Exploratory and agile behaviors with central dopaminergic activities in open field tests in Formosan wood mice (Apodemus semotus)

There are 13 species of wood mice belonging to the genus Apodemus that are mainly and extensively distributed in the Palearctic zone, including Japan, Korea and Taiwan. Apodemus semotus are commonly called Formosan wood mice, or Taiwan field mice, and are endemic to the southern and central areas of Taiwan. Formosan wood mice are broadly distributed in areas with altitudes between 1400 and 3700 m and are the dominant mammal species in these areas (Yu, 1993, 1994). Most studies in Formosan wood mice have focused on their growth curve, reproduction and ecobiology (Lin et al., 1993; Lin and Shiraishi, 1992b), and recent studies have focused on their behavioral responses in the laboratory (Shieh et al., 2008; Shieh and Yang, 2018). Wild Formosan wood mice were successfully inbred for more than 10 generations in the Laboratory Animal Center of Tzu Chi University. These laboratory-inbred Formosan wood mice show enhanced exploratory behaviors that enable them to jump down from the platform of an elevated plus maze, which is up to 65 cm in height. These increased exploratory behaviors in laboratory-inbred Formosan wood mice are similar to the behaviors of mice living in the natural environment according to our subjective observations and other studies (Hauser et al., 2009; Kennedy and Elwood, 1988; Kuwahara et al., 2000; Lejeune et al., 2000; Perrigo et al., 1993; Rosalino et al., 2013; Takechi and Hayashi, 2012; Tosh et al., 2012). The differences in the behavioral responses of wood mice compared with those of common laboratory mice have been shown in previous studies (Hauser et al., 2009; Kennedy and Elwood, 1988; Kuwahara et al., 2000; Lejeune et al., 2000; Perrigo et al., 1993; Rosalino et al., 2013; Takechi and Hayashi, 2012; Tosh et al., 2012), and these observations and results provide further understanding of the behavioral responses and potential dominance of wood mice.

Elevated plus maze and open field tests are frequently used as behavioral tests to determine anxiety-like responses in mammals, including in mice (Kazlauckas et al., 2005; Podhorna and Brown, 2002; van Gaalen and Steckler, 2000). Behavioral responses in open field and elevated plus maze tests include animals' responses to fear or potential threat and the demonstration of a passive avoidance reaction. Our recent study showed that Formosan wood mice exhibited more exploratory and locomotor activities than common C57BL/6J laboratory mice in an elevated plus maze test (Shieh and Yang, 2018). Because Formosan wood mice were not afraid to jump down from the platform of the elevated plus maze, as mentioned above, the rate for completing the procedure without escape in the elevated plus maze test was approximately 65% (Shieh and Yang, 2018). Our previous study cannot rule out the possibility that the will to escape in Formosan wood mice disrupted the results. In addition to examining anxiety-like responses, the open field test has another advantage in examining activities such as indices of locomotor activity (Kazlauckas et al., 2005; Podhorna and Brown, 2002; van Gaalen and Steckler, 2000); thus, one focus of this study was to determine whether the behavioral responses in the open field test were similar to those in previous findings in the elevated plus maze.

The cause of higher levels of locomotor activity in these laboratory-inbred Formosan wood mice remains unclear. Long-term inbreeding in the laboratory might induce abnormalities, including behaviors. We studied whether the behavioral responses in the open field test are different between newly inbred mice and mice that had been inbred for more than 10 generations. The learning and memory profile of Formosan wood mice was also explored in this study through novel object recognition and inhibitory avoidance discrimination tests. Formosan wood mice might show increased responses in novel environments and adapt quickly, as a previous study found that wood mice adapted to urban conditions (Pieniążek et al., 2017). The third focus was to examine whether the novel environment was a factor that induced higher levels of exploration in Formosan wood mice.

Finally, previous studies have shown that the mesostriatal, mesolimbic and mesocortical dopaminergic (DAergic) systems play important roles in exploratory or novelty-seeking behavior (Bardo et al., 1996; Graf et al., 2013; Kalivas and Stewart, 1991). The major projections of the mesostriatal, mesolimbic and mesocortical DAergic systems are the striatum (ST), nucleus accumbens (NA) and medial prefrontal cortex (MPFC). Therefore, whether associations existed between central DAergic activities and behavioral responses in the open field test was examined, and whether these potential associations were influenced by an anxiolytic, diazepam, was studied.

Male 4-month-old Formosan wood mice (Apodemus semotus Thomas 1908) (n=90; weighing 30–35 g; the Laboratory Animal Center of Tzu Chi University, Hualien, Taiwan) and male 4-month-old C57BL/6J mice (n=78; weighing 30–35 g; the National Laboratory Animal Center, Taipei, Taiwan) were used. All experimental procedures were approved by the Institutional Animal Care and Use Committee at Tzu Chi University. The institutional guidelines were followed for the care and use of animals and experiments were conducted in accordance with the European Community Council Directive of 24 November 1986 (86/609/EEC). The care and use of all animals abided and followed the institutional guidelines with 3Rs principles.

Animals were housed in a light- and temperature-controlled room (lights on from 06:00 to 18:00 h; 22±1°C), and had access to rodent chow and tap water ad libitum. Every six male C57BL/6J or Formosan wood mice were accommodated in a regular cage (30 cm in width, 45 cm in length and 30 cm in height).

EthoVision video tracking software (3.0, Noldus Information Technology, Wageningen, The Netherlands) was used to automatically record the distance, time spent, latency and durations of the movements in the behavioral responses, as performed in previous studies (Shieh et al., 2008; Shieh and Yang, 2018; Võikar et al., 2005; Yang et al., 2005). A video camera was used and connected to the computer as the part of the video tracking system. The apparatuses for behavioral tests were cleaned with 70% ethanol to eliminate odors between any sessions in the behavioral tests.

The open field apparatus was a square Plexiglas box (50 cm in width, 50 cm in length and 50 cm in height). The light intensity at the center (20 cm in width and 20 cm in length) was 300 lx. Animals were placed in the center of the open field for 30 min to launch the test. The entry of an animal into the center arena was defined as the placement of four paws in the central zone of the open field apparatus. The EthoVision software recorded the distances and time spent in the center and peripheral arenas. Additionally, the number of fecal boli was manually recorded after the test by another experimenter. Three weeks after the open field test, the animals underwent the inhibitory avoidance discrimination test in the first experiment.

In the second experiment, the naïve animals without any behavioral test were used. Immediately after the open field test, the brains were obtained for dopamine turnover analysis, as described in the following section. In the third experiment, six male Formosan wood mice each from among mice newly caught and inbred for two generations and from among mice inbred for more than ten generations were subjected to the open field test, as described above. In the fourth experiment, male C57BL/6J and Formosan wood mice were repeatedly subjected to open field tests three times every 4 days. In the fifth experiment, male C57BL/6J and Formosan wood mice were intraperitoneally injected with diazepam hydrochloride at a low dose (0.05 mg ml⁻¹ kg⁻¹ body mass; Sigma-Aldrich, St Louis, MO, USA) and a high dose (0.5 mg ml⁻¹ kg⁻¹ body mass). Saline containing 2% Tween 80 (Sigma-Aldrich) was used as the vehicle to dilute diazepam hydrochloride. One hour after diazepam injection, mice were subjected to the open field test, as described above.

The behavioral test of the inhibitory avoidance discrimination task assesses learning and memory. The apparatus consists of a heated plate (30 cm in width and 30 cm in length; Diagnostic & Research Instruments Co., Taoyuan, Taiwan) with a wooden platform (3 cm in width, 3 cm in length and 3 cm in height) set in the center of the heated floor. The heated plate was set at 52°C for use as a noxious stimulus. For testing, each mouse was gently placed on the wooden platform. When the animal stepped down from the wooden platform onto the heated floor, the most common response was to jump back up to the wooden platform. One and 7 days after training, each animal was gently placed on the wooden platform again. The latency of step-down reactions was measured as the behavioral response of inhibitory avoidance discrimination. A cut-off time of 5 min was chosen. One week after the inhibitory avoidance discrimination test, the animals underwent the novel object recognition test.

The behavioral test of novel object recognition is a nonforced test of learning and memory. Mice were placed in a Plexiglas cage 30 cm in width, 45 cm in length and 25 cm in height. Additionally, two identical novel objects (approximately 5–8 cm high) were adhered to the middle of the cage. Mice were allowed to explore for 10 min, and the time spent exploring each object was recorded. Exploration was considered to be the orientation of mouse's head facing or sniffing the object within 2 cm. After 1 and 7 days, mice were put back into the same testing cage to test the approach of novel and familiar objects. Results are expressed as the percentage of novel object touching time [(time of novel object/time of novel and familiar objects)×100]. Additionally, the number of fecal boli was manually recorded by another experimenter.

Naïve animals without any behavioral test were used in this experiment. Immediately after the open field test, brains were obtained and frozen after euthanasia through CO₂ inhalation. Frozen brains without fixative procedures were sectioned (300 μm) with a freezing microtome (Leica SM2000R, Leica Microsystems GmbH, Nussloch, Germany) on the same day. Brain slices corresponded to the regions of MPFC, ST and NA from each mouse, according to the mouse brain atlas (Franklin and Paxinos, 1997), and were dissected as previously described (Shieh and Yang, 2018; Yang and Shieh, 2007; Yang et al., 2005). The individual brain regions were stored at −20°C until assayed.

Using high-performance liquid chromatography coupled with electrochemical detection (BAS LC480, with a phase II ODS column at 3.2×100 mm with 3-μm spheres, and an LC-4C EC detector; Bioanalytical Systems, West Lafayette, IN, USA), concentrations of DA and 3,4-dihydroxyphenylacetic acid (DOPAC), a major metabolite of DA, were measured as previously reported (Shieh and Yang, 2018; Yang and Shieh, 2007; Yang et al., 2005). The mobile phase was 0.15 mol l⁻¹ sodium phosphate buffer containing 0.65 mmol l⁻¹ sodium octanesulphonate, 0.5 mmol l⁻¹ EDTA, and 12% methanol and had a pH of 2.8. The DOPAC contents and the DOPAC/DA ratio, i.e. DA turnover, were assessed as DAergic neuronal activities. Tissue pellets were dissolved in 1 mol l⁻¹ NaOH and assayed for their protein content (Lowry et al., 1951).

Data are expressed as means±s.e.m. Data were tested using the Kolmogorov–Smirnov test for a normal distribution before other statistical tests. Differences in the open field test between C57BL/6J and Formosan wood mice were determined using a two-tailed unpaired Student's t-test. Differences in the inhibitory avoidance discrimination and novel object recognition tests, and open field tests in the third and fourth experiments between C57BL/6J and Formosan wood mice were examined by two-way ANOVA with repeated measures. Differences between two groups were analyzed by the post hoc Bonferroni test. For all statistical analyses, a P-value less than 0.05 was considered significant.

Exploratory behaviors were evaluated in the open field test. Compared with C57BL/6J mice (n=12), Formosan wood mice (n=12) spent more time in the center arena and entered the center arena more often (t₂₂=4.37, P<0.001 and t₂₂=2.41, P<0.05; Fig. 1A,B), but no differences in defecation were observed (t₂₂=0.78, P=0.42; Fig. 1C). For distance traveled in both the center (t₂₂=11.30, P<0.001; Fig. 1D) and peripheral arenas (t₂₂=6.20, P<0.001; Fig. 1E), Formosan wood mice moved markedly more than C57BL/6J mice. The total distance moved in the open field test was also higher in Formosan wood mice than in C57BL/6J mice (t₂₂=15.48, P<0.001; Fig. 1F).

In terms of learning and memory tests, this study used inhibitory avoidance discrimination (hot plate step-down inhibitory avoidance) and novel object recognition tests (Fig. 2). Formosan wood mice (n=12) showed a lower latency to step down than C57BL/6J mice (n=12) on the first and seventh days of the inhibitory avoidance discrimination test (two-way ANOVA; species F_1,44=85.25, P<0.001; time F_1,44=127.50, P<0.001; interaction F_1,44=80.71, P<0.001; Fig. 2A). However, the percentages of novel object touching time were not different between C57BL/6J mice (n=12) and Formosan wood mice (n=12) on either the first or the seventh day of the novel object recognition test (two-way ANOVA; species F_1,44=0.25, P=0.62; time F_1,44=200.30, P<0.001; interaction F_1,44=0.38, P=0.54; Fig. 2B). According to these findings, Formosan wood mice exhibited higher levels of locomotion and exploration in the open field test than C57BL/6J mice, but they showed similar learning and memory performance.

To examine whether central DAergic activities were changed after the open field test, we used naïve mice to measure behavior; mice were immediately killed to examine DAergic turnover, DOPAC and DA contents and DOPAC/DA ratios in the NA, ST and MPFC. DOPAC content was increased in the NA (t₂₂=6.66, P<0.001;), ST (t₂₂=4.57, P<0.001) and MPFC (t₂₂=8.91, P<0.001) of Formosan wood mice (n=12) compared with C57BL/6J mice (n=12) (Fig. 3A–C). DA contents in the NA (t₂₂=0.70, P=0.49), ST (t₂₂=0.85, P=0.41) and MPFC (t₂₂=1.37, P=0.18) (Fig. 3D–F) of Formosan wood mice and C57BL/6J mice were similar, so the trends in DOPAC/DA ratios and DOPAC content were also similar. Higher ratios of DOPAC/DA were found in the NA (t₂₂=4.91, P<0.001), ST (t₂₂=4.45, P<0.001) and MPFC (t₂₂=8.69, P<0.001) of Formosan wood mice compared with C57BL/6J mice (Fig. 3G–I). Therefore, central DAergic activities were increased after the open field test in Formosan wood mice compared with C57BL/6J mice.

To examine whether the differences in exploratory activity existed across different generations of Formosan wood mice, we used a new second-generation group of Formosan wood mice (n=6) and a group of Formosan wood mice that had been inbred for more than 10 generations (n=6) to measure behavior in the open field test (Fig. 4). No difference was found in the time spent and number of entries in the center arena (t₁₀=0.08, P=0.94 and t₁₀=0.19, P=0.85; Fig. 4A,B) or defecation (t₁₀=0.14, P=0.89; Fig. 4C). The distance traveled in both the center (t₁₀=0.48, P=0.64; Fig. 4D) and peripheral arenas (t₁₀=0.10, P=0.92; Fig. 4E), and total distance traveled (t₁₀=0.08, P=0.94; Fig. 4F) in the open field test was also similar. Thus, there was no difference in exploratory activity in the open field test between the new second-generation group and the group of Formosan wood mice inbred for more than 10 generations.

Next, we tried to examine whether the difference in exploratory activity (Fig. 5) and central DAergic turnover (Fig. 6) still existed after repeated exposure to the open field test. Because exploratory activity was basically higher in Formosan wood mice than in C57BL/6J mice (Fig. 1), we only focused on the effects of repeated exposure to the open field test in Formosan wood mice and C57BL/6J mice (Fig. 5). For time spent in the center arena, a significant decline after repeated exposure to the open field test (two-way ANOVA; species F_1,42=56.67, P<0.001; times F_2,42=3.00, P<0.05; interaction F_2,42=0.29, P=0.75) was found in C57BL/6J mice (n=12, 6, 6 in each group; F_2,21=19.75, P<0.001) but not in Formosan wood mice (n=12, 6, 6 in each group; F_2,21=0.49, P=0.62) (Fig. 5A). For number of entries into the center arena, a significant decline after repeated exposure to open field tests (two-way ANOVA; species F_1,42=33.66, P<0.001; times F_2,42=4.42, P<0.05; interaction F_2,42=0.98, P=0.38) was found in C57BL/6J mice (F_2,21=11.06, P<0.001) but not in Formosan wood mice (F_2,21=0.50, P=0.61) (Fig. 5B). No difference in defecation was found in either Formosan wood mice or C57BL/6J mice (two-way ANOVA; species F_1,42=0.12, P=0.73; times F_2,42=0.16, P=0.85; interaction F_2,66=0.18, P=0.84; Fig. 5C). For distance traveled in the center arena, a significant decline after repeated exposure to the open field test (two-way ANOVA; species F_1,42=299.6, P<0.001; times F_2,42=12.30, P<0.001; interaction F_2,42=0.18, P=0.83) was found in C57BL/6J mice (F_2,21=69.43, P<0.001) but not in Formosan wood mice (F_2,21=3.06, P=0.07) (Fig. 5D). For distance traveled in the peripheral arena, a significant decline after repeated exposure to the open field test (two-way ANOVA; species F_1,42=114.8, P<0.001; times F_2,42=12.56, P<0.001; interaction F_2,42=0.60, P=0.55) was found in C57BL/6J mice (F_2,21=21.95, P<0.001) but not in Formosan wood mice (F_2,21=2.98, P=0.07) (Fig. 5E). For the total distance traveled, a significant decline after repeated exposure to the open field test (two-way ANOVA; species F_1,42=65.03, P<0.001; times F_2,42=783.9, P<0.001; interaction F_2,42=2.05, P=0.14) was found in both C57BL/6J mice (F_2,21=55.73, P<0.001) and Formosan wood mice (F_2,21=24.60, P<0.001) (Fig. 5F).

DOPAC and DA contents were measured to examine the effects of repeated exposure in the open field test on central DA activities (Fig. 6). For the DOPAC content in NA, a significant reduction after repeated exposure to the open field test (two-way ANOVA; species F_1,42=161.6, P<0.001; times F_2,42=22.11, P<0.001; interaction F_2,42=4.58, P<0.05) was found in C57BL/6J mice (F_2,21=47.35, P<0.001) but not in Formosan wood mice (F_2,21=2.76, P=0.09) (Fig. 6A). For the DOPAC content in ST, a significant reduction after repeated exposure to the open field test (two-way ANOVA; species F_1,42=295.4, P<0.001; times F_2,42=65.22, P<0.001; interaction F_2,42=14.97, P<0.001) was found in both C57BL/6J mice (F_2,21=77.37, P<0.001) and Formosan wood mice (F_2,21=10.72, P<0.001) (Fig. 6B). For the DOPAC content in MPFC, a significant reduction after repeated exposure to the open field test (two-way ANOVA; species F_1,42=179.0, P<0.001; times F_2,42=41.73, P<0.001; interaction F_2,42=5.51, P<0.01) was found in both C57BL/6J mice (F_2,21=74.65, P<0.001) and Formosan wood mice (F_2,21=23.83, P<0.001) (Fig. 6C). After repeated exposure to the open field test, the DA contents in the NA (two-way ANOVA; species F_1,42=0.52, P=0.60; times F_2,42=3.96, P=0.053; interaction F_2,42=0.001, P=0.99; Fig. 6D), ST (two-way ANOVA; species F_1,42=0.61, P=0.55; times F_2,42=1.05, P=0.31; interaction F_2,42=0.0005, P=0.99; Fig. 6E) and MPFC (two-way ANOVA; species F_1,42=0.27, P=0.77; times F_2,42=0.14, P=0.71; interaction F_2,42=0.14, P=0.87; Fig. 6F) of Formosan wood mice and C57BL/6J mice were similar. For the DOPAC/DA ratios in NA, a significant reduction after repeated exposure to the open field test (two-way ANOVA; species F_1,42=16.0, P<0.001; times F_2,42=124.6, P<0.001; interaction F_2,66=1.01, P=0.37) was found in C57BL/6J mice (F_2,21=32.73, P<0.001) but not in Formosan wood mice (F_2,21=2.98, P=0.07) (Fig. 6G). For the DOPAC/DA ratios in ST, a significant reduction after repeated exposure to the open field test (two-way ANOVA; species F_1,42=40.41, P<0.001; times F_2,42=165.3, P<0.001; interaction F_2,42=5.28, P<0.01) was found in both C57BL/6J mice (F_2,21=52.95, P<0.001) and Formosan wood mice (F_2,21=8.28, P<0.01) (Fig. 6H). For the DOPAC/DA ratios in MPFC, a significant decrease after repeated exposure to the open field test (two-way ANOVA; species F_1,42=41.23, P<0.001; times F_2,42=161.4, P<0.001; interaction F_2,42=6.68, P<0.01) was found in both C57BL/6J mice (F_2,21=56.01, P<0.001) and Formosan wood mice (F_2,21=25.75, P<0.001) (Fig. 6I). Behavioral exploration and central DAergic activities decreased in C57BL/6J mice and Formosan wood mice after repeated exposure to the open field test, but these decreases were more significant in C57BL/6J mice than in Formosan wood mice.

Finally, to examine whether exploratory activity (Fig. 7) and central DAergic turnover (Fig. 8) were influenced by the anxiolytic treatment diazepam and whether differences existed between Formosan wood mice and C57BL/6J mice, we administered low (0.05 mg kg⁻¹; n=6) and high (0.5 mg kg⁻¹; n=6) doses of diazepam 1 h before animals were exposed to the open field test. For time spent in the center arena in the open field test, a significant change was found after diazepam treatments in Formosan wood mice but not in C57BL/6J mice (two-way ANOVA; species F_1,30=4.24, P<0.05; dose F_2,30=13.63, P<0.001; interaction F_2,30=9.83, P<0.001; Fig. 7A). Formosan wood mice spent significantly less time in the center arena after a high dose of diazepam than the vehicle group (F_2,15=8.88, P<0.01; Fig. 7A). Under the same treatment conditions, Formosan wood mice that were administered vehicle or a low dose of diazepam spent significantly more time in the center arena than comparable C57BL/6J mouse groups (t₁₀=4.97, P<0.001 and t₁₀=5.65, P<0.05), but no difference was found between Formosan wood mice and C57BL/6J mice after a high dose of diazepam (t₁₀=1.23, P>0.05) (Fig. 7A). For entries into the center arena in the open field test, a significant change was found after diazepam treatment in Formosan wood mice but not in C57BL/6J mice (two-way ANOVA; species F_1,30=1.17, P=0.29; dose F_2,30=3.47, P<0.05; interaction F_2,30=8.76, P<0.001; Fig. 7B). The number of entries into the center arena was significantly lower in Formosan wood mice after a high dose of diazepam than in the vehicle group (F_2,15=7.97, P<0.01; Fig. 7B). Between species, the number of entries into the center arena was significantly higher in Formosan wood mice with vehicle treatment than that in the equivalent group of C57BL/6J mice (t₁₀=3.23, P<0.01), but no difference was found between the low and the high dose of diazepam between Formosan wood mice and C57BL/6J mice (t₁₀=1.24 and t₁₀=2.25, P>0.05) (Fig. 7B). No difference in defecation was found between C57BL/6J mice and Formosan wood mice (two-way ANOVA; species F_1,30=0.41, P=0.67; dose F_1,30=0.03, P=0.86; interaction F_1,30=0.09, P=0.92; Fig. 7C).

For the distances traveled in the center arena, there was a significant change after diazepam treatment in C57BL/6J mice and Formosan wood mice (two-way ANOVA; species F_1,30=14.47, P<0.001; dose F_2,30=61.40, P<0.001; interaction F_2,30=30.79, P<0.001; Fig. 7D). The distances traveled in the center arena by C57BL/6J mice was significantly higher after a high dose of diazepam than after vehicle or a low dose of diazepam (F_2,15=6.59, P<0.01; Fig. 7D). The distances traveled in the center arena was significantly lower in Formosan wood mice after a high dose of diazepam than after vehicle or a low dose of diazepam, and this distance after a low dose of diazepam was also lower than that after vehicle (F_2,15=25.94, P<0.001; Fig. 7D). When comparing C57BL/6J mice and Formosan wood mice under the same treated conditions, the distances traveled in the center arena was significantly higher in Formosan wood mice with vehicle or a low dose of diazepam than in C57BL/6J mice (t₁₀=9.78, P<0.001 and t₁₀=5.59, P<0.001), but no difference was found between C57BL/6J mice and Formosan wood mice after the high dose of diazepam (t₁₀=1.47, P>0.05) (Fig. 7D). For the distance traveled in the peripheral arena, a significant change was found after diazepam treatment in Formosan wood mice but not in C57BL/6J mice (two-way ANOVA; species F_1,30=5.26, P<0.05; dose F_2,30=33.81, P<0.001; interaction F_2,30=9.51, P<0.001; Fig. 7E). The distance traveled in the peripheral arena was significantly lower in Formosan wood mice after a high dose of diazepam than that in the vehicle control group (F_2,15=10.40, P<0.01; Fig. 7E). The distance traveled in the peripheral arena in Formosan wood mice after vehicle treatment and after a low dose of diazepam was significantly higher than that in C57BL/6J mice (t₁₀=6.52, P<0.001 and t₁₀=3.20, P<0.01), but no difference was found between C57BL/6J mice and Formosan wood mice for the high dose of diazepam (t₁₀=0.35, P>0.05) (Fig. 7E). For the total distance traveled in the open field test, a significant change was found in Formosan wood mice after diazepam treatment but not in C57BL/6J mice (Fig. 7F; two-way ANOVA; species F_1,30=11.69, P<0.001; dose F_2,30=59.69, P<0.001; interaction F_2,30=19.52, P<0.001). The lowest total distance traveled in the open field test was found in Formosan wood mice treated with a high dose of diazepam compared with that of mice receiving vehicle or a low dose of diazepam, and that for mice treated with a low dose of diazepam was significantly lower than that for mice receiving vehicle (F_2,15=27.25, P<0.001; Fig. 7F). Comparing C57BL/6J mice and Formosan wood mice, the total distance traveled by Formosan wood mice that received vehicle treatment or the low dose of diazepam was significantly higher than that traveled by C57BL/6J mice (t₁₀=8.68, P<0.001 and t₁₀=4.83, P<0.00;1 Fig. 7F), but no difference was found between C57BL/6J mice and Formosan wood mice receiving the high dose of diazepam (t₁₀=0.13, P>0.05; Fig. 7D). Exploration by Formosan wood mice in the open field test was significantly lower after diazepam treatment, but C57BL/6J mice showed fewer effects.

To determine whether central DAergic activities were also influenced by diazepam treatments, we examined DOPAC and DA contents (Fig. 8A–F) and DOPAC/DA ratios (Fig. 8G–I) in the NA, ST and MPFC. After diazepam treatments, the DOPAC contents in the NA (two-way ANOVA; species F_1,30=7.09, P<0.01; dose F_2,30=9.33, P<0.01; interaction F_2,30=3.58, P<0.05; Fig. 8A), ST (two-way ANOVA; species F_1,30=5.54, P<0.01; dose F_2,30=8.04, P<0.01; interaction F_2,30=5.58, P<0.01; Fig. 8B) and MPFC (two-way ANOVA; species F_1,30=41.82, P<0.001; dose F_2,30=185.1, P<0.001; interaction F_2,30=41.81, P<0.001; Fig. 8C) changed in Formosan wood mice but not in C57BL/6J mice. In Formosan wood mice treated with a high dose of diazepam, the DOPAC contents in the NA (F_2,15=12.82, P<0.001), ST (F_2,15=10.95, P<0.01) and MPFC (F_2,15=56.85, P<0.001) (Fig. 8A–C) were significantly lower than in those treated with vehicle or a low dose of diazepam in the same brain regions. Comparing C57BL/6J mice and Formosan wood mice, the DOPAC content in the NA was significantly higher in Formosan wood mice after vehicle treatment or the low dose of diazepam in similar groups of C57BL/6J mice (t₁₀=2.97, P<0.05 and t₁₀=2.75, P<0.05), and similar findings were also shown in the ST (t₁₀=2.55, P<0.05 and t₁₀=3.41, P<0.01) and MPFC (t₁₀=11.66, P<0.001 and t₁₀=11.51, P<0.001) (Fig. 8A–C). There was no difference between Formosan wood mice and C57BL/6J mice in DOPAC contents in NA (t₁₀=0.42, P>0.05), ST (t₁₀=1.05, P>0.05) or MPFC (t₁₀=0.39, P>0.05) after a high dose of diazepam (Fig. 8A–C). After diazepam treatment, the DA contents in the NA (two-way ANOVA; species F_1,30=3.30, P=0.051; dose F_2,30=0.008, P=0.93; interaction F_2,30=0.02, P=0.97; Fig. 8D), ST (two-way ANOVA; species F_1,30=2.59, P=0.09; dose F_2,30=1.17, P=0.29; interaction F_2,30=0.0002, P=0.99; Fig. 8E) and MPFC (two-way ANOVA; species F_1,30=0.33, P=0.72; dose F_2,30=3.02, P=0.09; interaction F_2,30=0.11, P=0.89; Fig. 8F) of Formosan wood mice and C57BL/6J mice were not significantly changed, so the trends in DOPAC/DA ratios (Fig. 8G–I) and DOPAC contents (Fig. 8A–C) were also similar.

After diazepam treatments, the DOPAC/DA ratios in the NA (two-way ANOVA; species F_1,30=6.61, P<0.01; dose F_2,30=30.15, P<0.001; interaction F_2,30=3.14, P=0.06; Fig. 8G), ST (two-way ANOVA; species F_1,30=5.27, P<0.05; dose F_2,30=45.57, P<0.001; interaction F_2,30=3.51, P<0.05; Fig. 8H) and MPFC (two-way ANOVA; species F_1,30=23.86, P<0.001; dose F_2,30=197.0, P<0.001; interaction F_2,30=28.02, P<0.001; Fig. 8I) were changed in Formosan wood mice but not in C57BL/6J mice. In Formosan wood mice treated with a high dose of diazepam, the DOPAC/DA ratios in the NA (F_2,15=6.41, P<0.01; Fig. 8G) and MPFC (F_2,15=29.07, P<0.001; Fig. 8I) were significantly lower than those in the same brain regions of mice treated with vehicle or a low dose of diazepam. The DOPAC/DA ratio in the ST of Formosan wood mice treated with a high dose of diazepam was significantly lower than that of mice treated with vehicle (F_2,15=4.94, P<0.05; Fig. 8H). Comparing C57BL/6J mice and Formosan wood mice, the DOPAC/DA ratio in NA was significantly higher in Formosan wood mice treated with vehicle or the low dose of diazepam than in C57BL/6J mice (t₁₀=4.31, P<0.001 and t₁₀=4.07, P<0.001), and similar findings were also shown in the ST (t₁₀=4.68, P<0.001 and t₁₀=5.25, P<0.001) and MPFC (t₁₀=10.83, P<0.001 and t₁₀=11.48, P<0.001) (Fig. 8G–I). There was no difference between Formosan wood mice and C57BL/6J mice in the DOPAC/DA ratios in NA (t₁₀=1.13, P>0.05), ST (t₁₀=1.76, P>0.05) or MPFC (t₁₀=2.00, P>0.05) after a high dose of diazepam (Fig. 8G–I). Central DAergic activities decreased in Formosan wood mice after a high dose of diazepam, but those in C57BL/6J mice were not decreased.

Wood mice are common rodents and are not an endangered or protected species worldwide. Most studies related to wood mice focus on ecology, parasitology or reproduction. Recently, more studies on wood mice have been related to environmental pollution (Okano et al., 2016) and stress or behavioral responses in the wild (Malkemper et al., 2015; Monarca et al., 2015; Navarro-Castilla and Barja, 2019; Wan-Long and Zheng-Kun, 2016). Some studies have focused on comparisons of the physiological and behavioral responses between wood mice and common laboratory mice (Lejeune et al., 2000; Shieh et al., 2008; Shieh and Yang, 2018; Tosh et al., 2012).

The Formosan wood mouse (A. semotus) is an endemic Taiwanese rodent, and most reports related to Formosan wood mice are field studies (Huang et al., 1997; Lee et al., 2001; Lin et al., 1993; Lin and Shiraishi, 1992a,b). Our recent study showed that Formosan wood mice had higher levels of locomotor activity and exploratory behavioral responses than C57BL/6J mice in a laboratory environment (Shieh and Yang, 2018). Previous studies have also reported that Formosan wood mice show high levels of exploration in the field (Huang et al., 1997; Lee et al., 2001; Lin et al., 1993). Because our previous findings used the elevated plus maze to examine behavioral responses, jumping-down behavior from open arms during the test was commonly found. Although we excluded the data from mice that showed jumping-down behavior from open arms in the previous study, whether the will of Formosan wood mice to escape could disrupt the findings was unclear. This issue could be resolved using the open field test, which prevents disruptions owing to escapes, and indeed, the open field test provides strong supporting evidence of exploratory behaviors in Formosan wood mice. The present study concentrated on the behavioral responses of the Formosan wood mouse and common laboratory C57BL/6J mouse in open field test without escape events in a laboratory environment.

Open field, light–dark exploration and elevated plus maze tests are commonly used to measure aspects of anxiety-like behavioral responses (Kulesskaya and Voikar, 2014; Lad et al., 2010; Mandillo et al., 2008; Nestler and Hyman, 2010). There are still some differences between these tests in terms of the levels of combined factors of agoraphobia, neophobia, exploratory drive and locomotor activity. Several previous studies have shown that C57BL/6J mice are more active than other mouse strains, such as 129S mice, in the open field test (Crabbe et al., 1999; Mandillo et al., 2008; Võikar et al., 2001). Therefore, in this study, the open field test was used to examine anxiety-like behavioral responses and locomotor activities in Formosan wood mice and C57BL/6J mice.

Formosan wood mice spent more time in the center arena and entered the center arena more often than C57BL/6J mice in the open field test (Fig. 1). These findings suggest that Formosan wood mice have fewer anxiety-like responses in situations that could induce agoraphobia and neophobia. Additionally, Formosan wood mice traveled longer distances in the center and peripheral arenas and traveled longer total distances in the open field test than C57BL/6J mice. These findings suggest that Formosan wood mice show higher levels of exploratory responses and locomotor activities without escape disruption. Our recent study showed similar findings in the elevated plus maze test in which Formosan wood mice displayed an increased distance traveled in both open and closed arms (Shieh and Yang, 2018). Based on the findings in this study and those from the elevated plus maze in our previous study (Shieh and Yang, 2018), we suggest that Formosan wood mice show relatively higher levels of locomotor activity, higher levels of exploration and fewer anxiety-like behaviors than C57BL/6J mice.

Furthermore, we used inhibitory avoidance discrimination and novel object recognition tests to examine learning and memory performance in Formosan wood mice. One and 7 days after nociceptive training in the inhibitory avoidance discrimination test, C57BL/6J mice showed a longer latency to step down to the hot plate. This result meant that C57BL/6J mice learned inhibitory avoidance. In the same tests, Formosan wood mice showed a shorter latency to step down to the hot plate. Common explanations for these findings among psychologists are that Formosan wood mice either show a deficiency in learning and memory performance or are less sensitive to nociceptive stimuli. Formosan wood mice showed similar responses in the tail-flick withdrawal test as C57BL/6J mice (Fig. S1), so we excluded the latter possibility. Owing to the higher levels of exploratory behaviors in the open field and elevated plus maze tests, we decided to use the novel object recognition test as another test to examine learning and memory performance. C57BL/6J mice and Formosan wood mice showed similar responses in this test. Two different tests of learning and memory showed different findings in a previous study (Kohara et al., 2014), and a possible explanation for the different results in the present study is the disruption by exploratory/willful behaviors and/or the will to escape. Formosan wood mice exhibited higher levels of locomotor activity, exploration and even the will to escape in the tests shown in this study and in previous studies (Shieh et al., 2008; Shieh and Yang, 2018). Therefore, Formosan wood mice might prefer to escape in the inhibitory avoidance discrimination test, as in the elevated plus maze test, rather than avoid the nociceptive stimulus. This result implies that the will to escape or willful behavioral responses could disrupt the interpretations of these findings, especially in animals with highly agile or emotional reactions. In the present study, we also confirmed that exposure to novel environments enhances the animal's level of exploration, but these increases in exploration do not promote or enhance learning abilities (Light et al., 2008).

Some animals with genetic mutations, such as neuropeptide Y4 receptor knockout mice, show higher levels of locomotion and less anxiety in the open field test (Tasan et al., 2009). To verify whether long-term inbreeding in the laboratory caused mutation and influenced locomotor activity and behavioral responses in the open field test, we used newly caught Formosan wood mice to inbreed a new second generation and compared them with Formosan wood mice that had been inbred for more than 10 generations (Fig. 5). Although there were only six male Formosan wood mice in the second generation, the behavioral responses in open field test, including the time spent, number of entries, travel distance in the center arena, and total distance traveled in the open field test, were similar between the second generation and those inbred for more than 10 generations. These findings suggest that higher levels of exploration in Formosan wood mice are not related to the long-term inbreeding conditions in the laboratory. A previous study also found that wood mice had higher levels of locomotion and exhibited more jumping behaviors even after four generations of inbreeding in the laboratory (Hendrie et al., 2001), and our present data support this result.

Novelty seeking or higher levels of exploration and locomotor activities are often related to the degree to which an animal probes a novel stimulus (such as an open field) (Piazza et al., 1990; Stead et al., 2006). Usually, animals repeatedly encounter the same conditions or environments without nociception, and their behavioral responses will be attenuated, or they show adaptive responses. A previous study found that C57BL/6 mice decreased their total distance traveled in the open field test after repeated exposure to the test (Kasahara et al., 2007). In this study, we found similar responses in C57BL/6J mice (Fig. 6). In contrast, Formosan wood mice still showed higher levels of exploration and locomotor activity after repeated exposure to open field test. Possible explanations for this result are either learning and memory deficiency or high alertness. Learning and memory deficiency make animals encounter the same conditions as if they were a bland new situation; however, Formosan wood mice did show good learning and memory performance in the novel object recognition test (Fig. 2). Thus, alertness was the better explanation, and we next used an anxiolytic treatment, diazepam, to test this idea.

Different doses of diazepam in mice have been shown to induce different levels of sedation or locomotion and exploration impairment (Belzung et al., 2001; Bradford et al., 2010; Irifune et al., 1998; Lee et al., 1987; Söderpalm et al., 1991; Young and Johnson, 1991). For instance, 0.1–2 mg kg⁻¹ diazepam increased locomotion or exploration (Belzung et al., 2001; Söderpalm et al., 1991; Young and Johnson, 1991) and 1–5 mg kg⁻¹ diazepam decreased locomotion or exploration (Barraco et al., 1984; Söderpalm et al., 1991; Young and Johnson, 1991), but no effect on locomotion was reported for 2 mg kg⁻¹ diazepam (Lee et al., 1987; Young and Johnson, 1991). Moreover, 3 mg kg⁻¹ diazepam attenuated the locomotor hyperactivity induced by MK-801 (Young and Johnson, 1991) or ketamine (Irifune et al., 1998), as shown previously. Therefore, 0.05 and 0.5 mg kg⁻¹ doses of diazepam in this study were used to eliminate the possibility of locomotion impairment or sedation induced by diazepam. Similar to the data in naïve animals (Fig. 1), Formosan wood mice receiving vehicle showed more behavioral responses in the open field test than C57BL/6J mice receiving vehicle (Fig. 7). The 0.05 and 0.5 mg kg⁻¹ diazepam treatments reduced the behavioral responses of Formosan wood mice in the open field test, including the time spent, number of entries and travel distance in the center arena, as well as the total distance traveled. These findings suggest that Formosan wood mice were alert, showing higher levels of exploration, and that anxiolytic treatments decreased alertness and exploration. The 0.5 mg kg⁻¹ dose of diazepam increased the number of entries and distance traveled in the center arena in C57BL/6J mice, but the 0.05 mg kg⁻¹ dose did not. This result was similar to previous findings showing that 0.25–1 mg kg⁻¹ diazepam had anxiolytic effects and increased locomotion or exploration (Belzung et al., 2001; Söderpalm et al., 1991). Interestingly, in the 0.5 mg kg⁻¹ diazepam treatment, the behavioral responses of Formosan wood mice were the same as those of C57BL/6J mice that received 0.5 mg kg⁻¹ diazepam. This result implies that alertness in different species had different effects on exploration and locomotion. Alertness might restrict exploration in C57BL/6J mice but enhance exploratory activity in Formosan wood mice. Additionally, locomotion in corticotrophin releasing factor-overexpressing transgenic mice in the open field test did not show adaptive responses after repeated exposure to the open field test (Kasahara et al., 2007), and these mice had higher basal corticosterone levels (Groenink et al., 2002). Male Formosan wood mice showed similar responses to those in a previous study (Kasahara et al., 2007), and our previous study also found that male Formosan wood mice had higher basal corticosterone levels than C57BL/6J mice (Shieh and Yang, 2018). We could not conclude that corticotrophin releasing factor levels were elevated in Formosan wood mice because the two strains have different genetic backgrounds and hormone levels. Whether the hypothalamic–pituitary–adrenal axis is involved in these findings remains unclear, and further study might be needed. Whether Formosan wood mice were more sensitive to diazepam treatment than C57BL/6J mice was also unclear. The pharmacological mechanism was not the main purpose of this study, and further study is needed to address this issue.

Central DAergic activities in the NA, ST and MPFC have been shown to play an important role in novelty-seeking and exploratory behaviors (Bardo et al., 1996; Graf et al., 2013; Kalivas and Stewart, 1991). After the open field test, DAergic turnover in the NA, ST and MPFC was higher in Formosan wood mice than in C57BL/6J mice (Fig. 3). DAergic turnover, measured by DOPAC contents or DOPAC/DA ratios, in the NA and ST of C57BL/6J mice showed a one-third decrease after repeated exposure to the open field test, and these measures in Formosan wood mice showed a one-tenth decrease (Fig. 6). Interestingly, a one-third decrease in MPFC DAergic turnover was found in both C57BL/6J mice and Formosan wood mice. Two doses of diazepam did not influence central DAergic turnover in C57BL/6J mice, but the higher dose of diazepam decreased central DAergic turnover in Formosan wood mice to the same level as that in C57BL/6J mice (Fig. 8). In this condition, with a higher diazepam dose, the locomotor activities and behavioral responses of Formosan wood mice were also similar to those of C57BL/6J mice. This finding suggests that anxiolytic treatment reduces alertness in Formosan wood mice and attenuates higher central DAergic activities. A previous study showed that ketamine (30 mg kg⁻¹) increased locomotor activity, and diazepam (3 mg kg⁻¹) inhibited this hyperlocomotion induced by ketamine (Irifune et al., 1998). Ketamine also increased DA turnover, the ratio of homovanillic acid (HVA, one of the DA metabolites) to DA, in the NA and ST, and 3 mg kg⁻¹ diazepam reversed these increases in the NA and ST (Irifune et al., 1998). Diazepam alone, at doses of 1 and 3 mg kg⁻¹, decreased the HVA/DA ratios in the ST, and at a dose of 3 mg kg⁻¹ decreased this ratio in mg kg⁻¹NA (Irifune et al., 1998). There were some differences between the previous study and the present study. The mouse strains (ddY versus C57BL/6J mice), indicators of DA turnover (HVA/DA versus DOPAC/DA ratios) and doses of diazepam (1, 3 and 10 mg kg⁻¹ versus 0.05 and 0.5 mg kg⁻¹) were different, but evidence for the role of DAergic activities in the regulation of reward-directed, locomotor, social and even aggressive behaviors has accumulated over the years (Baskerville and Douglas, 2010; Bergamini et al., 2018; Sershen et al., 1987). From findings of neurobiological and behavioral aberrations, the NA and MPFC act as a hub, and their DA signals transmit social motivational, reward and stress adaptations (Aragona et al., 2007; Bergamini et al., 2018). The DAergic activity in the ST plays an important role in the regulation of locomotor activity (Sershen et al., 1987). Furthermore, the locomotor activity and exploration behaviors in the open field test associated well with DAergic activity in the MPFC and ST, and the activity in the NA in the present study was confirmed by previous studies (Aragona et al., 2007; Bambico et al., 2015; Baskerville and Douglas, 2010; Bergamini et al., 2018; Cabib et al., 1988; Egashira et al., 2005; Sershen et al., 1987).

In summary, male Formosan wood mice revealed higher levels of exploratory activity and fewer anxiety-like behaviors than C57BL/6J mice in a laboratory environment, and these behavioral responses might be associated with higher levels of central DAergic activities. These exploratory and agile behaviors in Formosan wood mice were not caused by the long-term inbreeding conditions in the laboratory, but the mechanism was unclear. This study provides comparative findings, as two phylogenetically related species showed differences in behavioral responses.

We appreciate the technical assistance received from the Laboratory Animal Center of Tzu Chi University.