Six-week-old male C57BL/6J mice were purchased form Slaccas Lab Animal Ltd, Shanghai, China, weighing about 22 g. β2-AR KO mice were the gift from Prof. Gang Pei (SIBS, Chinese Academy of Sciences) and backcrossed onto a C57BL/6J background. Experiments were carried out in wild type (WT), HET and KO littermates. All animals were housed with a reversed 12 h light/dark cycle and with access to food and water available ad libitum. Treatments were strictly in accordance with the National Institutes of Health Guide for the Care and Use of Laboratory Animals and were approved by Animal Care and Use Committee of Shanghai Medical College of Fudan University. The male mice with 8–10 weeks’ age were used for all behavioral tests. The behavioral tests were performed in the following orders: open field (OF), light/dark box and elevated plus maze tasks were carried out with the first cohort of mice (WT: n = 26–27, HET: n = 17, KO: n = 27). TST and forced swimming test (FST) were taken out with the second cohort (WT: n = 12, HET: n = 18, KO: n = 16). The third cohort (WT: n = 9, HET: n = 6, KO: n = 9) were submitted to locomotion tests induced by acute cocaine administration and the forth cohort (WT: n = 18, HET: n = 8, KO: n = 20) were submitted to cocaine conditioned place preference (CPP; Figure 1). β2-AR KO mice and subsequent offspring were genotyped using the following primer sets: 5′-CAC GAG ACT AGT GAG ACG TG-3′; 5′-ACC AAG AAT AAG GCC CGA GT-3′; 5′-CCG GGA ATA GAC AAA GAC CA-3′.

Cocaine hydrochloride (Qinghai Pharmaceutical Firm, China) was dissolved in 0.9% saline at 2 mg/ml for mouse CPP, 4 mg/ml for locomotor response test. Betaxolol (Tocris Bioscience, UK), a selective β1-AR antagonist, was dissolved in saline at 2 mg/ml and administered at a dose of 10 mg/kg (i. p.; Vranjkovic et al., 2012; Al-Hasani et al., 2013). Clenbuterol (Tocris Bioscience, UK), a selective β2-AR agonist, was dissolved in saline at 1 mg/ml and administering at a dose of 5 mg/kg (i. p.; Heal et al., 1991; Mellios et al., 2014). Control animals received an equivalent volume of saline.

Mice brains were removed on ice, hippocampus and medial prefrontal cortex (mPFC) were rapidly dissected and the tissues were prepared as following. Briefly, brain samples were homogenized in ice-cold 0.32 M sucrose, 5 mM HEPES (pH 7.4), 0.1 mM EDTA, and protease inhibitors mixture in a glass homogenizer. Homogenates were centrifuged (1000× g, 10 min, 4°C), and the supernatants were spun at 14,000× g for 30 min in a centrifuge at 4°C. Then the pellets were resuspended in 200 μL of 1× P buffer (5.4 mM KCl, 0.8 mM MgSO₄, 5.5 mM glucose, 50 mM HEPES, 130 mM choline chloride, 1 mM BSA, and 0.01% CHAPS). The protein concentration was determined by BCA assay (Pierce, 23235). Equal amounts of total protein (30 μg) isolated from cell membrane of the hippocampus and mPFC from WT and β2-AR KO mice were loaded on 10% sodium dodecyl sulfate polyacrylamide gels and then transferred to polyvinylidene difluoride membranes (Amersham Pharmacia Biotech, Piscataway, NJ, USA). Then the membranes were incubated in primary antibody for β1-AR (1:100, Santa cruz), β2-AR (1:100, Santa cruz) or β-tubulin (1:2500; Sigma, St Louis, MO, USA) at 4°C overnight. The membranes were then incubated with corresponding secondary antibody (1:50,000, Jackson Immuno Research) for 2 h after washing in TBST for three times. Protein bands were visualized using odyssey (LI-COR Biosciences). The immunoblots were analyzed with Image-Pro Plus to measure the optical density of the bands of β1-AR and β2-AR. Data of β-AR protein levels for WT and KO mice were expressed as percentage of the averaged values of β-tubulin.

Three days prior to locomotor activity assessments, mice were habituated to the testing room for 30 min each day. During the test, mice were released from the center of an OF test chamber (20 × 20 cm², ENV-515; Med Associates, Inc., St. Albans, VT, USA) and were allowed to explore the arena freely for 30 min. Total distance, distance and duration in the center area, entrance times to the center area were analyzed using OF Activity Software package v. 4.0 (Med Associates, Inc.). Testing was conducted with 25 Lux illuminance in the chamber. In the test for locomotor activity induced by acute cocaine administration, mice were placed in the OF chamber for 30 min followed by another 60 min after an injection of cocaine (20 mg/kg i.p.).

The light/dark box (46 × 27 × 30 cm) is composed of two compartments: the dark compartment (one third of the box) and the light compartment (two thirds of the box). Testing was conducted with 25 Lux illuminance in the light box (Birkett et al., 2011; Lehmann and Herkenham, 2011; Bluett et al., 2014; Shonesy et al., 2014). Mice were placed in the center of the light box and allowed full access to both compartments for 6 min. The tasks were taped by a digital video camera and the time mice spent in each compartment was analyzed with Clever System software (CleverSys).

Anxiety-related behavior was also measured by EPM task as previously described (Pravetoni and Wickman, 2008). Briefly, the maze was elevated 50 cm above the ground and consisted of two open arms, and two closed arms, as well as an exposed center panel. The mice were placed in the center of the maze, facing the closed arm. The sessions were taped by a digital video camera for 6 min and the time spent in each arm was analyzed with Clever System software (CleverSys). Time spent in the EPM center was not included. Testing was conducted under standard room lighting conditions.

The TST as one of the most widely used models for assessing the antidepressant-like activity in mice was used in our experiment. Mice were suspended 50 cm above the floor by the tip of the tail (<2 cm) with adhesive tape from a horizontal bar and were positioned such that the base of their tails was vertical to the bottom of the bar. The tasks were taped by a digital video camera for 5 min. Immobility was analyzed with the Tail Suspension Software (CleverSys). Mice were considered immobile only when they hung passively and completely motionless.

FST is another widely used model for assessing the antidepressant-like activity in mice. Mice were gently released into a transparent Plexiglas cylinder (diameter 10 cm, height 25 cm) filled with water (25 ± 0.5°C, 20 cm high) for a 5 min session (under <500 lux ambient light). Immobility (i.e., cessation of limb movement except minor involuntary movement of the hind limbs) was scored with the ForcedSwimScan software (CleverSys). Each mouse was judged to be immobile when it ceased struggling and remained floating motionless in the water, making only those movements necessary to keep its head above water as previous reported (Kaster et al., 2007).

A two-chamber, unbiased CPP paradigm was applied in accordance with the method previously employed (Liu et al., 2015). The CPP apparatus consisted of two compartments with distinct flooring and walls. Before each session mice were habituated to the experimental room for at least 30 min. On day 1, mice were allowed free access to both sides of apparatus for 15 min, and the duration in the two different chambers were recorded respectively. Mice with an initial preference (>65% of total time) for either chamber were excluded from the experiment. On day 2, mice were subjected to the conditioning phase, they were confined to one of the conditioning compartments for 30 min after injection of saline (4 ml/kg, AM) and confined to the other compartment after injection of cocaine (10 mg/kg, PM). On day 3 the training were conducted with switched order of the saline and cocaine conditioning. On day 4 the conditioning was performed as day 2. On day 5, mice were allowed free access to the entire apparatus, and the duration in each side was recorded. During the conditioning, clenbuterol or betaxolol were injected 30 min before cocaine conditioning. The sessions were taped by a digital video camera and the time spent in each chamber recorded by a trained observer blind to the genotype and treatment. The CPP preference was determined as score with time (s) spent in cocaine paired side minus saline side.

Experimental data were presented as the mean ± SEM and analyzed by Sigmaplot 12.5. The behavioral results from OF, L/D box, EPM, TST, FST and CPP were analyzed with one-way ANOVA or two-way repeated measures (RM) ANOVA followed by the Bonferroni’s post hoc test. The WB results were analyzed with two-tailed Student’s t-test.

In general, β2-AR KO mice showed no genotype-dependent differences in body weight, no obvious developmental abnormalities, or deficits in touch, vision and hearing compared with their WT littermates. To verify the deletion of β2-AR, the hippocampus and mPFC were dissected from β2-AR KO mice and WT mice for Western blot analysis (Figures 2A,B). The results showed deficient expression of β2-AR and normal expression levels of β1-AR in the hippocampus and mPFC of the β2-AR KO mice (Figure 2A, β2-AR: p < 0.001, β1-AR: p = 0.732; Figure 2B, β2-AR: p = 0.024, β1-AR: p = 0.614, t-test).

In the L/D box and EPM tasks, both β2-AR KO and HET mice spent significantly less amounts of time in the light compartment and the open arms than WT mice (Figure 2C, F_(2,67) = 10.638, p < 0.001, one-way ANOVA; Figure 2D, F_(2,66) = 7.322, p = 0.001, one-way ANOVA). The results from both tests consistently showed that the anxious state and innate fear were increased by deletion of one or two copies of Adrb2 and sufficient β2-AR expression might have the anxiolytic effects.

Next, the performance of β2-AR KO mice was assessed in TST and FST, two behavioral assays widely used to examine behavioral despair in mice (Porsolt et al., 1977; Steru et al., 1985). In TST, β2-AR KO mice displayed significantly less immobility during the entire 5 min than WT and HET mice (Figure 3A, F_(2,43) = 6.315, p = 0.004, one-way ANOVA), however, immobility time during each 5-min were indistinguishable of the test (Figure 3B, F_{genotype × time(2,43)} = 1.344, p = 0.255, two-way RM ANOVA). In FST, immobility duration in total or each 5 min was comparable in β2-AR KO, HET and WT mice (Figure 3C, F_(2,43) = 1.137, p = 0.33, one-way ANOVA; Figure 3D, F_{genotype × time(2,43)} = 1.356, p = 0.219, two-way RM ANOVA). These results suggest that behavioral despair is suppressed by β2-AR ablation in TST.

To assess the locomotor activity in β2-AR KO mice, the OF task was performed. The data showed that the total distances traveled by β2-AR KO mice, HET and WT littermates during both the entire test period and each 5-min were indistinguishable (Figure 4A, F_(2,68) = 1.506, p = 0.229, one-way ANOVA; Figure 4B, F_{genotype × time(2,68)} = 1.506, p = 0.229, RM two-way ANOVA). β2-AR KO mice did not show difference in time spent in the center of the chamber and distance traveled, or number of entries into the center of the OF chamber from HET and WT mice (Figures 4C–E, F_(2,68) = 2.857, p = 0.064; F_(2,68) = 0.680, p = 0.510; F_(2,68) = 0.863, p = 0.462; one-way ANOVA). These results indicate that β2-AR KO does not change locomotor activity.

Acute cocaine (20 mg/kg, i.p.) administration significantly increased total distance traveled in WT, HET and β2-AR KO mice, but β2-AR KO mice displayed significantly higher locomotor activity than WT and HET mice (Figure 4F, F_{genotype × treatment(2,22)} = 7.604, p = 0.003, two-way RM ANOVA) during a 60-min period following cocaine injection in the OF test. Two-way RM ANOVA revealed a significant genotype by time interaction and Bonferroni’s post hoc revealed that acute cocaine treatment induced an enhanced locomotor activity in β2-AR KO mice, the increase in locomotor activity primarily took place during the first 35 min of the test (Figure 4G, F_{genotype × time(2,22)} = 4.132, p < 0.001). The results indicate that ablation of β2-AR enhances response to acute cocaine administration.

In the task of cocaine-CPP, all groups showed preference for cocaine-paired side under the dose of 10 mg/kg cocaine (Figure 5A). Two-way RM ANOVA revealed a significant genotype by session interaction (Figure 5B, F_{genotype × session(2,43)} = 3.827, p = 0.03), and Bonferroni’s post hoc revealed a significant higher preference for the cocaine-paired side developed in β2-AR KO and HET mice compared with WT mice, indicating β2-AR KO increased rewarding properties of cocaine. To further verify the role of β2-AR in cocaine rewarding effect, clenbuterol (5 mg/kg), was administrated 30 min before each cocaine conditioning session (Figure 5C). CPP scores were recorded 24 h after the last conditioning session. Clenbuterol treatment significantly reduced the preference for cocaine-paired side (Figure 5D, F_{treatment × session(1,24)} = 7.439, p = 0.012, two-way RM ANOVA). With regard to the normal expression level of β1-AR in β2-AR KO mice, the role of β1-AR in cocaine-CPP was tested. Betaxolol (10 mg/kg) was given 30 min before cocaine conditioning daily (Figure 5E). We found that betaxolol inhibited the preference for cocaine-paired side in WT and β2-AR KO mice, however, antagonism of β1-AR in β2-AR KO mice only reduced the CPP score to the levels of cocaine-CPP developed in WT mice rather than a completed disruption, further suggesting that β2-AR deletion enhanced rewarding effect of cocaine (Figure 5F left, F_{treatment × session(1,40)} = 4.176, p = 0.048, two-way RM ANOVA; Figure 5F right, F_{treatment × session(1,29)} = 4.667, p = 0.039, two-way RM ANOVA). The above results indicate that β2-AR might suppress rewarding effects induced by cocaine, producing effects opposite to the function of β1-AR.