1 (BAK1) and BCL2-associated agonist of cell death (BAD), were down-regulated. In contrast, in lithium non-responders, BCL2 and IRS2 were down-regulated, while BAK1 and BAD up-regulated at the one-month time-point.
BCL2 gene family members are key regulators of apoptotic cell death and include both pro- and anti-apoptotic genes [35, 36]. Collectively, the expression levels of the various BCL2 family members define thresholds for apoptosis in a given cell. The two pro-apoptotic proteins, BAX and BAK1 promote apoptosis by binding to the mitochondrial voltage-dependent anion channel (VDAC), and accelerating its opening, leading to a loss in membrane potential, the release of cytochrome c, and subsequent activation of the intrinsic caspase pathway. Anti-apoptotic members of the BCL2 family, including BCL2 itself, the related protein BCL-xL (encoded by the BLCL1 gene) and MCL-1, prevent apoptosis by binding to BAX and BAK1, preventing their interaction with VDAC. Other pro-apoptotic BCL2 family members, termed the ‘BH3-only’ proteins, including BAD, BID, and BMF, are thought to indirectly promote apoptosis by binding to anti-apoptotic family members and preventing their interaction with BAX and BAK1 [35, 36]. IRS2 is a cytoplasmic signaling molecule that mediates effects of insulin, insulin-like growth factor 1, and other cytokines by acting as a molecular adaptor between diverse receptor tyrosine kinases and downstream effectors. In addition, IRS2 has been shown to bind to the Bcl2 protein and block phosphorylation of Bcl2 induced by insulin and suppress apoptotic cell death . Thus, our finding that BCL2 and IRS2 are both increased in lithium responders at the one month time point, while BAK1 and BAD were down-regulated, suggests that in lithium responders there was a shift in the balance of expression among pro- and anti- apoptotic members of the BCL2 family favoring the anti-apoptotic genes. Conversely, in lithium non-responders there was a decrease in BCL2 and IRS2 and an increase BAK1 and BAD, suggesting that there was a shift in the opposite direction, favoring the pro-apoptotic members of the BCL2 family. Intriguingly, changes in the ratio of anti- to pro-apoptotic gene expression among both lithium responders and non-responders appeared to return to baseline by the 8-week time-point, although differences in clinical status were more marked at this point than at 4 weeks. This suggests that transient changes in gene-expression can have enduring effects on the state of the organism, even when those differences can no longer be directly observed. Additional studies will be required to determine if there are changes in the level or function of the proteins encoded by these genes that are longer lasting and perhaps more directly related to the clinical status of the patients.
While changes in the expression of BCL2 family genes in peripheral blood are unlikely to be directly related to the changes in mood symptoms, systemic differences in the way subjects with different genetic and epigenetic backgrounds respond at the biochemical level to treatment with lithium may underlie some of the heterogeneity in clinical response to lithium.
Studies of post-mortem tissue from human subjects has shown that activity of mitochondrial complex I is decreased, and oxidative damage is increased in the prefrontal cortex of patients with BD . Lithium has been shown to increase the activity of mitochondrial ETC. complexes in extracts from human post-mortem brain tissue at therapeutically relevant concentrations , while rats subjected to an experimental model of depression showed impaired mitochondrial function . Conversely, transgenic mice expressing the anti-apoptotic protein Bax Inhibitor 1, showed protection in the learned helplessness model of depression . These results have been interpreted in the context of a neurotrophic hypothesis of mood disorders [42, 43], indicating that increased expression of BCL2 and related genes is necessary for the therapeutic effects of lithium and other mood stabilizers. Our results indicate that in a substantial group of patients, this effect does not occur, and in fact the opposite effect was seen in patients who did not respond to lithium. Better understanding of the mechanisms underlying this difference may lead to improved methods for personalizing treatment for bipolar disorder in the future.
Limitations of this study include a relatively small sample-size, admixture of subjects with BD I and BD II, and the fact that BCL2 family members were assayed at the level of gene-expression rather than protein or functional assays. A further caveat is that while the average Li level in both groups was similar, and was within our target range of 0.6 to 1.2 mEq/L, there were several subjects in both groups whose levels were outside of that range, which may have affected both their response to treatment and the patterns of gene-induction observed in these subjects. In addition, these studies address only the acute anti-depressant properties of lithium, and do not address changes in gene-expression that may be related to lithium’s anti-manic, prophylactic or anti-suicide properties, each of which may be associated with a unique molecular profile. Due to the small sample size and limited information that was collected regarding prior episodes, we are unable to address the relationship, if any, between these molecular changes and the proposed specificity of lithium for “classic bipolar disorder” as opposed to the broader spectrum of bipolar illnesses . Future studies will be needed to confirm these findings in a larger cohort of patients and to determine the relevancy of these changes to other aspects of lithium’s clinical effects in patients with BD.