Development and Optimization of a Multilayer Rat Purkinje Neuron Culture

Neuronal Culture Preparation

All procedures were performed according to the National Institutes of Health Guidelines for the Care and Use of Laboratory Animals Norway (FOTS 20,135,149/20157494/20170001). Wistar Hannover GLAST rat pups (n = 328), embryonic day 18 (E18) to postnatal day 10 (P10), were used for neuronal culture preparation. Briefly, following anaesthesia and decapitation, the brains were rapidly transferred into preparation solution of ice cold EBSS (Gibco, #24,010,043) solution containing 0.5% glucose (Sigma, #G8769) and 10 mM HEPES (Gibco, #15,630,056). Under a dissection microscope, the meninges and medulla oblongata were carefully removed, and the cerebellum was separated from the pons and the midbrain. Depending on the culture, either only the cerebellum or the cerebellum including pons was transferred to a 15-mL tube containing 20 U/mL papain (Worthington, #LK003178) dissolved in preparation solution and warmed up to 36 °C. The tube containing the cerebellar tissue was placed into the incubator for 15 min at 36 °C with occasionally swirling to digest the tissue. The papain solution was carefully removed with a fire-polished Pasteur pipette, and the digestion reaction was stopped by addition of stop media (advanced DMEM/F12 solution (Gibco, #12,634,010) containing 0.5% glucose (Sigma, #G8769) and 10% foetal bovine serum (FBS, Gibco, #10,500,064)), pre-warmed to 36 °C. After 5 min of deactivation, the stop media was removed and 250 µL growth media containing 10% FBS per cerebellum was added. The tissue/media suspension was pipetted up and down with a fire-polished Pasteur pipette at least 100 times until cells were separated.

Support Cell Layer

To grow the support cell layer, first 375.000 cells per mL were isolated from the cerebellum and pons of embryonic (E18) or postnatal day 0 rat pups and seeded on coverslips pre-coated with poly-d-lysine (PDL; Neuvitro: #GG-12–1.5-PDL, 24 well, 500 µL/well; #GG-18–1.5-PDL, 12 well, 1 mL/well; #GG-25–1.5-laminin, 6 well, 2 mL/well). The support cell layer cultures were maintained in 6-, 12-, or 24-well plates in a medium based on the Furuya protocol [2] with modifications. The growth medium was made of 45% advanced DMEM/F12 solution (Gibco, #126,340,010), 45% NBM solution (Miltenyi Biotec, #130–093-570), 1.5% B-27 serum-free supplement (Gibco, #17,504,044), 1.5% NB-21 serum-free supplement (Miltenyi Biotec, #130–093-566), 1% sodium pyruvate (Invitrogen, #11,360,088), 1% heat-inactivated FBS (Invitrogen, #10,500,064), 2% Glutamax (Gibco, #35,050,038), 5 mg/mL D-glucose, and 10 mM HEPES (Invitrogen, #15,630,056). The media volume of each well depended on the plate size: 1500 µL for 6-well plates, 750 µL for 12-well plates, and 400 µL for 24-well plates. Half of the culture medium was replaced every 7 days.

Purkinje Neuron Layer

Neuronal cells can survive only within a narrow pH range. The effects of the extracellular and intracellular pH on numerous enzymes and channels is an extensive subject which is beyond the scope of this study. However, it has been found that acidification can inhibit important classes of synaptic channels such as VGCC [13], whereas alkalinisation increases neuronal activity [14]. Purkinje neurons and unipolar brush cells of the cerebellum are rich in acid-sensitive ion channels [15] and rich in proteins that control calcium and thereby activity. We have found that a neutral pH, in the slightly acid range of 6.8 to 7.0, was essential for cell survival of newly plated Purkinje neurons, and therefore, the support cell layer culture was fed 24 h prior to addition of the Purkinje neuron layer to ensure a pH in this range. E18- and P0-derived Purkinje neuron cultures were prepared by seeding 500.000 cells per mL from the vermis and the flocculus of the cerebellum onto support cell layers of different in vitro ages. The P10-derived Purkinje neuron culture was prepared by seeding 750.000 cells per mL from only the vermis of the cerebellum onto the support cell layers of different in vitro ages. Depending on the plate, a different volume of single-cell suspension was added to the support cell culture per well: 500 µL for 6-well plates, 250 µL for 12-well plates, and 100 µL for 24-well plates. The support cell layer growth media was supplemented with insulin (Invitrogen, #12,585,014; 1:250, stock 4 mg/mL), progesterone (Sigma, #P8783, 1:2000, stock 80 mM), insulin-like growth factor 1 (IGF1; Promokine, #E-60840, 1:40,000, stock 1 µg/µL), and protein kinase C inhibitor K252a (Alomone, # K-150; IC50 25 nM). In cultures that were maintained for more than 28 days in vitro, the IGF1 and progesterone concentrations were reduced to 10 ng/mL and 20 µM, respectively. K252a was supplemented for 21 days before the washout process started, its optimal concentration was experimentally determined for each culture type. In the first 10 days of culturing, the K252a concentration was 5 nM for E18-derived Purkinje neuron layers, 10 nM for P0-derived Purkinje neuron layers, and 25 nM for P10-derived Purkinje neuron layers. After 10 days, the K252a concentration was increased to 25 nM for all Purkinje neuron layers until day 21 in vitro. After 21 days in vitro (DIV), the washout phase of K252a was implemented by replacing half of the media with media that did not contain K252a. Half of the culture medium was replaced every 3.5 days for 6-well plates and every 2 days for 12- and 24-well plates. The experiments that determined different culture parameters and their impact on the Purkinje neuron yield were randomly performed, three to six times per experimental setting with five independent repeats for each group and condition.

Lentiviral Gene Editing

The full-length L7 promoter region (1005 bp) [16, 17] was custom cloned by SBI System Bioscience into construct pCDH-L7-MCS-copGFP (#CS970S-1), and viral particles with a yield of 2.24 × 10^9 infectious units per µL were produced. Two viral transduction approaches were tested. First, freshly prepared cells of E18 or P0 rat cerebellum (suspended in growth media containing no serum) were incubated for 10 min at 36 °C with 1.22 × 10^6 viral particles per mL cell suspension and were then seeded onto the support cell layer culture. To grow the transduced neurons, either 12-well plates containing coverslips or 35-mm µ-dishes for live-cell imaging (Ibidi, #80,136) were used. Media was replaced after 3 days, and transfection efficiency was evaluated by live-cell microscopy 24 h post transfection, and then daily until 21 DIV. At 21 DIV, the scan interval was changed from daily to every 3 or 4 days for cultures of E18 or P0 rat cerebellum, respectively, until 169 DIV. In the second experimental approach, the lentiviral transduction was performed on neurons in culture. Cells were fed on 15 DIV and 29 DIV, for cultures of E18 or P0 rat cerebellum, respectively, and 1 day later 2.5 × 10^6 viral particles per mL were applied to the culture media.

The neuronal morphology of GFP-expressing Purkinje neurons was analysed by capturing 10 independent 3 × 3 tile scans (Zyla camera configuration: 2048 × 2048; objectives: CFI Plan Apochromat Lambda dry objective 10 (NA 0.45, pixel size 603 nm) or dry objective 20 (NA 0.75, pixel size 301 nm)) on an Andor Dragonfly microscope system (Oxford Instruments). The viral transduction of the cultures was repeated three times.

Immunohistochemical Characterisation of Cell Types

Cultures were probed extensively by immunofluorescence microscopy to characterize their component cell types (see Table 1). Table 1 provides a detailed list of the used antibodies, indexing cell-specificity, concentration, antibody species, and staining conditions. Culture were washed with pre-warmed 0.1 M PBS (1xPBS; Gibco, #70,013,016) and fixed with 1.5–4% paraformaldehyde (pH 6–7.2; ThermoScientific, #28,908) containing 0.5% sucrose for 15 min at 36 °C. Tris-based or citric acid-based heat-induced antigen retrieval (pH 9 and pH 6, respectively; 45 min, 85 °C) [18] was performed for some targets (Table 1). The cultures were quenched with 1xPBS containing 50 mM NH4Cl (PBSN), permeabilised with 0.2% Triton X-100 (Sigma, #T9284) in PBSN (5 min, 36 °C), rinsed with PBSN containing 0.5% cold water fish gelatine (Sigma, #G7041; PBSNG, 3 × 15 min), and incubated with primary antibodies overnight at 4 °C in PBSNG containing 10% Sea Block (Thermo Scientific, #37,527), 0.05% Triton X-100, and 100 μM glycine (Sigma, #G7126). The coverslips were rinsed with PBSNG (three times, 20 min) and incubated with highly cross-absorbed donkey secondary antibodies conjugated to CF™ 488/594/647 dye (Biotium, #20,014, #20,115, #20,046, #20,015, #20,152, #20,047, #20,074, #20,075, #20,169, #20,170; 1:400) for 2 h at room temperature in PBSNG containing 2.5% Sea Block. To remove unbound secondary antibody, coverslips were rinsed with PBSN (three times, 20 min), and briefly dipped into MilliQ water before mounting onto microscope glass using Prolong™ Glass Antifade Reagent (Invitrogen, #P36981). The mounted cultures were kept 2 days at room temperature in the dark, followed by long-term storage at 4 °C until imaging.

Table 1 Primary antibodies. The signal to noise ratio for the antibodies were evaluated for the following conditions: 4% PFA at pH 7.2 diluted in 100 mM PBS; 1.5% PFA at pH 6 diluted in 100 mM natrium acetate buffer (NaAcB)); without heat-induced antigen retrieval (HIAGR); and with HIAGR either TRIS-based (pH 9) or citric acid-based (pH 6). The best conditions for each used antibody are described belowPurkinje Neuron Counting and Imaging

Purkinje neurons were visualized by immunohistochemical staining for calbindin (CALB1) or Purkinje cell protein 2 (PCP2) and counted manually by blinded investigators using a Leitz Diaplan fluorescence microscope (equipped with CoolLED pE-300white lamp). In addition, 10 Z-stack images were taken per coverslip (5 independent fields of view) on an Andor Dragonfly microscope system (Zyla camera configuration: 2048 × 2048; CFI Plan Apochromat Lambda objectives: S LWD water 40 × NA 1.14 (pixel size 151 nm), oil 60 × NA 1.20 (pixel size 103 nm), oil CFI SR HP Apo TIRF 100 × NA 1.49 (pixel size 60 nm)). The images were superimposed with the Fusion software (Oxford Instruments).

Dendritic Tree Branch Analysis

The dendritic development (length and order) of the Purkinje neurons was evaluated by analysing 10 Purkinje neurons per coverslip in 10 independent experiments using the ImageJ plugin Simple_Neurit_Tracer (SNT; Neuroanatomy) [19]. This free software plugin is available at http://fiji.sc/Simple_Neurite_Tracer. The plugin allows manually selection of points along tubular structures in confocal microscope z-stack images to trace the dendritic skeleton of neurons. The program marks the path between two branch points by determining the highest fluorophore emission intensity. This semi-automated process ensures that the right paths are chosen and traced. Each dendritic tree was evaluated with branching grades (orders) according to the Fujishima protocol [20].

The branch orders for each analyzed Purkinje neuron were determined as follows: Microscope Leica files (.lif) were converted to TIFF files and opened with the Fiji SNT plugin. The reconstruction of Purkinje neuron dendritic branching started with tracing of the primary dendrites by selecting the soma as the starting point and selecting the point of intersection between primary and secondary dendrite segments. Tracing non-primary dendrites was done by selecting a new point on the next dendrite intersection or the ending point of the dendrite. This was performed for the whole dendritic arbor of each Purkinje neuron. The data were exported to Excel where the average length of dendrites of the same order, as well as the number of segments of the same order, was evaluated for each analyzed Purkinje neuron.

3D Representation of Dendrites and Synapses

3D surface visualization of synapses was performed using Oxford Instruments analysis software IMARIS 9.3.1 (Bitplane) and the filament tracer tool [21]. The “surface” tool was applied to make a solid filled surface representation of the Purkinje neurons through the CALB1 or PCP2 staining, with the background subtraction option enabled. A threshold was selected that demarcated the neuron structure accurately while excluding background. For the synapse analyses, a 600 × 800 µm selection box was placed around the dendrite in each image, and surfaces were created for synapses within the selection box. The same threshold settings were used across all images, and individual surface data from each dendrite were exported for the 3D representation of the synapses.

Micro-electrode Array Recordings

E18 culture at a concentration of 500.000 cells per mL was plated onto 24-well plates of the Multiwell-MEA-System pre-coated with PDL (Multichannel Systems, Reutlingen). Each well contained 12 PEDOT-coated gold micro-electrodes (30 µm diameter, 300 µm space, 3 × 4 geometrical layout) on a glass base (#890,850, 24W300/30G-288). The amplifier (data resolution: 24 bit; bandwidth: 0.1 Hz to 10 kHz, modifiable via software; default 1 Hz to 3.5 kHz; sampling frequency per channel: 50 kHz or lower, software controlled; input voltage range: ± 2500 mV), stimulator (current stimulation: max. ± 1 mA; voltage stimulation: max. ± 10 V; stimulation pattern: pulse or burst stimulation sites freely selectable), and heating element (regulation: ± 0.1 °C) were integrated into the Multiwell-MEA headstage, which was driven by the MCS-Interface Board 3.0 Multiboot. The Multiwell recording platform was covered by a mini incubator (5% CO2; balanced air). Electrophysiological signals were acquired at a sampling rate of 20 kHz (Multiwell-Screen software). Plates were tested for spontaneous activity every second day from 5 DIV on. Raw voltage traces were recorded for 120 s to calculate spike rate and burst activity using offline MCS-Multiwell-Analyzer. The spontaneous activity of the culture was evaluated for two different conditions. In the first condition, cultures were maintained in Purkinje neuron culture media (45% advanced DMEM/F12 solution, 45% NBM solution, 1.5% B-27 serum-free supplement, 1.5% NB-21 serum-free supplement, 1% sodium pyruvate, 1% heat-inactivated FBS, 2% GlutaMAX, 5 mg/mL D-glucose, 10 mM HEPES, 16 µg/mL insulin, 25 ng/mL IGF1, 40 µM progesterone, 5 nM K252a) for 63 days. In the second condition, the Purkinje neuron culture media was exchanged to organotypic slice culture media [22] (30% advanced DMEM/F12 solution, 20% MEM solution (Gibco, #41,090,028), 25% EBSS solution (Gibco, #24,010,043), 25% heat-inactivated horse serum (Sigma, #H1138), 2% GlutaMAX, 5 mg/ml D-glucose, and 2% B-27 serum-free supplement) on the 28 DIV, and cultures were monitored for an additional 45 days.

Statistical Analysis

All conditions were tested in triplicate or quadruplicate, if not otherwise indicated. Data analysis and calculations were performed using the software Excel 2016 and Graph Pad Prism 7.0. Data are presented as means ± SEM. Statistical significance was determined using the non-parametric two-tailed paired Mann–Whitney’s U test. The level of significance is indicated with asterisks: *p < 0.05; **p < 0.01; ***p < 0.001.

留言 (0)

沒有登入
gif