Male Proechimys rodents (n = 18), originally from the Amazon rainforest, were bred in a colony established at the Neuroscience Laboratory’s facility (Federal Register-IBAMA number 1561643) of Escola Paulista de Medicina/Universidade Federal de São Paulo (EPM/UNIFESP). Male Wistar (n = 18) rats were acquired from CEDEME/UNIFESP. Throughout the study, the animals were maintained at a constant temperature of 22 ± 1°C, 12-h light-dark cycle and with free access to food and water. All animal procedures were carried out in accordance with the Ethical and Practical Principles of the Use of Laboratory Animals with the approval of the Ethical Committee of UNIFESP (CEUA 8615110915). Precautions were taken to minimize the number of rodents utilized in the experiments.

The PM-DAT is an useful test for investigating hippocampal-dependent memory involving emotionality and conveniently allows the simultaneous study of the learning, memory and anxiety (Silva and Frussa-Filho, 2000; Rachetti et al., 2013; Frussa-Filho et al., 2016; Leão et al., 2016b). The device used was a modified elevated plus-maze, 52 cm above the floor, composed of two closed arms (48 × 16 × 50 cm) opposite to two open arms (48 × 16 cm). Wistar rats (n = 10) and Proechimys rodents (n = 9) were conditioned to decide between the two enclosed arms (in one of them, a 100-W lamp and a 2200 W hair drier, as aversive stimuli, were both placed above the end of the arm) whilst avoiding the open arms of the device (Figure 1A). During the training session, which lasted 10 min, each rodent was positioned in the center of the device and each time the animal penetrated the aversive enclosed arm, the aversive stimulus was generated until the rodent left this arm. Then, the animals were submitted to two test sessions of 3 min each, which occurred 3 h and 24 h after the conditioning training session. In the test sessions, the rodents were put on device but aversive stimuli were not actioned, although the hair-dryer and the lamp stayed positioned. Rodents were randomly tested and, after each session, a 5% alcohol solution was used to clean the device. Learning was assessed by time the animal remained in the aversive enclosed arm during training session (Silva et al., 2004; Carvalho et al., 2006; Patti et al., 2006). In turn, memory was assessed by time spent in the aversive enclosed arm during the test sessions (Silva and Frussa-Filho, 2000; Rachetti et al., 2013; Leão et al., 2016b). Anxiety-like behavior was given by the percent time spent in the open arms (amount of time the animal remained in the open arms/total time spent in both open and closed arms) of the device (Silva et al., 2000; Calzavara et al., 2004; Leão et al., 2016a).

A rectangular acrylic box (66 × 40 × 30 cm) was used in the experiments. The objects to be recognized were made of glass and presented different shapes and textures, they were tested in advance and animals showed no preferential exploration. Exploration was established when rodents sniffed or touched the object with the nose and/or forepaws. Before the test, the Wistar (n = 8) and Proechimys (n = 9) were habituated to the experimental box by allowing them to freely explore it during 10-min every day for consecutive 5 days. No stimulus object was placed inside the arena during habituation. The rodents were individually inserted into the box with two distinct objects (A and B) and left to freely explore them for 5-min. Twenty four hours later, animals were placed again in the same experimental box for a 5-min, but now they were presented to the familiar object (A) and to a novel object (C).

Brain slice preparation was produced as described previously (Salar et al., 2016). Briefly, rodents were anesthetized with 1% isoflurane dissolved in 70% N₂O and 30% O₂. Then, adult Proechimys (n = 4) and Wistar rats (n = 4) were decapitated and brains quickly removed. Horizontal hippocampal slices were made in ice-cold artificial cerebrospinal fluid (aCSF) at 4 ± 0.5°C temperature containing (in mM): NaCl 129, NaHCO₃ 21, KCl 3, CaCl₂ 1.6, MgSO₄ 1.8, NaH₂PO₄ 1.25, and glucose 10, saturated with 95% O₂ and 5% CO₂. Using a vibratome (LEICA VT 1200S), 400 μm thick brain slices were generated and promptly moved to an interface chamber perfused with aCSF at 36 ± 0.5°C (flow rate: 1.5–2 ml/min, pH 7.4, osmolarity: 295–300 mOsmol/L) where they were allowed to recover for 2–3 h.

Extracellular field potentials (FPs) were recorded from the stratum radiatum of cornu Ammonis 1 (CA1) area. The recording electrode consisted of a chlorinated silver wire inserted into a glass pipette filled with 154 mM NaCl (5–10 MΩ). Schaffer collateral axons received the orthodromic stimuli through a bipolar twisted electrode (Ni-Cr wire with 50 μm diameter tip) positioned in the stratum radiatum of the CA1 hippocampal subfield. All data were low-pass filtered at 3 kHz, digitized at 10 KHz and stored on computer disk using a CED 1401 interface for off-line analysis using Spike 2 v6.09 (CED-1401, Cambridge, UK). Before starting the LTP protocol, input-output (I/O) relationships of stimulus intensity vs. field excitatory postsynaptic potentials (fEPSP) magnitude were performed in order to determine the maximal response after constant increments of stimulus intensity until no further increase in the fEPSP amplitude. From that, stimulus intensities were adapted to generate a fEPSP slope of 50%–60% of the maximum obtained from the I/O. Field EPSP slope values were accessed after the fiber volley in order to avoid the influence of other sources of current flow. Twenty minutes of stable baseline recording period was established before application of the high frequency stimulation, in which paired pulses were delivered with an inter-stimulus interval of 50 ms. LTP was induced by Theta Burst Stimulation (TBS) protocol (Liu et al., 2016), which consists in 10 bursts repeated at 200 ms intervals, with four pulses at 100 Hz for each burst, at the same stimulus intensity as the baseline, delivered to the Schaffer collaterals and the responses were recorded during 180 min.

Electrophysiological data were analyzed using custom-made scripts (© Jan-Oliver Hollnagel, MATLAB R2013b; Salar et al., 2016). The fEPSP slope was measured between 20%–80% of its maximal amplitude. In order to verify the potentiation after TBS, field EPSP slopes were normalized relative to the averaged baseline response. Statistical analysis of behavioral and electrophysiological data was reported as mean ± standard error of the mean (SEM). Shapiro–Wilk test was used to determine normality. On behavioral tests, the statistical significance was assessed by non-parametric Mann-Whitney test, since less than 30 animals were used in the study, therefore the sample is not reliable to perform parametric approach (Ghasemi and Zahediasl, 2012). Statistical analysis for LTP was performed using two-way ANOVA (fEPSP slope at different times) followed by Bonferroni post hoc test. p < 0.05 (*) was considered to indicate a significant difference. Statistical analysis performed using Prism 5.00 (GraphPad Software, Inc., San Diego, CA, USA).

The wild Amazon rodents exhibited high anxiety-like behavior in the training session. The Wistar group spent more time on the open arms of the equipment than the Proechimys group (U = 18.00; p < 0.05; Figure 1B), throughout the training session. However, during test sessions, both species of animals exhibited similar patterns of anxiety-related behavior. No differences, between Wistar e Proechimys, were observed in the amount of time spent in the open arms of the device during the test sessions that occurred 3 h (U = 36.00; p > 0.05) and 24 h (U = 45.00; p > 0.05) after the training session (Figures 1C,D).

In the training session both species of rodents spent less time in the aversive closed arm in comparison to the amount of time spent in the non-aversive closed arm. Wistar (U = 0.00; p > 0.0001) and Proechimys (U = 0.00; p > 0.0001; Figures 2A,B).

In the test session, after 3 h of the training, both groups spent significantly less time in the aversive enclosed arm when compared to the time spent in the non-aversive enclosed arm. Wistar (U = 23.00; p < 0.05) and Proechimys (U = 17.50; p < 0.05; Figures 3A,B), suggesting that both animal species remembered the aversive situation.

In the test session, 24 h after training, Wistar rats remained less time in the aversive closed arm in comparison with the non-aversive arm (U = 21.50; p < 0.05; Figure 3C), although Proechimys rodents exhibited no difference in the amount of time spent in both enclosed arms (U = 19.00; p > 0.05; Figure 3D), suggesting that long term memory duration did not persist for 24 h in Proechimys.

The comparison between the time spent to explore both objects (A and B) revealed an equivalent exploration time between the different groups of rodents in the acquisition trial, Wistar (U = 27.50 and p > 0.05) and Proechimys (U = 31.50 and p > 0.05; Figures 4A,B). Wistar rats, 24 h later of the acquisition phase, spent significantly more time exploring the novel object compared with the familiar object (U = 12.50 and p < 0.05; Figure 4C). In contrast there was no significant difference in exploration of the novel object when compared to familiar object in Proechimys rodents (U = 33.00 and p > 0.05; Figure 4D), suggesting that long term memory duration did not persist for 24 h in Proechimys.

In order to evaluate LTP under physiological condition, animals that underwent behavioral tests were not used. Eight slices per animal group (n = 16) were used in the experiments (Figure 5A). There was no significant effect of interaction between groups (F_(2,42) = 0.5335; p > 0.05). Bonferroni’s post hoc test showed that the potentiation in the slope of fEPSP was significantly different (F_(1,42) = 64.84; p < 0.0001) between, Proechimys (17.2 ± 8.4%) and Wistar (54.4 ± 1.2%) at 60 min after TBS (Figure 5C). In Proechimys, fEPSP slope potentiation decayed over time reaching basal levels at 90 min after TBS, contrasting to Wistar rats (10.1 ± 7.1% vs. 40.1 ± 4.1%, respectively) in which potentiation was stable during recordings (e.g., fEPSP slope potentiation at 180 min after TBS, Proechimys 1.5 ± 3.1%, Wistar 33.3 ± 6.0%; Figures 5B,D). The results revealed differences associated with the maintenance phase of LTP between Proechimys and Wistar, since LTP was successfully induced in both animal species after TBS protocol.