Login processing...

Trial ends in Request Full Access Tell Your Colleague About Jove


A Scalable Method to Study Neuronal Survival in Primary Neuronal Culture with Single-cell and Real-Time Resolution

Published: July 26, 2021 doi: 10.3791/62759
Ángela Rodríguez-Prieto*1, Ana González-Manteiga*1, Yaiza Domínguez-Canterla1, Carmen Navarro-González1, Pietro Fazzari1
* These authors contributed equally


Neuronal loss is at the core of many neuropathologies, including stroke, Alzheimer's disease, and Parkinson's disease. Different methods were developed to study the process of neuronal survival upon cytotoxic stress. Most methods are based on biochemical approaches that do not allow single-cell resolution or involve complex and costly methodologies. Presented here is a versatile, inexpensive, and effective experimental paradigm to study neuronal survival. This method takes advantage of sparse fluorescent labeling of the neurons followed by live imaging and automated quantification. To this aim, the neurons are electroporated to express fluorescent markers and co-cultured with non-electroporated neurons to easily regulate cell density and increase survival.

Sparse labeling by electroporation allows a simple and robust automated quantification. In addition, fluorescent labeling can be combined with the co-expression of a gene of interest to study specific molecular pathways. Here, we present a model of stroke as a neurotoxic model, namely, the oxygen-glucose deprivation (OGD) assay, which was performed in an affordable and robust homemade hypoxic chamber. Finally, two different workflows are described using IN Cell Analyzer 2200 or the open-source ImageJ for image analysis for semi-automatic data processing. This workflow can be easily adapted to different experimental models of toxicity and scaled up for high-throughput screening. In conclusion, the described protocol provides an approachable, affordable, and effective in vitro model of neurotoxicity, which can be suitable for testing the roles of specific genes and pathways in live imaging and for high-throughput drug screening.


The study of neuronal disorders requires experimental models that are amenable to genetic, molecular, and cellular analyses. Primary cortical neurons are a very potent model for studying neuronal survival and toxicity1,2,3,4. Under the appropriate conditions, primary neurons will progressively develop synaptic contacts and present hallmarks of mature neurons. Therefore, this model is more reliable than immortalized cell lines in modeling the physiology of the neurons and more amenable to manipulations than animal models. However, in comparison to cell lines, primary cortical neurons are difficult to transfect and relatively fragile. Moreover, the intricate morphology of cortical neurons may complicate the imaging and analysis in high-density cultures. Conversely, low-density primary neuronal cultures are easier to analyze but tend to be more fragile.

Taking all of this into account, this paper presents an affordable, versatile, and scalable in vitro workflow to model and investigate neuronal survival (Figure 1). Key points of this protocol are i) the method of in vitro electroporation of the neurons with a fluorescent marker to allow sparse labeling and imaging of live cells; ii) its versatility in different models of cytotoxicity; and iii) the semi-automated or automated analysis.

The electroporation system (see the Table of Materials) provides an open and inexpensive procedure that does not involve kits and specific solutions4,5,6. Moreover, this method can be easily adapted to obtain an optimal transformation efficiency and density of transfected neurons, mixing electroporated with non-electroporated cells. The co-culture with naïve (non-electroporated) neurons significantly improves the viability of the electroporated cells compared to low-density cultures. Moreover, it enables sparse labeling with an adjustable density of the electroporated neurons, maintaining a consistent level of gene expression. It is important to aim for 3-5% of electroporated neurons in survival assays.

Having sparse labeling essentially facilitates the automation of the image analysis because the cells are well-separated and easily distinguishable. Notably, this experimental paradigm may be adapted to test multiple genes of interest by simultaneously co-culturing neurons co-electroporated with different cytosolic fluorescent markers (e.g., cyan fluorescent protein, red fluorescent protein (RFP), green fluorescent protein (GFP)) in the same well. Similar to other cytotoxicity assays (e.g., propidium iodide or lactate dehydrogenase), the assay described herein is based on the fact that neuronal death is accompanied by rupturing of the cell membrane. This provokes the release of cytosolic proteins by diffusion and, consequently, the loss of GFP fluorescence.

An in vitro model of stroke, namely the OGD model, is presented as an example of neurotoxicity7,8,9. This protocol entails exposing the primary neurons to a salt buffer similar to an artificial cerebrospinal fluid but deprived of oxygen and glucose. Although this model has been presented as an example of neurotoxic stress, different cytotoxic conditions can be tested with the same workflow8,9. Finally, sparse labeling easily enables the development of automated imaging analysis. Here, a protocol was established based on standard immunofluorescence and ImageJ analysis in a smaller setup. Next, this workflow was adapted and scaled up using a cell imaging system that allows an automatized analysis in live imaging in mid- and high-throughput modes. In conclusion, this paper presents a flexible, affordable, and scalable methodology to study neuronal survival in different experimental models of toxicity using live imaging and automated quantification.

Subscription Required. Please recommend JoVE to your librarian.


All procedures using animals should be supervised by the bioethical animal committee of the institute and performed in compliance with local regulations. The procedures presented herein were approved by the delegated authority and comply with the regulations in Spain and Europe.

1. Primary neuronal culture

NOTE: All the steps are performed inside the culture hood, using sterile materials and solutions to maintain sterile conditions.

  1. Poly-L-lysine (PLL) coating
    1. Wash the coverslips (CVs, 15 mm) in 70% ethanol overnight (o/n) in an orbital shaker.
      NOTE: The speed of the orbital shaker should be enough to move the CV gently but not too fast to avoid damaging or scratching the CVs. The specific RPM will vary depending on the model of the shaker. CVs can be stored in ethanol for long periods (months).
    2. Remove the ethanol and wash twice with deionized water for 10 min each.
    3. Let the CVs dry out at room temperature (RT) in a 100 mm Petri dish (13 CVs/plate approximately). Ensure that the CVs do not touch each other or the edges of the plate.
      NOTE: Dry CVs can be stored at RT for long periods (months).
    4. Prepare the 5 mg/mL PLL stock solution in water, make aliquots, and store them at -20 °C. Dilute the stock solution in 1x PBS to a working concentration of 0.1 mg/mL.
      NOTE: Higher concentrations of up to 1 mg/mL PLL may be tested.
    5. Add the minimal volume of PLL to cover the surface of the CVs (see optimal volumes of PLL in Table 1). Add the PLL solution dropwise to facilitate the formation of a droplet.
      NOTE: Plating neurons on CVs is required for immunofluorescence. For time-lapse imaging (e.g., for the IN Cell), cells are plated directly over the well without CVs because the CVs may move during the time-lapse. In that case, add the minimal volume of PLL to cover the surface of each well.
Surface PLL volume (mL)
Coverslips 15 mm 0.2
Coverslips 13 mm 0.18
96 well plate 0.07
24 well plate 0.3
6 well plate 2
60 mm plate 4.5

Table 1: Optimal volumes of poly-L-lysine according to the surface area.

  1. Incubate for a minimum of 1 h at 37 °C.
    NOTE: Incubation period could be as long as o/n.
  2. Remove the PLL from the CVs. Wash the CVs twice with deionized water for 10 min each. Let the CVs dry at RT.
    NOTE: The PLL solution can be stored at 4 °C for reuse (3 times maximum).
  3. Store the dried CVs inside a Petri dish at 4 °C (for up to 7-10 days) or place them inside the wells with plating medium (see composition in Table 2 and working volumes in Table 3). Add plating medium directly to the wells of 12-well plates (Table 3).
    NOTE: Neurons are very sensitive to the quality of the fetal bovine serum (FBS). Different batches of FBS should be tested to identify the batch in which the cells look healthier. Once a good lot of FBS is identified, it is recommended to obtain a large amount and store it at -80 °C. This protocol could be adapted for other types of CVs and plates (see the different working volumes in Table 3).
Plating Medium
MEM 42.5 mL
10% Horse serum or FBS* 5 mL
30% glucose (0,6% final concentration) 1 mL
PS (penicillin, 10000 U/mL; streptomycin, 10 mg/mL); 0.5 mL
Electroporation medium
Neuronal Medium
Neurobasal medium 48.5 mL
B-27 1 mL
Glutamax 200 mM 0.125 mL
PS (penicillin, 10000 U/mL; streptomycin, 10 mg/mL); 0.5 mL
OGD Medium (Tasca, et al. 2014)
CaCl2 1 mM
KCl 5 mM
NaCl 137 mM
KH2PO4 0.4 mM
Na2HPO4 0.3 mM
MgCl2 0.5 mM
MgSO4 0.4 mM
NaHCO3 4 mM

Table 2: Composition of the different media used.

Surface area of wells and plates Working Volume
96 well plate 0.32 cm2 0.1 mL
24 well plate 1.9 cm2 0.5 mL
12 well plate 3.8 cm2 1 mL
6 well plate 9.5 cm2 2.5 mL
60 mm plate 21.5 cm2 6 mL

Table 3: Surface area of the different types of plates and the working volumes.

  1. Preparation of the neuronal culture
    1. Prewarm (37 °C) and equilibrate (in a 5% CO2 atmosphere) the plating medium, neuronal medium, and electroporation medium (see composition in Table 2) in a partially opened bottle inside the cell incubator for a minimum of 2 h. Let the trypsin warm up at RT. Pre-warm the 1x Hank's Balanced Salt Solution (HBSS) at 37 °C in a water bath.
    2. Preset the electroporation parameters (Table 4).
Length V Interval N D. Rate Polarity
2 ms 175 50 2 10 plus
Length V Interval N D. Rate Polarity
50 20 50 5 40 plus/minus

Table 4: Electroporation settings for NEPA21.

  1. Brain dissection and harvesting CD-1 mouse embryos at E15 post-coitum
    1. Sacrifice the pregnant female by cervical dislocation without sedation. Sterilize the belly with 70% ethanol; cut and open the abdomen with sharp scissors.
    2. Extract the uterus and place it in a Petri dish with cold 1x HBSS. Leave it on ice inside the cell culture hood with common forceps.
    3. Extract an embryo from the uterus with the aid of forceps and scissors and decapitate it with a single cut using scissors.
    4. Hold the head and introduce forceps (e.g., Dumont fine forceps) into the eye orbit.
    5. Using the needle (30 G needle syringe) as a blade, cut the bregma and break through the posterior part of the head.
    6. Introduce sharp forceps through the incision vertically into the bregma, and break the thin layer of skin of the anterior region. Pull the forceps to follow the longitudinal fissure between the two hemispheres without damaging the cortex.
    7. Use the small forceps at a 45° angle, from the anterior to the posterior part, to remove the brain from the head.
    8. Move the brain to a 3.5 cm Petri dish with ice-cold 1x HBSS under a stereomicroscope.
  2. Dissection of cortices
    NOTE: The protocol for neuronal culture is adapted from the procedure described by Dotti and Banker10.
    1. Use the Dumont forceps (11 cm) to hold the thalamus and a needle tilted at a 45° angle to separate the cortex of each hemisphere from the midbrain. Ensure that the cortex does not show traces of other tissues; use sharp forceps to remove the meninges.
      NOTE: The meninges form a thin elastic membrane enveloping the brain and are easily recognized because of their red color. Making soft movements in zig-zag will help remove the meninges. Make sure all the meninges have been removed as they are toxic for the culture.
    2. Move the clean cortex to a clean 3.5 cm Petri dish with ice-cold 1x HBSS and place it on ice.
  3. Disaggregation
    1. Chop each cortex into two halves and use forceps or a sharp needle (30 G) to facilitate the effect of trypsin in the next step.
    2. Transfer the pieces to 7 mL of trypsin-EDTA buffer at room temperature in a 15 mL tube using a P1000 pipette (cut off 3-4 mm of the tip). Wait until they sink to the bottom of the tip before releasing them from the pipette tip. Mix by tilting.
    3. Incubate at 37 °C for 12 min in a water bath. Mix by tilting every 4-5 min.
      NOTE: Incubation time could vary from 12 to 15 min.
    4. Let the tissue sink by gravity and carefully remove the trypsin. Wash the tissue twice with 8 mL of 1x HBSS and agitate gently by tilting. Remove the 1x HBSS carefully using a P1000 pipette but using an uncut pipette tip.
    5. Wash the tissue once with 5 mL of plating medium. Agitate by tilting. Remove the plating medium carefully using a P1000 pipette.
    6. Add 1.5 mL of plating medium. Disaggregate the mix mechanically using a P1000 pipette by aspirating the medium from the bottom of the tube and releasing it gently over the walls for a maximum of 8-12 times.
    7. Filter through a sterile cell strainer (70 µm) placed on a 50 mL tube. Add 10 mL or other desired volume of plating medium to the tube.
    8. Introduce 5 µL of the cell suspension into the cell counting chamber, add 5 µL of trypan blue (0.4% w/v), and count the cells.
    9. Plate the cells (see section 1.6) by transferring an adequate volume of the cell suspension to have the desired number of cells for electroporation (e.g., 106 cells) in a 15 mL centrifuge tube. Centrifuge the cells at 250 × g for 5 min. Proceed to section 1.6.
  4. Plating cells
    1. Plate the desired density of the uncentrifuged cell suspension from step 1.5.7 in each well containing the plating medium.
      NOTE: In this protocol, 100,000 to 200,000 cells/well were seeded in a 12-well plate. As a reference, plate at least 3 CVs per experimental condition to account for experimental variability. A typical experimental setting may entail imaging 4-5 fields per CV (e.g., using a 20x objective, each field would be expected to be ~1.3 mm2 wide).
    2. Move the plate in the cell culture hood in all the cardinal directions to distribute the cells uniformly.
  5. Electroporation
    NOTE: Other transfection systems may be used. The NEPA21 system used in this protocol has the advantage of being an "open system." The protocol can be modified by the user, and it does not require specific reagents for the transfection of special cuvettes used for the electroporation. This protocol was modified from 5.
    1. Continue from step 1.5.9. Resuspend the cell pellet with 5 mL of electroporation medium in a 15 mL tube. Centrifuge the tube at 250 × g for 5 min.
    2. Repeat step 1.7.1. Resuspend with electroporation medium to obtain a concentration of 106 cells/mL in each of the electroporation cuvettes. Mix the cells with the desired amount of DNA in the cuvette.
      NOTE: For GFP expression, 3 µg of plasmid was used per million cells. Because of the high variability, it is recommended to test the optimal concentration for each plasmid. Between 2 and 24 µg of plasmid per million cells have been tested. Diverse plasmids can be combined in the same cuvette to perform a double-transfection. In the case of co-transfection, transfect the gene of interest in excess as compared to GFP (e.g., 3:1 is an optimal ratio that should be tested) to ensure that most GFP-expressing cells co-express the gene of interest.
    3. Place the cuvette into the cuvette chamber (Figure 2). Press the Start button to execute the preset program (see electroporation conditions in Table 4). Record the values of currents and joules displayed in the Measurements frame.
      NOTE: The user may change specific conditions. For instance, decreasing the voltage may reduce the efficiency of transfection but increase neuronal survival.
    4. Take the cuvette out of the chamber. Add 500 µL of the plating medium into the cuvette immediately after the electroporation and transfer the sample from the cuvette into a 1.5 mL centrifuge tube using a pipette.
      NOTE: The faster the addition of medium, the better the survival after electroporation as it helps to avoid cell damage.
    5. Plate the desired number of electroporated cells in each well.
      NOTE: Plate from 20,000 to 100,000 cells in a 12-well plate. Depending on the conditions, only 5-10% of the electroporated cells may survive. The exact number should be tested and adapted by the user.
    6. Incubate for 2-4 h in standard conditions (37 °C, 5% CO2, saturated humidity).
      NOTE: After the incubation period, observe the cells under the microscope. If the culture is healthy, most cells would have adhered to the surface and started sprouting visible neurite extensions.
    7. Change the plating medium to neuronal medium (volumes and composition in Table 1 and Table 2, respectively).
  6. Maintaining cells
    1. To refresh the medium, remove half of the medium and add an equal volume of fresh pre-equilibrated medium after 7 days of culture.
    2. On day 10 of culture, replace half of the volume with fresh neuronal medium without the glutamine supplement (see the Table of Materials). Repeat this step every other day after day 10.
      ​NOTE: After 10 days of culture, cells produce glutamate by themselves.

2. Construction of the hypoxic chamber

  1. Use any hermetic container with a screw lid or an equivalent sealing system for this purpose (Figure 3A). Drill two holes in the wall of the plastic container and introduce a plastic tube in each one. Seal them with silicone so that the container is hermetically sealed (Figure 3).
    ​NOTE: Here, a cylindrical plastic container (size 1 liter) with a screw lid was used to construct the chamber. Ensure that the container is big enough to accommodate the CVs required for the experiment. To treat more CVs in the same space, plastic platforms can be stacked on top of each other (Figure 3A).
  2. To prepare the bubbling chamber (see 3.2), drill a hole into the cap of a flask, using any standard mechanical drill with a drill slightly smaller than the tube, and introduce a plastic tube through the hole into the flask. Seal the hole with parafilm to maintain the hermetic seal (Figure 3B).
  3. Sterilize the container and the flask with 70% ethanol and irradiate with ultraviolet light for one cycle before using them.

3. Oxygen-glucose deprivation (OGD)

NOTE: The protocol for OGD is adapted from Tasca et al.7

  1. Prepare the OGD solution with phosphate solution (Table 2) that has been previously supplemented with antibiotics.
    NOTE: For each 35 mm plate, use 5 mL; for a 100 mm plate, use 10 mL.
  2. Add the OGD solution inside the bubbling chamber (Figure 3B) and place it inside a water bath (already prewarmed at 37 °C). Connect the flask with the nitrogen source (N2, 100%).
    NOTE: Please note that because the OGD solution contains HEPES as the buffer, it does not require CO2. If the OGD solution is based on a bicarbonate buffering system, use 5% CO2 and 95% N2 to saturate the hypoxic solution and chamber.
  3. Saturate the OGD solution with gently bubbling nitrogen. Leave the solution in a mildly bubbling state for 20-30 min.
  4. Immediately transfer the flask to the cell culture hood and add 5 mL of the OGD solution without oxygen in 35 mm plates or 10 mL in 100 mm culture plates.
  5. Move the CVs with the cultured cells to the plates with OGD. Without covering the plates, move them into the hypoxia chamber and close them firmly but carefully.
    NOTE: Steps 3.4 and 3.5 must be completed as quickly as possible. Maintain the control CVs (Neuronal Medium in normoxia) in the well plates inside the incubator. Alternatively, expose the control neurons to the normoxic OGD medium supplemented with glucose (10 mM) to examine the effects of hypoxia and glucose deprivation more selectively.
  6. Connect the nitrogen system to the hypoxia chamber and check that the tubes are open. Saturate the chamber with N2 for a few minutes at a pressure of 2-3 bars (e.g., 2-3 min; 1 bar = 750 mmHg) (Figure 3C).
    NOTE: Do not forget to bubble the OGD solution again for 15 min between multiple time points.
  7. Reduce the nitrogen pressure to 1.3-1.5 bars and close both tubes of the hypoxia chamber (Figure 3C). Leave the chamber in the incubator for the required time (e.g., 60 min) at 37 °C, 5% CO2.
    NOTE: OGD duration may require adjustment depending on the experimental condition and the research goals. A short incubation time (e.g., 30 min) would simulate mild hypoxia. Two hours of OGD are usually very harsh toward the neurons. It is recommended that an OGD duration of 45 to 90 min be tested before starting the study. Of course, the duration of the OGD will impact the total survival and the relative onset of neuronal death after recovery from OGD. If the goal is to analyze neuronal death in real time, aim to achieve OGD conditions that would provoke cell death a few hours after recovery from OGD for a more effective analysis.
  8. Open the chamber and move the CVs to their original multiwell plate for recovery. After this step, consider two alternatives: 1) for small-scale experiments, proceed to section 4 for immunofluorescence and image analysis; 2) for high content screening, proceed to section 5 for the analysis in the cell analyzer (or equivalent platform).
    ​NOTE: Live imaging provides a more flexible approach because it allows real-time monitoring of the different experimental points.

4. Analysis by standard immunofluorescence and ImageJ

NOTE: Recovery time may vary from 1 to 24 h depending on the conditions.

  1. Wash the neurons with 1x PBS for 10 min.
    NOTE: This step is optional as there are several ways to proceed and analyze the images.
  2. Fix the neurons with 4% paraformaldehyde for 10 min at RT. Wash them three times with 1x PBS for 10 min each wash.
    NOTE: As paraformaldehyde is toxic, work under a chemical hood.
  3. Permeabilize the neurons by adding 1x PBS containing 0.1% Triton X-100 for 10 min. Wash the cells quickly with 1x PBS. Repeat the washing step twice for 10 min each wash.
  4. Add 350 µL of blocking solution (1x PBS containing 2% bovine serum albumin) to the CVs and leave them for at least 30 min. Spin the primary antibody (AB) stock (5 × g for 30 s) and mix it with the blocking solution at the required dilution, e.g., anti-GFP at 1:600.
    1. Transfer the coverslips to parafilm in a humidified chamber (a box with water-soaked tissue to maintain humidity and avoid evaporation). Depending on the size of the CVs, carefully add 60 or 80 µL of the primary antibody dilution to each CV, and ensure that the cells are in contact with the antibody. Incubate the CVs o/n at 4 °C.
      NOTE: A droplet should be formed on the CV surface to cover the entire CV surface. Avoid disturbing the cells while pipetting the antibody solution on the CVs. Incubation time could be reduced to 2 h at RT.
  5. Wash the CVs with 1x PBS for 10 min over the parafilm. Repeat the washing step twice.
  6. Dilute the secondary AB, e.g., Alexa 488 (1:500), and 4′,6-diamidino-2-phenylindole (DAPI, 1:2,000) in blocking solution. Depending on the size of the CVs, add a 60-80 µL droplet of the secondary AB solution on the cells. Incubate the CVs for 1 h in the dark at RT. Wash the secondary AB and DAPI from the neurons with 1x PBS for 10 min and repeat this step twice.
  7. Mount the CVs in mounting medium (see the Table of Materials) for immunofluorescence analysis and store them in the dark at 4 °C after drying.
  8. Image capturing and analysis.
    NOTE: An inverted microscope was used to take the images at 20x (Figure 4A, see the Table of Materials).
    1. Use the open-source ImageJ software to analyze the images and obtain the mask for the GFP channel (Figure 4C).
      NOTE: There are several ways of imaging and analyzing the neurons, depending on equipment availability and the experimental format. The following workflow is provided as a basic example wherein the objective is to identify the soma of individual GFP-expressing neurons present in the image. The specific values presented here, such as the threshold value and particle size, should be tested by the user based on their specific experimental conditions. In addition, different filters and processing steps may be added. For instance, the ImageJ filters Subtract Background and Despeckle often help in cleaning the image at the beginning of the processing. Optimization of the workflow for specific experimental conditions requires previous experience in image analysis and the knowledge of basic concepts of quantitative image analysis. As an in-depth guide on image analysis is out of the scope of the current protocol, we recommend the following resources: ImageJ user guide, https://imagej.nih.gov/ij/docs/guide/; for examples and tutorials, https://imagej.nih.gov/ij/docs/examples/index.htm; MBF ImageJ bundle, https://imagej.net/mbf/index.htm.
    2. For segmentation, select the Intensity segmentation with the necessary parameters to isolate the soma, which is the brightest region of the neurons.: Menu | Image | Adjust | Threshold | Set min threshold (e.g., 100) | click on Apply.
    3. Repeat the threshold from 4.8.2. to all the images to see the mask (i.e., the segmented image) for the somas of the neurons (Figure 4E).
    4. To obtain the number of surviving cells in each image and export the results to a spreadsheet to perform calculations, select Menu | Analyze | Analyze particles and enter the necessary parameter values (e.g., Size: 50-infinity pixel; Circularity: 0.5-1.00). Consider that at t = 0 h, the number of cells counted are the total number of cells identified, and calculate the percentage of cells that survived at the different time points.
      NOTE: A lower size threshold is useful for filtering out smaller objects than the soma (e.g., 60-80 µm2).
    5. Use the Record Macro function of ImageJ to automatize and repeat the analysis: Menu | Plugins | Macros | Record | click on Create to generate the macro.

5. Real-time analysis with IN Cell Analyzer 2200

  1. Insert the 12-well plate in the cell analyzer, connect the CO2 system, and set the temperature at 37 °C.
    ​NOTE: This step can be easily scaled up in this platform, performing a similar experiment in neurons plated in a 96 wells plate.
  2. Select the 20x objective, randomly select 5 fields per well (covering 2.33% of the well in this example), and focus them automatically for each fluorescence channel (fluorescein isothiocyanate [FITC] in this case). Select the parameters of acquisition (30 min time-lapse for 16 h). Press Start and wait.
    NOTE: The number of imaging fields must be determined by the investigator and should be sufficient to account for the variability inside a single well.
  3. Use IN Cell Developertoolbox v1.9 to set up the appropriate workflow for analysis.
  4. Analyze the neuron images (FITC channel) to generate a mask for the soma of the neurons (Figure 4).
    NOTE: The objective is to identify the soma of individual neurons present in the image. This analysis follows steps that are equivalent to those describes in section 4 for ImageJ. Please refer to section 4 for the workflow and general considerations. The intensity threshold is determined to select the soma, which is the brightest region of the neurons.
    1. For segmentation, select the Intensity segmentation with the necessary parameters (in this case, a minimum threshold of 20005.40).
    2. For postprocessing, select the Sieve option, keeping targets with an area greater than a predetermined value (here, 189 pixels (~80 µm2)).
    3. Select the Sum parameter to count and sum all the neurons identified in the mask (surviving neurons) for each field and for all the fields of a well at each time point of the time course.
    4. Export the results to a spreadsheet. At t = 0 h, consider that the number of cells counted are the total cells identified, and calculate the percentage of cells that survived at the different time points. Calculate the time at which 50% of the cells are dead to facilitate the comparison between conditions.
      NOTE: The parameters for the segmentation and postprocessing steps must be adjusted by the user for different fluorescent channels depending on the structures to be identified.

Subscription Required. Please recommend JoVE to your librarian.

Representative Results

This protocol aims to establish an in vitro model of stroke. It is important to obtain an adequate neuronal density, which will allow the recognition of individual electroporated neurons to analyze them individually. The stage of the neuronal culture after plating is also crucial. The maturation of neurons in culture is progressive. The dependence on growth factors, neurite outgrowth, connectivity, and electrophysiological activity will vary greatly depending on the stage. In these specific conditions at 4-6 days in vitro, although the neurites are well-extended, the neurons are still very immature, highly dependent on growth factors, and possess a relatively low electrophysiological activity. At this stage, OGD may simulate conditions closer to embryonic development. Dendritic spines and clear synaptic clusters can be observed by days 8-10 in vitro11. On day 14 in vitro under these experimental conditions, cortical neurons are mature and have reached a full complexity of connectivity and electrophysiological activity3. Here, we challenged cortical neurons with OGD at day 14 to simulate the conditions of an adult brain.

After 60 min of OGD, the neurons were incubated in the IN Cell Analyzer 2200. Control neurons were imaged in parallel and did not show any signs of distress. In the control, the somas and neurites of GFP-expressing neurons are visible at the later time points (Figure 5A). Neurons subjected to OGD begin showing signs of neurite degeneration at early time points, followed by loss of fluorescence in the soma due to rupturing of the plasma membrane and diffusion of GFP out of the neurons (Figure 5A). Because of the higher signal intensity, round morphology, and sparse labeling, the simple analytical workflow designed for this study allows robust and fast analysis of neuronal survival (Figure 5B). The results show the high reproducibility between the fields in the same CV or well and between the CVs of the same condition (Figure 5B, Supplemental Table S1, and Supplemental Table S2).

For comparison, Figure 5C shows manual quantification of the surviving neurons in a subset of the same images that validate the automatized analysis (Figure 5B,C; the raw data are available on request). Hence, this procedure is a reliable model of stroke in vitro, simulating the cellular death that underlies ischemic events. Moreover, this model opens a wide number of possibilities, as drugs or the overexpression of certain proteins (co-transfection with the electroporation method) can be tested to study the response to hypoxia or other toxic factors. Taken together, the described protocol is proposed for the establishment of an adequate model to perform medium-scale screening.

Figure 1
Figure 1: Schematic outline of the experimental paradigm. The schema summarizes the main steps of this experimental paradigm. First, the establishment of the primary neuronal culture with the electroporation step. Next, mature neurons are subjected to OGD to reproduce a stroke in vitro. Finally, for the image analysis, two different options are proposed: for small scale, a semi-automatic approach with ImageJ in immunostained samples; for high-content screening, automatic analysis using IN Cell Analyzer 2200. Abbreviations: OGD = oxygen-glucose deprivation; RT = room temperature; CVs = coverslips. Please click here to view a larger version of this figure.

Figure 2
Figure 2: The electroporation system. (A) The device used for electroporation with NEPA21. (B) Magnification of the device where the cuvettes are inserted for electroporation. (C) The cuvette used for electroporation. Please click here to view a larger version of this figure.

Figure 3
Figure 3: The custom-made hypoxic chamber. (A) The different parts of the hypoxic chamber. As noted, it is formed by a hermetic container in which platforms can be introduced to treat different samples. (B) The bubbling chamber and a cell culture flask in which the OGD solution is saturated with nitrogen. (C) This scheme shows the gas flux to saturate the chamber with N2 to reach oxygen deprivation. Abbreviation: OGD = oxygen-glucose deprivation. Please click here to view a larger version of this figure.

Figure 4
Figure 4: Comparative images and workflow of the analysis. Representative images obtained at 20x with (A) an inverted microscope and (B) Cell analyzer. (C) Workflow generated to create a mask and analyze the images obtained with a microscope using ImageJ software. (D) Workflow generated to create a mask and analyze the images obtained with the cell analyzer using the associated developer toolbox v1.9. (E) Representative image of the mask generated with ImageJ. (F) Representative image of the mask generated with the developer toolbox v1.9. Scale bars = 50 µm. Abbreviations: FITC = fluorescein isothiocyanate. Please click here to view a larger version of this figure.

Figure 5
Figure 5: Kinetics of neuronal cell death after hypoxia. (A) Representative images of neurons expressing GFP in control conditions or subjected to hypoxia for 60 min and imaged at different time points (up to 360 min (')). Images were taken with the IN Cell Analyzer 2200 in a time-lapse mode (every 30 min). White arrowheads point to neurons that are alive at the beginning of the time course and dying because of the hypoxic insult during the time-lapse. (B) Percentage of live neurons at each time point of the time course for control cells and for neurons subjected to hypoxia. These data were automatically calculated after setting the analysis workflow in the IN Cell Developer Toolbox v1.9 software. (C) Percentage of live neurons at each time point of the time course for control cells and neurons subjected to hypoxia. These data were manually calculated after setting the analysis workflow in the ImageJ software. Scale bars = 50 µm. Abbreviations: Ctrl = control; GFP = green fluorescent protein. Please click here to view a larger version of this figure.

Supplemental Table S1: Raw and processed data obtained after setting the analysis parameters in the IN Cell Developer Toolbox v1.9 software. Red highlighting indicates that more than 50% of the neurons in the analyzed field are dead. Please click here to download this Table.

Supplemental Table S2: Raw and processed data obtained after setting the analysis workflow in the ImageJ software. Red highlighting indicates that more than 50% of the neurons in the analyzed field are dead. Please click here to download this Table.

Subscription Required. Please recommend JoVE to your librarian.


This protocol shows an effective way of modeling a stroke in vitro. To achieve this goal, we proposed sparse labeling of cortical neurons using the electroporation system NEPA215. This is an open system that allows customization of the protocol with minimal operative cost compared to other systems that employ kits or specific devices. Mixed culture of naïve and electroporated neurons allows more flexibility and robustness as compared to low-density neuronal culture. This allows the study of the influence of different genes, constructs, and proteins and is easily scalable. For instance, we used a similar approach to determine the role of Nrg1, a neurotropic factor expressed by cortical neurons, in intracellular signaling in stroke4. Moreover, the electroporated cells can be mixed at different densities with non-electroporated cells to obtain sparse and adjustable numbers of fluorescent cells depending on the goal of the study.

For example, it is recommended to use a lower concentration for cellular morphology analysis. Thus, each neuron can be seen isolated, especially in the late stages where the neurites are considerably extended. As a rule of thumb, aim to have sufficient fluorescent neurons for consistent quantification in any given field of imaging (e.g., average n > 10) but not too many so that single cells can still be distinguished (e.g., average n < 50). Some variability in the number of fluorescent neurons per field is expected given the stochastic distribution of the sparse cells in the well. Although the methodology proposed here is robust across different cell densities, an excessive cell density is more problematic because it does not allow the user to distinguish between different cells in the image. While the NEPA21 electroporation system is a flexible and cost-effective platform12,13, other electroporation systems, such as the Amaxa Nucleofector or the Neon Transfection System, may also be used.

This paper describes the OGD method as a cytotoxic model, an established in vitro model of stroke4,7. OGD is one of the several models of neurotoxicity available, and it challenges the neurons in vitro with multiple hits: absence of glucose and oxygen and lack of growth factors, vitamins, anti-oxidants, hormones, and other important components that are typically present in the neuronal culture medium and supplement (NB27). Here, naïve neurons were used as healthy controls to evaluate the neuronal response to cytotoxic stress, as described previously4.

This control will ensure that the viability of the primary culture is optimal and that the imaging process is not harmful to the neurons in case of time-lapse microscopy. In addition, to selectively evaluate the response to lack of glucose and oxygen, the control neurons may be treated with OGD normoxic medium supplemented with glucose7. The choice of the most appropriate controls should be considered carefully depending on the specific aims of the experiment. Also described in detail is how to build customized and inexpensive hypoxic chambers.

Other methods may be used to model stroke in vitro or other neurotoxic conditions. For example, inhibitors of the electron transport chain (rotenone, antimycin, and sodium azide) or glutamate receptor agonists (N-methyl-D-aspartate) have been widely used8. Finally, this paper describes the imaging and analysis workflow as a representative example. The optimal workflow may vary greatly depending on the experimental conditions and imaging. Investigators can adapt and optimize this basic protocol according to the specific requirements of their experiments.

Sparse labeling and visualization of the neurons greatly simplify the analysis that can be easily streamlined and automatized in different platforms. This allows setting up the protocol on a small scale for image analysis and easily adapt these conditions to high-throughput screening with IN Cell or similar platforms for drug screening. The IN Cell system has multiple and significant advantages over conventional microscopy for image analysis. It allows the user to customize the settings, perform repeated measurements of the same region of interest, and analyze all the pictures semi-automatically, facilitating data processing. Other live-imaging systems, as recently shown14, may also be adapted to this workflow. Notably, GFP may be co-transfected with specific markers to visualize cellular structures, such as mitochondria or synaptic markers, for more refined analysis.

Live-imaging and real-time analyses have two major advantages: first, they do not require the use of antibodies or other expensive reagents; second, they provide temporal information about cell death that allows a more precise comparison between different experimental conditions. For instance, a specific gene of interest may be co-electroporated with a different fluorescent cytosolic marker (e.g., RFP) and co-cultured with GFP-expressing internal control neurons. In conclusion, this protocol offers an effective way of reproducing this in vitro model of neurotoxicity given its simplicity, versatility, and scalability. It is particularly apt for testing the roles of specific genes and pathways in live imaging and for high-throughput drug screening.

Subscription Required. Please recommend JoVE to your librarian.


The authors have no conflicts of interest to disclose.


We would like to thank Carlos Dotti for sharing his expertise in neuronal culture. We also thank Alicia Martínez-Romero from the Cytomics Core Facility of the Centro de Investigación Principe Felipe (CIPF), which is supported by European FEDERER funding. The project is supported by the Spanish Ministry of Economy and Competitiveness for (SAF2017-89020-R) reagents, materials, and the salaries of YDC and PF. PF is also supported by the grant RyC-2014-16410. CGN and PF are supported by Conselleria de Sanitat of the Generalitat Valenciana, as well as AGM (ACIF/2019/015). Ángela Rodríguez Prieto is supported by the Spanish Ministry of Science, Innovation and Universities, with the grant PRE2018-083562.


Name Company Catalog Number Comments
10 cm petri dishes FISHER SCIENTIFIC, S.L. 1130-9283
3.5 cm petri dishes Sterilin
75 cm2 flask Corning 430641U
B-27 Life Technologies 17504-044
Cell culture plates Corning incorporation Costar® 3513
Cell incubator Thermo Electron Corporation Model 371
Cell strainer 70 µm Falcon 352350
Cold lights Schott 223488 KL 1500 LCD
Coverslips (15 mm) Marienfeld 111530
CU500 Cuvette Chamber Nepa Gene
CU600 Cuvette Stand Holder Nepa Gene
DAPI Sigma-Aldrich Quimica, S. L. D9542-10MG 1:2000
DMSO Panreac A3672
Dumont Fine Forceps FST 11254-20
Dumont Fine Forceps FST 11252-00
EC-002S NEPA Electroporation Cuvettes, 2mm gap Nepa Gene
Filter strainer Falcon 352350
Fine Scissors-Sharp-Blunt FST 14028-10
Fine Scissors-ToughCut r FST 14058-09
GFP chicken IgY Aves Labs GFP-1010 1:600
Glucose Sigma 68769-100ml
GlutaMAX-I Supplement 200 mM 100 mL Fisher Scientific 35050-061
HBSS Thermofisher 14175-095 https://www.thermofisher.com/es/es/home/technical-resources/media-formulation.156.html
Hepes 1 M ThermoFisher 15630-080
Horse Serum Invitrogen 26050088
MEM Thermofisher 11095080 https://www.thermofisher.com/order/catalog/product/11095080#/11095080
Microscope slide (polilysine) VWR 631-0107 Dimension: 25 x 75 x 1 mm
Mowiol 4-88 Sigma-Aldrich Quimica, S. L. 81381-250G
Needles yellow, 30 gauge BD Microlance™ 3 304000
NEPA21 electroporator Nepa Gene
Neubauer chamber Blau Brand 717805
Neurobasal Medium ThermoFisher 21103-49
Opti-MEM Invitrogen 31985-062
Parafilm Cole-Parmer PM996
Paraformaldehyde (PFA), 95% Sigma-Aldrich Quimica, S. L. 158127-500G Use solution: 4%
PEI Polysciences 23966-1
Plasmid for GFP pCMV-GFP-ires-Cre, described in Fazzari et al., Nature, 2010
Poly-L-Lysine SIGMA P2636
PS ( Penicillin, Streptomycin) ThermoFisher 15140122
Serrate forceps FST 11152-10
Syringes BD Plastipak 1ml 303176
Triton X-100 Sigma-Aldrich Quimica, S. L. MDL number: MFCD00128254 Non-ionic
Trypsin-EDTA ThermoFisher 25300054
Tubes 15 mL Fisher 05-539-4
Tubes 50 mL VWR 21008-242
Tupperware - - Hermetic tupperware with screw lid. SP Berner - Taper 1 L Redondo con Rosca. Any equivalent hermetic Tupperware may be purchased in any supermarket.
Water bath SHELDON LABORATORY MODEL 1224 1641951



  1. Salazar, I. L., Mele, M., Caldeira, M., Costa, R. O., Correia, B., Frisari, S., Duarte, C. B. Preparation of primary cultures of embryonic rat hippocampal and cerebrocortical neurons. Bio-protocol. 7, (18), 2551 (2017).
  2. Kaech, S., Banker, G. Culturing hippocampal neurons. Nature Protocols. 1, (5), 2406-2415 (2006).
  3. Fazzari, P., et al. Cell autonomous regulation of hippocampal circuitry via Aph1b-γ-secretase/neuregulin 1 signalling. eLife. 3, 02196 (2014).
  4. Navarro-González, C., Huerga-Gómez, A., Fazzari, P. Nrg1 intracellular signaling is neuroprotective upon stroke. Oxidative Medicine and Cellular Longevity. 2019, 3930186 (2019).
  5. NEPA21 transfection of cell suspensions using cuvettes. NEPAGENE Co. Ldt. Available from: http://2017.igem.org/wiki/images/3/3a/T--TECHNION-ISRAEL--Electroporation-NEPA2_for_HPC-7.pdf (2017).
  6. Larson, M. A., Gibson, K. A., Vivian, J. L. In vivo validation of CRISPR reagents in preimplantation mouse embryos. Methods in Molecular Biology. 2066, 47-57 (2020).
  7. Tasca, C. I., Dal-Cim, T., Cimarosti, H. In vitro oxygen-glucose deprivation to study ischemic cell death. Methods in Molecular Biology. 1254, 197-210 (2015).
  8. Holloway, P. M., Gavins, F. N. E. Modeling ischemic stroke in vitro: the status quo and future perspectives. Physiology and Behavior. 47, (2), 561-569 (2016).
  9. Sommer, C. J. Ischemic stroke: experimental models and reality. Acta Neuropathologica. 133, (2), 245-261 (2017).
  10. Dotti, C. G., Banker, G. A. Experimentally induced alteration in the polarity of developing neurons. Nature. 330, (6145), 254-256 (1987).
  11. Fazzari, P., et al. Control of cortical GABA circuitry development by Nrg1 and ErbB4 signalling. Nature. 464, (7293), 1376-1380 (2010).
  12. Sadiq, A. A., et al. An overview: Investigation of electroporation and sonoporation techniques. 2015 2nd International Conference on Biomedical Engineering. 1-6 (2015).
  13. Shi, J., et al. A review on electroporation-based intracellular delivery. Molecules. 23, (11), 3044 (2018).
  14. Fazzari, P., Mortimer, N., Yabut, O., Vogt, D., Pla, R. Cortical distribution of GABAergic interneurons is determined by migration time and brain size. Development. 147, (14), 185033 (2020).
This article has been published
Video Coming Soon

Cite this Article

Rodríguez-Prieto, Á.,González-Manteiga, A., Domínguez-Canterla, Y., Navarro-González, C., Fazzari, P. A Scalable Method to Study Neuronal Survival in Primary Neuronal Culture with Single-cell and Real-Time Resolution. J. Vis. Exp. (173), e62759, doi:10.3791/62759 (2021).More

Rodríguez-Prieto, Á., González-Manteiga, A., Domínguez-Canterla, Y., Navarro-González, C., Fazzari, P. A Scalable Method to Study Neuronal Survival in Primary Neuronal Culture with Single-cell and Real-Time Resolution. J. Vis. Exp. (173), e62759, doi:10.3791/62759 (2021).

Copy Citation Download Citation Reprints and Permissions
View Video

Get cutting-edge science videos from JoVE sent straight to your inbox every month.

Waiting X
Simple Hit Counter