Animals

The study involved 40 male Sprague‒Dawley rats selectively bred from the original lineage obtained from Charles River Laboratories (Charles River Laboratories, Research Models and Services, Germany GmbH). The rats were housed at the University of Health Sciences, Laboratory Animals Production and Research Center, with controlled humidity (50% ± 10) and temperature (22 ± 1 °C) under a 12:12-h light‒dark cycle. The procedures were reviewed and approved by the Animal Experimentation Ethics Committee at the University of Health Sciences (protocol no.: 2021–02/02).

Several studies indicate a sex-dependent anxiety response and PNN formation following early-life adversities, such as maternal separation and neonatal isolation which are used as models for early-life adversities [4, 20, 21]. Additionally, hormonal fluctuations during the estrous cycle can influence the anxiety response and overall behavior of female rats [22,23,24]. To minimize the effects of these factors on behavior, only male rats were used in this study.

The sample size was determined by separate statistical power analysis for each experiment, and the largest sample size was selected as the basis for determining the number of animals used (G*Power 3.1.9.4; Cohen’s d = 0.5, alpha = 0.05, and power = 0.80) [25].

Modeling Early-Life Adversity Through Maternal Separation and Neonatal Isolation

Maternal separation (MS) and neonatal isolation (NI) were used to model early-life adversities (ELA). From postnatal day (PND) 2 to 14, the rat pups from 5 litters were separated from their dams daily for 180 min, housed in a room different from their mothers, and isolated from each other by a separator [26, 27]. The temperature was adjusted to 30 ± 3 °C by a homeothermic blanket (Harvard Apparatus, Holliston, MA, USA). After PND14, the pups were returned to normal housing until they were weaned on PND23, and only males were chosen for this study. All subsequent experiments were performed during adulthood (15 weeks).

Experimental Design

The overall experimental design, including animal rearing, is illustrated in Fig. 1. Initial experimental groups were as follows:

Control group (n = 16): normal rearing

Early-Life Adversity group (ELA, n = 24): MS and NI

Fig. 1

Experimental design. Workflow diagram illustrating the in vivo procedures carried out throughout the study (Created with BioRender.com)

Immediately prior to cannulation, the control group was divided into two subgroups: the control and sham control groups. Meanwhile, the ELA group was divided into three subgroups: ELA0, ELA1, and ELA2. Animals were randomly assigned to groups while ensuring that the average body weight was comparable across all groups.

These new groups and treatments of animals in these groups are as follows:

Control group (n = 8): normal rearing, no cannulation

Sham control group (n = 8): normal rearing, cannulation

Early-Life Adversity group (ELA0, n = 8): MS and NI and no cannulation

Early-Life Adversity + Phase 1 Hyase group (ELA1, n = 8): MS and NI and Hyase injection before phase 1 of the two-way active avoidance test

Early-Life Adversity + Phase 2 Hyase group (ELA2, n = 8): MS and NI and Hyase injection before phase 2 of the two-way active avoidance test

Behavioral Tests

Elevated plus maze (EPM) and open-field (OF) tests were conducted on the same day during the diurnal phase (lights on) of the 12:12-h light–dark cycle, beginning with the OF test. Testing took place in a sound-attenuated room under consistent lighting (48 lx at maze level). Animals were acclimated to the testing room for 30–60 min prior to placement on the apparatus. Each apparatus was cleaned with a 20% ethanol solution and dried after each animal to prevent residual moisture and scent exposure. To mitigate the effects of same-day testing and test order, we maintained a 3-h interval between tests, allowing for recovery and reducing variability. Testing environments were standardized in lighting, noise, and other contextual factors to minimize external influences on behavior.

Elevated Plus Maze Test

The elevated plus maze was used to examine the anxiety behavior of ELA (total n = 24) and control (total n = 16) rats [2, 28]. The number of entries into the open arms and the distance traveled in the open arms were detected by ANY-Maze object tracking software for 300 s (ANY-Maze Video Tracking System, Stoelting Europe, Dublin, Ireland). The anxiety index was calculated via the following formula: 1 − [([Time in open arms/Test duration] + [number of entries into open arms/Total number of entries into arms])/2] [29]. The test was repeated after cannulation to examine the effect of the procedure on the stress response.

Open Field Test

The open-field test was used to compare the locomotor activity of the ELA (total n = 24) rats with that of the control (total n = 16) rats. The animals were placed into the open-field test setup with dimensions of 100 × 100 × 30 cm split into 16 equal squares from a predetermined corner, and their behaviors were recorded for 300 s (ANY-Maze Video Tracking System, Stoelting Europe, Dublin, Ireland). The total distance traveled, number of entries into the center, and time spent in the center were considered indicators of locomotor activity and anxiety [30,31,32]. The test was repeated after cannulation to examine the effects of the procedure on locomotor activity.

Intracortical Cannulation and InjectionCannulation

The rats in the ELA1, ELA2, and sham control groups were implanted with i.c.v. guide cannulas (C313GRL/SPCguide 22GA, Plastic One Inc., Roanoke, VA, USA) bilaterally under 90 mg/kg (i.p.) ketamine hydrochloride (Ketalar, Pfizer, Istanbul, Turkey) and 10 mg/kg (i.p.) xylazine hydrochloride (Xylazinbio, Biyoveta, Istanbul, Turkey) in combination, via a stereotaxic apparatus (Stoelting Instruments, Wood Dale, IL, USA). Cannulas were placed in the temporal cortex (AP + 2.0 mm, ML ± 3.5 mm, and DV + 3.5 mm by bregma). Coordinates were determined according to the rat brain atlas (Paxinos & Watson, 1986) and validated by implantation into same-age rats that were not included in the main study. The guide cannula was closed with a stainless-steel wire dummy cannula (C313DC/SPCdummy 0.014/0.36 mm, Plastic One Inc., Roanoke, VA, USA) when not injected. The infusion cannulas were attached to the skull via three stainless-steel screws and dental acrylic. Although cannulation can cause short-termr ECM disruption, this effect does not persist long term [33, 34]. Following cannulation, the rats were housed in plexiglass cages for a recovery period of 1 week, allowing the brain ECM to return to baseline levels. Additionally, a sham control group was included to account for any temporary ECM changes due to cannulation.

Hyase Injection

ELA1 rats were injected (i.c.) with the ECM-degrading enzyme hyaluronidase (Hyase, Sigma-Aldrich HX0514) before the initial discrimination phase, and ELA2 group rats were injected before the reversal learning phase of the two-way active avoidance test. Microinjections were administered bilaterally with a 25 µl Hamilton microinjector (702 N; Hamilton Bonaduz, Switzerland) through previously implanted cannulas and a flexible polyethylene tube into the auditory cortex (ACx) (AP + 2.0 mm, ML ± 3.5 mm, and DV + 3.5 mm by bregma), delivering a volume of 10 ml (750 U) of Hyase solution. The use of Hyase in this study was preferred over other enzymes due to its distinct advantage of allowing the ECM to return to its normal state within approximately 2 weeks after administration into the brain [35, 36]. This characteristic reduces the risk of the enzyme’s effects extending into subsequent experimental phases, which was critical for our experimental design.

Two-Way Active Avoidance Test with FM Discrimination

Adult Sprague–Dawley rats were trained once a day in a two-compartment shuttle box (Ugo Basile, Comerio VA, Italy) to discern the direction of linear frequency modulation employed as conditioned go/no-go stimuli (70 dB; CS + : 2–4 kHz; CS–: 4–2 kHz; duration: 4 s). Individual foot shock intensities (0. 1–0.4 mA) were adjusted and delivered through a metal floor grid as the unconditioned stimulus (US) (Fig. 2A).

Fig. 2

Task design (A) and behavioral outcomes (B) for the CS + and CS − trials in a two-way active avoidance test with a frequency-modulated discrimination task (Created with BioRender.com)

The training was conducted in daily sessions of 40 trials (20 CS + , 20 CS −) presented in randomized order after fixed intertrial intervals of 15 s. A compartment change after CS + onset within 4 s was classified as a “hit” response. The absence of a hit response was counted as a “miss,” and the US of 10 s was immediately delivered following the CS + and terminated by a change in the compartment (Fig. 2B).

For the CS- trials, a compartment change within 4 s was classified as a “false alarm” response, and the US was applied for up to 10 s following the movement. The US was not delivered after the CS − if the animal stayed in the compartment, and the response was considered “correct rejection.” Hit rates (CR +) and false alarm rates (CR −) were calculated as the percentage of hits and false alarms in each 40-trial session. After 14 days of the initial learning phase, the rats were trained with a reversed contingency as the reversal learning phase.

Sacrifice and Tissue Collection

The rats were sacrificed by cervical dislocation under 90 mg/kg (i.p.) ketamine hydrochloride (Ketalar, Pfizer, Istanbul, Turkey) and 10 mg/kg (i.p.) xylazine hydrochloride (Xylazinbio, Biyoveta, Istanbul, Turkey) in combination 2 weeks after the experiments.

Blood samples were collected through an insulin injector from cardiac blood and allowed to coagulate in a microcentrifuge for 30–60 min. The supernatants were collected after centrifugation at 1500 × g for 10 min.

The brains were rapidly removed, fixed in 4% PFA for 24 h, and washed 3 × 10 min in PBS before being cryoprotected in 15% and 30% sucrose. After the tissues sank in sucrose, they were frozen and embedded in a matrix (Bioblock).

Serum Corticosterone Measurement

The serum corticosterone (CORT) concentration was determined via competitive ELISA (CEA540Ge, USCN Life Science Inc., Wuhan, China). Tests were performed according to the manufacturer’s instructions.

Immunohistochemistry

The brain tissues were cut into 50-µm-thick coronal sections via a cryostat (Cryostar NX50, Thermo Scientific) and collected in PBS-containing wells. Four to six sections containing the temporal cortex were stained according to the free-floating tissue staining protocol of Potts et al. with necessary adjustments [37]. The tissue sections were washed and blocked in PBS supplemented with 0.3% Triton-X (#1,003,243,275, Sigma–Aldrich), 5% goat serum (Cat# GOA-1B, Capricorn Scientific) for 1 h at RT before being stained for 2 h at 37 °C in a primary antibody cocktail containing anti-PV (1:200, Sigma‒Aldrich Cat# SAB4200545, RRID: AB_2857970) and WFA-FITC (1:200, Invitrogen Thermo Fisher Scientific, Cat# L32481, RRID: AB_308666666). The sections were washed 3 × 10 min in PBS, transferred to secondary antibody (goat anti-mouse Alexa Fluor® 594, 1:500, Abcam, Cat#ab150120, RRID: AB_2631447) and incubated for another 2 h at 37 °C. The tissue sections were mounted onto slides and covered with mounting medium (Fluoroshield, Cat# F6182; Sigma Aldrich).

All imaging and analyses were performed by an investigator blind to the experimental groups. Images were captured with an Axiocam-105 camera connected to a Zeiss Axio Vert.A1 fluorescence microscope and Zen Blue 2 software (RRID: SCR_013672). For analysis, three images were taken from the ACx and motor cortex (MCx) of each animal. Image analysis was conducted via ImageJ (https://imagej.net/ij/, RRID: SCR_003070) and CellProfiler Image Analysis Software (www.cellprofiler.org, RRID: SCR_007358).

Statistical Analysis

Data are presented as medians with interquartile ranges (IQRs) and means ± SEMs depending on the normality of the distribution. SPSS (RRID:SCR_002865) and Jamovi (RRID:SCR_016142) software were used for the analyses. Statistical significance levels are expressed as *p < 0.05, **p < 0.01, and ***p < 0.001.

Serum CORT levels and the results of the EPM and OF tests were compared between the control group (n = 16) and the ELA group (n = 24). Owing to the nonnormal distribution according to the Shapiro–Wilk test, comparisons between these two groups were performed via the Mann‒Whitney U test. Histological analyses were also performed between these two groups via paired-sample Student’s t tests.

FM discrimination was evaluated in five groups: the control (n = 8), sham control (n = 8), ELA0 (n = 8), ELA1 (n = 8), and ELA2 (n = 8) groups. Daily learning performance was calculated as the difference between hit rates and false alarm rates, where hit rates = hits/number of trials and false alarm rates = false alarms/number of trials. Discrimination sensitivity was assessed via d′ values via signal detection theory [38], where a d′ of 1.0 indicates a signal discrimination strength of one standard deviation above the noise level. Discrimination performance within each group was evaluated via the Wilcoxon signed-rank test. Learning and discrimination performance differences between groups were analyzed via mixed-model ANOVA (with Greenhouse–Geisser and Huynh–Feldt correction of sphericity when necessary) since ANOVA is generally robust to nonnormal data when sphericity is met [39, 40].