Sertraline modulates hippocampal plasticity via sigma 1 receptors, cellular stress and neurosteroids

Abstract

In addition to modulating serotonin transport, selective serotonin reuptake inhibitors (SSRIs) have multiple other mechanisms that may contribute to clinical effects, and some of these latter actions prompt repurposing of SSRIs for non-psychiatric indications. In a recent study of the SSRIs fluvoxamine, fluoxetine and sertraline we found that, unlike the other two SSRIs, sertraline acutely inhibited LTP at a low micromolar concentration through inverse agonism of sigma 1 receptors (S1Rs). In the present studies, we pursued mechanisms contributing to sertraline modulation of LTP in rat hippocampal slices. We found that sertraline partially inhibits synaptic responses mediated by N-methyl-D-aspartate receptors (NMDARs) via effects on NMDARs that contain GluN2B subunits. A selective S1R antagonist (NE-100), but not an S1R agonist (PRE-084) blocked effects on NMDARs, even though both S1R ligands were previously shown to prevent LTP inhibition. Both NE-100 and PRE-084, however, prevented adverse effects of sertraline on one-trial learning. Because of the important role that S1Rs play in modulating endoplasmic reticulum stress, we examined whether inhibitors of cellular stress alter effects of sertraline. We found that two stress inhibitors, ISRIB and quercetin, prevented LTP inhibition, as did inhibitors of the synthesis of endogenous neurosteroids, which are homeostatic regulators of cellular stress. These studies highlight complex effects of sertraline, S1Rs and neurosteroids on hippocampal function and have relevance for understanding therapeutic and adverse drug actions.

Fig. 1. Sertraline inhibits LTP and partially depresses NMDAR responses.

A In control slices, a single 100 Hz × 1 s HFS (arrow) reliably induces LTP in the CA1 region (white circles). A 30 min administration of 1μM sertraline (red bar) completely inhibits LTP induction (black circles). B Sertraline partially depresses NMDAR-mediated EPSPs. Addition of a low concentration of D-APV (5 μM) produces additive depression of NMDAR responses. C The GluN2B inhibitor, ifenprodil (10 μM) partially depresses NMDAR responses and addition of sertraline results in no further depression. D Despite inhibiting GluN2B NMDARs, sertraline failed to block induction of LTD by 1 Hz LFS (hashed bar). E The bar graphs show summary results from A–D; **p < 0.01 as indicated. Traces to the right of the graphs show representative EPSPs under baseline conditions (dashed traces) and 60 min following HFS (A) or LFS (D) (solid traces). In panels B and C, baseline NMDAR EPSPs are dashed traces and solid traces are in the presence of drugs. Scale bar: 1 mV, 5 ms.

Fig. 2. Effects of sertraline on NMDAR responses involve S1Rs.

A At 1 μM, the S1R antagonist NE-100 (white bar) had no effect on NMDAR EPSPs but blocked the effects of sertraline (red bar). A low concentration of APV (as in Fig. 1, green bar) resulted in partial suppression of these responses. B The S1R agonist, PRE-084 (1 μM), a concentration that overcame the effects of sertraline on LTP in our prior study [19] had no effect on NMDAR responses alone and failed to inhibit the effects of sertraline. C The bar graphs depict summary results from A and B; **p < 0.01 as indicated. As in Fig. 1, low APV produced additive depression of responses. Traces show representative NMDAR EPSPs at the times indicated. Calibration bar: 1 mV, 5 ms.

Fig. 3. Sertraline administration results in persistent LTP inhibition that is not altered by co-administration of saturating APV.

A Administration of sertraline for 30 min (red bar) followed by 30 min washout prior to HFS (arrow) resulted in complete suppression of LTP induction. B In contrast to what we have observed with agents that produce NMDAR-dependent metaplastic LTP inhibition [37], administration of sertraline in the presence of saturating APV (50 μM, green bar) followed by washout of both agents 30 min prior to APV did not prevent LTP inhibition. C. The bars show results from (A, B). LTP was suppressed in both experiments. Traces show EPSPs as in Fig. 1A. Calibration: 1 mV, 5 ms.

Fig. 4. Inhibitors of cellular stress prevent the effects of sertraline on LTP.

A Administration of 1 μM ISRIB (purple bar), an agent that inhibits the integrated stress response, overcame the acute effects of sertraline (red bar) on LTP. B Quercetin (50 μM, pink bar), an agent that inhibits ER stress, also prevented the effects of sertraline (red bar) on LTP when perfused before and during sertraline. C. The bar graph depicts results from (A, B); **p < 0.01 compared to sertraline alone. Traces show representative EPSPs; dashed traces are baseline and red traces are 60 min after HFS. Scale bar: 1 mV, 5 ms.

Fig. 5. Inhibitors neurosteroid synthesis prevent the effects of sertraline on LTP.

A The 5-alpha reductase (5AR) inhibitor finasteride prevented the effects of sertraline on LTP when administered at 10 μM (green bar) but not 1 μM (light green bar). The two concentrations of finasteride were administered in completely separate experiments. B The more potent broader spectrum 5AR antagonist, dutasteride (1 μM), completely prevented the effects of sertraline when perfused before and during sertraline. C At 1 μM, the GABA_AR antagonist picrotoxin (PTX) failed to alter LTP inhibition by sertraline. D. The bar graph shows results from A–C; **p < 0.01, *p < 0.05 compared to sertraline alone. Traces show EPSPs. Calibration: 1 mV, 5 ms.

Fig. 6. Sertraline acutely dampens one-trial inhibitory avoidance learning in an S1R-dependent fashion.

The graph shows that control rats treated with vehicle (DMSO) alone readily learn the task and remain in the lit compartment for the full 300 s trial when tested 24 h following conditioning. Sertraline (10 mg/kg ip) administered an hour prior to conditioning produced defects in learning that were completely reversed by either an S1R agonist (PRE-084,10 mg/kg ip) or S1R antagonist (NE-100, 10 mg/kg ip) administered 1 h prior to sertraline. S1R ligands were administered 1-h prior to sertraline as described in Methods. **p = 0.0064 by Dunn’s multiple comparison test following Kruskal–Wallis test.