Animal subjects
CD1 mice of both sexes were used throughout this study. Animals were housed under controlled conditions of temperature (21–23°C) and light (LD 12:12) with free access to food (AO4 standard maintenance diet, SAFE, Épinay-sur-Orge, France) and water. Care and handling of animals were performed according to the Guidelines of the European Union Council (2010/63/EU), and the Spanish regulations (65/2012 and RD53/2013) for the use of laboratory animals.
The number of animals used in each experiment was determined based on previous studies (17–19). Adult male mice were randomized during the first week after birth by cross-fostering and used when they became two months old. The protocol used has been authorized by the Ethics Committee of the “Consejería de Agricultura, Ganadería, Pesca y Desarrollo Sostenible de la Junta de Andalucía”, in Spain with the approval numbers 15/01/2020/002, 04/03/2020/033 and 10/03/2020/039. All studies involving animals are reported in accordance with the ARRIVE guidelines 2.0 for reporting experiments involving animals (20, 21). The number of animals used in each experiment is indicated in the figure legends.
Design of the study
Mice used in the study were lesioned in the primary motor cortex by generating controlled mechanical injuries while anesthetized by a cocktail of ketamine (100 mg/kg) and xylazine (20 mg/kg) For the migration studies, mice were injected with lentiviral vectors prior to the injury in the same surgical act. Once injured, vehicle consisting on saline solution was used in control animals and diterpene EOF2 was used as treatment. Either vehicle or EOF2 were administered daily by intranasal infusions for either 14, 28, or 56 days as we describe in the paragraphs below. Upon the completion of the treatments, mice were anesthetized with a cocktail of 100 mg/kg ketamine and 20 mg/kg xylazine and cerebrospinal fluid (CSF) was extracted as explained below, then a dose of Dolethal® (Ventoquinol, Lure, France) containing a lethal 50 mg dose of pentobarbital to euthanized the animals was applied followed by either brain perfusion with paraformaldehyde (for histological studies) or brain extraction (for molecular biology and studies). See description of the different procedures below. In the case of electrophysiological studies mice were anesthetized with a lethal dose of anesthetic previous to perfusion with artificial CSF.
In all in vivo experiments, the experimental groups used were saline treated (control) or EFO2 treated mice (EOF2). In postmortem studies, the experimental unit was the single animal and sample size for each experiment is indicated in the figure legends. In electrophysiological studies the experimental unit was the single cell and the sample size is indicated in supplementary tables S4-S7.
Injection of ZsGreen expressing lentiviral vectors in the SVZ and unilateral mechanical cortical brain lesions.
Controlled unilateral mechanical cortical brain injuries were performed in the primary motor cortex of the right brain hemisphere of anesthetized mice. Mice were anesthetized using an anesthetic mixture composed of ketamine (100 mg/mL) and xylazine (20 mg/Kg) in sterile physiological saline. Using a stereotaxic frame (Harvard Apparatus), a longitudinal incision was made in the skin of the head to expose the skull and proceed to inject the ZsGreen expressing lentiviral vector. At Bregma/-0.8 mm, a small craniotomy was performed using a hand drill and a 0.9 mm drill bit (Meisinger, Neuss, Germany). Then a Hamilton Gastight syringe (Hamilton Company, NV, USA) was introduced attached to the stereotaxic device through the incision made, and 1 µL of the ZsGreen virus was injected into the lateral ventricle at a rate of 0.1 µL/min. The lentiviral construct was produced by us as described previously (14). HEK Lenti-XTM 293T were used as packaging cell lines to produce lentiviral supernatant as previously described (22). Cells were co-transfected with the pHRSincPPT- SEW transfer vector expressing the green fluorescent protein ZS-Green, together with plasmids pCMV∆R8.91, coding for HIV-1 GAG / POL proteins and pMD2.G for pseudotyping with the Vesicular Stomatitis Virus G protein (VSVG). Cells were transfected in OptiMEM™ medium (Thermo Fisher Scientific Inc., Carlsbad, CA, USA) by polyethylenimine (PEI)-mediated transfection (23) and after one hour the medium was replaced by DMEM supplemented with 10% fetal calf serum (GIBCO; www.thermofisher.com/gibco). Supernatants were collected at 48 and 72 h, centrifuged at 2100 g for 5 minutes to remove cell debris and subjected to two concentration rounds using Lenti-X™ Concentrator (Clontech;Mountain View, CA, USA) to obtain a clean high-titer virus-containing pellet. Briefly, viral supernatants were incubated for 30 min at 4°C with 3 volumes of Lenti-X concentrator reagent, centrifuged at 1500×g for 50 min and the pellet resuspended in 1 mL PBS. Lenti-X was further added and upon 30min incubation at 4ºC and an additional centrifugation, the pellet was snap frozen in liquid nitrogen and stored at -80°C until use. Viral titers were determined, by evaluating their efficiency in transducing Jurkat cells by means of a Cyto-flex™ flow cytometer (Beckman, Indianapolis, IN) 48h after transduction. Viral titers were always above 2x105 transducing units (TU) per mL.
In the same surgical act, mice were unilaterally lesioned in the right hemisphere of the primary motor cortex. They were craniotomized with a manual drill at -1.1 mm rostral and + 1.5 mm lateral to Bregma. Thereafter, a controlled mechanical lesion was performed in the underlying primary motor cortex using a manual drill (0.9 mm diameter). This drill was allowed to penetrate 1 mm below the bone surface. Mice were injured and placed into a controlled cage during the required days post injury (dpi) that depended on the treatment and experimental design. Lesions were performed unilaterally; the injured hemisphere was considered the ipsilateral side, while the intact hemisphere was considered the contralateral hemisphere and was used as a control. After performing the injury, craniotomies were sealed by surgical cement (Fisher Scientific) and the incision made in the skin was sutured. Subsequent analgesic and aseptic measures were taken to ensure animals welfare. This procedure was previously stablished by our research group and has been used elsewhere (6, 9, 14, 24).
Intranasal administration of EOF2
EOF2 (CAS number 2230806-06-9) was produced by us as previously described (9) and was delivered intranasally while the animal was placed in a standing position with an extended neck as previously described (25). Eighteen microliters of each solution (5 µM EOF2 in saline, or saline as vehicle) was delivered over both nasal cavities alternating 3 µL/each using a micropipette. Mouse was maintained in this position for 10 additional seconds to ensure all fluid was inhaled. In all experiments, mice were coded and treatment (vehicle or EOF2) was assigned randomly to code numbers and applied. The treatment was administered daily until 90 dpi.
Brain processing for immunohistochemistry studies
At the end of the treatment, brains were perfused with paraformaldehyde (PFA) and sliced using a cryotome into 30 µm sections. Immunohistochemistry was performed as previously described (18, 24, 26). See antibodies in supplementary tables S1-S3.
Brain slices obtention for electrophysiological studies
Brain slices were acquired from previously treated mice by anesthetizing them and perfusing with a modified artificial CSF (ACSF) or cutting solution. Following perfusion, the brain was swiftly extracted, and after removal of the cerebellum and rostral telencephalon, coronal slices of 300 µm thickness were obtained using a vibratome (Leica VT1000S, Leica Biosystems, United Kingdom). These slices were then incubated in a chamber containing cutting solution at 34º C for 10 minutes, followed by transfer to another chamber filled with recording solution at room temperature for at least 1 hour before further use. The composition of the different ACSF used was as follows (data in mM): i) Recording solution: 126 NaCl, 2 KCl, 1.25 NaH2PO4, 26 NaHCO3, 10 glucose, 2 MgCl2, and 2 CaCl2; ii) Cutting solution: 92 NMDG, 2.5 KCl, 1.2 NaH2PO4, 30 NaHCO3, 20 HEPES, 25 Glucose, 4 MgCl2, 0.1 CaCl2. Both solutions were bubbled with 95% O2 – 5% CO2 (pH 7.4, adjusted with HCl; 295–305 mOsmol/kg).
Whole-Cell patch clamp recordings and analysis.
After one hour of incubation in the holding chamber, the slices were transferred to the recording chamber of the microscope. To visualize the cells, a Nikon Eclipse FN1 microscope equipped with infrared differential interference contrast (IR-DIC) optics, a 40x water immersion objective and an infrared camera WAT-902H2 is used. In the holding chamber, the slices were constantly perfused with ACSF aerated at room temperature and at a rate of 1 mL/min using a peristaltic pump (Harvard Apparatus MPII, Holliston, MA, USA). The micropipettes used to perform the patch-clamp were obtained from borosilicate glass capillaries (od 1 mm, id 0.58 mm, length 10 cm, Sutter) stretched with a vertical puller (PC-10, Narishige, Tokyo, Japan) adjusted to get a resistance between 3–6 MΩ. The micropipettes were filled with a K-gluconate based solution with the following composition (in mM): 120 K-Gluconate, 10 KCl, 10 phosphocreatine disodium salt, 2 Mg-ATP, 0.3 Na-GTP, 0.1 EGTA, 10 HEPES. pH was adjusted to 7.3 using KOH and the osmolality to 285 mOsmol/kg with sucrose with the help of an osmometer (Osmomat 300, gonotec). To perform the recordings, the micropipettes were placed using a micromanipulator (MP-225, Sutter Instrument, CA, United States) in the injured area. In this area, Zs-Green labelled cells were identified using the fluorescence microscopy system coupled to the Nikon microscope (see fig S1A). To achieve the whole-cell patch-clamp configuration, we use an amplifier (MultiClamp 700B) and a Digidata 1550 analog-to-digital converter (Axon Instruments, Molecular Devices, Sunnyvale, CA, United States). Recordings were acquired with the pCLAMP 10.4 software (Molecular Devices), lowpass Bessel-filtered at 3 kHz and the data were digitized at 20 kHz. For the data analysis, Clampfit 10.4 software (Molecular Devices). Series resistances was typically 10–20 MΩ, and the experiments were discarded if higher than 25 MΩ. Liquid junction potentials were compensated automatically.
Current clamp studies. In this study, both passive and active membrane properties of the cells were studied as previously described (27). In brief, the resting membrane potential was calculated by subtracting intracellular from extracellular potential after the removal of the recording electrode. Input resistance was measured through the injection of hyperpolarizing and depolarizing square current pulses (500 ms, 1 Hz) with 10 pA increments between each one, and then calculated as the slope of the current-voltage relationship, following Ohm's Law. Rheobase, defined as the minimum intensity of current necessary to provoke an action potential, was determined by applying square pulses of 100 ms, 1 Hz, with 10 pA increments. Voltage threshold and depolarization voltage were computed relative to the resting membrane potential. To ascertain the spike threshold, action potential recordings were differentiated, with the spike onset identified as the membrane potential at which the first derivative exceeded 10 V/s (28, 29).
Action potential amplitude and duration were calculated based on peak voltage and width at half amplitude. Repetitive firing properties were assessed by applying depolarizing current steps (1 s, 0.5 Hz) with 20–50 pA increments. Maximum firing frequency was defined as the highest number of spikes achieved during repetitive discharge, regardless of current intensity, while frequency gain was determined as the slope of the relationship between firing frequency and applied current. Cancellation current represented the intensity at which the neuron ceased firing during maximal discharge.
Voltage clamp studies. Voltage-dependent currents were elicited by 50 ms square depolarizing pulses ranging from − 60 to + 40 mV, in 10 mV steps. No leak substraction was performed. Current amplitudes were measured in the peak for inward currents and at the end of the pulse for outward currents. Conductances were calculated as chord conductance (30). Thus, G = I/(V-VE), being G, conductance, I, the measured current, V, command voltage and VE the theorical Nernst potential for potassium (outward currents) or sodium (inward currents). TTX (Tocris), TEA, (Sigma-Aldrich) and 4-AP (Sigma-Aldrich) were diluted in the bath solution for blocking the different conductances.
To investigate spontaneous postsynaptic currents (sPSC), 60-second continuous recordings were conducted following previously described methods (31). The holding potential was clamped at 0 mV for spontaneous inhibitory postsynaptic currents (sIPSCs) or -60 mV for spontaneous excitatory postsynaptic currents (sEPSCs). Synaptic events were detected and analyzed using EasyElectrophysiology software. A template was created by fitting a function to a single event, and subsequent events were extracted using varying detection thresholds. False positives were manually removed, and real events were fitted with a biexponential function to determine percentage appearance, frequency (events per minute), and amplitude (baseline to peak). To assess the nature of synaptic events, CNQX (50 µM), APV (20 µM), and SR95531 (gabazine, 20 µM) were used, all of which were purchased from Tocris. The protocol involved initially superfusing each slice with normal ACSF for control recordings, followed by ACSF containing the drugs to record voltage responses.
Morphological study of the newly generated neurons.
To carry out the morphometric study of the new generated cells, dye-filling technique was used. For this purpose, iontophoretic injection of 0.2% neurobiotin (Vector Laboratories, Burlingame, CA, USA) contained in the internal pipette solution was carried out by applying current steps of 400 pA of 500 ms at 0.5 Hz for 20 min (32). Slices containing labeled cells were deposited in a 4% paraformaldehyde solution at 4º C overnight and then transferred to 30% sucrose in phosphate buffer at 4º C to maintain them. Thereafter, dye-filled neurons were revealed with a goat-polyclonal antibiotin Texas Red conjugated antibody (1:900) from Rockland (Pennsylvania, USA). A Zeiss LSM 900 Airyscan 2 confocal microscope was used for the visualization and reconstruction of the labeled cells. Stacks of 30–70 photographs were performed, using a 1 µm interval. Images were processed with ZEN 3.2 Blue Edition software and the cells were reconstructed using the Neurolucida 360 version 2020 3.1 system (MicroBrightField, Williston, VT, United States).
Analysis were made according to previous works (32, 33). The quantitative morphometric data presented in this study were generated using Neuroexplorer software. Each dendritic segment was systematically assigned an order in a centrifugal manner, from the soma to the terminal segments (see supplementary Fig. S1). A node, or intersection, was defined as any bifurcation along the dendrite. Furthermore, a segment represented the portion of the dendrite linking either the soma to an intersection or two intersections, while a terminal segment denoted the part connecting the last branching point to the dendrite's terminal ending. This study focused on three key aspects of neurons. Firstly, neuron surface area was examined, with total dendritic surface area defined as the sum of each dendrite's surface area and total surface area calculated as the sum of somatic and dendritic areas. Secondly, dendritic length was investigated, specifically targeting total dendritic length, which encompasses the sum of individual dendrite lengths. Lastly, neuronal complexity was analyzed. Branch order was scored to measure dendritic complexity, representing the highest order reached in each neuron. Additional complexity measures included the total number of segments, representing the entirety of segments within a neuron; the number of nodes, indicating the number of intersections; and the number of terminals, defined as the sum of terminal segments within a neuron. Sholl diagrams were constructed for the reconstructed cells. Neurons were oriented along dorsal and lateral axes, with the soma positioned at the center of concentric circles progressively increasing by 50 µm in radius. The number of dendrites intersecting each circle was recorded as part of the analysis.
The experimental animal groups for electrophysiological recordings and morphometric analysis were as follows: Group 1 consisted of animals analyzed at 7–14 dpi, Group 2 at 15–28 dpi, Group 3 at 29–56 dpi, Group 4 at 57–90 dpi, and Group 5, or control group, comprised pyramidal neurons from layer V recorded from the contralateral side to the injury in 3-month-old animals.
RNA isolation, reverse transcription and real-time quantitative PCR
For RT-qPCR analysis, RNA was isolated from the SVZ; intact SVZ were processed for RNA extraction using the TRIzol (Cat. 15,596,026, Invitrogen, Carlsbad, CA, USA), separation method, following the manufacturer’s instructions and resuspended in purified nuclease-free water. RNA was quantified using a BioTek’s Synergy Mx fluorimeter (BioTek Instruments, Inc, Winooski, VT, USA). cDNA was prepared from 500 ng RNA using iScriptTM cDNA Synthesis Kit (Cat.1708890, Bio-Rad Laboratories Inc, Hercules, CA, USA) on a Techne Genius thermal cycler (Techne Ltd., Cambridge, UK). The 15 µl RT-qPCR reaction mix contained 7.5 µl 2X iTaq Universal SYBR Green Supermix (Cat. 1,725,122, Bio-Rad Laboratories Inc, Hercules, CA, USA), 10 nmol of both the forward and the reverse primers, and 1 µl of the sample. The PCR thermal profile included 40 cycles of denaturation at 95°C for 10 s, an annealing temperature according to each set of primers for 15 s, and extension at 72°C for 20 s, followed by a melting curve analysis. Each sample was analyzed in triplicate. The mRNA level of rRNA18S was used as internal control. Relative quantification values of mRNA expression were calculated as 2 (Livak Method). Oligonucleotides used in this study were designed by BLAST and were obtained from Merck (Madrid, Spain). Primer sequences (5 ́-3 ́) for detecting expression of mouse mRNA were the following: for NRG1, FW: CGCTGTTCTGGTCTCATCCG, RW: GCGGTGGAGTGGAGTGTAAG; for ErbB4, FW: TACCTCCTCCCATCTACACATCC, RW: CCTCTGGTATGGTGCTGGTTG.
SVZ Cell Isolation and Culture
NPCs were obtained from the SVZ of 7-day postnatal mice following the same procedure described in (19). Neurosphere cultures were maintained in defined medium (DM) composed of Dulbecco’s modified Eagle’s medium/F12 medium (1:1vol/vol) plus 1mg/L gentamicin (GIBCO) and the B27 supplement (Invitrogen, Carlsbad, CA). EGF (20ng/mL) and bFGF (10ng/mL; both from PeproTech, Frankfurt, Germany) were added to DM for culture expansion.
Neurosphere Differentiation Assay
Neurosphere cells were centrifuged, resuspended in defined medium without growth factors, and seeded. NRG1 (R&D Systems) at different concentrations (1, 5 or 10 ng/mL) were added and cells were maintained for 72 hours before being fixed for immunocytochemistry.
Cloning of human TGFα and NRG1 cDNA fused to eGFP and mCherry
Full-length cDNA encoding the membrane-bound isoform of human pro-neuregulin-1 β1-type (NRG1, NCBI reference sequence: NP_039250.2) with mCherry cDNA inserted between nucleotides 93 and 94 of NRG1 open reading frame was cloned into pEGFP-N1 to add EGFP cDNA to the 3′ end. Construct was synthesized by GeneCust (Boynes, France) to generate the mCherry-NRG1-GFP construct.
Time-lapse experiments and fluorescence analysis of recombinant mCherry-NRG1-eGFP protein in the culture medium of NPC
NPC were plated in µ–dishes (35 mm high; Ibidi) and transfected with mCherry-NRG1-eGFP construct. After overnight incubation, cells were left for 30 min in serum-free Fluorobrite DMEM (Thermo Fisher Scientific) and used either in time-lapse experiments. Cells were treated with EOF2 compound (5 µM) and images were taken every 2 minutes.
Statistical Analysis.
The data and statistical analysis comply with the recommendations on experimental design and analysis in pharmacology (34). All statistical analyses were conducted on raw data. Results are presented as the mean ± standard error of the mean (SEM), where 'n' denotes the number of cells or animals included. Statistical calculations were performed using GraphPad Prism software. Initially, the normality of the data distribution was assessed using the Shapiro-Wilk test. For comparisons of means between groups, a repeated measures analysis of variance (ANOVA) was applied. If significant differences were detected, the Tukey test was used for pairwise comparisons between groups. When comparing only two unpaired experimental groups, the Student's t-test was employed. To determine statistically significant differences in the percentage of cells firing action potentials, repetitive discharge, or synaptic inputs between groups, the Chi-square test for independence was used. If the expected frequencies were small, Fisher's exact test was utilized. A 95% confidence interval was applied in all analyses, and groups were considered statistically different if p ≤ 0.05. In all figures and supplementary tables, asterisk (*) indicates statistical differences between groups, and crosses (†) indicate differences between the various groups and the 57–90 dpi group.