Animal

C57BL/6 J male mice (7 weeks old) were purchased from the Experimental Animals Center of Tongji Medical College, Huazhong University of Science and Technology. All mice experiments in this research were carried out following the recommendations of, and were approved by the Animal Welfare Committee of Huazhong University of Science and Technology.

The mouse strains used included wild-type C57BL/6 J, PV-Cre mice. C57BL/6 male mice (7 weeksold) were purchased from the Experimental Animals Center of Tongji Medical College, Huazhong University of Science and Technology. PV-Cre mice were described previously (Wen et al. 2010), PV-Cre mouse line (Stock No. 017320, Jackson Laboratory) used was kindly provided by Dr. Lin Mei (Case Western Reserve University). Tail genomic DNA was used for genotyping by PCR. A 300-bp fragment was detected for the PV-Cre allele.

Mice aged between 7 and 14 weeks were used for all experiments. Animals were housed in less than 5 mice per cage at 22–24 °C with 40–70% humidity, on a 12 h/12 h light/dark cycle schedule, with water and food available ad libitum. Mice were all backcrossed with C57BL/6 J mice for over 10 generations.

Experimental Design

Protocols for animal experiments were approved by the Animal Experimental Ethics Committee of the Huazhong University of Science and Technology (Approval Code: 82071508) on September 27, 2020. Each experiment and the statistical calculations described below were carried out in a randomized order by the experimenter blinded to the group. A schematic experimental design is shown in Fig. 1. AG1478 (HY-13524, MCE) was solubilized in DMSO and diluted to a concentration of 10 mg/kg mouse weight in sterile saline. AG1478 or saline (1% DMSO in normal saline), at a dose of 50 mg/kg body weight, was chronically administered intraperitoneally (i.p.) for 2 h before running every other day for four weeks. The timing and dose of AG1478 administration was based on a previous study (Weglicki et al. 2012).

Fig. 1

AG1478 neutralized treadmill running effect on ErbB4 phosphorylation in the hippocampus. A Schematic experimental design. B The body weight of both the vehicle:run group and AG1478: run group weighs less than the relevant control group (vehicle: static group and AG1478: static group), while AG1478 does not affect it. Data are expressed as mean ± SEM, Three way ANOVA: F (4, 180) = 862.7, p < 0.0001; Tukey’s multiple comparisons test: in the 2nd week: vehilce: static vs. vehicle:run: p = 0.0096, AG1478: static vs. vehicle: run: p = 0.0477, in the 3rd week: vehilce: static vs. vehicle: run: p < 0.0001, AG1478: static vs. AG1478: run: p = 0.0038, AG1478: static vs. vehicle: run: p = 0.0070, in the 4th week: AG1478: static vs. AG1478: run: p < 0.0001, AG1478: static vs. vehicle: run: p < 0.0001, vehicle: static vs. AG1478: run: p < 0.0001, vehicle: static vs. vehicle: run: p < 0.0001, *p < 0.05, ****p < 0.0001. n = 8. C Representative western blots of total ErbB4 and phosphate-ErbB4 in the hippocampus. a-Tubulin was used for normalization. n = 5. D Quantitative analysis of the expression of total ErbB4 in the hippocampus. Two way ANOVA: F (1, 16) = 0.008286, P = 0.9286, Tukey’s multiple comparisons test: AG1478: static vs. vehicle: run: p = 0.0385. E Quantitative analysis of the relative phosphate-ErbB4 level to total ErbB4 in the hippocampus. n = 5, Two way ANOVA: F (1,16) = 0.5757, p = 0.4590, Tukey’s multiple comparisons test: vehicle: static vs. AG1478:static: p = 0.0242, vehicle:static vs. vehicle: run: p = 0.0245, vehicle:run vs. AG1478: static: p < 0.0001, AG1478:static vs. AG1478:run: p = 0.0026

The procedures have been described previously (Yi et al. 2020). The mice in running groups underwent adaptive run-training sessions in individual lanes of a treadmill (FT-200, Taimeng, China) (5 m/min for 45 min) for 5 days, and running sessions (5 m/min for 10 min, 8 m/min for 30 min, 5 m/min for 10 min) for the next 4 weeks to prevent stress-induced inhibition of hippocampal neurogenesis (Leem et al. 2018). Mice in the static groups were placed on the stationary treadmill for the same duration, and they often did not move or move by themselves. The running time is fixed in the morning.

BrdU Injections

5′-bromo-2′-deoxyuridine (BrdU; Sigma) in saline was administered i.p. 1 h before running every 4 hours (100 mg/kg), twice 1 day, for 6 days. Animals were sacrificed five weeks after the last BrdU injection.

Behavioral Analysis

All behavioral tests were performed during light periods, and all mice were handled for at least 5 min twice a day for 3 days before the behavioral test (Albarran-Zeckler et al. 2012). Behavioral analysis was performed on 12-week-old mice by investigators unaware of their genotype and groups.

Animals were tested in sequential order of least disruptive (absence of noxious stimulation: open field and T-maze) to most disruptive (e.g., MWM, fear conditioning test, and contextual fear discrimination learning, which involved physical stimuli, such as loud noise, foot shock, or cool water).

Open Field Test (OFT)

To provide measures of locomotor activity and general anxiety-like behavior, locomotor activity was recorded in the open field, made of a rectangular chamber (45 × 45 × 45 cm). The floor illumination level is set to 200 lux by LED lights on both sides of the wall. The mouse was gently placed on the center square and allowed to freely explore the arena for 5 min. The total distance traveled during a session, the time spent moving, and the total duration spent in the center were measured by an automated video tracking system (TMV-100S, TaiMeng, China) above the open field. After each trial, the apparatus was swept out with 75% alcohol to avoid the presence of olfactory cues. General activity and anxiety were assessed by calculating the total distance traveled, total movement time, and time spent in the center of the field, respectively.

Forced-Choice Spontaneous Alternation in T-Maze Test

A forced-choice paradigm was employed to encourage a higher level of alternation (Hughes 2004). The apparatus was an enclosed maze with 3 arms, a start arm (38 × 7 cm), a central choice area (7 × 7 cm), and two symmetrically choice arms (30 × 7 cm). In brief, each trial consisted of a 5-min acquisition phase, an inter-trial interval (ITI; 2 min), and a final 5-min test phase. ITIs of 2 min were included to assess the persistence of short-term spatial memory (Lalonde et al. 2003). During the acquisition phase, the mouse was placed in the starting arm facing the wall and allowed to explore the apparatus. As soon as the animal entered (with all four paws) one of the two choice arms, the compartment door was closed for 30 s. Then, the mouse was gently removed from the maze to the home cage for the ITI. At the test, the block was removed, and the mouse was placed back in the starting arm for a second-choice trial. The novel arm in the second trial was the right choice, the arena was thoroughly cleaned using 70% ethanol to remove any scent cues, that could identify the novel arm. The test was conducted in the room where the animals were housed and consisted of 10 trials.

$$}\;}\;\left( \% \right)\; = \;\frac}\;}\;}\;}\;}\;}\;} }}}\;}}}\; \times \;100\%$$

Contextual and Cued Fear Conditioning Test

The contextual and cued fear conditioning test is the behavioral paradigm used to evaluate associative fear learning, hippocampus-dependent, and hippocampus-independent memory function memory in rodents (Shoji et al. 2014). Briefly, the mouse was placed in the conditioning chamber using the Freeze Monitor system (San Diego Instruments, San Diego, CA, USA) for 3 min as an accommodation period and then a tone-foot-shock pairing (tone, 30 s, 65 dB, 1 kHz; foot shock, 2 s, 0.75 mA) was delivered. The mouse was allowed to explore the chamber for another 30 s after the shock to study postshock freezing.

The contextual fear conditioning test was assessed 24 h after training by placing the mice back in the same test chamber for 3 min to assess short-term memory. The floor illumination level is set to 100 lux by LED lights. The cued fear conditioning test was assessed 2 h after the contextual fear conditioning test in a novel acrylic triangular chamber with a flat, white floor, covered with a transparent lid chamber (context C) changed smell (A drop of lemon juice on one wall) and tone of the training for 3 min. The floor illumination level is set to 30 lux by LED lights.

Freezing behavior, defined as the absence of all visible movement of the body except the movement necessitated by respiration, was scored by a observing software. At the end of each test, the chamber was cleaned with 75% alcohol to avoid the presence of olfactory cues.

Morris Water Maze

To assess hippocampal-dependent spatial learning and memory, mice were tested in the Morris water maze (MWM,XR-XM101; Shanghai Softmaze Information Technology Co., Ltd., Shanghai, China) (130 cm diameter, 45 cm high) containing opaque water (24–26 °C) and a circular platform (10 cm diameter, approximately 1 cm below the water surface) located in the center of the target quadrant. The MWM was virtually divided into four equal imaginary quadrants by the AnyMaze software. The test was executed as previously described (Montag-Sallaz and Montag 2003). Briefly, the test consisted of a 4-day hidden platform training test and a one-day single probe test. In the training test, the mouse was allowed to face the pool wall in a random starting place to find the hidden platform. The mouse was allowed 120 s to find the platform they sat for 20 s. If the mouse did not find the platform within 120 s, it was gently guided there and allowed to stay on it for 30 s. Twenty-four hours after the last training session, the platform was removed from the pool, the mouse was placed in the opposite quadrant, and a 60-s probe trial was performed. The escape latency, the percentage of time spent on the target, and the number of platform crossings were recorded.

Contextual Fear Discrimination Learning

Pattern separation is a fundamental computational function of DG (Faghihi and Moustafa 2015; McNaughton et al. 1986), which depends on normal adult neurogenesis. This paradigm tests the animal's ability to distinguish between similar contexts (Sahay et al. 2011). The shock-associated training context A (with foot shock) and the similar context B (without foot shock) shared many features, including an exposed stainless-steel grid floor and roof. Four black and white inserts were applied to cover the walls, and a plexiglass floorboard was above the stainless-steel grid floor in context B. A non-alcoholic antiseptic solution was used to clean the grids between trials. In pilot experiments, mice were exposed to context A, where they received a single 2 s 0.75 mA foot shock, 185 s following placement in the soundproof chamber (29 × 29 × 24 cm; Coulbourn instruments, Allentown, PA, USA, model H10-11 M-7C-SF) and the floor's illumination level is set to 100 lux by the LED lights. In discrimination learning, mice were exposed to the training context A. One hour later, mice were placed in a similar context C, and left for 180 s without foot shocks. When no mouse movement is detected for more than 2 s, its behavior is counted as “freezing.”

Freezing behavior during the test is tracked and measured as an index of fear memory using video tracking software daily. Context discrimination ratio was calculated as follows: a score of 0 indicated a complete lack of discrimination, i.e., freezing levels were the same in similar and training contexts (Freezing similar context = Freezing training context).

$$}\;}\; = \;\frac}\;}\;}\,}\; - \;}\;}\;}\,}}}}\;}\;}\;}\; + \;}\;}\;}\;}}}$$

In Vitro and In Vivo Calcium Imaging

To assess the effect of AG1478 on intracellular Ca2+ levels in PV neurons, calcium signals from cells were observed in brain slices. First, the brain slice was quickly removed and placed in 4 °C artificial cerebrospinal fluid (ACSF) composed of (in mM) choline chloride 110, KCl 2.5, NaH2PO4 1.25, NaHCO3 26.0, CaCl2 0.5, MgCl2 7, d-glucose 10, Na-ascorbate 11.6, Na-pyruvate 3.1, and atropine sulfate 0.01. The coronal slices (300 μm) were cut using a Leica VT1000S vibratome and then incubated in standard ACSF containing (in mM) NaCl 126, KCl 3, NaH2PO4 1.25, NaHCO3 26.0, CaCl2 2, MgSO4 2, d-glucose 11 saturated with 95% O2 and 5% CO2. Next, hippocampus slices were kept at 32 °C for approximately 30 min before recording. The Ca2+ signals were visualized by an inverted fluorescent microscope (Olympus, USA) with a water-immersion objective lens (40X, LUMPlanFL, 0.80 numerical aperture; OLYMPUS). Brain slices from one animal were divided into two groups. One group was incubated for 30 min at 32 °C in ACSF as the control group, whereas another in 10 μmol/L AG1478 as the experiment group. Upon excitation at 488 nm, calcium-complexed GCaMP6 was collected at ten frames per second using the Image-pro plus 7.0 software. The Ca2+ levels in slices were calculated as following using calcium imaging in vitro assay.

To further assess the effect of AG1478 on intracellular Ca2+ levels in PV neurons of the running mice, calcium signals from cells in awake animals with head-fixed animals were observed by in vivo calcium imaging. Firstly, 300 nL of AAV-EF1α-DIO-GCaMP6m-WPRE-hGH-pA viruses (BrainVTA, catalog# PT-0283) were injected into the DG of PV-Cre mice using a stereotaxic instrument. The stereotactic technology procedures have been described previously (Yi et al. 2020). Adult PV-Cre mice were anesthetized with chloral hydrate (400 mg/kg, i.p.) and head-fixed in a stereotaxic device (RWD Life Science; 68,025). Viruses were unilaterally injected (0.3 mL per side, 0.05 nL/min) with a glass pipette (Cetin et al. 2006) (tip size, ~ 20 mm) at the following coordinates relative to bregma: anteroposterior, − 1.94 mm; dorsoventral, − 2.14 mm; and mediolateral, ± 1.5 mm. After injection, the glass pipette was left in place for 15 min before slowly removing it. The titers of AAV-EF1α-DIO-GCaMP6m-WPRE-hGH-pA (BrainVTA, catalog #: PT-0283) were 2 × 1012 genome copies per mL. Secondly, 4 weeks after virus injection, the gradient-index (GRIN) lens (0.5 mm in diameter and 5.901 mm in length, Gofoton, USA, GRIN Tech) was firmly mounted to a stereotaxic holder. The GRIN lens were inserted into the DG (either the left or the right side) and positioned at 0.10 mm to 0.20 mm above the highest viral injection site. Next, dental cement (New Century Dental Materials Co. Ltd., Shanghai, China) was applied to secure the GRIN lens. During recording, the distance from the microscope to the GRIN lens was adjusted by a micro-manipulator until the field of view was in focus. The gain was set at 16, and the LED (470 nm) power was maintained at 100%. Images were collected at 30 frames per second using the MiniScope V2.0 software.

In vivo Ca2+ imaging experiments, we divided the mice into two groups. Both groups received a dose of vehicle, and the Ca2+ signals in the static state were recorded 30 min later. Subsequently, one group continued to receive vehicle, while the other group was administered AG1478. Ca2+ signals were recorded once more after another 30 min. Finally, following a period of running exercise, we recorded the Ca2+ signals again. All comparisons were based on the Ca2+ signal intensity recorded during the static state.

Images were processed using MATLAB R2020b (Mathworks, Natick, MA) with customized code (https://github.com/thinkertech333/analysisforminiscope, Thinkertech, Nanjing, China), and Cellsort 4.0 software. After motion correction (set at 10) and denoising (set at 100), the region of interest was manually selected according to the fluorescence intensity. The Ca2+ signals in the first 5 s were considered baseline, and the average Ca2+ signal in this period was used as a reference (F0) to normalize the fluorescence signal (DF/F). The formula is as follows:

$$ \mathord F}} \right. \kern-0pt} F} = \frac}_}}} - }}}}}}$$

Fsignal is the real-time Ca2+ signal intensity of the cells of interest, F0 is the average Ca2+ signal intensity in the baseline period.

Immunofluorescence

Mice were anesthetized with chloral hydrate and perfused transcardially with 4% paraformaldehyde (PFA) in phosphate-buffered saline (PBS, pH7.4), and tissues were fixed overnight in 4% PFA at 4 °C. After cryoprotected in 30% sucrose, brain tissues were frozen in OCT and cut into 40 μm by a cryostat (Thermo Scientific, HM550).

The free-floating sections were rinsed three times in PBS, blocked in PBS with 0.3% Triton and 10% goat serum, and 0.1% Triton-X 100 (PBST) for 60 min at room temperature. Sections were then incubated with primary antibodies in PBST at 4 °C overnight. After washing with PBS 3 times, samples were incubated with second antibodies in PBST for 1 h at room temperature. Samples were mounted with mounting medium, antifading (with DAPI) (S2110, Beijing Solarbio Science & Technology), and images were taken with a Zeiss Axioplan light microscope. Label quantification of labeling was determined by counting all fluorescent cells in 3–5 sections per animal. Adobe Photoshop CS6 (PS) software was used to calculate the areas of the DG. Next, we used the counting tool to count the number of cells in the target area. The density of positive cells was calculated by the number of positive cells divided by the area of the DG.

For BrdU staining, sections were incubated with 2 N HCl for 30 min at 37 °C to denature the DNA, followed by neutralization with 0.1 M borate buffer (pH 8.5) for 10 min at room temperature. After neutralization, sections were rinsed with PBS several times before incubation with primary antibodies.

Immunocytochemistry used following antibodies: rabbit anti-KI67 (1:500, Abcam, Cambridge, UK, catalog #: ab15580, PMID: 34420159), rat anti-BrdU (1:300, FITC conjugated, Abcam, Cambridge, UK, catalog #:ab74545, PMID: 32117963), mouse anti-NeuN (1:500, Abcam, Cambridge, UK, catalog #: ab104224, PMID: 32117963), rabbit anti-PV (1:100, Abclonal, Shanghai, CN, catalog #: A2791,PMID: 32117963), donkey anti-rabbit IgG conjugated with Alexa Fluor 594 (1:200, Jackson ImmunoResearch, USA, catalog #: R37119, PMID: 32117963), goat anti-rabbit IgG conjugated with Alexa Fluor 488 (1: 250, Proteintech, CN, catalog #: SA00006-2, PMID: 30208760), goat anti-mouse 488(1:250, Proteintech, CN, catalog #: SA00013-1, PMID: 31849642).

Western Blot

One day after behavioral tests, the mice were anesthetized with 5% chloral hydrate (8 ml/kg), and tissues were rapidly collected. The tissue homogenates were prepared on ice in RIPA buffer containing 50 mM Tris-HCl (pH 7.4), 150 mM NaCl, 5% sodium deoxycholate, 1% NP40, 1 mM PMSF, and 1 μg/ml protease inhibitor cocktail. For immunoblotting p-ErbB4, homogenates were subjected to immunoprecipitation with ErbB4 antibody and protein-A (Roche) at 4 °C overnight. Homogenates or bound proteins were resolved on SDS/PAGE and transferred to PVDF membranes, which were incubated in Tris–Phosphate buffer (TBS) containing 0.1% Tween-20 and 5% milk (TBST) for 1 h at room temperature before the addition of primary antibodies for incubation overnight at 4 °C. After washing, the membranes were incubated with HRP-conjugated secondary antibodies (1: 30000, Biosharp, catalog #: BL003A, PMID: 32117963) in TBS for 1 h at room temperature. Primary antibodies used and blotting conditions were: rabbit anti-phosphate-ErbB4 (1:1000, Abcam, Cambridge, UK, catalog #: Ab76132, PMID: 31043588), ErbB4 (1:1000, catalog #: ab19391, Cambridge, UK), rabbit polyclonal anti-a-Tubulin (AC003, 1:500, Abclonal, Abclonal, Shanghai, CN, PMID: 32117963). These bands were visualized using enhanced chemiluminescence (Biorad, UK, catalog #: 1705060), scanned by MicroChemi 4.2 (DNR Bio-imaging Systems, ISRAEL), and quantified with Image J software (National Institutes of Health, Bethesda, MD, USA). Protein levels were quantified by measuring the density of each band using Image J software and then normalized to the α-Tubulin level.

Statistical Analysis

All statistical analyses were performed by GraphPad Prism 8.0 Software (GraphPad Software, San Diego, CA) and detailed in the corresponding figure legends. The quantitative data were expressed as Mean ± SEM. Group comparisons were made using two-way ANOVA followed by Tukey’s post hoc tests for multiple comparisons. A mixed 3-way ANOVA was applied in case of repeated measurements (for instance, for the body weight in Fig. 1B, the open field data in Fig. 3B, and pattern separation in Fig. 5H) with “running” and “treatment” as between groups factors and “time bins” as within groups factor. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001. Comparisons with no asterisk or ‘NS’ had p > 0.05 and were considered not significant.