Estradiol accelerates the effects of fluoxetine on serotonin 1A receptor signaling

Abstract

A major problem with current anti-depressant therapy is that it takes on average 6-7 weeks for remission. Since desensitization of serotonin (5-HT)1A receptor signaling contributes to the anti-depressive response, acceleration of the desensitization may reduce this delay in response to antidepressants. The purpose of the present study was to test the hypothesis that estradiol accelerates fluoxetine-induced desensitization of 5-HT1A receptor signaling in the paraventricular nucleus of the hypothalamus (PVN) of rats, via alterations in components of the 5-HT1A receptor signaling pathway. Ovariectomized rats were injected with estradiol and/or fluoxetine, then adrenocorticotropic hormone (ACTH) and oxytocin responses to a 5-HT1A receptor agonist (+)-8-hydroxy-2-dipropylaminotetralin (8-OH-DPAT) were examined to assess the function of 5-HT1A receptors in the PVN. Treatment with estradiol for either 2 or 7 days or fluoxetine for 2 days produced at most a partial desensitization of 5-HT1A receptor signaling, whereas 7 days of fluoxetine produced full desensitization. Combined treatment with estradiol and fluoxetine for 2 days produced nearly a full desensitization, demonstrating an accelerated response compared to either treatment alone. With two days of combined treatments, estradiol prevented the fluoxetine-induced increase in 5-HT1A receptor protein, which could contribute to the more rapid desensitization. Furthermore, EB treatment for 2 days decreased the abundance of the 35 kD Gαz protein which could contribute to the desensitization response. We found two isoforms of Gαz proteins with molecular mass of 35 and 33 kD, which differentially distributed in the detergent resistant microdomain (DRM) and in Triton X-100 soluble membrane region, respectively. The 35 kD Gαz proteins in the DRM can be sumoylated by SUMO1. Stimulation of 5-HT1A receptors with 8-OH-DPAT increases the sumoylation of Gαz proteins and reduces the 33 kD Gαz proteins, suggesting that these responses may be related to the desensitization of 5-HT1A receptors. Treatment with estradiol for 2 days also reduced the levels of the G-protein coupled estrogen receptor GPR30, possibly limiting to the ability of estradiol to produce only a partial desensitization response. These data provide evidence that estradiol may be effective as a short-term adjuvant to SSRIs to accelerate the onset of therapeutic effects.

Figure 1. Effects of fluoxetine and/or EB treatment for 2 and 7 days on hormonal response to 8-OH-DPAT

Plasma levels of oxytocin (A) and ACTH (B) were measured 15 min after a saline or 8-OH-DPAT injection in ovariectomized rats pre-treated with fluoxetine and EB, alone or in combination, for 2 or 7 days. The data are presented as the mean ± SEM (n = 8–14). *: significantly different from saline-challenged group with same pre-treatment; #: significantly different from 8-OH-DPAT-challenged vehicle group; &: significantly different from 2 day pretreatment groups with same treatment. A line between groups indicates significant difference between the groups. All of the differences were determined by ANOVA and Student Newman-Keuls post-hoc tests. Vehicle: rats treated with saline and oil for 7 days; EB: rats treated with saline and EB, Flx: rats treated with fluoxetine and oil and EB plus Flx: rats treated with EB plus fluoxetine.

Figure 2. Effects of fluoxetine and/or EB treatment for 2 days on protein levels of 5-HT_1A receptor in the PVN membrane of saline- or 8-OH-DPAT-challenged rats

A. Characterization of 5-HT_1A receptor antibody: Left panel shows that the rabbit-anti-5-HT_1A receptor antibody detected 5-HT_1A receptors expressed in HEK 293 and A1A1V cells that were transfected with pcDNA-5HT_1A receptors as well as endogenously expressed 5-HT_1A receptors. Right panel shows the detection of 5-HT_1A receptors in the cytosol and membrane of rat hippocampus. B: Immunoblot assay for protein levels of 5-HT_1A receptors in the PVN membrane from fluoxetine and/or EB-treated rats challenged with saline (top) and 8-OH-DPAT (bottom). The data are presented as the mean ± SEM (n = 6–8). *: significantly different from saline/oil group with same challenge by Student Newman-Keuls post-hoc test, P < 0.05. Saline/oil: rats treated with saline and oil; Saline/EB: rats treated with saline and EB; Flx/oil: rats treated with fluoxetine and oil and Flx/EB: rats treated with combined fluoxetine and EB.

Figure 3. Effects of fluoxetine and/or EB treatment for 2 days and acute 8-OH-DPATon the Gαz protein and sumoylated Gαz in the DRM

A: Immunoblot of sucrose fractions from Triton X-100 treated cortex membrane to determine the distribution of Gαz protein in the DRM. Fractions containing the DRM were identified by the DRM marker, flotillin; Total: Triton X-100-treated membrane preparation without sucrose gradient centrifugation. B: To verify that both the 35 and 33 kD bands detected by anti-Gαz antibody are Gαz proteins, we unilaterally injected recombinant adenovirus containing Gαz-antisense sequence (Ad-Gαz-AS) into the PVN. Immunoblotting of Gαz protein performed using tissue from the viral injection site and its contra-lateral side of the PVN, showed that injection of Ad-Gαz-AS reduces both 35 and 33 kD Gαz proteins. C: Effect of acute treatment with 8-OH-DPAT on the ratio of 35/33 kD Gαz protein (left y-axis) and protein level of Gαz protein (right y-axis). The data were presented as the mean ± SEM (n=4–5). *: significantly different from saline treated group by Student Newman-Keuls post hoc test, P < 0.05. D: Effect of EB and/or fluoxetine treatment for 2 days on the ratio of 35/33 kD Gαz protein (left y-axis) and protein level of Gαz protein (right y-axis). The data are presented as the mean ± SEM (n = 5–6). *: significantly different from saline/oil group by Student-Newman-Keuls post-hoc test. E and F: Sumoylation of Gαz protein is verified by imunoprecipitation with anti-SUMO1 antibody (E) and anti-Gαz antibody (F). Flotillin 1 shown in (E) was prepared using 8-OH-DPAT-treated fractions. IP: immunoprecipitation, IB: Immunobloting. IgG: Rabbit IgG. Input: sample without IP. G: Sumoylated Gαz protein in the hypothalamus without PVN was increased by acute treatment with 8-OH-DPAT. The data were normalized to sumoylated β-actin and calculated as described in Methods. The data are presented as the mean ± SEM (n=5–6). *: significantly different from saline treated group by t-test, P < 0.05.

Figure 4. Effects of fluoxetine and/or EB treatment for 2 days on protein levels of active form of GPR30 in the membrane of the PVN

A: The majority of non-modified GPR30 (~32 kD is located in the DRM; Total: Triton X-100-treated membrane preparation without sucrose gradient centrifugation. B: Effect of EB and/or fluoxetine treatment for 2 days on the protein level of GPR30. The data are presented as the mean ± SEM (n = 5–6). Two-way ANOVA showed a significant difference on the effect of EB, although no significant difference between individual groups was detected with the Student Newman-Keuls post-hoc test.

Figure 5. Subcellular distribution of Gαz protein and GPR30 in rat hippocampus

Immunoblots shows the co-localization of Gαz protein and GPR30 with plasma membrane markers, Na⁺/K⁺ ATPase α1 and flotillin 1; ER marker, calreticulin; ER-Golgi intermediate compartment marker, LMAN1 (Lectin, Mannose-binding 1, or ERGIC 53); trans-Golgi network marker, TGM 38 (Trans-Golgi network protein2); endosome marker, EEA1(early endosome antigen 1) and cytosol marker, LDH (lactate dehydrogenase). Mem and Cyto indicate the total membrane fraction and cytosolic fraction of the cortex tissue prepared as described previously (Creech et al., 2012).