Mice deficient in synaptic protease neurotrypsin show impaired spaced long-term potentiation and blunted learning-induced modulation of dendritic spines

Animals

All experiments and behavioral procedures were conducted in accordance with animal research ethics standards defined by German law and approved by the Ethical Committee on Animal Health and Care of the State of Saxony-Anhalt (TVA 2502–2-1159 and 42,502–2-1343).

Mice constitutively lacking exons 10 and 11 from the NT gene (NT−/−) (Reif et al., 2007) and their wild-type littermates (NT+/+) were backcrossed to C57BL/6 J mice for > 9 generations. NT−/− and NT+/+ mice for experiments were obtained by mating male and female NT+/− mice. NT mice were kindly provided by Dr. Peter Sonderegger from the University of Zurich. Heterozygous neurotrypsin (NT+/−) mice were crossbred with Thy1-EGFP-M+/− mice, which were purchased from Jackson Laboratory (www.jax.org/strain/007788). The NT+/+/Thy1-EGFP-M+/* and NT−/−/Thy1-EGFP-M+/* mice (+ /* stands for + / + or +/−) used for spine analysis were obtained by mating male and female NT+/−/Thy1-EGFP-M+/− mice.

C57BL/6 J, NT, Thy1-EGFP-M, and NT/Thy1-EGFP-M mice were bred at the animal facility of DZNE Magdeburg. For electrophysiological experiments, we used 4-week-old NT−/− and NT+/+ mice of both sexes. For behavioral experiments, we used “juvenile” (3- to 5-week-old) male littermates or a cohort of aged (11- to 24-month-old age-matched; on average, 18 months old) NT−/− and NT+/+ male mice. For immunohistochemistry and spine imaging, we used 3- to 4-week-old NT−/−/Thy1-EGFP-M+/* and NT+/+/Thy1-EGFP-M+/* mice of both sexes [for spine analysis associated behavioral test (contextual fear conditioning test), data collected in male juvenile mice are presented in Fig. 2i]. For viral injections, we used NT−/−/Thy1-EGFP-M+/* P7 mice of both sexes.

Mice were kept in a reversed light–dark cycle (12:12 h, light on at 9:00 pm) with access to food and water ad libitum and were transferred to fresh cages weekly. All behavioral experiments were carried out during the dark phase of the cycle, i.e., when mice are active.

For behavioral and spine imaging experiments, mice were individually housed 7 days before the start of the experiments. For electrophysiological experiments, mice were housed in groups of three-to-four mice per home cage. Behavioral analysis in matured mice was performed by an experimenter blinded to group identity. After the open field test, a few juvenile mice of both genotypes were excluded from further cognitive behavioral tasks, as they were not properly habituated to the arena and showed signs of nervousness, anxiety and agitation, most likely due to their young age. The numbers of mice used for each experiment are given in the figure legends. Outliers were excluded from graphs and subsequent statistical analysis using GraphPad outlier calculator software (www.graphpad.com/quickcalcs/Grubbs1.cfm). For spine imaging experiments, mice from the same litter were randomly allocated into three experimental groups (naïve, contextual fear-conditioned, and treated for extinction). For viral injections, littermates were randomly allocated into two experimental groups (AAV-Ag15 or AAV-Ag22). Injections were performed as follows: 1 mouse was injected with AAV-Ag15, and then, 1 mouse was injected with AAV-Ag22, etc. On postnatal day 7, mice were randomly picked up by their tail from the nest.

Electrophysiological recordings in hippocampal slices

Acute hippocampal slices were prepared from 4-week-old NT−/− and NT+/+ mice. Each mouse was killed by cervical dislocation, followed by decapitation. The brain was removed from the skull and transferred into ice-cold artificial cerebrospinal fluid (ACSF) saturated with carbogen (95% O2/5% CO2) containing (in mM) 250 sucrose, 25.6 NaHCO3, 10 glucose, 4.9 KCl, 1.25 KH2PO4, 2 CaCl2, and 2.0 MgSO4 (pH = 7.3). Both hippocampi were dissected out and sliced transversally (400 µm) using a tissue chopper with a cooled stage (custom-made by LIN, Magdeburg, Germany). Slices were kept at room temperature in carbogen-bubbled ACSF (95% O2 /5% CO2) containing 124 mM NaCl instead of 250 mM sucrose for at least 2 h before recordings were initiated.

Recordings were performed in the same solution in a submerged chamber that was continuously superfused with carbogen-bubbled ACSF (1.2 ml/min) at 32 °C. Recordings of field excitatory postsynaptic potentials (fEPSPs) were performed in CA1a and CA1c with a glass pipette filled with ACSF to activate synapses in the CA1b stratum radiatum. The resistance of the pipette was 1–4 MΩ. Stimulation pulses were applied to Schaffer collaterals via a monopolar, electrolytically sharpened and lacquer-coated stainless-steel electrode located approximately 300 mm closer to the CA3 subfield than to the recording electrode. Basal synaptic transmission was monitored at 0.05 Hz and collected at 3 pulses/min. The spaced LTP protocol was performed as previously described (Kramár et al., 2012). LTP was induced by applying 5xTBS with an interval of 20 s. One TBS consisted of a single train of ten bursts (four pulses at 100 Hz) separated by 200 ms and the width of the single pulses was 0.2 ms. To induce spaced LTP, we applied two trains of TBS (TBS1/TBS2) separated by 1 h. The stimulation strength was set to provide baseline fEPSPs with slopes of approximately 50% of the subthreshold maximum. The data were recorded at a sampling rate of 10 kHz and then filtered (0–5 kHz) and analyzed using IntraCell software (custom-made, LIN Magdeburg, Germany).

Behavioral tests

All experiments were performed under uniform illumination (30 lx, unless otherwise stated), and all behavior was video recorded using a USB video camera and analyzed using ANY-maze software (ANY-maze, version 4.99, Stoelting Co., Wood Dale, IL). All recorded movies were analyzed by a trained observer blinded to the groups.

In juvenile male mice, the following behavioral tests were performed: open field test, novel object recognition test, three-chamber sociability test, and conventional and spaced contextual fear conditioning (using two different cohorts of mice). To characterize the persistence of behavioral changes found in juvenile mice and to further extend the behavioral characterization of NT−/− mice, a battery of eight behavioral tests was performed using one cohort of aged male mice. The test battery included the open field test, novel object location test, novel object recognition test, temporal order recognition test, three-chamber sociability test, three-chamber social recognition test, elevated plus maze test, and contextual fear conditioning (CFC). The order of the tests was optimized according to the degree of invasiveness to reduce the chance that prior tests would influence animal performance in later tests [63]. Because the cohort of matured mice included mice of varying ages, covariance analysis with age as a covariate was performed for all behavioral tests. Because no effect of age was revealed (Suppl. Data 2), data for all ages were pooled and analyzed by ANOVA.

Open field test

The open field apparatus was made out of white polyacrylics and consisted of a white-square arena (50 × 50 × 30 cm). Experimental subjects were carried to the testing room in their home cages at least 30 min before the beginning of the experiment. Mice were placed in the center of the open field and allowed to freely explore it for 10 min. In juvenile mice, the first 5 min were used for the analysis of the following behavioral parameters: time spent in the inner area (central zone, 30 × 30 cm) and in the outer area (periphery) of the open field arena, locomotor activity (total travel distance), average speed, immobility time, grooming activity (including washing or mouthing of forelimbs, hind paws, face, body, and genitals), and number of defecations (number of fecal boli produced). The protocol for the open field test in matured mice was the same as that used for young mice. However, since the elevated plus maze test was performed to evaluate the anxiety status of the mice, parameters that highly depend on manual counting were not analyzed. Before the start of each session and between animals, the open field test apparatus was carefully wiped with a 70% alcohol solution.

New object location test

The test started 24 h after the open field test in matured mice. The same apparatus (50 × 50 × 30 cm) used for the open field test was used in the NOLT. Two pieces of A4 paper with stripe patterns were stuck to the upper middle area on two adjacent walls and served as landmarks. Experimental subjects were carried to the testing room in their home cages at least 30 min before the beginning of the experiment. The experiment included two phases: the encoding phase and the retrieval phase. During the encoding phase, two identical objects were placed at the adjacent corners of the central area (30 × 30 cm) of the open field arena; one of them was close to the corner which had landmarks on both walls. During the retrieval phase, this object was moved to the adjacent corner of the central part in the open field area to have both objects in a diagonal configuration. In both phases, mice were given 10 min for free exploration. In the same trial, objects were counterbalanced between mice. In different trials, different pairs of objects were used. The interval between the encoding and retrieval phases was 24 h. Before placing the next animal in the arena, the apparatus was carefully wiped with a 70% alcohol solution. The exploration time for each object was automatically counted by ANY-maze software. Exploration was considered to occur when the animal’s head was at a distance < 2 cm from the object but excluded time intervals when the animal climbed onto the object. The discrimination ratio was calculated as [time exploring the object in a novel position—time exploring the object in a familiar position]/[time exploring the object in a novel position + time exploring the object in a familiar position] × 100%.

Novel object recognition test

The same apparatus used for the open field test was used in the novel object recognition test. Experimental subjects were carried to the testing room in their home cages at least 30 min before the beginning of the experiment. The test was performed using a standard protocol [64] that included two phases: a familiarization/encoding (F) phase and a test/retrieval (R) phase. Juvenile mice were habituated to the apparatus 2 days before familiarization for 10 min each day. In matured mice, the test started 24 h after the novel location recognition test, which served as the encoding phase of the novel object recognition test. During familiarization, mice were placed in the arena for 10 min and were allowed to freely explore two identical objects separated by 25 cm in the center of the arena. During the retrieval phase, one familiar object and one novel object were placed in the center of the arena and mice were allowed to explore the apparatus for 10 min. In the same trial, objects were counterbalanced between mice; between trials, different sets of objects were used. The interval between the encoding and retrieval phases was 24 h. Before the start of each session and prior to bringing the next animal into the arena, the apparatus was carefully wiped with a 70% alcohol solution. The exploration time for each object and the total exploration time were manually estimated. Exploration was considered to have occurred when the orientation of the animal’s nose was at a distance < 2 cm from the object and included the time spent sniffing and directly touching the object. Novelty detection was evaluated by calculating the discrimination ratio as follows: [novel object time – familiar object time]/[novel object time + familiar object time] × 100%.

Temporal order recognition test

Matured mice were evaluated by the temporal order recognition test. The test was initiated 24 h after the novel object recognition test using the same test chamber and location of objects. This test included two encoding phases and a retrieval phase. In each encoding phase, a pair of novel identical objects were introduced to animals; in the retrieval phase, one object from each pair was used to test if animals could recognize the temporal order of objects. Objects and the relative position of objects did not change in the retrieval phase and were counterbalanced for animal genotype. The interval between any two consecutive test phases was 1.5 h. In all three phases, animals were given 10 min for free exploration. The same criteria as in NOLT that were used to define exploration time were applied for automatic analysis by ANY-maze software. The discrimination ratio was calculated as [exploration time of the object less recently shown – exploration time of the object recently shown]/[exploration time of the object less recently shown + exploration time of the object recently shown] × 100%.

Sociability test

Sociability levels were assessed using a three-chamber apparatus (60 × 30 × 30 cm) made out of white polyacrylics and that had connecting doors between chambers. The test in juvenile mice was performed using a standard protocol [65]: mice were habituated to the apparatus 2 h before the test for 10 min. Subsequently, one “stimulus” mouse was placed inside a small cage at one end of the compartments and an identical empty cage was placed in the opposite compartment. The mouse performing the test was allowed to explore the whole apparatus for 10 min. The exploration times spent near the cage containing the stimulus mouse and near the empty cage were estimated. The time spent sniffing and directly touching the mouse and the empty cage was considered as exploration time. In juvenile mice, as active social interaction is difficult to track automatically and hard to distinguish from time spent merely in proximity to the social partner, social interaction was analyzed and scored manually by an experimenter who was blind to the mouse genotype. In the three-chamber sociability test, the discrimination ratio was calculated as follows: [stimulus mouse time – empty cage time]/[stimulus mouse time + empty cage time] × 100%. Before the start of each session and before placing the next animal into the apparatus, the three-chamber sociability test apparatus was carefully wiped with a 70% alcohol solution; stimulus mice were changed every two sessions to avoid anxiety and stress.

In matured mice, the sociability test was initiated 24 h after the temporal order recognition test. The protocol was similar to that used for young mice. A few settings were optimized to improve the quality and reliability of automatic analysis: 1) the identical small cages that would be used as a potential container of “stimulus animals'' were set on both terminal compartments in the phase when an animal was allowed to familiarize itself with the whole test environment (to reduce the exploration triggered by small cages); 2) during the test phase, a novel object was put into the small cage which had no animal inside as a stimulus (to better control the preference of subject mice for alive mice than for a nonanimated object); 3) a novel stimulus mouse was introduced after every four consecutive trials (to maintain the mental status of stimulus mice); and 4) the small cages were covered from above (to reduce the chance that the ANY-maze software might mistakenly track stimulus mouse movement rather than subject mouse movement, leading to the creation of an artifact). After these optimizations, the automatic analysis was precise enough to track the movements of subject mice and reflect their interests in other animals. Exploration was considered to have occurred when the orientation of the animal’s head was at a distance < 2 cm from the cages but excluded instances when the animal climbed onto the small cages.

Social recognition test

The test was performed in matured mice in the three-chamber maze and was initiated 4 h after the sociability test using similar basic settings. The familiar stimulus mouse was placed inside the same small cage at the same end of the compartments as in the sociability test, and a novel stimulus mouse was placed inside the opposite compartment. Mice performing the test were given 10 min to freely explore the three-chamber apparatus. Before placing the next subject animal in the apparatus, the apparatus was carefully wiped with a 70% alcohol solution. Stimulus animals were changed every four consecutive trials. The exploration time toward cages including novel or familiar stimulus animals was automatically measured by ANY-maze software using the same criteria as was used in the sociability test. The discrimination ratio was calculated as [time exploring the cage containing the novel stimulus animal—time exploring the cage containing the familiar stimulus animal]/[time exploring the cage containing the novel stimulus animal + time exploring the cage containing the familiar stimulus animal] × 100%.

Elevated plus maze test (EPM)

The EPM was performed on the day after SRT in matured mice. The maze consisted of four arms in the shape of a cross (length: 30 cm, width: 5 cm, and height from floor: 50 cm). Two arms were enclosed by 15.25 cm-high walls and faced each other on opposite sides; the other two arms were open borders [66]. To reduce behavioral bias introduced by the light–dark difference between open and enclosed arms, LED lamps were used to guarantee that the light was reflected from the ceiling over enclosed arms. The final light intensity readouts at the end of all four arms were 15 lx. The movement of mice was evaluated in a single 10-min session. The heads of the mice were tracked by ANY-maze software, so that entry into open arms and enclosed arms could be noted. The time animals’ heads were located in different arms as well as the discrimination ratio [(time in enclosed arms—time in open arms)/(time in closed arms + time in open arms) × 100%] were used to evaluate anxiety status in mice.

Spaced contextual fear conditioning (CFC)

A spaced contextual fear conditioning (CFC) paradigm was performed as previously described [67] but with two conditioning sessions separated by 1 h instead of having only one session. Before CFC training, mice were handled and habituated to the experimental room, and the conditions were maintained in a home cage for 3 days for 5 min each day. During the training day (day 0, d0), CFC was performed as follows: mice were placed into a neutral context (CC-) (the freezing level during this time interval was taken as the baseline value for the CC-) for 5 min; 2 h later, mice were placed into a conditioned context (CC +) and 3 × medium intensity footshocks were applied (0.5 mA, 1 s) with an interval of 30 s. This procedure was repeated 1 h later (hence, spaced learning was employed). The protocol included 1 min exploration in the CC + before the first shock was administered (the freezing level during this time interval was taken as the baseline value for the CC +), 30 s after the second shock, and again 30 s after the third shock. Mice were then left for additional 30 s in the CC + before being transferred to a home cage. The CC + was a chamber (20 × 20 × 30 cm) with a contrast black-and-white chess-like pattern on the walls and a metal grid on the floor. The neutral context (CC −) was the same chamber, but with gray walls and a gray plastic floor. Before the start of each session and before placing the next animal in the apparatus, the fear conditioning apparatus was carefully wiped with a 75% alcohol solution (CC +) or with Meliseptol® having a different smell (CC-) to facilitate discrimination between both contexts. In juvenile mice, memory retrieval was tested at d2, and mice were placed in the CC + for 5 min to assess the retention of contextual memory. Subsequently, 9 × memory extinction sessions were performed on 3 consecutive days (d5-d7, 3 × sessions per day). In each session, mice were placed in the CC + for 5 min. At d9, mice were placed in the CC + again for 5 min (second memory retrieval), and freezing was assessed to evaluate fear memory extinction. A computerized fear conditioning system (Ugo Basile, Gemonio, Italy) was used for analysis. The total freezing time was manually calculated as the percentage of 5 min (in either of the two contexts, CC + and CC-) when animals showed no movement except for breathing. The discrimination ratio was calculated as follows: [freezing time in the CC + – freezing time in CC-]/[freezing time in the CC +  + freezing time in CC-] × 100%.

Contextual fear conditioning

Classical (non-spaced) CFC was performed as above but with the following modifications. In matured mice, CFC was performed on the day after EPM. During the training day (d0), mice experienced the CC- and CC + for 5 min each, only once. The interval between the CC- and CC + was 2 h. The 0.6 mA footshock with a duration of 1 s was administered in the CC + at 120, 180, and 240 s. During the first recall day (d1), mice were placed in the CC + and CC- for 5 min to confirm the contextual fear memory. From day 2 to day 4, 9 × memory extinction sessions were performed (3 × sessions per day) by placing mice in the CC + for 5 min. On the second recall day (d5), mice were placed in the CC + and CC- for 5 min to confirm the extinction of contextual fear memory. For juvenile mice, the intensity of foot shocks were slightly lower than that administered to mature mice (0.5 mA), and no CC- was present.

Spine analysisSample collection, perfusion, and tissue processing in juvenile mice

Juvenile mice were individually anesthetized with 100 mg/kg ketamine and 5 mg/kg xylazine and transcardially perfused with 0.1 M phosphate buffer solution (PBS, pH 7.4) followed by 4% formaldehyde diluted in 0.1 M phosphate buffer solution for 15 min. Brains were removed and postfixed for 24 h in 4% formaldehyde-PBS at 4 °C. The brains were then transferred to a sucrose solution (1 M in 0.1 M NaH2PO4 buffer) until the solution had infiltrated into the whole brain (~ 48 h) to cryoprotect the tissue. Finally, the brains were frozen in 100% 2-methylbutane at -80 °C and cryosectioned in 50-μm-thick coronal sections. Floating sections were kept in cryoprotective solution (1 part ethyl glycol, 1 part glycerin, and 2 parts PBS, pH 7.4). All sections were washed 3 × in 0.1 M phosphate buffer solution (PBS, pH 7.4) for 10 min with gentle shaking. Subsequently, sections were briefly washed in ddH2O to remove salts from PBS and mounted on SuperFrost glasses with Fluoromount (Sigma F4680).

Golgi–Cox staining in aged mice

Tissue preparation started 24 h after the end of extinction test in matured mice. For each genotype, 6 mice were used. Mice were anesthetized with 3% isoflurane and decapitated. Brains were quickly removed from the skull and washed with ddH2O to remove blood from the surface. Golgi–Cox impregnation of neurons was performed using the FD Rapid GolgiStain™ kit (FD NeuroTechnologies, # PK401)[68]. After a 3-week incubation, dye-impregnated brains were rapidly frozen in isopentane at -50 °C and then stored at -80 °C. For cryosection, brains were embedded in TissueTek O.C.T. compound (Sakura Finetek, # 4583) and coronally cryosectioned in 100-µm thickness and directly mounted on gelatin-coated slides (FD NeuroTechnologies, # PO101) with the help of solution C provided in the kit. Sections were stained according to the manufacturer’s protocol and mounted using the ROTI®Histokitt embedding medium (Carl Roth, # 6638)[68].

Spine imaging and deconvolution

For imaging, dendrites were selected from slices containing the dorsal hippocampal area. Images were acquired by an experimenter blinded to genotype information using a confocal laser-scanning microscope (LSM 700 and LSM 780 for samples from juvenile and aged brains, respectively; Carl Zeiss, Germany) and Zen software (Carl Zeiss, Jena, Germany). Secondary apical dendrites from CA1 pyramidal neurons were imaged for spine analysis.

For analysis of spines in juvenile brains, we used NT/Thy1-EGFP-M mice that express EGFP in a fraction of pyramidal cells. Z-stacks were collected with 0.21 µm interval using a 488 nm laser and a 63 × oil objective (NA = 1.4) with a 2.6 × optical zoom. The voxel size was 0.0644 × 0.0644 × 0.2065 µm. Deconvolution of images was performed using Huygens deconvolution software (Scientific Volume Imaging). The images were deconvolved using the “Classic Maximum Likelihood Estimation (CMLE)” algorithm implemented in Huygens software (Scientific Volume Imaging), set with 50 iterations, a quality threshold of 0.01, and a signal-to-noise ratio value of 25. A theoretical point spread function was used.

For analysis of spines in aged brains, Z-stacks were collected with 0.25 µm interval using a 405 nm laser, a 63 × oil objective (NA = 1.4), and 2 × digitally zoomed. The final voxel size was 0.066 × 0.066 × 0.250 μm. Images were rotated and cropped in ImageJ software to focus on target secondary apical dendrites and also set the root of each dendrite at the right edge of image for further analysis. To facilitate visual inspection of the Z-projections in Fig. 6A, the background was removed using the rolling ball method (radius = 30 px) and images were corrected using an unsharp mask filter (radius = 2 px, mask weight = 0.7) in Fiji software.

Spine density and morphology analysis

To identify, classify, and count GFP-labeled dendritic spines in juvenile mice, images were morphometrically analyzed using NeuronStudio software (CNIC, Mount Sinai School of Medicine, New York, NY, USA) and a custom-written Excel worksheet template to analyze the parameters provided by the NeuronStudio software. The analysis was performed by an experimenter blinded to group identity. Measurements started with some interval after the branching point after which the spine density appeared as stable. Spines along the dendrites were assessed using standard parameters for the distinction of stubby-, filopodia-like/thin-, and mushroom-type spines, as previously described [69, 70]. Parameters were set as suggested [69]. Only protrusions with a clear connection of the head of the spine to the shaft of the dendrite were counted as spines. In addition, a visual examination was also used to detect false ‘‘spine calls’’. This systematic approach was chosen to account for possible changes in spine distribution along dendrites.

Morphometric analysis of Golgi–Cox labeled dendritic spines in aged mice was performed as described previously [71,72,73,74,75]. Briefly, the images were semiautomatically analyzed using SpineMagick software [71] (the code is available in https://doi.org/10.5281/zenodo.6114928). To perform morphological analysis of images, we inverted the image intensity scale, and enhanced the contrast to 0.1% saturated pixels using ImageJ software. Next, spatial Gaussian blurring with Sigma(radius) = 1 was applied to decrease the level of noise. Subsequently, we performed image projections onto z plane, cropping the dendrite segments in the z-direction, to reduce the amount of projected artifacts from the background. For some images, it was possible to crop the entire dendrite by the single cube; in other cases, the dendrite segment was not straight in the z-direction and it was necessary to use few cubes to crop it properly. The spine density was calculated by dividing the number of spines by the length of the marked dendritic segment. Dendritic spines were counted manually by scrolling through the z-stacks of 3D images. That dendrite length value was interpolated using the SpineMagick.

Length (L), head width (H), and neck width (N) measured from SpineMagick were adopted for dendritic spine classification [76,77,78]. Spines were classified into mushroom spines (H/N > 1.3); stubby spines (H/N ≤ 1.3, and L/N ≤ 1.1); branched spines were excluded from quantitative morphological analysis; the rest spines were recognized as thin spines. The cutoff values were decided based on previous publications when genotype information was blinded [76,77,78].

Generation of agrin-expression vectors and adeno-associated viral (AAV) particles

Full-length (GeneID: 11,603) agrin constructs were obtained from Dharmacon (accession: BC150703). The DNA sequence corresponding to the 22 kDa C-terminus of agrin was used to induce spinogenesis and filopodia as previously described [7], while the 15 kDa C-terminus sequence was used as a control [10]. The cDNA was amplified using primers (sequences of the primers can be found in supplementary data, Table S2) and cloned into an AAV vector where the gene was expressed under the synapsin promoter and fused at the N-terminus of the red fluorescent reporter protein scarlet [79]. To secrete agrin fragments into the extracellular environment, we additionally cloned a secretion signal sequence from the receptor protein tyrosine phosphatase sigma at the N-terminus of the agrin sequence as described previously [80].

AAV particles were produced as previously described [81] with minor modifications. Briefly, HEK 293 T cells were transfected using the calcium phosphate method with an equimolar mixture of the expression plasmid, pHelper plasmid and RapCap plasmid DJ. After 48 h of transfection, cells were lysed using freeze–thaw cycles and treated with benzonase at a final concentration of 50 units/ml for 1 h at 37 °C. The lysate was centrifuged at 8000 g at 4 °C. The supernatant was collected and filtered with a 0.2-micron filter. The filtered supernatant was passed through pre-equilibrated Hitrap Heparin columns (Cat no. 17–0406-01; Ge HealthCare Life Science), followed by a wash with wash Buffer 1 (20 mM Tris, 100 mM NaCl, pH 8.0; filtered sterile). Columns were additionally washed with wash Buffer 2 (20 mM Tris 250 mM NaCl, pH 8.0; filtered sterile). Viral particles were eluted with elution buffer (20 mM Tris 500 mM NaCl, pH 8.0; filtered sterile). Amicon Ultra-4 centrifugal filters (100 kD cutoff) were used to exchange the elution buffer with sterile PBS. Finally, viral particles were filtered through a 0.22 μm syringe filter (Sigma-Aldrich, product no. Z741696-100EA), aliquoted, and stored at -80 °C until required.

AAV intrahippocampal injections

NT−/−/Thy1-EGFP-M± mice of both sexes were anesthetized at postnatal day P7 with 3% isoflurane (Baxter, Germany) delivered as a mixture with O2 through a vaporizer (Matrx VIP 3000, Midmark, Versailles, USA) and a custom-made mouse breathing mask that was a suitable size for P7 mice. The cranial skin was locally disinfected and incised, the skull was exposed by displacement of the skin and muscles, and a small hole was drilled into the skull at the injection site. Craniotomy was performed on both hemispheres using stereotaxic information for external landmarks on the skull, such as lambda and bregma and to other distinct landmarks, such as characteristic blood vessels of the bone and the brain [82], which had to be adapted to the smaller size of the young skull and brain. The following coordinates were used to target the CA1 area: ML, 1 mm and DV, 1.2 mm. A total of 500 nl of viral suspension (1.84 × 1011 particles/ml) was injected per hemisphere using a pulled glass micropipette (World Precision Instruments, WPI, glass capillaries with product no. 4878) and a nanoliter injector (WPI, Nanoliter 2010). To prevent backflow, the micropipette was left in the brain for 5 min before it was pulled out. The scalp was closed and sutured, and then, the animals were allowed to recover on a heated pad. P7 pups were separated from the mother for a maximum of 3 h to prevent them from being rejected.

Immunohistochemistry

Sample collection, animal perfusion, and tissue processing were performed as described above for spine analysis. For α3NKA, VGLUT1, and PSD95 immunolabeling, 40 μm free-floating sections were washed in PBS [3 × 10 min, at room temperature (RT) with gentle shaking] and incubated for 1 h (at RT with gentle shaking) in a blocking and permeabilizing solution containing 5% normal goat serum (NGS, Gibco, 16,210–064), 0.5% Triton X-100 (Sigma-Aldrich, T9284), and 0.1% Tween-20 (Roth, 9127.1) in PBS. Subsequently, slices were treated for 24 h (at 4 °C with gentle shaking) with the primary antibody in PBS containing 5% NGS, 0.5% Triton X-100, and 0.1% Tween-20. Anti-sodium potassium ATPase alpha 3 (mouse, dilution 1:250, XVIF9-G10, Novus Biologicals), anti-VGLUT1 (guinea pig, dilution 1:1000, 135,304, Synaptic Systems), and anti-PSD95 (mouse, dilution 1:500, Ab2723, Abcam) were used as primary antibodies. The slices were then washed 3 × for 10 min at RT in PBS containing 0.1% Triton X-100 and 0.1% Tween-20 (washing buffer) and incubated on a shaker for 3 h at RT with the secondary antibody. Secondary antibodies conjugated with Alexa 405 and 488 (Life Technologies) against the respective primary antibody were used with a dilution of 1:800 for Alexa 488 and 1:500 for Alexa 405. Afterward, slices were washed 3 × 10 min at RT with washing buffer and 1 × 10 min at RT with PBS, and then mounted on SuperFrost glass with Fluoromount (Sigma F4680).

Image capturing and analysis of immunohistochemical data and presynaptic boutons

To analyze the size of VGLUT1-positive puncta colocalizing or not with Ag22-Scarlet, three independent images were selected for counting. Images were acquired using a confocal laser-scanning microscope (LSM 700, Carl Zeiss, Germany) and Zen software (Carl Zeiss, Jena, Germany). ImageJ 1.46 software (NIH, USA) was used for image analysis. For each image, channels were separated (Image > Color > Split channels). Then, thresholds were manually adjusted for each channel (Image > Adjust > Threshold (using Yen and Over/Under functions)). Subsequently, binary maps were created (Process > Binary > Make binary), and VGLUT1 puncta were recognized automatically as particles greater than 0.02 μm2 in the VGLUT1 channel (Analyze > Analyze particles > 0.02-Infinity). The size of each ROI was measured, and ROIs were superimposed on the binary map of the Ag22-Scarlet channel. ROIs were divided into two populations: ROIs with or without colocalization with AAV-Ag22 particles.

To analyze axonal bouton density and size after AAV-Ag15/22 injections in Thy1-GFP mice, Z-stack images were acquired in dorsal hippocampal CA1 area using SP8 microscope (Leica, Germany) using a 488 nm laser and a 63× oil objective (NA = 1.4). The voxel size was 0.045 × 0.045 × 0.25 µm. ImageJ was used for image processing and analysis. For each image, about three axons with clear bouton structures were selected, and then, images were maximumly projected and cropped for each axon. After the Gaussian blur filter was applied (sigma: 1.0), axons were traced with the neuroanatomy module of ImageJ (Plugins > Neuroanatomy > SNT). Bouton width was automatically masked with ImageJ [Analyze > Local Thickness > Local Thickness (masked, calibrated, silent)]. Based on this mask, binary maps were created for en passant boutons more than 0.45 μm (Process > Binary > Make binary); the boutons were further masked automatically as particles greater than 0.2 μm2 (Analyze > Analyze particles > 0.2-Infinity). The size of each ROI was measured, and ROIs were superimposed on the “Local Thickness” mask to read bouton width data. The bouton density of each axon was computed as ROI counts per 10 μm axonal length.

Statistics

Statistical analysis of the results from behavioral tests in juvenile and matured mice was performed with SigmaPlot 13.0. A normality test (Shapiro–Wilk method) and an equal variance test (Brown–Forsythe method) were applied to determine which parametric test should be used. Grubbs' test (the extreme studentized deviate method) was applied to determine whether one of the values in the list is a significant outlier from the rest. For data obtained from repeated measures, a two-way RM ANOVA with the Holm–Šidák post hoc test was applied (CFC, LTP). For data not repeatedly acquired from many groups, two-way ANOVA with the Holm–Šidák post hoc test was applied (LTP, CFC associated spine analysis). For data collected from a single test based on novelty recognition (NOLT, NORT, TORT, Sociability, SRT), a two-sided paired t test was applied for analysis of exploring time. For other comparisons between two groups (discrimination ratios, datasets in OF, EPM, spine analysis in rescue experiment), a two-sided unpaired t test was applied. For comparison of datasets failed in the equal variance test (Figs. 6C, D, 7D, 8F (upper), Fig. 8G (upper), Fig. 8H (upper), Fig.S2B, and Fig.S4A), Welch's t test was applied. P < 0.05 was used to reject the null hypothesis. For comparison of cumulative distributions of dendritic spine parameters and axonal bouton analysis, the Kolmogorov–Smirnov (KS) test was used.

留言 (0)

沒有登入
gif