Rosmarinic Acid Ameliorates Depressive-like Behaviors In A Rat Model Of CUS And Its Underlying Mechanisms

OBJECTIVE:The purpose of this study is to explore the anti-depressive effects of Rosmarinic acid(RA, a primary constituent of a Chinese herbal medicine) and explore its underlying mechanisms. METHODS:In the precent study, all the experiments were performed in rats, and were designed for 4 parts:Experiment I:To investigate the therapeutic effects of the RA in a CUS rat model, SD rats were randomly assigned to the following six groups(n = 16 for each group): sham + vehicle, sham + low-dose or high-dose RA(L) or(H), CUS + vehicle, CUS + RA(L) and CUS + RA(H). The rats in the CUS + vehicle group were subjected to CUS for 21 days and received saline as a vehicle for 14 days(1 ml/kg daily) from the last week of CUS. The CUS + RA(L) and CUS + RA(H) group also experienced CUS, but they received RA(5 and 10 mg/kg daily, respectively) for 14 days. The rats in sham, sham + RA(L) and sham + RA(H) groups were not experienced CUS, and they received saline or RA(5 and 10 mg/kg daily, respectively) for 14 days at the same time as the former three groups. After the treatment of RA or vehicle, some rats(n = 8 for each group) were subjected to open-field and forced swim tests, then sacrificed, and hippocampal ERK1/2, p ERK1/2 and BDNF, GDNF levels were determined by western blots and enzyme-linked immunosorbent assays(ELISAs). The remaining animals(n = 8 for each group) were subjected to the Morris water maze(MWM) test. RA(536954, Sigma-Aldrich, St. Louis, MO, USA) was dissolved in saline, and all treatments were administered intraperitoneally for all groups at the same times(from 9:00 to 10:00 A.M.).Experiment II: To assess the effects of RA and U0126 on ERK1/2 signaling in depressive-like behavior and hippocampal BDNF levels, rats were randomly assigned to 6 groups: sham + vehicle, sham + U0126, CUS + vehicle, CUS + U0126, CUS + RA(H) and CUS + RA(H) + U0126(n = 14 per group). The rats in CUS + vehicle, CUS + U0126, CUS + RA(H) and CUS + RA(H) + U0126 groups were experienced CUS, and received drug treatment for 14 days from the last week of CUS. The rats in sham+ vehicle, sham + U0126, and CUS + RA(H) groups were not experienced CUS, and they received drug treatment for 14 days at the same time as the former 4 groups. The animals in the sham + vehicle, CUS + vehicle, and CUS + RA(H) groups were intraperitoneally injected with 0.6 ml/kg vehicle(DMSO: saline = 1: 4) and then received saline or RA(10 mg/kg daily) 30 min later once a day for 14 days. The rats in the sham + U0126 and CUS + U0126 groups were intraperitoneally injected with 0.5 mg/kg U0126 and then received saline 30 min later once a day for 14 days. The rats in the CUS + RA(H) + U0126 group were intraperitoneally injected with 0.5 mg/kg U0126 once a day for 14 days 30 min prior to RA administration. The dose of U0126 in this experiment was chosen based on previous studies. The animals were subjected to behavioral tests 7 days later then sacrificed for subsequent western blot and ELISA analysis(as in Experiment I). U0126(#9903, Cell Signaling Technology, Danvers, MA USA), an ERK1/2 phosphorylation inhibitor, was dissolved in dimethyl sulfoxide(DMSO) and diluted with saline(DMSO: saline = 1: 4) before use.Experiment III: We designed an experiment to assess the effect of RA on BDNF release of hippocampal-derived astrocytes and the involvement of ERK1/2. Astrocytes were obtained by an astrocyte culture assay and exposed to different concentrations of RA(sham, 1, 5, 10, 20, and 40 μg/m L) or U0126(2 μM as a final concentration) + RA(sham, 1, 5, 10, 20, and 40 μg/m L) for 48 h before ERK1/2, p ERK1/2 and BDNF levels were assessed by Western blot and ELISA. In addition, the effect of different dose of RA on the proliferation of astrocytes for 48 h was also detected by using Brd U detection assay. RA was dissolved in Dulbecco’s modified Eagle’s medium(DMEM), and U0126 was dissolved in DMSO and diluted with DMEM(200 μM, 100-fold dilution as a final concentration) so that the final DMSO concentration was 2%. All experiments were carried out using the second passage of astrocytes.Experiment Ⅳ: We further designed this experiment to assess the effect of RA on GDNF release, cell viability and the proliferation of astrocytes. Astrocytes were obtained by an astrocyte culture assay and exposed to different concentrations of RA(sham, 1, 5, 10, 20, and 40 μg/m L) for 48 h or 72 h, and then the levels of GDNF were assessed. In addition, the effect of different dose of RA on the cell viability and proliferation of astrocytes for 72 h were also detected by using WST-1 or Brd U detection assay. RA was dissolved in Dulbecco’s modified Eagle’s medium(DMEM), and all experiments were carried out using the second passage of astrocytes. RESULTS:1. RA ameliorated depressive-like behavior and upregulated hippocampal BDNF, GDNF and p ERK1/2 levels in CUS rats.Two-way ANOVA revealed significant differences of stress factor(sham vs. CUS) in horizontal distance traveled in the OFT(F = 12.184, P = 0.001), the percent of floating time in FST(F= 9.987, P = 0.003), and the percentages of time spent in the target quadrant in the MWM(F = 8.336, P = 0.006). In addition, there were also significant differences of drug dose factor(vehicle, RA(H) and RA(L)) in horizontal distance traveled in the OFT(F = 3.287, P = 0.047), the percent of floating time in FST(F = 4.142, P = 0.023), and the percentages of time spent in the target quadrant in the MWM(F = 3.532, P = 0.038). There was a significant interaction between stress and drug dose main factors, for the percent of floating time in FST(F = 4.666, P = 0.015), but not for the distance traveled in the OFT(F = 1.562, P = 0.222) or the percentages of time spent in the target quadrant in the MWM(F = 2.949, P = 0.063). Post hoc comparisons further showed that there were significant differences between RA treatment(10 mg/kg daily) and vehicle treatment in all the three parameters(P < 0.05). However, there were no significant differences between the vehicle and RA(L) groups. There were significant differences of stress factor(sham vs. CUS) in hippocampal levels of p ERK1/2(F = 29.812, P = 0.000), BDNF(F = 12.322, P = 0.001) and GDNF(F = 26.1, P < 0.01).There were also significant differences of drug dose factor(vehicle, RA(H) and RA(L)) in hippocampal levels of p ERK1/2(F = 3.881, P = 0.031), BDNF(F = 3.580, P = 0.037) and GDNF(F = 2.881, P = 0.041). But there were no significant differences of stress factor(F = 0.098, P = 0.756) and drug dose factor(F = 0.118, P = 0.889) in hippocampal levels of ERK1/2. In addition, there was significant difference of interaction between stress factor and drug dose factor in the expression of p ERK1/2(F = 4.364, P = 0.021). Post hoc comparisons further showed that RA treatment(10 mg/kg daily) significantly increased these parameters compared to vehicle treatment groups(P < 0.05). However, there were no significant differences between the vehicle and RA(L) groups.2. U0126 abolished the antidepressive effect and inhibited RA-induced increases in p ERK1/2 and BDNF.Two-way ANOVA revealed significant differences of stress factor(sham vs. CUS) in horizontal distance traveled in the OFT(F = 27.088, P = 0.000), the percent of floating time in FST(F = 15.924, P = 0.000), and the percentages of time spent in the target quadrant in the MWM(F = 28.233, P = 0.000). In addition, there were also significant differences of treatment factor(vehicle, U0126, RA(H) and RA(H) + U0126) in horizontal distance traveled in the OFT(F = 3.247, P = 0.031), the percent of floating time in FST(F = 3.981, P = 0.014), and the percentages of time spent in the target quadrant in the MWM(F = 3.800, P = 0.017). There were no significant differences of interaction between stress factor and treatment factor in the percent of floating time in FST(F = 0.036, P = 0.850), the distance traveled in the OFT(F = 0.072, P = 0.789) and the percentages of time spent in the target quadrant in the MWM(F = 0.158, P = 0.693). Post hoc comparisons further showed that there were significant differences between RA(H) and RA(H) + U0126 treatment in all the three parameters(P < 0.05).There were significant differences of stress factor(sham vs. CUS) in hippocampal levels of p ERK1/2(F = 30.712, P < 0.001) and BDNF(F = 18.920, P = 0.001). There were also significant differences of treatment factor(vehicle, U0126, RA(H) and RA(H) + U0126) in hippocampal levels of p ERK1/2(F = 4.772, P = 0.008) and BDNF(F = 2.646, P = 0.041). But there were no significant differences of stress factor(F = 0.361, P = 0.554) and treatment factor(F = 0.138, P = 0.936) in the expression of ERK1/2. In addition, there were no significant difference of interaction between stress factor and treatment factor in the expression of p ERK1/2(F = 0.387, P = 0.539) and the levels of BDNF(F = 0.182, P = 0.672). Post hoc comparisons further showed that there were significant differences between RA(H) and RA(H) + U0126 treatment in all the these two parameters(P < 0.05).3. The expression of p ERK1/2 and the level of BDNF of astrocyte were promoted by RA but inhibited by U0126.By using astrocytes culture and identification assay, the most majority of re-plated cells were expressed GFAP(a marker for astrocyte), and there were no significant differences in cell viability(F5, 30 = 0.456, P = 0.806) and proliferation(F5, 30 = 0.766, P = 0.912) after treatment with different doses of RA administered for 48 h.Significant differences were observed in terms of p ERK1/2(F5, 24 = 3.147, P = 0.025) and BDNF(F5, 30 = 2.615, P = 0.045) levels between the different RA dosage groups, and there was no significant differences on the expression of ERK1/2 between different RA dosage groups(F5, 24 = 0.098, P = 0.992, data not show). Post hoc comparisons further showed that p ERK1/2 expression was significantly higher in the 20 μg/m L and 40 μg/m L groups compared to the sham group(P < 0.05). Moreover, the BDNF level in the 20 μg/m L group was also significantly higher than in the sham group(P < 0.05). Two-way ANOVA analyses revealed that exposure to U0126 significantly decreased both p ERK1/2(F = 0.591, P < 0.001) and BDNF(F = 22.277, P < 0.001), but there was no significant differences on the expression of ERK1/2(F = 0.245, P = 0.992). There were also no significant differences between the U0126 + sham and U0126 + RA(1, 5, 10, 20, and 40μg/m L) groups(P > 0.05).4. The cell viability, cell proliferation and the levels of GDNF of astrocyte were promoted by RA treatment for 72 h.Significant differences were observed in terms of the cell viability(F5, 30 = 0.351, P = 0.0306), cell proliferation(F5, 25 = 6.693, P < 0.01) and GDNF level(F5, 30 = 3.251, P = 0.016) between the different RA dosage groups. Post hoc comparisons further showed that the cell viability and GDNF level were significantly higher in the 20 μg/m L compared to the sham group(P < 0.05). Moreover, proliferation of astrocyte in the 20 μg/m L and 40 μg/m L group were also significantly higher than in the other group(P < 0.05). CONCLUTIONS:Though 4 parts of experiments, we found: ①Rats subjected to 3 weeks of CUS decreased the distance traveled in the OFT and increased immobility time during the FST compared to non-stressed rats. Additionally, chronic RA treatment(10 mg/kg) significantly suppressed these behavioral changes. We further found that in the MWM test, CUS-treated animals spent less time in the target quadrant without the platform compared to non-stressed rats, and chronic RA treatment(10 mg/kg) suppressed this behavioral change. Taken together, the behavioral study results indicate that although the low dose(5 mg/kg) of RA had no effects, the high dose(10 mg/kg) exerted an antidepressant-like action. ②CUS exposure decreased BDNF, GDNF and p ERK1/2 levels, and chronic RA treatment(10 mg/kg) attenuated these effects. Collectively, our findings indicate that a suitable dose of RA could ameliorate the depressive-like behaviours of CUS rats, and these effects were correlated with elevations in p ERK1/2 and BDNF. ③There were no improvements in depressive-like behaviors or p ERK1/2 and BDNF levels in the hippocampus of rats that received intraperitoneal injections of U0126 prior to RA administration. ④Astrocytic p ERK1/2 levels were elevated after RA treatment for 48 h at doses of 20 and 40 μg/m L, and BDNF was also increased at the dosage of 20 μg/m L. Moreover, we found that the effect was inhibited by U0126. These results indicate that RA could promote astrocytic BDNF expression by activating ERK signalling.⑤The astrocytic cell viability, proliferation and the released level of GDNF were also increased by RA treatment for 72 h.In conclusion, our present study indicates that RA exerts an antidepressant-like effect that may be related to an increase in astrocytic BDNF and GDNF, this effect maybe modulation by the ERK phosphorylation. Moreover, astrocyte may be one of the cell targets for RA.