Forty-two outbred male 10-weeks old Mongolian gerbils (50–70 g; Meriones unguiculatus), obtained from the National Institute of Health at Bogotá-Colombia, were used in this study. Animals were housed in groups of 2–3 in polycarbonate cages (42 × 20 × 20 cm) which contained dust free wood shaving bedding, and were kept in an animal room under a 12 h light/dark cycle (lights on at 08:00 h) with water and standard rodent pellets available ad libitum. The room temperature was maintained at 23°C with 55% relative humidity. Animal handling was limited to home cage-cleaning time and the animals were not habituated to a novel environment before the first test session. All experimental procedures were performed in accordance with the United States National Institute of Health Guide for the Care and Use of Laboratory Animals and were approved by our Institutional Animal Care and Use Committee (CICUAL-KL/COM43-2016).

An EPM similar to that previously described elsewhere (Starkey and Bridges, 2010; Rico et al., 2016) was used in this study. The apparatus consisted of two open arms (40 cm × 8 cm) at right angles with two closed arms of the same size, with 30-cm high black Plexiglas walls. Arms emerged from a central platform (8 cm × 8 cm). The floor of the plus-maze was made of black smooth Plexiglas and the entire apparatus was elevated 50-cm from the ground. A raised Plexiglas edge (0.5-cm) surrounded the open arms to prevent gerbils from falling. The level of illumination of the test room was adjusted to 30 lux measured at the central area of the plus-maze.

The benzodiazepine DZP (0.5 mg/kg; Roche, São Paulo) was dissolved in physiological saline (NaCl 0.9%; Vehicle). Thirty minutes before placing the gerbil in the EPM, DZP or Vehicle was administrated intraperitoneally (i.p.) in a volume of 1 ml/kg. Drug doses and administration protocol were based on those used in similar studies using gerbils (Varty et al., 2002a; Bradley et al., 2007a,b, 2011).

In rats, the aversion to the open arms seems to be influenced by procedural factors, such as the time of the day at which testing occurs (Griebel et al., 1993). In gerbils, testing in the EPM has been performed between 8:00 and 16:00 h (Bridges and Starkey, 2004; Starkey and Bridges, 2010). Accordingly, our experimental sessions were carried out during the light phase (08:00–13:00 h). In order to characterize the spatiotemporal behavioral pattern during the session and to obtain behavioral dimensions that emerge from factor analysis, the sample of control group was expanded. Thirty minutes prior to testing, each animal was removed from the home cage, weighed and injected intraperitoneally with DZP (0.5 mg/kg; n = 9) or with Vehicle (n = 33). Then, gerbils were placed in the central area of the EPM facing one of the open arms, and could freely explore the instrument for 20-min. At the end of each session, the maze was cleaned with a 10% ethanol solution and dried with a cloth. All behavioral tests were recorded with a video camera placed above the EPM and connected to a digital video recorder in an adjacent room. A trained observer, blind for treatment, analyzed the videos (intra-observer agreement >90%). Behavioral measures were scored using the ethological free software X-PloRat (Tejada et al., 2017). The frequency of entries and the time spent into the open and the closed arms, the number of crossed squares into each arm and the frequency of stretching attend posture (SAP; a movement where gerbil leans forward with a flattened back followed by retraction to original position), were analyzed. An entry into an arm or into a square (8 cm × 8 cm) within an arm was scored after all four paws of the gerbil entered it. When an animal jump from the EPM, it was replaced on the central area as soon as possible facing one of the open arms, and additional time proportional to the time the subject was off the maze was added to the session. In these cases, video scoring was paused until the animal was repositioned in the EPM.

Gerbils occasionally exhibit seizures when exposed to the EPM, often followed by a period of immobility (Bridges and Starkey, 2004; Rico et al., 2016). Six animals from the control group that displayed seizures were excluded from the analyses.

In order to characterize the spatiotemporal behavioral pattern across an extended EPM session, data from 33 animals of the vehicle group were analyzed minute-by-minute and 5-min bins. Then, a factor analysis was performed for each of the 5-min bins to capture the behavioral dimensions that emerge across different segments analyzed with this session. Finally, the effects of DZP on the behavioral response of gerbils during the session were analyzed according to the factors that were identified. Behavioral parameters from vehicle group were expressed as mean ± SEM and submitted to Friedman repeated measures (RM) analysis of variance (ANOVA) on ranks. The factor analysis was performed by principal-component followed by an orthogonal Varimax rotation. Factors with eigenvalues greater than 1.0 and loadings greater than 0.7 were kept. The measures analyzed with this procedure were the percentage of time spent in the open arms and in the central area; the percentage of entries in the open-arms and the closed-arms entries; the distance traveled in the closed-arms and the frequency of SAP in the central area. Data from the effects of DZP were expressed as mean ± SEM and submitted to Friedman RM-ANOVA on ranks for intra-group analysis. For each minute or for each 5-min bin, comparisons between the DZP and vehicle group were made by means of Mann-Whitney rank sum test. Whenever necessary, the SNK post hoc test were used. In all cases, the significance level was set at p < 0.05.

Results of temporal analysis on vehicle group data are summarized in Figure 1 and Table 1. Eleven animals jumped from the open arms of the EPM during the first minute of exposure to the apparatus. As described in method section, subjects were replaced in the central area and session time was adjusted. For the percentage of time spent and percentage of entries in the open and closed arms, RM-ANOVA on ranks showed differences between the intervals in which the session was divided (min-by-min and 5-min bins; Table 1).

Post hoc minute-by-minute tests indicate that animals spent more time in the open arms during the first 3 min of the session, as compared to the segment between the 11th and 19th minute (Figure 1A). Similarly, the number of entries to the open arms was higher during the first 3 min than during minute 7 through 19 (Figure 1B). Post hoc minute-by-minute tests also showed an opposite pattern compared to the closed arms; namely, less time spent and fewer numbers of entries during the first 3 min of the session as compared to later segments. A further analysis based on 5-min segments showed significant differences between the first bin and the rest of the segments for both measures (time spent and entries) and both types of arms (open and closed) – See Figure 1D,E. RM-ANOVA on ranks also showed differences in the distance that the gerbils ran in the EPM across different moments of the session (Table 1). Post hoc minute-by-minute tests showed that the gerbils reduced the locomotion in the open arms during the last 11 min of the session as compared to the segment between the 1th and 5th minute (Figure 1C). Post hoc tests also indicated that across the de 5-min bins, gerbils’ locomotion gradually decreased (Figure 1F).

Factor analysis results are summarized in Table 2. For each of the four 5-min bins of the session, a factor analysis was performed. Behavioral parameters related to open and closed arms exploration, as well as time spent and stretching attended posture (SAP) in the central area of the maze were processed using principal-component analysis. After an orthogonal Varimax rotation, a single factor representing the 80.9% of the variance emerged for the first 5-min of the session. For that single factor, open-arms exploration measures were negatively correlated with closed-arms and central area exploration. For the following three bins, three factors emerged explaining 88.3, 86.1, and 85% of the variance, respectively. A similar pattern of factor solution was observed for the last three 5-min bins of the session. One factor grouped the open-arms exploration measures, while the other two factors grouped measures related to closed-arms exploration and central area activity (Table 2).

The effects of the anxiolytic compound DZP on the behavioral response of gerbils are summarized in Figure 2–4 and Table 1 (Vehicle vs. DZP). RM-ANOVA indicated differences between the bins into which the session was divided for the open arms (percentage of time and entries), the closed-arms (entries and distance run) and the central-area activity-related measures (percentage time and SAP; Table 1). Post hoc tests of minute-by-minute data showed that starting in the fourth minute of the session, gerbils treated with vehicle and DZP significantly reduced open-arms exploration in terms of percentage of time and entries (Figure 2A,C, respectively).

Similarly, a 5-min bin analysis showed significantly lower time allocation and entries to the open-arm during the last three bins as compared to the first bin (see Figure 2B,D). Post hoc tests also showed that compared to the first 5-min, gerbils treated with DZP increased both closed-arms exploration and central-area activity during the last 15 min of the session (Figure 3, 4). A Mann-Whitney test was used to compare vehicle and DZP groups. It showed that during the second minute and the first 5-min bin, DZP-treated gerbils increased significantly the percentage of entries to the open arms (p < 0.05) compared to vehicle group (Figure 2C,D). In addition, during the first 5-min bin DZP also increased SAP frequency (p < 0.05). A Mann-Whitney test for the last 15-min of the session, indicated that DZP-treated gerbils reduced significantly (p < 0.05) time allocation and entries to the open-arms, as compared to vehicle group (Figure 2). Lastly, DZP significantly increased (p < 0.05) gerbils’ entries and distance run in the closed arms, and time spent and SAP frequency in the central area (Figure 3, 4).

This is the first study to report a factor analysis of a 5-min EPM session for gerbils. Whereas behavioral measures in rats and mice (Wall and Messier, 2001) tend to group into at least three independent factors related to anxiety (exploration of open arms), locomotion (exploration of closed arms), and decision making/risk assessment (time in the central area and SAP), in our study a single factor emerged. The factor solution observed in gerbils suggests that during the first 5 min of the session, the EPM test is measuring a single defensive behavior characterized by a negative relationship between exploration of open arms (percentage and number of entries) and closed-arms (number of entries and distance traveled), and time in the central area and SAP. Similar to the escape response described by Ellard (1996), during the first 5 min of the session gerbils showed vigorous locomotion in the open arms, which included bursts of high-speed running in the absence of risk-assessment and visits to the closed arms. Accordingly, it seems that during the standard 5-min session the measures obtained in the EPM are mostly related to the escape behavior and not to anxiety-related responses.

Animals in the vehicle group exhibit a high percentage of entries to the open arms (69.9%), which was increased by the DZP treatment (80.3%). Our findings suggest that DZP facilitates the escape response through the open arms during the first 5 min of the session. Though Bradley et al. (2007a) also reported an increase in the percentage of entries to the open arms in DZP-treated gerbils, Varty et al. (2002a) reported that the same anxiolytic increased the time spent but not the number of entries to the open arms. Again, these inconsistencies with Varty et al’s results may be related to the characteristics of the EPM that was utilized (clear closed arms) and the sex of the gerbils (females).

Factor analysis of the subsequent 5-min bins of the session revealed three independent factors that grouped measures of open- and closed-arms exploration, and activity in the central area, respectively. These findings suggest that after a 5-min exploration period, the EPM test measures three different behavioral dimensions. According to our interpretation, once habituated to the instrument, gerbils may face a conflict between finding shelter in the closed arms – notwithstanding these areas do not represent an entirely safe refuge – or continue trying to escape via the open arms. We propose that the identified factors possibly correspond to three independent defensive behaviors, namely anxiety (% of time and entries to the open arms), approach to the closed arms (entries and distance traveled), and risk assessment (% of time in the central area and SAP). During the last three 5-min bins of the session, gerbils treated with DZP reduced their time spent and entries to the open arms, while increased their exploration of closed-arms, the time in the central area, and incidence of SAP. These effects of DZP on the defensive behavior of gerbils are different to those reported for mice and rats (for a review see Carobrez and Bertoglio, 2005).

It has been proposed that the EPM is a conflict test in which a novel situation produces both approach (open-arms) and unconditioned avoidance (closed arms), interpreted as curiosity and caution, respectively (Handley and McBlane, 1993; Ohl, 2005). In this context, anxiolytic compounds such as DZP can shift the balance of this conflict from avoidance toward approach (Gray and McNaughton, 2000/2003). Although further tests of our interpretations are needed, we propose that the conflict produced by the presentations of open and closed areas appears later in the session, once gerbils have habituated to the EPM. This conflict would consist on the tendency to continuous attempts to escape via the open arms and finding shelter in the closed arms. Accordingly, DZP would rapidly reduce the escape response through the open arms, while increasing approximations to the closed arms and risk-assessment that precedes the visits to the open arms.

Our results also indicate that after 5 min in the EPM, DZP-treated gerbils increase their locomotion in the closed arms. Apparently, this effect of DZP is related to the gerbils’ defensive response to novelty; it has been shown that gerbils’ reactivity to novel environments interferes with the effect of diverse compounds, included DZP (Babcock et al., 2002; Okano et al., 2005; Prinssen et al., 2006). Moreover, gerbils treated with DZP increase locomotion in new environments, but not in familiar settings (Prinssen et al., 2006).

Our results indicate that male naïve gerbils require sessions of at least 10 min in the EPM in order to properly measure anxiety-related behavior in this model. Data from this session should be analyzed in two 5-min bins; during the first 5 min (standard-session length), the obtained measures seem to be related to a vigorous escape response via the open arms triggered by the novel environment. Once habituated to the instrument, the escape response decreases, and approximations to the closed arms increase. Gerbils seem to face a conflicting situation between continue attempting to escape via the open arms or find shelter in the closed arms, notwithstanding these areas do not offer entirely safe refuge as they are not covered. This conflict appears to be reduced by DZP treatment, which produces a reduction in the escape behavior and increases both approximations to the closed arms and risk assessment. Accordingly, anxiety-related measures obtained in the EPM for gerbils could be only observed after the 5-min habituation period. Unlike in rats and mice, decrements in exploration of open-arms and increments in risk-assessment could be interpreted in gerbils as an anxiolytic effect. Further studies are necessary to test the role of procedural factors such as the effect of DZP in EPM-experienced gerbils, with or without prior habituation to the experimental room or the experimenter (manipulation).