The present study demonstrates statistically significant decreases in protein and mRNA levels of anti-apoptotic factors (Bcl-2, BDNF) and of synaptic markers (synaptophysin and drebrin), and significant increases in pro-apoptotic factors (Bax, BAD, active Caspase- 3 and -9) in postmortem prefrontal cortex from BD compared with control subjects.
Recent brain imaging studies have revealed that the volumes of the hippocampus, amygdala, and frontal cortex are decreased in BD patients (42
), and that numbers and sizes of glia and neurons are reduced in discrete brain areas (45
). Several studies have also demonstrated mitochondrial dysfunction and increased pro-apoptotic activity in serum of BD patients (8
). Although studies implicate the association of apoptosis in the pathophysiology of BD, a few studies have investigated apoptosis directly in the postmortem brain of BD patients.
Members of the Bcl-2 family play important roles in the regulation of apoptosis. The representative member of this family is Bcl-2, an inner mitochondrial membrane protein with anti-apoptotic activity (47
). The Bcl-2 homologue, Bax, a monomeric cytosolic protein, displays a pro-apoptotic function. Bax can homodimerize and trigger the activation of terminal caspase by altering mitochondrial function, which results in the release of apoptosis-promoting factors into the cytoplasm. The ratio between Bax/Bcl-2 appears to be essential in deciding the life or death of a cell (48
). In our current study, we showed an increased ratio of Bax/Bcl-2 and increased Caspase-3 and -9 active protein and mRNA levels. These results suggest that there might be an aberration in the apoptotic pathway of the BD brain.
We recently reported significant increases of mRNA and protein levels of calcium-dependent phospholipase A2
), which releases arachidonic acid (AA) from membrane phospholipids, in postmortem brain of BD (23
). AA can bind 14-3-3ζ protein, which has important roles in preventing apoptosis by retaining the pro-apoptotic protein BAD, and by reducing the binding of 14-3-3ζ to phosphorylated BAD (49
). Release of 14-3-3ζ from BAD allows dephosphorylation of BAD and allows BAD to move from the cytoplasm to the mitochondria, where it can displace Bax, leading to apoptosis (50
). In this study, we observed increased BAD protein and mRNA levels in postmortem brain of BD. Increased cPLA2
expression may induce AA release which can promote early steps in the apoptotic pathway, through the dissociation of 14-3-3ζ from phosphorylated BAD.
Serretti et al. (51
) used linkage and association methods to identify genes that are involved in BD, which included the BDNF gene. BDNF is a primary neurotrophic factor, and plays important roles in cell survival, cell plasticity, and in the growth and differentiation of new neurons and synapses (52
). Animal models that demonstrated upregulated AA signaling and bipolar-like behaviors have been reported to have downregulated brain BDNF expression (53
). Furthermore, several drugs approved for treating BD show a neuroprotective effect ascribed to increased BDNF expression (55
). In our study, we demonstrated decreased BDNF mRNA and protein levels in the frontal cortex of BD. These data indicate that decreased BDNF may be part of the pathophysiology of BD.
Mood stabilizers utilized in BD, when given long-term to rats, downregulate brain expression of cPLA2
or COX-2, which are key enzymes of AA metabolism (56
). Furthermore, cPLA2
and COX-2 are increased in prefrontal cortex of the postmortem BD brain (23
). Thus increased cPLA2
would release more AA, which may interrupt the anti-apoptotic action of 14-3-3ζ in the brain. Consistent with these findings, a recent study showed that the inhibition of cPLA2
-mediated AA release reduced apoptosis in astrocytes (57
). Additionally, mood stabilizers are reported to suppress caspase-3 activity, stimulate Bcl-2 and BDNF expression and enhance neurogenesis in rat hippocampus (14
). Our findings suggest that deregulation of apoptosis may be involved in BD.
Altered pro- and anti-apoptotic factors may cause changes in neuronal markers. Reports have demonstrated loss of synaptic integrity, associated with decreased expression of the postsynaptic marker drebrin and presynaptic marker synaptophsyin, in the Alzheimer disease brain (58
). We observed a significant decrease in protein levels of synaptophysin and drebrin in BD brain compared with control. These decreases may be responsible for cognitive deficits that have been reported in BD patients.
The limitation of the present study is non-availability of medical diagnosis, and lack of information on whether the patients were in the manic or depressive phase at the time of death. However, since several BD patients died by suicide, they may have been in the depressed phase of their illness. Also, the BD patients had been exposed to various drugs not experienced by the control subjects, which may have confounded the results. Therefore, our findings may be related to differences in drug exposure, rather than the BD trait. However, no statistical differences were found in all genes studied in the present study when the BD subjects were compared with the subgroup of BD subjects that were treated with lithium (data not shown). Also, no statistical significance was found when the BD subjects were compared to the BD subjects that died by suicide. However, future studies should examine apoptotic and synaptic markers in brains of patients with schizophrenia using roughly comparable drug exposure as a control, or with unipolar major depression, or with Alzheimer disease to test for disease specificity.
In summary, postmortem frontal cortex from BD patients compared with control cortex showed significantly decreased anti-apoptotic factor (Bcl-2 and BDNF) protein and mRNA levels, and reduced protein levels of synaptic markers (synaptophysin and drebrin), but increased protein and mRNA expression of pro-apoptotic factors (Bax, BAD and caspase-9/-3). These alterations may enhance apoptosis in the frontal cortex of BD patients. Apoptosis and synaptic loss may occur in the presence of neuroinflammation and excitotoxicity in the BD brain, and may be triggered or interact with these process (24
). These multiple pathological processes may be the basis of disease progression, evidenced by reports of progressive mood disturbance, brain atrophy and cognitive decline. Therapeutic strategies aimed at downregulating apoptotic processes and neuronal degeneration might be effective in slowing the progression of BD. Mood stabilizers may help to do this (56