Diverse antidepressants increase CDP-diacylglycerol production and phosphatidylinositide resynthesis in depression-relevant regions of the rat brain

Abstract

Background: Major depression is a serious mood disorder affecting millions of adults and children worldwide. While the etiopathology of depression remains obscure, antidepressant medications increase synaptic levels of monoamine neurotransmitters in brain regions associated with the disease. Monoamine transmitters activate multiple signaling cascades some of which have been investigated as potential mediators of depression or antidepressant drug action. However, the diacylglycerol arm of phosphoinositide signaling cascades has not been systematically investigated, even though downstream targets of this cascade have been implicated in depression. With the ultimate goal of uncovering the primary postsynaptic actions that may initiate cellular antidepressive signaling, we have examined the antidepressant-induced production of CDP-diacylglycerol which is both a product of diacylglycerol phosphorylation and a precursor for the synthesis of physiologically critical glycerophospholipids such as the phosphatidylinositides. For this, drug effects on [3H]cytidine-labeled CDP-diacylglycerol and [3H]inositol-labeled phosphatidylinositides were measured in response to the tricyclics desipramine and imipramine, the selective serotonin reuptake inhibitors fluoxetine and paroxetine, the atypical antidepressants maprotiline and nomifensine, and several monoamine oxidase inhibitors.

Results: Multiple compounds from each antidepressant category significantly stimulated [3H]CDP-diacylglycerol accumulation in cerebrocortical, hippocampal, and striatal tissues, and also enhanced the resynthesis of inositol phospholipids. Conversely, various antipsychotics, anxiolytics, and non-antidepressant psychotropic agents failed to significantly induce CDP-diacylglycerol or phosphoinositide synthesis. Drug-induced CDP-diacylglycerol accumulation was independent of lithium and only partially dependent on phosphoinositide hydrolysis, thus indicating that antidepressants can mobilize CDP-diacylglycerol from additional pools lying outside of the inositol cycle. Further, unlike direct serotonergic, muscarinic, or alpha-adrenergic agonists that elicited comparable or lower effects on CDP-diacylglycerol versus inositol phosphates, the antidepressants dose-dependently induced significantly greater accumulations of CDP-diacylglycerol.

Conclusion: Chemically divergent antidepressant agents commonly and significantly enhanced the accumulation of CDP-diacylglycerol. The latter is not only a derived product of phosphoinositide hydrolysis but is also a crucial intermediate in the biosynthesis of several signaling substrates. Hence, altered CDP-diacylglycerol signaling might be implicated in the pathophysiology of depression or the mechanism of action of diverse antidepressant medications.

Figure 1

Effects of classic antidepressants on [³H]CDP-diacylglycerol accumulation. Tissue slices prepared from indicated brain regions were prelabeled with [³H]cytidine and incubated with various concentrations of either imipramine (IMI), desipramine (DES), fluoxetine (FLU), paroxetine (PAR), maprotiline (MAP), or nomifensine (NOM). After 90 min, tissue contents of [³H]CDP-diacylglycerol were assayed. Each bar is the mean ± SEM (N = 9). Each drug stimulated significant concentration-dependent accumulations of CDP-diacylglycerol (ANOVA, p < 0.001 for each drug). Based on posthoc Dunnett tests, all agents induced statistically significant CDP-diacylglycerol responses at the 3 or 10 μM concentrations, except for paroxetine in the hippocampus and imipramine in the striatum where the drug effects were not significant until the 30 μM and higher concentrations.

Figure 2

Effects of various classic antidepressants on [³H]phosphatidylinositide synthesis. Tissue slices prepared from indicated brain regions and prelabeled with [³H]inositol were incubated with various concentrations of either imipramine (IMI), desipramine (DES), fluoxetine (FLU), paroxetine (PAR), or maprotiline (MAP). After 90 min, [³H]inositol phospholipids were extracted and assayed as a total pool of extractable phosphatidylinositides. Each bar is the mean ± SEM (N = 9). Each drug stimulated significant and concentration-dependent increases in [³H]inositol phospholipid synthesis (ANOVA, p < 0.001 for each drug). From the subsequent posthoc Dunnett tests, all agents induced statistically significant CDP-diacylglycerol responses at the 3 or 10 μM and higher concentrations.

Figure 3

Effects of the monoamine oxidase inhibitors on [³H]CDP-diacylglycerol accumulation and phosphatidylinositide synthesis. Cortical slices were tested with indicated concentrations of phenelzine or hydralazine. CDP-diacylglycerol (CDP-DG) levels were analyzed as outlined under Figure 1, while phosphatidylinositide (PI) synthesis was measured as the total pool of extractable [³H]inositol-labeled phospholipids as in Figure 2. Each point is the mean ± SEM (N = 9). Each drug stimulated significant concentration-dependent accumulations of CDP-diacylglycerol (ANOVA, p < 0.001 for each drug) and phosphatidylinositide synthesis (ANOVA, p < 0.001 for each drug). Note that if individually compared to the corresponding control measure, each of the tested concentrations from 0.1 – 300 μM gave statistically significant effects on CDP-DG and PIs. Note also the greater relative accumulation of PIs compared to the accumulation of CDP-DG.

Figure 4

Effects of diverse antidepressants on [³H]inositol phosphate accumulation. Experiments were conducted as outlined in the legend to Figure 2, except that tissue contents of [³H]inositol phosphates were assayed by Dowex anion exchange chromatography as detailed in Methods. Each bar is the mean ± SEM (N = 9). Each antidepressant agent stimulated significant concentration-dependent accumulations of inositol phosphate (ANOVA, p < 0.001 for each drug).

Figure 5

Inhibition of antidepressant-induced [³H]CDP-diacylglycerol production and phosphoinositide synthesis by neomycin. Slices of rat prefrontal cortex or hippocampus prepared from the same rats were pre-labeled in parallel with [³H]cytidine or [³H]inositol and incubated with indicated concentrations of neomycin, followed by addition of 100 μM fluoxetine (FLU) or 300 μM imipramine (IMI), desipramine (DES), paroxetine (PAR), maprotiline (MAP), or nomifensine (NOM). Accumulated [³H]CDP-diacylglycerol (A) or [³H]phosphoinositides (B) were measured after 90 min. While only the hippocampus data are shown for CDP-diacylglycerol and the cortical data for phosphatidylinositides, each analyte was assessed in each brain region with similar results. Each bar is the mean ± SEM (n = 6). Neomycin significantly and concentration-dependently inhibited drug-induced [³H]CDP-diacylglycerol production and [³H]inositol phospholipid synthesis (ANOVA, p < 0.01 for each drug). The effects of neomycin alone on CDP-diacylglycerol were not significant, whereas it exerted a slight but significant increase at the 0.3 mM concentration (Dunett's, p < 0.01) and decrease at the 3 mM concentration (Dunnett, p < 0.01) in [³H]phosphatidylinositide levels. #p < 0.05; *p < 0.05, compared to respective effects of antidepressant alone by Dunnett test.

Figure 6

Effects of the PLC inhibitor U73122 on antidepressant-mediated CDP-diacylglycerol production and inositol phosphate accumulation. Cerebrocortical or hippocampal slices labeled with either [³H]inositol or [³H]cytidine were incubated in parallel with buffer alone or the indicated concentrations of U73122, followed by addition of 100 μM fluoxetine (FLU) or 300 μM imipramine (IMI), paroxetine (PAR), maprotiline (MAP), nomifensine (NOM), or SKF38393 (SKF) as indicated. Accumulated [³H]CDP-diacylglycerol (A) or [³H]inositol phosphates (B) were determined after 90 min. While only the hippocampus data are shown for inositol phosphates and the cortical data for CDP-diacylglycerol, each analyte was assessed in each brain region with similar results. Each bar is the mean ± SEM (n = 6). U73122 completely blocked [³H]inositol phosphate accumulation stimulated by either antidepressant agent (ANOVA, p < 0.001). Conversely, U73122 only partially reduced antidepressant drug-induced [³H]CDP-diacylglycerol production. The phospholipase C inhibitor by itself showed slight but significant effects on either analyte. *p < 0.01 compared to effects of antidepressant alone (Dunnett test).

Figure 7

Effects of LiCl on antidepressant-induced [³H]CDP-diacylglycerol production. Slices of prefrontal cortical or hippocampal tissues were labeled with [³H]cytidine and incubated in the presence or absence of 5 mM LiCl. Indicated concentrations of fluoxetine or imipramine were added, and after 60 min accumulated [³H]CDP-diacylglycerol was measured. While the frontal cortex data are shown, similar observations were made in the hippocampus. Each bar is the mean ± SEM (n = 6). The presence of LiCl did not significantly alter the stimulatory effects of fluoxetine or imipramine on [³H]DCP- diacylglycerol accumulation (ANOVA, p > 0.05).

Figure 8

Ratios of antidepressant-induced CDP-diacylglycerol versus inositol phosphates. The data in Figures 1-4 were recalculated by dividing the CDP- diacylglycerol effects of each concentration of each antidepressant agent by the corresponding effects of the agent on inositol phosphate accumulation to yield the CDP- diacylglycerol/inositol phosphate (CDP-DG/IP) ratios shown. To facilitate merging of data from multiple experiments, these values were converted to percentages relative to the ratio values in the respective control samples and then averaged to give the mean ± SEM shown for the classic agents (A), phenelzine (B), and hydralazine (C). Data for the hippocampus are shown, but similar observations were made in the cortical tissues. Data for each agent were analyzed by One-way ANOVA followed by posthoc Dunnett tests. ^#p < 0.05; *p < 0.01; compared to the respective control (zero drug concentration).

Figure 9

Effects of α-methylserotonin, carbachol, and phenylephrine on CDP-diacylglycerol production, inositol phosphate accumulation and phosphatidylinositide synthesis. Slices of prefrontal cortical or hippocampal tissues were pre-labeled with [³H]cytidine or [³H]inositol in parallel, and then incubated in the presence of 5 mM LiCl. Indicated concentrations of the agonists were added for 90 min, followed by assay of the levels of [³H]CDP- diacylglycerol, [³H]inositol phosphates, and [³H]phosphoinositides. Calculated ratios of CDP- diacylglycerol over inositol phosphates (CDP-DG/IP ratios) are shown in the rightmost panel relative to the scale on the right y-axis. Each bar is the mean ± SEM (n = 15 for α-methylserotonin, 12 for carbachol, 9 for phenylephrine). Data were separately analyzed by One-way ANOVA for each receptor agonist. *p < 0.01, Dunnett test compared to the respective control (zero drug concentration).

Figure 10

Effects of SKF38393 on CDP-diacylglycerol production, inositol phosphate accumulation and phosphatidylinositide synthesis. Prefrontal cortical slices pre-labeled with [³H]cytidine or [³H]inositol in parallel were incubated with indicated concentrations of SKF38393 in the presence of 5 mM LiCl for 90 min, followed by assay of [³H]CDP-diacylglycerol (A), [³H]phosphoinositides (B) and [³H]inositol phosphates (C). Calculated ratios of CDP-diacylglycerol over inositol phosphates (CDP-DG/IP ratio), phosphatidylinositides (CDP-DG/PI ratio), or the sum of the inositol phosphates and phosphatidylinositides (CDP-DG/IPPI ratio) are shown in the right panel (D-E, respectively). Each bar is the mean ± SEM (n = 12). Data were analyzed by One-way ANOVA and posthoc Dunnett tests. #p < 0.05; *p < 0.01, compared to the respective control by Dunnett test.