Rebound activation of 5-HT neurons following SSRI discontinuation

Abstract

Cessation of therapy with a selective serotonin (5-HT) reuptake inhibitor (SSRI) is often associated with an early onset and disabling discontinuation syndrome, the mechanism of which is surprisingly little investigated. Here we determined the effect on 5-HT neurochemistry of discontinuation from the SSRI paroxetine. Paroxetine was administered repeatedly to mice (once daily, 12 days versus saline controls) and then either continued or discontinued for up to 5 days. Whereas brain tissue levels of 5-HT and/or its metabolite 5-HIAA tended to decrease during continuous paroxetine, levels increased above controls after discontinuation, notably in hippocampus. In microdialysis experiments continuous paroxetine elevated hippocampal extracellular 5-HT and this effect fell to saline control levels on discontinuation. However, depolarisation (high potassium)-evoked 5-HT release was reduced by continuous paroxetine but increased above controls post-discontinuation. Extracellular hippocampal 5-HIAA also decreased during continuous paroxetine and increased above controls post-discontinuation. Next, immunohistochemistry experiments found that paroxetine discontinuation increased c-Fos expression in midbrain 5-HT (TPH2 positive) neurons, adding further evidence for a hyperexcitable 5-HT system. The latter effect was recapitulated by 5-HT_1A receptor antagonist administration although gene expression analysis could not confirm altered expression of 5-HT_1A autoreceptors following paroxetine discontinuation. Finally, in behavioural experiments paroxetine discontinuation increased anxiety-like behaviour, which partially correlated in time with the measures of increased 5-HT function. In summary, this study reports evidence that, across a range of experiments, SSRI discontinuation triggers a rebound activation of 5-HT neurons. This effect is reminiscent of neural changes associated with various psychotropic drug withdrawal states, suggesting a common unifying mechanism.

Fig. 1. Effect of paroxetine discontinuation on tissue levels of 5-HT and 5-HIAA in mouse brain regions.

Regions were (A) hippocampus, (B) striatum, (C) frontal cortex, and (D) midbrain. Mice received either saline (SAL) or 10 mg/kg s.c. paroxetine (CON) once daily for 12 days and then paroxetine was discontinued for 2 days (DIS 2) or 5 days (DIS 5). Each column is a mean ± SEM value of the individual points shown, and each data point is derived from a single animal. Experiments for 2 day and 5-day discontinuation were run as 3-arm studies (saline, continuation, discontinuation) and values for saline and continuous paroxetine groups were pooled across experiments for clarity of illustration (open and closed symbols are matched across experiments). Data (nmol/mg tissue) were analysed using one-way ANOVA with post hoc Fisher’s LSD, *p < 0.05, **p > 0.01.

Fig. 2. Effect of paroxetine discontinuation on extracellular 5-HT and 5-HIAA in hippocampus as measured by microdialysis in anaesthetised mice.

Mice received either saline (SAL) or 10 mg/kg s.c. paroxetine (CON) once daily for 12 days and then paroxetine was discontinued for 2 days (DIS 2) or 5 days (DIS 5). Mice were then subject to the experimental design illustrated (A). Baseline levels of 5-HT (B) and 5-HIAA (C) at individual time points (left) and averaged over the time course (right). D Effect of perfusion with 56 mM KCl (left) or 100 mM KCl (middle) together with 100 mM: 56 mM KCl ratio (right). Mean ± SEM values are shown, and each data point is derived from a single animal. Data analysed using either repeated measures ANOVA with post-hoc Bonferroni’s multiple comparisons test (time course) or one-way ANOVA with Fisher’s LSD (time course averages), *p < 0.05, **p < 0.01.

Fig. 3. Effect of paroxetine discontinuation on c-Fos/TPH2 double-labelled neurons in the DRN.

Mice received either saline (SAL) or 10 mg/kg s.c. paroxetine (CON) once daily for 12 days and then paroxetine was discontinued for 2 days (DIS 2) or 5 days (DIS 5). A Illustration of DRN localisation (pink shading) together with images of c-Fos, TPH2 and NeuN immunoreactivity and their merger at low (x10) and high (x40) resolution (scale bar 100 μm; white arrows show c-Fos/TPH2 co-labelled neurons). Dotted lines indicate the medial longitudinal fasciculi. B Quantified data showing number of showing c-Fos/TPH2 co-labelled neurons. Mean ± SEM values are shown and each data point is derived from a single animal. Data analysed using one-way ANOVA with post-hoc Fisher’s LSD, *p < 0.05, **p < 0.01.

Fig. 4. Effect of paroxetine discontinuation on anxiety-like behaviour on the EPM.

Mice received either saline (SAL) or 10 mg/kg s.c. paroxetine (CON) once daily for 12 days and then paroxetine was discontinued for 2 days (DIS 2) or 5 days (DIS 5). A Experimental design with illustration of EPM apparatus. The study comprised a 3-arm design with groups being repeated saline, repeated paroxetine, and repeated paroxetine discontinued for either 2 or 5 days. EPM parameters (time on open arms, open arm entries, open arm head dips, distance travelled on maze) are shown for day 2 (B) and day 5 (C) of discontinuation. Mean ± SEM values are shown and each data point is derived from a single animal. Data analysed using one-way ANOVA with Fisher’s LSD, **p < 0.01.