Experimental subjects
All the procedures for handling and sacrificing animals followed the European Commission guidelines for the welfare of experimental animals (2010/63/EU) and were approved by the local Bioethics Committee (2013/53/RD). Animals were housed in standard laboratory cages with ad libitum access to food and water, under a 12-hour dark–light cycle in temperature-controlled rooms. Male C57BL/6 wild-type mice and Ip3r2−/− (RRID: MGI:3640970) were used in the present study from 1-3 months old. Mice studied for longitudinal analysis of corticosterone treatment effects (Extended data Fig.1a) were 5 months old. C57BL/6 mice were purchased from Jackson Laboratory. Ip3r2−/− mice were generated by crossing germline-heterozygous-null mutant Ip3r2+/− mice89.
Corticosterone treatment
Corticosterone (Cort, Cat# C2505; Sigma-Aldrich) was dissolved in commercial mineral water46. Decreasing Cort concentrations were presented to male C57BL/6 wild-type mice for 28 days: 30 μg/ml during 15 days (resulting in a dose of approximately 6.6 mg/kg/day), followed by 15 μg/ml (2.7 mg/kg/day) during 3 days, and 7.5 μg/ml (1.1 mg/kg/day) during 10 days; for a gradual recovery of endogenous corticosterone plasma levels44,46. Cort solution was available ad libitum in drinking water (dark bottles) and was renewed every 72 hours. To verify Cort consumption, bottles´ weight was controlled each time the solution was renewed. Control mice (naïve) followed the same experimental approach without Cort in their bottles. At the end of the treatment, mice from both groups were subjected to behavioral tests. Forced swimming test (FST) was routinely evaluated to highlight the level of animal despair in order to guarantee the existence of a reliable mouse model of depression44,46. In a subset of mice, long-term effects of Cort treatment were analyzed after 3 months (Extended Data Fig. 1a). After two weeks of treatment, naïve and Cort-mice were injected with viral vectors.
Surgeries and viral injections
Mice (1-2 months) were anesthetized via isoflurane (5% for induction, 2% for maintenance) in oxygen and place in a custom adapted stereotaxic frame. Depth of anaesthesia was determined by testing toe-pinch reflexes. Body temperature was continuously monitored and maintained at 37°C. Buprenorphine (0.1 mg/kg; Buprenex, 0.1 mg/ml) was subcutaneously injected before surgeries. The hair of the scalp was shaved and cleaned with 70% ethanol. Once bregma and lambda were exposed, target coordinates were taken from Paxinos atlas90. Medial prefrontal cortex (mPFC) coordinates: 1.78 mm anterior, ±0.3 mm lateral from bregma, and from 2.5 to 0.8 dorso-ventral; Dorsal raphe nucleus (DRN) coordinates: posterior coordinate was scaled using bregma-lambda distance x 0.98 for every mouse91, 1.1 mm lateral from lambda, and from 3.3 to 2.8 dorso-ventral, 20º angle. A craniotomy was made at the injection site using a small burr (Ref. 19007-05, Fine Science Tools), powered by a surgical drill (NSK V-Max Volvere Lab System). Saline solution (0.9%) was applied to keep the skull cold and to maintain hydration. Intracranial injections were made using a borosilicate micropipette (World Precision Instuments) at 50 nl/min infusion rate. The following viral vectors were used: AAV5-gfaABC1D-cyto-GCaMP6f (Addgene 52925; viral titer 1.3 x 1013 GC/ml), AAV5-GFAP-hM3Dq-mCherry (Addgene 50478; viral titer 2 x 1013 GC/ml), AAV8-GFAP104-mCherry (UNC Vector Core; viral titer 2.7 x 1012 GC/ml), AAV9-hSyn-ChrimsonR-tdTom (UNC Vector Core; viral titer 4.1 x 1012 GC/ml), AAV5-GFAP-mcherry-cre (UNC Vector Core; viral titer 4.3 x 1012 GC/ml), AAV5-CAG-flex-iSeroSnFR (Addgene 128486; viral titer 5 x+1012 GC/ml, Unitat de Vectors Virals, UAB), AAV5-GFAP-eGFP-WPRE-hGH (Addgene 105549; viral titer 1.3 x 1013 GC/ml). After injection, the micropipette was held in place for 5 min prior to retraction to prevent leakage. The skin was sutured and the mice were monitored, kept on a heating pad while recovering and returned to the home cage. Buprenorphine (0.05 ml, 0.1 mg/ml), was given once daily for 48 h post-surgery. Experiments were performed 2–3 weeks post-injection.
Cortical slice preparation
Animals were sacrificed and their brains were quickly and carefully removed92. The brains were placed in an ice-cold artificial cerebrospinal fluid (aCSF) containing [in mM]: NaCl 124, KCl 2.69, KH2PO4 1.25, MgSO4 2, NaHCO3 26, CaCl2 2, and glucose 10, and was gassed with carbogen (95% O2/5% CO2, pH = 7.3). Slices 350 μm thick were obtained with a vibratome (Leica Vibratome VT1200S, Germany) and incubated (> 1 hour) at room temperature (RT, 22–24°C) in aCSF continuously bubbled. Slices were then transferred to an immersion recording chamber superfused at 2 ml/min with gassed aCSF and visualized under an Olympus BX50WI microscope (Olympus Optical, Japan) coupled with a 40X water immersion lens and infrared-DIC optics.
To improve slice viability in adult mice (>2 months old), ice-cold (4ºC) NMDG-HEPES solution was perfused prior to brain extraction. Subsequently, slices were placed in the same NMDG-HEPES solution at 37ºC for 10 min. Afterwards, slices were kept in ACSF until use (>1h). NMDG-HEPES contained the following [in mM]: NMDG 92, KCl 2.5, NaH2PO4 1.2, NaHCO3 30, HEPES 20, glucose 25, thiourea 2, Na-ascorbate 5, Na-pyruvate 3, CaCl2·2H2O 0.5, and MgSO4·7H2O 10 (95% O2/5% CO2, pH = 7.3)93. A modified Mg2+-free aCSF was used to monitor slow inward currents (SICs) in order to optimize NMDA receptor activation, which contained the following [in Mm]: NaCl 124, KCl 2.69, KH2PO4 1.25, NaHCO3 26, glucose 10, CaCl2 4 and glycine 0,01 (95% O2/5% CO2, pH = 7.3).
Ex vivo calcium imaging and analysis
The genetically encoded calcium indicator (GECIs) AAV5-Gfap-cytoGCaMP6f was bilaterally injected in mPFC, and after 2-3 weeks astrocytes from cortical slices were analyzed. Astrocytes of mPFC layer 2/3 were imaged using a CCD camera (ORCA-235, Hamamatsu, Japan) attached to the microscope. Cells were illuminated for 100-200 ms at 490 nm using LED system (CoolLED pE-100), and images were acquired at 1 Hz during 2 min. The LED system and the camera were controlled and synchronized by NIS Elements software (Nikon, Japan). Spontaneous Ca2+ events were monitored during 2 min, and Ca2+ evoked responses were analyzed by recording baseline activity for 30 s, followed by local application of 5-HT (1 mM; 10 s, 1 bar), ATP (1 mM; 10 s, 1 bar), and clozapine-N-oxide (CNO, 1 mM; 2 s, 1 bar) restricted to 60 s after stimuli. Local application of agonists was delivered by pressure pulses through a micropipette (Picospritzer II, Parker Hannifin, Mayfield Heights, OH, USA). To isolate the specific response of astrocytes to 5-HT, the following drug cocktail was included in aCSF: TTX 1 µM, picrotoxin 50 µM, AM251 2 µM, MRS 2179 10 µM, CGP 55845 5 µM, LY367385 100 µM. For ATP experiments, MRS 2179 was excluded of the cocktail. For the experiments combining Ca2+ imaging and optogenetic stimulation, baseline recordings were acquired for 60 s, and astrocyte-evoked responses were considered up to 60 s after DRN fibers stimulation.
Regions of interest (ROI) were selected using ImageJ software. All pixels within each ROI were averaged to obtain a single time course F[t] per ROI. Custom-written software in MATLAB (MATLAB R2020a; Mathworks, Natick, MA) was used for further processing (modified from Mederos et al., 2020). Artifacts in the fluorescence signal produced by mechanical movement were removed from the analysis. Then, signals were low-pass filtered with a Chebyshev II filter. Photobleaching was adjusted and the ΔF/F0 was calculated for each ROI. Events were considered when their ΔF/F0 > 2-3 times the noise variance and had at least > 3% of relative change (0.03). Frequency, amplitude, area under the curve, and duration were analyzed for each ROI. In a subset of experiments, mPFC slices were incubated with Fluo-4 AM (1 μl of 2 mM dye was dropped over the mPFC, attaining a final concentration of 2-10 μM) dissolved in 0.02% pluronic and 0.04% DMSO for 15-20 min at RT, and Ca2+ signal analysis was restricted to the cell soma (cf.94).
In vivo calcium recordings and analysis
AAV5-gfaABC1D-cyto-GCaMP6f was injected in the right hemisphere of mPFC, followed by implantation of 2 mm borosilicate fiber-optic cannulas (fiber core Ø of 400 µm; 0.66 NA; ref. MFC_400/430-0.66_2.0_MF1.25_FLT, Doric Lenses). Cannulas were secured to the skull using a base layer of adhesive dental cement (Meron, Voco). 2-3 weeks after the surgery, behavioral testing started to allow for viral expression and animal recovery.
Doric GCaMP Fiber Photometry System (FPS_1S_GCaMP, Doric Lenses) was used, with a 405 nm LED as the isosbestic point, and a 465 nm LED as the excitation dependent GCaMP fluorescence. Blue light was delivered to the brain at 20-50 µW. Signals were interleaved and collected at 100 Hz. Raw signals were demodulated and analyzed with custom-written software in Matlab, with a cut-off frequency of 20 Hz and an attenuation of 20 dB, followed by a 1 s moving mean window. Isosbestic signals were fitted to Ca2+-dependent signals and subtracted to eliminate motion related artifacts95,96. GCaMP6 fluorescence signals across animals were standardized as follows: DF(F-F0)/F0, where F0 was computed by linearly interpolating between the local minima of the fluorescence signal across different time windows (window size: 45 s) to account for any remaining photobleaching. Ca2+ event was defined as a period in which fluorescence showed a local maximun > 2 times the noise variance of the signal97. Events whose maximum value was below 0.01 (1% of relative change), or whose prominence were below 0.001 (0.1%) were excluded. Events were expanded towards the closest local minima (both before and after the peak), to designate the start and end of putative Ca2+ events. To account for multipeak events, Gaussians were fitted to each Ca2+ event to infer their real duration and area under the curve.
For behavioral testing, spontaneous Ca2+ signals were analyzed in the open field test (OF), and mice showing <2 events in OF were removed from the analysis. Spontaneous Ca2+ events detected during the first 5 minutes were selected and analyzed. For social recognition test, analysis was restricted to the first 5 explorations for both the neutral object and unfamiliar mouse, to avoid the exponential decay shown after several explorations98. To analyze Ca2+ events during explorations, events whose peak occurred in the interval defined from 1 s before exploration onset up to 3 seconds after the end of an exploration were selected. To compare across subjects, signals were Z-score transformed, by computing the ratio of the DF/F0 signal over the standard deviation of the signal during the first 5 minutes when mice were in the neutral chamber 99. Ca2+ signals were time-aligned from 5 s prior to exploration onset up to 20 s after exploration onset (Fig. 1h). Animal speed was evaluated to discard any possible confounding between astrocytic activity and mouse running speed. For the open field, the mean velocity during the entire duration of each Ca2+ event was computed. Then, a linear regression model linking Ca2+ event amplitude and mean velocity was fitted using the fitlm function in MATLAB. For the social recognition test, the mean speed associated to each Ca2+ event was computed as described above, and only events associated to the first 5 explorations (object and mouse) were considered for downstream analysis. Then, a linear regression was fitted separately for Ca2+ events associated to object explorations and for events associated to mouse exploration.
In vivo serotonergic recordings and analysis
Mice were injected with either AAV5-CAG-flex-iSeroSnFR + AAV5/GFAP-mcherry-cre, or AAV5-GFAP-eGFP-WPRE-hGH virus in the right hemisphere of mPFC, followed by implantation of fiber-optic cannula (fiber core Ø of 400 µm; 0.50 NA; ref. FP400URT Thorlabs) following the same surgical procedure as for GCaMP Fiber Photometry. In addition, AAV9-hSyn-ChrimsonR-tdTom was injected in DRN, and fiber-optic cannula implanted (fiber core Ø of 400 µm; 0.50 NA; ref. FP400URT, Thorlabs). 2-3 weeks after surgery, behavioral testing started to allow for viral expression and animal recovery. Fiber photometry recordings were performed using FPS_1S_GCaMP system. A 465 nm LED delivered at 70-130 µW was used for iSeroSnFR excitation, and a 590 nm LED (M590F3 - 590 nm, Fiber-Coupled LED, 1000 mA, SMA- LEDD1B - T-Cube LED Driver) at 5 mW was used for optogenetic stimulation of DRN. iSeroSnFR and eGFP signals were recorded while the mouse was freely moving in the OF arena. Each animal underwent between 2-6 trials of DRN stimuli (40 Hz, 10 s) with 50 s inter-intervals, with 3 min of baseline recordings previous to DRN stimulation. For the analysis, the first min of recordings was discarded to account for signal photobleaching effects. Photometry signals were collected interleaved at a sampling frequency of 50Hz and analyzed as described above. Signals were low-pass filtered with a Chebyshev Type II filter with a 30 Hz cut-off frequency. DF/F0 signal was computed for iSeroSnFR and eGFP fluorescence measurements. For traces representation (Extended Data Fig. 5e, g), signals were low-pass filtered with a cut-off frequency of 20Hz to reduce noise, and Z-scored was computed to compare across subjects. Analysis was restricted to the signals recorded 30 s before (baseline) and the 30 s after the stimulation onset.
Ex vivo electrophysiological recordings
Whole-cell patch-clamp recordings from layer 2/3 pyramidal neurons and astrocytes of mPFC were performed. Neuronal currents were recorded by borosilicate capillaries (3-6 MΩ) filled with an intracellular solution that contained [in mM]: K-gluconate 135, KCl 10, HEPES 10, MgCl2 1, and ATP-Na2 2 (pH = 7.3). In some experiments, intracellular solution was modified containing GDPβS 2 mM. Astrocytic whole-cell recordings were performed (8-10 MΩ) using an intracellular solution containing [in mM]: BAPTA-K4 40, NaCl 8, MgCl2 1, HEPES 10, GTP-tris salt 0.4 and ATP-Na2 2 (pH = 7.3). Astrocyte recordings lasted ≥ 30 min to allow the dialysis of BAPTA through the gap-junction connected astrocytic network100. Recordings were obtained with PC-ONE amplifiers (Dagan Corporation, Minneapolis, MN) in voltage-clamp conditions and the membrane potential was held at -70 mV. Series and input resistances were monitored throughout the experiment using -5 mV pulses. Recordings with access resistance change >20% were rejected. Signals were fed to a Pentium-based PC through a DigiData 1440 interface board (Axon Instruments). Signals were filtered at 1 kHz and acquired at 10 kHz sampling rate. The pCLAMP 10.7 software (Axon Instruments) was used for stimulus generation, data display, acquisition, and storage. Experiments were performed at RT.
Slow inward currents (SICs) were recorded in the presence of TTX (1 µM) and distinguished from miniature synaptic currents (mEPSCs) by their slower time courses55,101. Excitatory postsynaptic currents (EPSC) were elicited by theta capillaries (2-5 μm tip diameter) located in layer V and filled with aCSF. Paired pulses (250 μs duration; 75 ms interval) were continuously delivered at 0.33 Hz by stimulator S-900 (Dagan Corporation). Baseline of synaptic activity was measured 5 min before local application of 5-HT/CNO. aCSF included picrotoxin (50 μM) to block GABAA-dependent inhibitory synaptic activity. For those neuronal recordings that did not last >25 min, the period of stable responses was considered for further analysis (short responses).
Optogenetic stimulation
Light stimulation with the CoolLED illumination system, 550 nm light pulses of 50 ms at 5 Hz (1 mW) was applied for the electrophysiological recordings in mPFC slices, which activated ChrimsonR-expressing fibers and induced the endogenous release of 5-HT. For ex vivo astrocyte Ca2+ recordings, 640-660 nm light stimulation (10 s, continuous light, <1 mW) through external laser was used to activate ChrimsonR-expressing fibers.
Immunohistochemistry and confocal microscopy
Mice were euthanized by sodium pentobarbital i.p. injections and transcardially perfused with phosphate-buffered saline (PBS: 137 mM NaCl, 2.7 mM KCl, 10 mM Na2HPO4, 2 mM KH2PO4, pH 7.4; 15714 Electron Microscopy Sciences, EM Grade) followed by ice-cold 4% paraformaldehyde (PFA). Brains were removed and postfixed overnight (o/n) at 4°C in 4% PFA. Coronal brain slices (50 μm thick) were obtained with a VT1000S vibratome (Leica) and collected as floating sections. For immunostaining, slices were first washed with PBS and permeabilized with 0.2% Triton/PBS. Nonspecific binding was blocked with PBS containing 1% goat serum and 0.3% Triton-X 100 for 1 h. Samples were then incubated with the corresponding primary antibodies in blocking solution overnight at 4 °C: rabbit anti-S100-β (1:200, Abcam, Cambridge, UK; RRID: AB_306716), mouse anti-SERT (1:500, Synaptic Systems CI.64G6) mouse anti-Neuronal Nuclei (NeuN, 1:500, Merck, MAB377). After three 20-min washes in blocking solution at RT, floating sections were incubated for 1 h RT with specific secondary antibodies: Alexa Fluor 488 (goat anti-mouse; 1:200, Bioss Inc., Woburn, MA, RRID:AB_10892893); Alexa Fluor 647 (goat anti-rabbit; 1:200, Thermo Fisher Scientific, RRID:AB_2535813); Alexa Fluor 488 (goat anti-rabbit; 1:200, Thermo Fisher Scientific, RRID:AB_143165). After three 20-min of PBS washes containing 0.1% Triton X-100, slices were incubated with DAPI (1.5 μg/mL, Sigma-Aldrich) for 10 min. Finally, sections were washed, three times 20-min each, in PBS and mounted with Vectashield antifade mounting medium (H-1000, Vector Laboratories, Burlingame, CA), and images acquired using a Leica SP-5 confocal microscope (Leica Biosystems). Quantification was performed using Fiji software (ImageJ 1.53i, NIH). All the antibodies used in the study have been satisfactorily validated by commercial vendors.
Colocalization of ChrimsonR-tdTom expressing fibers and serotonergic projections labelled with anti-SERT in mPFC was analyzed. Maximal projections of z-stacks (10 μm thickness) obtained with a 63 × 1.40 NA oil immersion objective (single optical sections 1 μm) were used. After thresholding the ChrimsonR-tdTom image a mask was created. This mask was superimposed over anti-SERT image and ROIs were automatically detected and measured. 5 background ROIs were manually selected in the anti-SERT image. Colocalization was considered when the mean intensity value of anti-SERT ROIs was above the background average plus 3 times the standard deviation. For each field of view, the colocalization percentage was calculated.
Behavioral assays
Handling period was performed for 5 min during 5 consecutive days before behavioral testing. Mice were transferred to the testing room for at least 30 min before the experiment to reduce stress68. All tasks were performed between 08:00 am and 3:00 pm. Arenas and maze were cleaned with a 0.1 % acetic acid dissolved in water between the sessions
Forced swimming test (FST). FST was performed in a clear acrylic cylinder (29 cm height, 12 cm diameter) filled with warm water (22–23°C). Mouse behavior was video-recorded for 6 min. Immobility score was analyzed during the last 4 min using EthoVision XT 7 software (Noldus Information Technology, Inc.; Leesburg, VA). Time when mice were immobile was used as indicator of hopelessness, which has been related with depressive-like phenotypes in rodents102,103.
Elevated plus maze (EPM) test. EPM is used as a reference of mouse anxiety levels104. The maze had two closed and two open arms (30 × 10 × 5 cm each) and is placed 1 m above the ground. At the beginning of the session (5 min total duration), the animal was placed at the intersection of the arms. The time spent in the open and enclosed arms was recorded by EthoVision XT 7 software. The exploration index was computed as the time spent in open arms vs the total time spent in open and closed arms. An entry was considered when the mouse had all four paws inside the arm of the maze.
Object in place (OIP) test. Acrylic open field arena (40 x 40 x 40 cm) with 4 non-identical objects placed near the corners of the arena105 was used. Mouse behavior was recorded by EthoVision XT 7 software. 3 days prior testing, animals were individually habituated to explore the empty arena for 30 min. In the first trial, animals were allowed to freely explore the different objects for 5 min. After 3 min of inter-trial delay, mice were allowed to explore the arena where two objects were reallocated (new object location). The index was computed as the time spent exploring the objects in novel locations of the total time exploring. Exploration was defined as reaching the object with the nose.
Novel object recognition (NOR) test. NOR test was conducted using the same arena as OIP. Once mice were habituated to the arena, they were allowed to explore two identical objects for 5 minutes. After one hour, mice are reintroduced into the arena where an object has been replaced by a novel one106. The index was computed as the total time spent exploring novel object versus the total time of exploration (novel + familiar).
Social recognition (SR) test. Sociability measurements were performed in a clear acrylic three-chamber cage (60 x 42 x 20 cm each)107,108. The middle chamber was used as a resting point, and the chambers on the side hold two small cylindrical cages that contained an unfamiliar mouse or a neutral small object109,110. Unfamiliar male mice were habituated to remain into the cylinder cages 2 days prior testing. On the testing day, mice were placed in the middle chamber (to prevent access to the side chambers, clear acrylic sliding doors were used), and were allowed to explore it for 5 min. Afterwards, the doors were opened and the animal was able to freely explore the “social chamber” (holding the unknown mouse) or the “non-social chamber” (holding the object) for 10 min. An exploration was considered when mouse’s nose was in contact with the cage.
For CNO or AIDA experiments, i.p. injections were conducted 20-30 minutes before the beginning of each task. Animals with no exploratory behavior in a particular test were eliminated from the analysis of that test.
Drugs and chemicals
The following reagents were bath-applied during ex vivo recordings for at least 15 min before testing: picrotoxin (50 µM, Sigma, Cat#P1675; CAS:124-87-8), D-AP5 (50 µM, Tocris, Cat#0106; CAS: 79055-68-8), LY367385 (100 µM, Tocris, Cat#1237; CAS: 198419-91-9), SB 216641 hydrochloride (50 µM, Tocris, Cat#1242; CAS 193611-67-5) AM251 (2 µM, Tocris, Cat#1117; CAS: 183232-66-8), MRS 2179 tetrasodium salt (10 µM, Tocris, Cat#0900; CAS: 1454889-37-2), CGP 55845 hydrochloride (5 µM, Tocris, Cat#1248; CAS 149184-22-5), Ketanserine tartrate (10 µM, Tocris, Cat#0908 CAS 83846-83-7), RS127445 hydrochloride (1 µM, Tocris, Cat#2993 CAS 199864-86-3), RS102221 hydrochloride (1 µM, Tocris, Cat#1050 CAS 187397-18-8), tetrodotoxin (TTX, 1 µM, Alomone labs, Cat#T-550; CAS: 18660-81-6). A constant flow of fresh ACSF plus selected drugs was continuously perfused into the recording chamber. Serotonin hydrochloride (1 mM, Tocris, Cat# 3547 CAS 153-98-0), Clozapine N-oxide (1 mM, Tocris, Cat# 4936 CAS 34233-69-7), and Adenose 5’triphosphate disodium salt hydrate (ATP, 1 mM, Sigma-Aldrich, Cat# A7699) were locally applied by a micropipette. The following inhibitors were added to the intracellular solutions: 1,2-bis(2-aminophenoxy)ethane-N,N,N′,N′-tetraacetic acid (BAPTA, 40mM, Sigma-Aldrich, Cat# A4926), and guanosine 5′-[β-thio]diphosphate (GDPβS), trilithium salt (2mM, Merck, Cat# G7637) 1-aminoindan-1,5-dicarboxylic acid (AIDA, 5 mg/kg, Tocris, Cat# 0904 CAS 168560-79-0) and Clozapine N-oxide (3 mg/kg, Tocris, Cat# 4936 CAS 34233-69-7) were administered via i.p. Fluo-4 AM (Invitrogen, Cat# F14201), Pluronic® F-127 (Merck, Cat# P2443), Dimethyl sulfoxide (DMSO, Sigma-Aldrich, Cat# D8418).
Statistical analysis
All animal samples and biological replicate numbers in this study are in line with well-accepted standards from the literature for each method. All data presented in this work were obtained from experimental replicates; that is, multiple animal cohorts from different litters, at least three experimental repeats for each assay, and production of biological replicates. All attempts of replication were successful. Each statistical test was used according to the design of the experiment and the structure of the data. Two-group comparisons were performed using one-way ANOVA with Dunn´s method, Tukey test and Holmes Sidak post-hoc analysis; or paired T-test or Wilcoxon matched-pair tests according to the normality of the data distribution. No statistical methods were used to predetermine sample sizes in this study, which were determined according to the accepted practice for the applied assays 9,111. Experiments, except the behavioral test, were not performed with blinding to the conditions of the experiments. However, data analyses were performed blinded to the scorer or did not require manual scoring. Descriptive statistics are reported as the mean ± s.e.m., and box and whisker plots. In BW plots the central mark indicates the median, and the bottom and top edges of the box indicate the 25th and 75th percentiles, respectively. The whiskers extend to the maximum and minimum data points (not considered outliers). In scatter dot plot graphs, the central mark indicates the median, and the top and bottom edges, so-called range, correspond to the maximum and minimum values, respectively. Statistically significant differences were established at *P < 0.05, **P < 0.01 and ***P < 0.001, two-sided.