The effect of escitalopram, desipramine, electroconvulsive seizures and lithium on brain-derived neurotrophic factor mRNA and protein expression in the rat brain and the correlation to 5-HT and 5-HIAA levels

Abstract

The reported increase in brain-derived neurotrophic factor (BDNF) mRNA expression after antidepressant treatment is a cornerstone of the BDNF hypothesis of antidepressant action. However, if this increase becomes manifest on the BDNF protein level is unknown. In the present study we performed parallel measurements of BDNF mRNA and protein expression in the frontal cortex and hippocampus of the rat after chronic treatment with electroconvulsive seizures (ECS), lithium, desipramine or escitalopram. ECS increased BDNF mRNA and protein in the hippocampus and BDNF protein in the frontal cortex. Desipramine moderately increased BDNF mRNA expression in the dentate gyrus but did not change BDNF protein in neither region. Escitalopram did not affect BDNF mRNA expression, but decreased BDNF protein in the frontal cortex and the hippocampus. Lithium increased BDNF protein levels in the hippocampus and frontal cortex, but overall decreased BDNF mRNA expression. Thus, here we report a striking non-correspondence between changes in BDNF mRNA and protein expression induced by the antidepressant treatments and lithium. Further, increased expression of BDNF mRNA or protein was not a common action of the treatments. We also investigated if treatment-induced modulations of the tissue contents of 5-hydroxytryptamine (5-HT) and its metabolite, 5-hydroxy-indoleacetic acid (5-HIAA), were related to changes in BDNF mRNA or protein expression. No correlation was found. However, all treatments increased 5-HT levels in the hippocampus.

Introduction

In the brain, brain-derived neurotrophic factor (BDNF) has been implicated in development, neural regeneration, synaptic transmission, synaptic plasticity and neurogenesis [40], [57], [64], [76], [87]. A number of findings collectively points to a role for BDNF in depression and/or the action of antidepressants. In the hippocampus, BDNF mRNA expression was decreased by stress and glucocorticoids [71], [78] while a range of antidepressants increased BDNF and trkB (the receptor for BDNF) mRNA [50]. Furthermore, the stress-induced decrease in BDNF mRNA expression in the hippocampus was blocked by chronic pre-treatment with antidepressants [50]. When taking the purported roles of BDNF into consideration, the opposing effects of stress and antidepressants on BDNF mRNA expression are intriguing. Specifically, depressed patients may have reduced neural activity in fronto-cortical areas [29], [81] and hippocampal atrophy [10], [65]. Conceivably, these findings could be related to decreased BDNF function and antidepressants may reverse brain dysfunctions by enhancing BDNF function. Indeed, increased hippocampal BDNF immunoreactivity post-mortem was reported in subjects treated with antidepressants [16]. However, a recent study could not replicate these findings [74].

Another line of evidence, implicating BDNF in depression, is the trophic effect of BDNF on the 5-hydroxytryptamine (5-HT) system. Intracerebral BDNF infusion stimulated 5-HT turnover, synthesis and sprouting of 5-HT axons [42], [41], [68], [70] and demonstrated antidepressant-like activity in animal models of depression [67], [69]. Conversely, heterozygous BDNF knock-out mice developed 5-HT deficits with age [36] and were unresponsive to antidepressants in the forced swim test [62].

Whatever the attractiveness of the BDNF hypothesis of antidepressant action [20], recent findings have been mixed. Chronic fluoxetine has been reported to increase [51], decrease [45] or have no effect [4], [18], [21] on BDNF mRNA expression in the hippocampus. Chronic desipramine has been reported to increase [18], [50] or not affect [19], [61] BDNF mRNA expression in the hippocampus. Electroconvulsive seizures (ECS) is consistently found to stimulate BDNF mRNA expression [3], [5], [50], [85]. In a recent report chronic fluoxetine or tranylcypromine decreased BDNF mRNA in the dentate gyrus 4 h after the last injection, but increased BDNF mRNA at 24 h [19]. These findings could, at least partly, explain the discrepant findings on antidepressants and BDNF mRNA expression. However, use of different animal strains, dosing regimens and drug doses could certainly play a role as well [4], [11], [18], [19], [45], [50], [51], [61], [79], [86].

The effect of antidepressants on BDNF protein has been less studied. Chronic 5 mg/kg/day amitryptyline or 5 mg/kg/day venlafaxine increased BDNF immunostaining in the hippocampus [83]. However, 10 mg/kg/day amitryptyline had no effect, while 10 mg/kg/day venlafaxine decreased BDNF immunostaining [83]. In another study, chronic 10 mg/kg/day amitryptyline increased BDNF protein in the frontal cortex and hippocampus [52]. In apparent contrast to previous findings on BDNF mRNA expression [50], [51], chronic fluoxetine or desipramine had no effect on BDNF protein in the hippocampus [3]. Moreover, chronic treatment with the mood stabilizer lithium increased BDNF protein in the hippocampus and cortex [24]. The latter finding is interesting, since BDNF may be essential for lithium's neuroprotective effect [26].

Increased BDNF mRNA expression after antidepressants, particularly in the hippocampus, is central in the BDNF hypothesis of depression [20]. However, it is not certain if BDNF mRNA modulations always become manifest on the BDNF protein level, at least not in the same region. In fact, there is precedence that antidepressant-induced modulations of mRNA and protein may not correspond [33]. Moreover, BDNF mRNA and protein may have different origins since BDNF protein can be anterogradely transported along axons [2].

In the present study, we examined if antidepressants modulated BDNF mRNA and protein in concert. TrkB mRNA expression was assessed in parallel. In addition, we examined if drug-induced modulations of BDNF mRNA and/or protein correlated to modulations in tissue levels of 5-HT and 5-hydroxy-indoleacetic acid (5-HIAA; the main metabolite of 5-HT). Tentatively, if BDNF function is a major determinant for 5-HT function, 5-HT or 5-HIAA might reflect BDNF levels. We focused on the hippocampus and frontal cortex, two brain areas that may be dysfunctional in depression [10], [29], [65], [81]. We chose three different classes of antidepressant treatments with distinct modes of action: escitalopram (a selective serotonin reuptake inhibitor [SSRI]), desipramine (a tricyclic antidepressant antagonizing the noradrenaline transporter) and ECS. In addition, we included lithium to investigate if the increase in BDNF protein previously reported [24], [26] was paralleled on the BDNF mRNA level. Where possible, we used modes of drug-delivery minimizing fluctuations in plasma drug levels. We delivered escitalopram via minipumps and lithium via the diet. Desipramine delivered via minipumps caused skin necrosis and had to be administered by injections.