3. Results
3.1.B AA attenuated naloxone-induced withdrawal signs in morphine physical dependence . Six groups of mice (n=10 per group) were subjected to bi-daily subcutaneous injections of normal saline (10 mL/kg), BAA (300 μg/kg) or morphine (escalating doses of 5, 10, 20, 40, 80 and 100 mg/kg) for 7 days. On the 7th day, mice received intraperitoneal injection of naloxone (5 mg/kg) 4 hours post the last injection of morphine (100 mg/kg) to induce physical dependence, and their withdrawal signs were observed immediately for 30 minutes. For the BAA inhibitory effects, the mice received a single injection of saline (10 mL/kg) or BAA (30, 100 or 300 μg/kg) 40 minutes prior to the intraperitoneal injection of naloxone. As shown in Fig. 1, intraperitoneal injection of naloxone did not induce any abnormal behaviors in bi-daily saline- or BAA-treated mice. In contrast, naloxone in bi-daily morphine injected mice induced profound withdrawal signs including shakes (Fig. 1A), jumps (Fig. 1B), genital licks (Fig. 1C), fecal excretion (Fig. 1D) and body weight loss (Fig. 1E) (P<0.05, by one-way ANOVA followed by the post-hoc Student-Newman-Keuls test). Moreover, pretreatment with subcutaneous injection of BAA (30, 100 and 300 μg/kg) dose-dependently attenuated naloxone-induced withdrawal signs in bi-daily morphine-treated mice, with maximal inhibition of around 70-100% in each different sign (P<0.05, by one-way ANOVA followed by the post-hoc Student-Newman-Keuls test). The dose-response analysis was performed in shakes with an ED50 of 74.4 μg/kg which was yielded after data transformation (Fig. 1A), and in the body weight loss with an ED50 of 105.8 μg/kg after data transformation (Fig. 1E).
3.2.BAA attenuated morphine-induced CPP acquisition . Four groups of mice (n=12 per group) were subjected to the preconditioning phase of three days and the place preference test on the 4th day followed by alternate daily subcutaneous injections of normal saline (10 mL/kg), BAA (300 μg/kg) or morphine (10 mg/kg) for 5 days. On the 10th day, mice received single subcutaneous injection of saline (10 mL/kg) or BAA (300 μg/kg) 50 minutes prior to the last injection and the place preference test was conducted immediately afterwards. As shown in Fig. 2, all four groups of mice did not show any obvious CPP acquisition in the preconditioning phase. Moreover, bi-daily subcutaneous injections of saline did not exhibit significant CPP acquisition in the post-conditioning phase. In contrast, bi-daily injections of morphine but not BAA showed remarkable CPP acquisition. However, pretreatment with subcutaneous injection of BAA completely attenuated morphine-induced CPP acquisition (P<0.05, by one-way ANOVA followed by the post-hoc Student-Newman-Keuls test).
3.3. BAA specifically stimulated microglial dynorphin A expression in NAc andhippocampus in morphine-multiply treated mice . Two groups of mice (n=10 per group) daily treated with morphine (10 mg/kg) for 5 days received subcutaneous injection of normal saline (10 mL/kg) or BAA (300 μg/kg). Mice were sacrificed 1 hour after subcutaneous injection and NAc and hippocampus were obtained for the prodynorphin mRNA detection using RT-PCR. As shown, treatment with BAA significantly increased prodynorphin gene expression by 1.9-fold in NAc (Fig. 3A) and 1.7-fold in hippocampus (Fig. 3B), respectively (P<0.05, by unpaired and two-tailed Student t-test). The stimulatory effects of BAA on dynorphin A protein expression were also measured in NAc and hippocampus in the same mice using the commercial fluorescent ELISA kit. As exhibited, subcutaneous BAA significantly increased dynorphin A expression in NAc (P<0.05, by unpaired and two-tailed Student t-test; Fig. 3C) and hippocampus (P=0.07, by unpaired and two-tailed Student t-test; Fig. 3D).
Dynorphin A is known to be localized in neurons, astrocytes and microglia in the central nervous system (Wahlert et al., 2013; Ayrout et al., 2019). To verify cell types that specifically upregulate dynorphin A expression in NAcSh and hippocampal CA1 following BAA treatment, dynorphin A was immunofluorescent labeled with the microglial cellular marker Iba-1, astrocytic cellular marker GFAP and neuronal cellular marker NeuN. Two groups of mice (n=6 per group) daily treated with morphine (10 mg/kg) for 5 days received subcutaneous injection of saline (10 mL/kg) or BAA (300 μg/kg). Mice were sacrificed 1 hour after the subcutaneous injection and NAcSh and hippocampal CA1 were obtained for fluorescent immunostaining. As shown, dynorphin A was colocalized with Iba-1, GFAP and NeuN in NAcSh of saline-treated mice (Fig. 4A-4F). Subcutaneous BAA specifically increased co-labeling of dynorphin A/Iba-1 (Fig. G, H) but not dynorphin A/GFAP (Fig. I, J) or dynorphin A/NeuN (Fig. K, L) at 10× and 30× magnification. In addition, the ImageJ software was used to quantify immunofluorescence intensity of dynorphin A with Iba-1, GFAP or NeuN at 10× magnification. Treatment with subcutaneous BAA significantly increased dynorphin A/Iba-1 by 2.4-fold (P<0.05, by unpaired and two-tailed Student t-test; Fig. 4M), but not dynorphin A/GFAP (Fig. 4N) or dynorphin A/NeuN (Fig. 4O). Furthermore, the same specific stimulatory effects of BAA on microglial dynorphin A expression were observed in hippocampal CA1 from the same mice as above (Fig. 5A-5L), with increased immunofluorescence intensity of dynorphin A/Iba-1 by 1.9-fold (P<0.05, by unpaired and two-tailed Student t-test; Fig. 5M), but not dynorphin A/GFAP (Fig. 5N) or dynorphin A/NeuN (Fig. 5O).
3.4. Brain microglial expression of dynorphin A mediated BBA-inhibited morphine dependence . To verify the causal relationship between the microglial expression of dynorphin A in brain and BAA-inhibited morphine physical dependence, the microglial activation inhibitor minocycline (Wu et al., 2002; Kobayashi et al., 2013), dynorphin A antiserum (Li et al., 2016) and κ-opioid receptor antagonist GNTI (Zhang et al., 2007; Liu et al., 2013) were intracerebroventricularly injected separately. Four groups of morphine physical dependence mice (n=10 per group) received the first intracerebroventricular injection followed by the second subcutaneous injection of 1) saline (6 μL) + saline (10 mL/kg), 2) minocycline (10 μg) + saline (10 mL/kg), 3) saline (6 μL) + BAA (300 μg/kg), and 4) minocycline (10 μg) + BAA (300 μg/kg). The second subcutaneous injection was 4 hours post the first intracerebroventricular injection. Withdrawal signs were precipitated by intraperitoneal injection of naloxone (5 mg/kg) 40 minutes after subcutaneous injection. As shown in Fig. 6A-6E, subcutaneous injection of BAA in morphine physical dependence mice significantly attenuated naloxone-induced withdrawal signs including shakes, jumps, genital licks, fecal excretions and body weight loss; whereas intracerebroventricular minocycline failed to influence morphine-induced physical dependence. However, pretreatment with intracerebroventricular minocycline nearly completely restored systemic BAA-suppressed withdrawal signs (P<0.05, by one-way ANOVA followed by the post-hoc Student-Newman-Keuls test).
Additional four groups of morphine physical dependence mice (n=10 per group) received the first intracerebroventricular injection followed by the second subcutaneous injection of 1) saline (6 μL) + saline (10 mL/kg), 2) dynorphin A antiserum (1:30 dilution, 6 μL) + saline (10 mL/kg), 3) saline (6 μL) + BAA (300 μg/kg) and 4) dynorphin A antiserum (1:30 dilution, 6 μL) + BAA (300 μg/kg). The second subcutaneous injection was 30 minutes post the first intracerebroventricular injection. Withdrawal signs were precipitated by intraperitoneal injection of naloxone (5 mg/kg,) 40 minutes after subcutaneous injection. Subcutaneous injection of BAA in morphine physical dependence mice attenuated naloxone-induced withdrawal signs. Intracerebroventricular injection of the dynorphin A antiserum did not significantly affect baseline morphine physical dependence, but reemerged naloxone-induced withdrawal syndrome from BAA inhibition (P<0.05, by one-way ANOVA followed by the post-hoc Student-Newman-Keuls test; Fig. 6F-6J).
Further four groups of morphine physical dependence mice (n=10 per group) received the same pattern as above except that intracerebroventricular injection of the dynorphin A antiserum was replaced with GNTI (5 μg). As displayed in Fig. 6K-6O, intracerebroventricular injection of GNTI predominantly restored BAA-suppressed morphine physical dependence (P<0.05, by one-way ANOVA followed by the post-hoc Student-Newman-Keuls test), although it did not significantly alter naloxone-induced withdrawal signs in baseline physical dependence.
3.5. Brain microglial expression of dynorphin A mediated BAA-inhibited CPP acquisition. Same as in the physical dependence model, minocycline, dynorphin A antiserum and GNTI were intracerebroventricularly injected separately into the morphine CPP acquisition mice in order to determine whether microglial expression of dynorphin A in brain contributed to BAA-inhibited morphine CPP acquisition. Four groups of morphine-induced CPP mice (n=10 per group) were first intracerebroventricularly injected followed 4 hours later by subcutaneous injected with 1) saline (6 μL) + saline (10 mL/kg), 2) minocycline (10 μg) + saline (10 mL/kg), 3) saline (6 μL) + BAA (300 μg/kg), and 4) minocycline (10 μg) + BAA (300 μg/kg). The place preference test was assessed 50 minutes subsequent to subcutaneous injection. As shown in Fig. 7A, subcutaneous injection of BAA but not intracerebroventricular minocycline completely attenuated morphine CPP acquisition. However, pretreatment with intracerebroventricular minocycline entirely restored BAA-suppressed CPP acquisition (P<0.05, by one-way ANOVA followed by the post-hoc Student-Newman-Keuls test).
In addition, four groups of morphine CPP mice (n=10 per group) were first intracerebroventricularly injected 0.5 hours later followed by subcutaneous injected with 1) saline (6 μL) + saline (10 mL/kg), 2) the dynorphin A (1:30 dilution, 10 μL) + saline (10 mL/kg), 3) saline (6 μL) + BAA (300 μg/kg), and 4) the dynorphin A (1:30 dilution, 10 μL) + BAA (300 μg/kg). The place preference test was assessed 50 minutes subsequent to subcutaneous injection. Subcutaneous injection of BAA but not intracerebroventricular the dynorphin A antiserum totally inhibited morphine CPP acquisition. However, pretreatment with intracerebroventricular injection of the dynorphin A antiserum completely attenuated BAA-suppressed morphine CPP acquisition (P<0.05, by one-way ANOVA followed by the post-hoc Student-Newman-Keuls test; Fig. 7B).
Last four groups of morphine CPP mice (n=10 per group) received the same regimen as above except that intracerebroventricular injection of the dynorphin A antiserum was replaced with GNTI (5 μg). As exhibited in Fig. 7C, intracerebroventricular injection of GNTI did not have a significant inhibitory effect on morphine CPP acquisition, but almost totally restored systemic BAA-suppressed morphine-induced CPP acquisition (P=0.09, by one-way ANOVA followed by the post-hoc Student-Newman-Keuls test).