High-throughput neurite outgrowth assay using GFP-labeled iPSC-derived neurons

Abstract

Due to the potential neurotoxicity caused by an increasing number of drugs and untested environmental chemicals, there is a need to develop reliable and efficient in vitro methods to identify chemicals that may adversely affect the nervous system. One process that has been implicated in neurodevelopment is neurite outgrowth, the disruption of which can result in developmental neurotoxicity. Currently, neurite outgrowth assays mainly rely on staining, which require multiple sample processing steps, particularly washing steps, which may introduce variation and limit throughput. Here, we describe a neurite outgrowth assay that uses induced pluripotent stem cell (iPSC)-derived human cortical glutamatergic neurons and/or spinal motor neurons labeled with green fluorescent protein (GFP) for testing compounds in a high-content and high-throughput format. The method enables live and time-lapse imaging of the GFP-labeled neurons, where the assay plate is continuously imaged at multiple times (e.g., 24 h and 48 h) after chemical treatment. In this protocol, we describe how to thaw the frozen GFP-labeled neurons, culture them, and treat them with a compound of interest, as well as how to use a high-content imaging platform to measure and analyze neurite outgrowth. The GFP-labeled iPSC-derived human neurons represent a promising tool to identify and prioritize compounds with developmental neurotoxicity potential for further hazard characterization.

INTRODUCTION

Some environmental chemicals, including heavy metals, pesticides, and food additives (Alavanja, Hoppin, & Kamel, 2004; Aschner et al., 2017; Lau, McLean, Williams, & Howard, 2006; Wang et al., 2010), are known to exert neurotoxicity and developmental neurotoxicity, which are linked to an increasing incidence of neurological defects, such as autism, attention deficit, and hyperactivity disorders. With concerns about a large number of chemicals being introduced into our environment and the increasing prevalence of neurological disorders, there is a great need to develop reliable and efficient high-throughput screening methods to identify chemical compounds that could potentially affect the neurological system.

Several in vitro assays have been developed for evaluation of developmental neurotoxicity, which are typically based on analyzing basic neuronal processes, including differentiation, proliferation, migration, and neurite outgrowth (Frimat et al., 2010; Harrill, Robinette, & Mundy, 2011; Krug et al., 2013; Radio & Mundy, 2008; Zimmer et al., 2012). Neurite outgrowth, representing defined neuronal morphology, is characterized by axonal and dendritic processes. Neurite outgrowth can be measured by a variety of methods, including determining neurite length and area, and these assays have been extensively used for the evaluation of developmental neurotoxicity and neurotoxicity (Radio & Mundy, 2008). High-content imaging enables the assessment of multiple morphological parameters for neurite outgrowth, and can also differentiate between specific effects on neurite outgrowth and non-specific cytotoxicity (Harrill, Freudenrich, Machacek, Stice, & Mundy, 2010; Stiegler, Krug, Matt, & Leist, 2011). As such, a high-throughput, high-content imaging-based neurite outgrowth assay can be a promising tool to rapidly profile a large number of chemicals for neurotoxicity and developmental neurotoxicity.

Several in vitro assay models, ranging from non-human primary cells to human induced pluripotent stem cell (iPSC)-derived neurons, have been used to examine the effects of chemicals on neurite outgrowth (Radio & Mundy, 2008; Schmidt et al., 2017). iPSC-derived neurons are more directly relevant to human biology, and maintain the morphology and physiology of their in situ counterparts (Chai, Dage, & Citron, 2012; Haythornthwaite et al., 2012; Whitemarsh et al., 2012).

iPSC-derived neurons have been used in high-content, high-throughput screening assays to evaluate compounds for developmental neurotoxicity and neurotoxicity (Ryan et al., 2016; Sirenko, Hesley, Rusyn, & Cromwell, 2014). The neurite outgrowth assay commonly utilizes β-III tubulin immunostaining or Calcein AM staining (Krug et al., 2013; Radio, Breier, Shafer, & Mundy, 2008). Tubulin staining, however, requires multiple washing steps, which is time-consuming and increases well-to-well variations. The neurite outgrowth assays using Calcein AM staining also requires a staining step, and this assay cannot be used for time-lapse imaging, due to cytotoxicity.

To address some of these issues, we have recently developed and validated a robust high-throughput, high-content imaging-based neurite outgrowth assay, which can be used for compound evaluation (Li et al., 2021). The assay uses commercially available human iPSC-derived cortical glutamatergic neurons and spinal motor neurons labeled with GFP, to enables direct, live, and time-lapse imaging without wash steps. This assay can be performed in 96-, 384-, and 1536-well plates, greatly increasing throughput. Using this platform, neurons develop neurites rapidly and consistently, as quantified by automated high-content imaging.

Here, we provide a detailed neurite outgrowth assay protocol, including thawing and seeding iPSC-derived neurons (Basic Protocol 1), culturing neurons and treating with test compounds (Basic Protocol 2), and measuring and analyzing neurite outgrowth (Basic Protocol 3). Figure 1 provides an outline of the entire workflow. Overall, this assay is screening friendly, reproducible, and robust, and show promise as a reliable high-throughput screening tool for compounds with developmental neurotoxicity potential.

Figure 1. Overview of the protocol for performing the high-throughput, high-content neurite outgrowth assay described in this article.

Basic Protocol 1 outlines the steps for thawing and seeding iPSC-derived neurons. Basic Protocol 2 describes the culturing of neurons and the treatment of cells with test compounds. Basic Protocol 3 describes measuring and analyzing neurite outgrowth data.

STRATEGIC PLANNING

It is important to have all the materials and reagents ready before the experiment. Users should pay special attention to the specific requirements of neurons, media, plate types, plate dispenser, compound transfer system (e.g., Pintool station), and high content image system. It is especially important that the plates should be pre-coated with PDL, Laminin, or Collagen I, so that the neurons can attach firmly to the plate. Eight different coating conditions, PDL, laminin, PEI, collagen I, fibronectin, laminin, BME, PDL-BME, and uncoated, were tested previously. Our optimization results suggest the PDL-coated plates provide the highest quality for imaging data. In addition, low base PDL-coated plates improve image quality even further in our hands.

Before performing the experiment, all the materials used in this assay, including the seeding medium that is also used as the thaw/culture medium, should be made freshly and carefully. Geltrex, one of the seeding medium components, helps cell attachment. It needs to be thawed on ice before making the seeding medium. Users can first thaw Geltrex and aliquot it for use or store it at −20°C for future use. Alternatively, users can dilute Geltrex with cold DMEM/F12 medium (1:10) for immediately use or store at −20°C for future use. DMEM/F12 medium should be taken straight out of 4°C for maximum activity.

One critical step to perform this assay is thawing the neurons. Prior to being thawed, the neurons should be stored in a liquid nitrogen tank or at −150°C for optimal freeze conditions. BioRAPTR Flying Reagent Dispenser or Multidrop Combi Reagent Dispenser can be used to dispense neurons into 1536-well plates.

CyBi-Well Vario Multichannel Automated Pipetting System or other similar equipment can be used to make a control compound plate. A Pintool station can be used to transfer test compounds including controls from compound plate to the assay plate.

Since this is a high content imaging assay, the user needs to have a high content imaging machine with the proper analysis software system to capture images in each well, and then convert image data to numerical data. Here, we use the Operetta CLS High Content Analysis System (PerkinElmer, HH16000000), with Harmony software v. 4.6.

BASIC PROTOCOL 1: Thawing and Seeding iPSC-derived Neurons

This protocol explains how to prepare specific seeding medium for neurons, and provides detailed steps for thawing and seeding them into a 1536-well plate. The iPSC-derived, GFP-labeled human neurons used here (cortical glutamatergic neurons and spinal motor neurons) are commercially available (BrainXell). Cortical glutamatergic neurons have high neuronal purity (>90%) and comprise predominantly excitatory neurons. Spinal motor neurons have high neuronal purity (>90%) and consist mostly of motor neurons. We have successfully tested both types for use in the neurite outgrowth assay described in this article. Users can choose which type to use based on their need.

The components for the seeding medium, according to BrainXell’s recipe, are listed in Table 1. Brain Derived Neurotrophic Factor (BDNF), Glial-Derived Neurotrophic Factor (GDNF), and Transforming Growth Factor-β1 (TGF-β1), supplied as lyophilized powders, are reconstituted according to the manufacturer’s instructions. BDNF is a neurotrophic growth factor that supports neuron proliferation and survival. GDNF is a disulfide-linked, homodimeric neurotrophic factor that specifically promotes dopamine uptake and survival, and morphological differentiation of midbrain neurons. TGF- β1 regulates cell proliferation, growth, differentiation, and motility, as well as synthesis and deposition of the extracellular matrix. For this assay, the recommended stock solution for BDNF and GDNF is 10 μg/ml, while for TGF-β1 it is 1 μg/ml. All these stock solutions should be aliquoted and kept in a −20°C freezer to avoid multiple freeze-thaw cycles, as those could compromise the integrity of the solution.

Table 1.

Seeding medium components and example for preparing 20 ml

Component	Stock Solution	Final Concentration	Volume from stock
DMEM/F12 Medium	1x	0.5x	9.6 ml
Neurobasal Medium	1x	0.5x	9.6 ml
B27 Supplement	50x	1x	400 μl
N2 Supplement	100x	1x	200 μl
Geltrex	15 mg/ml	15 μg /ml	200 μl (of 1:10)
GlutaMAX	200 mM	0.5 mM	50 μl
BDNF	10 μg/ml	10 ng/ml	20 μl
GDNF	10 μg/ml	10 ng/ml	20 μl
TGF-β1	1 μg/ml	1 ng/ml	20 μl
Neuron Supplement	1000x	1x	20 μl
Penicillin/Streptomycin*	200x	1x	100 μl

^*Final concentration: 50 U/ml penicillin, 50µg/ml streptomycin.

Materials

GFP-labeled human cortical glutamatergic neurons and supplement (BrainXell, cat. no. BX-0300)

GFP-labeled human spinal motor neurons and supplement (BrainXell, cat. no. BX-0100)

DMEM/F12 Medium (Thermo Fisher Scientific, cat. no. 11039021, no phenol red)

Neurobasal Medium (Thermo Fisher Scientific, cat. no. 12348017, no phenol red)

B27 Supplement (Thermo Fisher Scientific, cat. no. 17504-044)

N2 Supplement (Thermo Fisher Scientific, cat. no. 17502-048)

GlutaMAX (Thermo Fisher Scientific, cat. no. 35050-061)

Geltrex (Thermo Fisher Scientific, cat. no. A1413201)

BDNF (Peprotech, cat. no. 450-02)

GDNF (Peprotech, cat. no. 450-10)

TGF-β1 (Peprotech, cat. no. 100-21C)

Pen Strep (Thermo Fisher Scientific, cat. no. 15140-122)

15-ml and 50-ml conical tubes

Corning Sterile Cell strainer (Fisher Scientific, cat. no. 07201430)

Multichannel pipettes (5, 10, 20, and 200 µl) and tips

Poly-D-lysine (PDL)-coated 1536-well black clear bottom low base assay plates (Aurora Microplates, cat. no. E8-1B1-210000-PDL)

Stainless steel lid with holes (for cell breath) as cover for assay plate

70% Ethanol

Hemocytometer

Trypan blue

BioRAPTR Flying Reagent Dispenser (FRD, Beckman Coulter cat. no. BioRAPTR2)

Multidrop Combi Reagent Dispenser (Thermo Fisher Scientific, cat. no. 5840300)

Laminar flow hood (Labconco Purifier BSC Class II, or equivalent)

CO₂ incubator for cell culture

Tissue culture microscope (Nikon or equivalent)

Protocol Steps

Preparing seeding medium, and thawing and seeding neurons

1. Gather the components shown in Table 1 for preparing seeding medium. Note that BDNF, GDNF, and TGF-β1 are supplied as lyophilized powders. Follow the manufacturer’s instructions for reconstitution. We recommend creating stock solutions of 10 μg/ml for BDNF, 10 μg/ml for GDNF, and 1 μg/ml for TGF-β1, respectively, and keeping in −20°C freezer (see Strategic Planning).

2. Use the Motor Neuron Seeding Supplement when making seeding medium for motor neurons. For preparing seeding medium for cortical neurons, use the Cortical Neuron Seeding Supplement.

3. To prepare the Geltrex, take an aliquot of frozen Geltrex and place it on ice for thawing. After thawing, add cold DMEM/F12 medium to obtain a 1:10 dilution.

   For example, if an aliquot of Geltrex has a volume of 100 μl, add 900 μl of cold DEME/F12.

4. Working in a cell culture hood (biological safety cabinet), combine all components listed in Table 1 in a sterile 50-ml conical tube, using separate tubes for motor neurons and cortical neurons media. Each vial of cells will require approximately 20 ml of seeding medium, so scale accordingly. Allow the medium to equilibrate to room temperature for 15 min. Do not warm the medium in a 37°C water bath.

5. Remove a cryovial of neurons from the liquid nitrogen and place in a 37°C water bath. To minimize contamination, avoid submerging the cap. Gently move the vial within the water bath to increase the rate of thawing.

   If users are using two neuron types, do the same for all cryovials.

6. As soon as the last piece of ice melts, remove the vial from the water bath. Disinfect the vial by spraying it with 70% ethanol and transfer it to the cell culture hood.

7. Slowly add 800 μl of the corresponding seeding medium (from Step 4) to the vial at a rate of ~1 drop/s using a 1 ml pipette tip. This process should take about 30 sec.

8. Gently transfer all contents (1 ml total) from the vial to a new sterile 50-ml conical tube, and then gently add another 1 ml of the corresponding seeding medium to the vial to collect the residual cells.

9. Slowly add an additional 3 ml of the corresponding seeding medium to the 50-ml conical tube containing the cells using a 10-ml serological pipette. Gently swirl the conical tube while adding the medium. This process should take about 1 min. Mix completely and filter with a cell strainer to a separate 50-ml conical tube.

10. Using the trypan blue exclusion method to test cell viability. To do this, gently swirl the conical tube from Step 9 again, take 10 μl of the cell suspension, and mix with the same amount (i.e. 10 μl) of 0.4% trypan blue, in a small plastic tube. Incubate mixture for about 3 min at room temperature. Apply a drop of the trypan blue/cell mixture to a hemacytometer. Place the hemacytometer on the stage of a binocular microscope and focus on the cells. Count the unstained (viable) and stained (nonviable) cells separately in the hemacytometer. To obtain the total number of viable cells per ml, multiply the total number of viable cells by 2 (the dilution factor for trypan blue). To obtain the total number of cells per ml, add up the total number of viable and nonviable cells and multiply by 2. Calculate the percentage of viable cells as follows:

    % Viable cells = (Total number of viable cells per ml of aliquot/Total number of cells per ml of aliquot) x 100.

    The viable cells percentage should be more than 80% for a good experimental result.

11. Transfer the required number of cells to a separate 50-ml conical tube, and then add the proper volume of the corresponding seeding medium to adjust cell density to 12–16 × 10⁴ cells/ml.

    Users also need to optimize cell-seeding density based on their experimental setting. Here, we recommend 600–800 viable cells/well for a 1536-well plate, 2000–2500 viable cells/well for 384-well plate, and 10,000 viable cells/well for a 96-well plate.

12. Dispense 5 μl of the 12–16 × 10⁴ cells/ml cell suspension (containing 600–800 cells) into each well of a PDL-coated 1536-well plate using a multi-drop Combi Dispenser or BioRAPTR. Cover the plate with a stainless-steel lid.

    Throughout the seeding process, be careful not to move or agitate the plate, as this may lead to uneven attachment.

13. After seeding, leave the plate at room temperature for 15 min to allow the cells to attach to the bottom of the well. Then, transfer the assay plate gently to a humidified incubator at 37°C with 5% CO₂.

    Do not immediately transfer the assay plate to the incubator after seeding, to avoid disturbance.

14. Incubate the assay plate for 40–42 hr to allow neurite formation. In the meantime, proceed to Basic Protocol 2.

    Before placing the assay plate into incubator look at cell density and distribution using a microscope. You should also check the neurite forming status each day using a microscope.

BASIC PROTOCOL 2: Plate preparation and compound treatment

To test compounds in a neurite outgrowth assay in 1536-well plates, the user needs to prepare two 1536-well plates: a control plate and a test compound plate. The control plate should contain a known positive control and a negative/vehicle control. These controls can be used to normalize the data of the test compounds. The Pintool station will transfer 23 nl control compounds or test compounds to the assay plate concurrently. This protocol describes how to prepare a positive control plate, including a serial concentration dilution, as well as steps to transfer compounds into the assay plate containing the cells.

As soon as the assay plate containing neurons is placed into an incubator (Basic Protocol 1), users can prepare a control compound (e.g., rotenone) plate. As shown in Fig. 2, rotenone, a neurotoxicant, was used here as a positive control for neurite outgrowth and users can pick one concentration of rotenone that gives maximum inhibition, to normalize test compound activity. In addition, tetraoctylammonium bromide was used as a positive control for cytotoxicity and DMSO was used as a vehicle control. The test compounds (dissolved in DMSO) were placed in columns 5–48. For compound treatment, 23 nl of positive, negative, or test compound is transferred to the assay plate using the Pintool station. The final concentration of DMSO in each assay well is generally less than 0.5%, which has minimal toxic effect on the cells.

Figure 2. Control compound (i.e., known positive compounds and negative/vehicle control) plate map in 1536-well format used in the protocols described in this article.

In this example outline, column 1 contains a sixteen-concentration titration ranging from 46 μM to 1.4 nM in duplicates of rotenone. Columns 2 and 3 contain 16 replicates for each concentration of Rotenone: 46 μM in upper half column 2, 11.5 μM in bottom half column 2, and 23 μM in upper half column 3, as well as 92 μM tetraoctylammonium bromide (as positive control of cytotoxicity) in the bottom half column 3. Additionally, DMSO, as a negative control, is dispensed into column 4. Note that the test compounds (grey) in columns 5–48 are in a different plate.

Materials

Assay plate (Basic Protocol 1)

Single-channel and multichannel pipettes (5, 10, 20, and 200 µl) and tips

CAPP 16-channel 0.5–10 µl pipettes (CAPP, cat. no. C10-16) and tips (CAPP, cat. no. 5030004)

96-well clear round bottom (Greiner, cat. no. 650001) and 384-well plates (Costar, cat. no. 3656), for making serial dilutions

1536-well plate (Greiner, cat. no. 789270-C), for preparing compounds plate

Rotenone (Sigma, cat. no. R-8875)

Tetraoctylammonium bromide (Sigma, cat. no. 14866-33-2)

DMSO (Sigma, cat. no. D2650)

Balance (Mettler Toledo)

CyBi-Well Vario Multichannel Automated Pipetting System (Analytik Jena, Beverly, MA, USA)

PinTool Compound Transfer Workstation (Wako Automation, San Diego, CA, USA)

Protocol Steps

Preparation of control plate

1. Weigh the positive control compound, such as rotenone or tetraoctylammonium bromide, and dissolve it in DMSO or the proper solvent to make a 10 mM stock solution.

   The stock solution can be stored at −20°C for future use.

2. To prepare the control plate in a 1536-well plate format, use two columns (16 wells) of a 96-well plate to prepare a serial dilution (16 concentrations in total) of the positive control, and then transfer them to one column in a 384-well plate.

   Please note that if treating neurons in 96-well or 384-well assay plate, the user needs to use medium or PBS instead of DMSO to dilute the compound, to avoid the effect of high concentration of DMSO on neurons. Since the amount of DMSO transferred by Pintool to 1536- well plate is just 23 nl, the influence can be ignored.

3. Transfer 5 µl of the solutions from the 384-well plate from Step 2 to a new 1536-well plate using CAPP Multichannel pipettes or CyBi-Well Vario Pipettor. Centrifuge the plate at 1000 rpm for 10 seconds at room temperature and seal the plate with film for later use.

   See Figure 2 for an example of 1536-well control compound plate map.

4. Prepare the compound plate by following the same Steps 2–3, on a different 1536-well plate.

5. Take the assay plate from Basic Protocol 1, Step 14, out of the incubator after 40–42 hr, and use a microscope to visualize neurite formation status. Remove the plate lid and put the assay plate inside the Wako Pintool station, as well as the control compound and test compound plates. Then transfer 23 nl of the mix from each well of the compound plates (e.g., rotenone and DMSO in the control plate, or test compound in the compound plate) to the assay plate using a Wako Pintool station.

   To avoid cross contamination, it is important to wash pins with appropriate reagents (e.g., DMSO and methanol) before and after use.

   The user can add 10 µl of compound solution diluted with PBS or assay medium on top using multichannel pipettes if performing the assay in 96-well or 384-well assay plate.

6. Cover the assay plate with the stainless-steel lid again and place it back in the incubator for another 24–48 hr.

BASIC PROTOCOL 3: High-content Imaging and Analysis

High-throughput imaging enables visualization and quantification by capturing many cellular features on a large scale, using automated microscopy and image analysis platforms. High-content screening, or automated microscope-based screening, is one of such high-throughput imaging approaches, combining multi-parametric microscopy with quantitation of a large dataset of morphological features. Measuring the length of neurite outgrowth is an important aspect of determining neurotoxic effects of a chemical, and one can do this by using a high-content imaging platform to visually examine the cells in a time-dependent manner (Li & Xia, 2019). Here, we describe how to measure neurite outgrowth and analyze the imaging data using Operetta CLS High Content Analysis System. In our experience, use of an objective of 20x water and selecting focus mode as confocal result in better image quality than 20x air and non-confocal, for 1536-well plate assays.

After taking an image of each well, the image data can be analyzed using the platform’s image analysis software (we use here Harmony 4.6) by converting the image into data for comparison between treatment and control groups. This protocol also describes a way to analyze the data generated from this complex high-throughput high-content imaging assay, to extract the relevant information regarding the effect of test chemicals on neurite outgrowth.

Materials

Treated plate from Basic Protocol 2

Breathable Microporous Sealing Film (USA Scientific, cat. no. 2920-1010)

Operetta CLS High Content Analysis System (PerkinElmer, cat. no. HH16000000) with Harmony software 4.6 or equivalent.

Protocol Steps

Image Measurement

• 1Open Harmony software and login in the computer connected to Operetta.
• 2Remove the assay plate from the incubator (Step 6 in Basic Protocol 2) and replace the lid with a sterilized breathable adhesive film. Before loading the plate, clean the bottom of the assay plate with fine paper towel to avoid particles interfering with the imaging and place it onto the plate holder of the Operetta. Click the “load” button on the Harmony screen so that the assay plate is loaded into the Operetta.

  Using the sterilized breathable adhesive film to maintain the neuron alive during imaging and keep the assay plates usable for subsequent imaging at the later time.

• 3Under “Setup”, select “Plate type”. In this experiment we selected Aurora 1536-well low base PDL coated low base microplate. Users can select different plate type based on what they have used.
• 4Select “Objective” as 20x Water, “Mode” as Confocal, and “Binning” as 2 for a 1536-well plate. Then, select channel as EGFP (460–490 nm excitation, 500–550 nm emission for EGFP) for “Channel Selection”.

  Use 10x air for 96-well plate and 20x air for 384-well plate.

• 5On the “Navigation” menu, select one well and one field, and then adjust exposure parameters such as time, power, and height to focus the image.

  Test a few other wells to confirm the focus.

• 6Under “Layout Selection”, select the outcome parameters for neurite outgrowth, such as total neurite length per well, root number per well, neurite segments per well, and maximum neurite length per well
• 7Highlight all wells and fields you want to image from Navigation, give a file name for saving, and start measurement (run experiment).

  The user can select one to four fields to image per each well in a 1536-well plate (four fields can cover all images in a well and take about 90 min for one plate; for a quick screening one field can be selected to measure). Save each plate measurement with a file name.

  See sample data in Figure 3.

• 8After measurement is completed, remove the breathable adhesive film, place the lid back on top of the plate, and put the plate back into an incubator for next measurement, if needed (e.g., as part of a time course).

  To evaluate compound cytotoxicity, a cytotoxicity test, such as Calcein Red™ AM (AAT Bioquest) (Li et al., 2021), can be done at the end of the experiment.

• 9Analyze images using the module for neurite outgrowth called CSIRO Neurite Analysis Method within the Harmony 4.6 High Content Imaging and Analysis software package.

Figure 3. Representative images of cortical neurons treated with rotenone.

Neurons in PDL-coated 1536-well plates were treated with different concentration of rotenone (0, 0.36, 6, and 23 µM) for 48 hr in incubator at 37°C with 5% CO₂. Image is one field, 20x water, confocal. Scale bar: 100 µm.

Image Analysis

Four parameters (algorithmic outputs), namely, total neurite length, root number, neurite segments, and maximum neurite length are used for quantitative image analysis of neurite outgrowth. Number of objects is used to indicate cell number measured by the imaging system through a step that first involves finding the nucleus. The following steps describe how to analysis image.

• 10Under the “Image Analysis function” in Harmony, select a measurement file and then select an image and field from a well.
• 11When the image shows on the computer monitor, find nuclei through EGFP channel for cell body and name it as “All Cells” (Number of objects), for data output.
• 12Find neurites through EGFP channel and use “Population of All Cells”, “Region of Nucleus”, and “Method of CSIRO Neurite Analysis” tools.

  See sample data in Figure 4 (four fields, 20x water, confocal) or Figure 5 (one field).

• 13Select the following outputs from the software: neurite segments, total neurite length, root number, and maximum neurite length per well per cell.
• 14Highlight fields in each well, select the images per well you analyze, and start analyzing each image by clicking on “Evaluation”. The image data are converted to a table of numerical numbers. The concentration-response curve of a compound and its EC₅₀ will show up when you highlight the wells treated with the positive control compound.

  Conversion of image data to numerical numbers for one 1536- well plate (four fields/well) takes about 15 minutes.

• 15Select the table with numerical data, and then save as an Excel file for further data analysis. See Understanding Results for a discussion on data analysis and data presentation

Figure 4. Example of four-field image analysis steps shown for neurite outgrowth in protocols described in Basic protocol 3.

(Left) The figure shows cortical neuron 90 hr post seeding (four fields of 1536-well plate, 20x water, confocal). The middle panel shows how to find nuclei, and the right panel shows how to find neurites. Scale bar: 100 µm.

Figure 5. Example of one field image and analysis steps shown for neurite outgrowth.

Left: Cortical neuron 90 hr post seeding, one field of 1536-well plate, 20x water, confocal. Also, finding nuclei step to show cell body (middle) and finding neurite step to show neurite (right). Scale bar: 100 µm.

COMMENTARY

Background Information

Neurotoxicity and developmental neurotoxicity can alter normal activity of the nervous system, which is a major cause of neurological diseases. To evaluate environmental chemicals for their neurotoxicity potential, several in vitro neurotoxicity assays, including cell viability, neuronal differentiation, neurite outgrowth, and neuron functions (e.g., spontaneous electrical activity of neuronal networks) have been developed (Aschner et al., 2017). A limitation of the viability assay is that it only measures cell death but cannot assess phenotypic changes of neurons. Neuronal functional assays (e.g., MEA), on the other hand, are not easy to use in the context of in a high-throughput screening platform. Neurite outgrowth measurement is considered a critical assay for evaluating an important cellular event within the development of the nervous system (Aschner et al., 2017). Indeed, utilizing neurite outgrowth assays in a high-throughput format and using high-content platform to screen compounds for developmental neurotoxicity or neurotoxicity, has previously been reported (Ryan et al., 2016; Sirenko et al., 2014). In general, either immunostaining or dye staining has been used to measure neurite outgrowth. For example, Radio et al., used immunostaining β-III tubulin to measure neurite outgrowth assay in a PC-12 rat neuronal cell line (Radio et al. 2008). Tubulin staining method requires multiple washing steps, which is time-consuming and easily causes well-to-well variation. Another type of neurite outgrowth assay, using Calcein AM to stain live cell structures, has been developed for high-content imaging (Krug et al. 2013a). However, the Calcein AM staining method still needs staining and causes cytotoxicity if the staining remains for an extended period of time. As such, this method is not amenable to time-lapse readouts.

The assay described here is a neurite outgrowth assay that can measure neurite outgrowth in real time because it uses GFP-labeled human neurons (Li et al., 2021). This robust method does not require washing steps within the experiment; thereby, it greatly improves screening throughput and reduces well-to-well variation (Li et al., 2021). In addition, eliminating the washing steps reduces cell disturbance as well as the amount of time the assay will take to complete. More importantly, this assay enables a live and time-lapse imaging of neurite outgrowth, allowing for the observation of chemical effects over time. Consequently, we have been able to obtain neurite outgrowth readouts at 24 and 48 hrs after compound treatment within one continuous experimental assay. The details about development, optimization, and validation of this assay have been reported previously (Li et al., 2021), demonstrating the efficiency and reliability of the described protocol. Using this assay, we have shown that the well-known neurotoxic compound, rotenone, displayed concentration-dependent neurite outgrowth inhibition in both cortical and motor neurons. The relative activity was compared to the vehicle control, DMSO, across multiple endpoints (total neurite outgrowth, neurite segments, maximum neurite length, and number of objects) in 1536-well plates. The data demonstrated that human iPSC-derived neurons are amenable for high-throughput assays of neurite outgrowth using high-content imaging, supporting the use of this platform for testing the effect of test compounds with neurotoxicity potential.

When performing the neurite outgrowth assay, the number of objects in the same well can be measured and used to identify and count the number of cells, in order to evaluate neuronal cell viability (cytotoxicity) due to compound treatment. Additionally, at the end of imaging, cytotoxicity can also be evaluated using Calcein Red™ AM (AAT Bioquest) (Li et al., 2021).

Critical Parameters

The neurite outgrowth assay using GFP-labeled human iPSC-derived neurons enables direct, live, and time-lapse imaging of neurites without wash steps. To get high quality images and reproducible results, proper thawing and seeding of neurons is critical. Before thawing neurons, users should prepare thawing and seeding medium freshly and carefully according to the components listed in Table 1, and perform thawing procedure gently and slowly, as described in Basic Protocol 1. The thawed neurons should be filtered through a cell strainer to obtain a single-neuron suspension and avoid cell clumping before counting. If a Multidrop Combi Dispenser or BioRAPTR FRD is used to dispense cells, it is very important to keep a single-neuron suspension to minimize dispenser-tubing clogging. The BioRAPTR FRD is a liquid handling system that can transfer 0.2–10 µl of up to four different reagents or cells simultaneously into a 1536 well plate. Alternatively, a Multidrop Combi Reagent Dispenser, a high-speed dispenser capable of emitting one reagent or cell suspension at a time using an eight-channel detachable dispensing cassette, may be utilized.

Viable neuronal cells should be counted with a hemocytometer to determine their density. The culture density should be adjusted to 12–16 × 10⁴ neuronal cells/ml to obtain 600–800 cells/well when using 1536-well plates. However, cell number per well should be optimized before a formal experiment. Usually, one vial (around 5×10⁶ cells) of neurons encompasses enough for three 1536-well plates. To get quality results, cell viability should be more than 80%.

To protect the cells from contamination during the dispensing process, Pen Strep can be added to the seeding medium (at half the concentration used in cell culture), which contains 11 components in total (see Table 1).

Stainless steel lids used for the assay described in the Materials section should contain small, evenly-placed holes, to allow for air exchange necessary in cellular assays. This cellular assay lid also contains a rubber gasket that sits around the top outer edge. The weight of the lid allows the gasket to form a strong barrier around the plate, virtually eliminating edge effects. Disinfect the lid with 70% v/v ethanol before use, to avoid contamination. The assay plate should be kept in an incubator at 37°C with 5% CO₂ and 95% humidified atmosphere. Check the plate under the microscope every day, to evaluate neuron and neurite conditions.

A CyBi-Well Vario Pipettor, (96-, 384-, and 1536-channel simultaneous pipette) is the most suitable piece of equipment when preparing the positive control plate in a 1536-well plate format. This pipette requires the use of disposable tips.

The PinTool station (Wako Automation) is needed to transfer 23 nl of test compounds, and positive and negative controls, from the 1536-well compound and control plate to a 1536-well assay plate with neurons.

We commonly use 10 mM compound concentration as a stock solution because most of compounds are soluble at this concentration. When making multiple concentrations, we usually use 1:2 or 1:3 serial dilution of the compounds with DMSO.

During the imaging process, it is critical to select the right plate type at “Setup”, and to use confocal and 20x magnification with water for a 1536-well plate. Since the well of 1536-well plate is tiny, focusing is important for image quality. Adjust exposure parameters such as time, power, and height to make the image clear. Using low base 1536-well plate can help increase the image quality, and 20x water is better than 20x air.

Troubleshooting

This robust high-throughput, high-content imaging neurite outgrowth assay is reliable and reproducible. It is not a difficult assay to perform when done carefully. However, great attention must be given to each step of the procedure, especially the thawing and seeding neurons step. Thaw and seed the neuron cells slowly and gently, as mentioned in Critical Parameters. Common problems with the protocol, their causes, and potential solutions are listed in Table 2.

Table 2.

Troubleshooting guide

Problem	Possible Cause	Solution
Neurons not growing or dying	Inappropriate seeding medium DMSO toxicity	Use correct medium according to manufacturer’s protocol Test DMSO concentrations on the neurons during assay optimization
Neurites not formed	Inappropriate Plate Inappropriate thawing and seeding process Inappropriate culture conditions	Use coated plate. Check thawing and seeding medium; thaw neurons slowly and gently. Check incubator’s temperature, humidity, and CO2
No GFP signal	Select wrong channel during imaging Inappropriate plate Image out of focus	Select right channel (e.g., 460–490 nm excitation, 500–550 nm emission) Use plate with clean and clear bottom Adjust setup parameters to focus the bottom of the plate
Image too crowded	Too many cells seeded per well	Adjust cell density. For a 1536- well plate, seeding 600–800 neuron cells per well is recommended.
Neuron clump showing in image	Neurons clump together when thawing	Filter neuron suspension through a cell strainer to get a single-cell neuron suspension.
Image with spotty background	Contamination	Add penicillin/streptomycin into seeding medium.
Image not clear	Lack of focus	Check plate type and readjust the exposure parameters such as time, power, and height etc.
Large well-to-well variation	Different neuron numbers in the well Use inappropriate algorithmic outputs	Use cell dispenser (e.g., BioRAPTR) to plate cells Optimize algorithmic outputs (e.g., total neurite length, root number, neurite segments, and maximum neurite length) and select the best one to measure neurite outgrowth

Understanding Results

When performing data analysis, we measure neurite outgrowth using the following parameters: total neurite length, neurite segments, maximum neurite length, and number of objects. The raw data are saved in Microsoft Excel and then normalized using positive and DMSO controls. The normalized data can be transferred to any plotting software (e.g., GraphPad) for calculating compound IC₅₀ (concentration of half-maximal inhibition) and efficacy (% inhibition). Figure 6 shows total neurite length per cell per well measured at different times from the same plate, which demonstrates that neurites grow quickly in the first three days of culturing, and then slows down. We selected the treatment time and length (e.g., 40–42 hr culturing; 24 or 48 hr treatment) based on this observation. Figure 7A shows the inhibitory effect of rotenone, the positive control, on neurite outgrowth.

Figure 6. Time-course of neurite outgrowth for cortical neurons in 1536-well plates.

Neurite outgrowth of cortical neurons in 1536-well plates. Cortical neurons labeled with GFP were imaged at 22, 42, 66 and 90 hr continuously. Total neurite length was used to evaluate neurite outgrowth. The unit of neurite length is expressed as pixels. Each value represents the mean ± SD of 668 assay wells.

Figure 7. Concentration-response of rotenone in cortical cells in a 1536-well plate.

(A) neurite growth and (B) cytotoxicity. Cortical neurons were treated with rotenone at multiple concentrations ranging from 1.4 nM to 46 µM for 48 hr. The concentration-response curve was generated by GraphPad software. Each value represents the mean ± SD from one experiment in duplicate.

For comparison, number of objects can be used when evaluating cell viability of each compound treated in the neuronal cells. For instance, Figure 7 shows the inhibitory effect of rotenone to neurite outgrowth (Fig. 7A) as well as cytotoxicity generated from the assay treatment (Fig. 7B). The neurite outgrowth IC₅₀ of rotenone is around 1.6 μM, whereas rotenone only affects the number of neurons at high concentrations (46 μM).

These results can also be examined qualitatively, directly through the visualization of images of neurite outgrowth. Figure 3 shows the images of neurite outgrowth treated with rotenone at three different concentrations at a single time point (i.e., 48 hr), showing rotenone inhibiting neurite outgrowth in a concentration-dependent manner.

Time Considerations

It takes 4–5 days to complete the assay, including incubation time for cell attachment and neurite formation. We suggest thawing neurons on Monday afternoon (Day 1), treating with compounds on Wednesday morning (Day 3, 40–42 hr after growing in incubator), and taking images on Thursday morning (Day 4, 24 hr in incubator after treatment) or Friday morning (Day 5, 48 hr treatment).

It takes 90 min to measure an image of four fields in each well, from a 1536-well plate, with a 20x water confocal. Subsequently, it takes 10–15 min to then evaluate the images. For quick image measurement, one field can also be considered, cutting the time down to 24 min per plate.

Data Availability Statement

Data sharing is not applicable to this article as no new data were created or analyzed in this study.