2024 Volume 47 Issue 2 Pages 366-372
Neuronal regrowth after traumatic injury is strongly inhibited in the central nervous system (CNS) of adult mammals. Cell-intrinsic and extrinsic factors limit the regulation of axonal growth and regrowth of fibers is minimal despite nearly all neurons surviving. Developing medical drugs to promote neurological recovery is crucial since neuronal injuries have few palliative cares and no pharmacological interventions. Herein, we developed a novel in vitro axonal regeneration assay system to screen the chemical reagents using human-induced pluripotent stem cell (hiPSC)-derived neurons. These neurons were cultured in a 96-well plate to form a monolayer and were scraped using a floating metal pin tool for axotomy. The cell number and plate coating conditions were optimized to score the regenerating axon. Treatment using the Rho-associated kinase (ROCK) inhibitor Y-27632 enhanced axonal regeneration in this regeneration assay system with hiPSC-derived neurons. Therefore, our novel screening method is suitable for drug screening to identify the chemical compounds that promote axonal regeneration after axotomy under in vitro conditions.
No medical intervention exists to promote neural recovery after a stroke or spinal cord injury. Neurological injuries, including spinal cord trauma, cause devastating and persistent disabilities despite nearly all neurons surviving. Neuronal network disconnection due to axonal damage causes functional deficits. Cell-autonomous and extrinsic inhibitors limit the regeneration of axons and absence of stimulatory factors. Extrinsic inhibitors, including oligodendrocyte-derived proteins, Nogo-A (RTN4A), myelin-associated Glycoprotein (MAG), and oligodendrocyte myelin glycoprotein (OMgp),1–4) and glia-derived proteins, such as chondroitin sulfate proteoglycans (CSPGs) from reactive astrocytes,5,6) are targeted in the development of pharmacological interventions after injury. Phosphatase and tensin homolog (PTEN),7) Krüppel-like factors (KLFs),8) or suppressors of cytokine signaling 3 (SOCS3)9) are well-studied cell-intrinsic growth inhibitors after axotomy. Despite considerable progress, a complete mechanism of neural repair after trauma to the central nervous system (CNS) has not yet been unveiled.
Therapeutic drugs for CNS injury are absent; therefore, exploring and developing pharmacological compounds are necessary. Several in vitro models have been conducted to study neuronal regeneration.10) One of the cell-based in vitro models, the neurite outgrowth assay, is the most commonly used test to evaluate genetic or pharmacological effects on axonal growth in rodents. Distinguishing between neurite initiation and axonal elongation from injured neurons is challenging since dissociating neurons from mouse or rat tissues is thought to cause extensive severing and induce injury responses. The scraping assay is beneficial in quantifying preexisting axonal regrowth. Therefore, we developed a scraping-assay-based screening system to evaluate axonal regeneration in human-induced pluripotent stem cell (hiPSC)-derived neurons.
Previously, we have demonstrated an in vitro scraping assay as an axon injury model in a 96-well format using rat or mouse embryonic cortical cultured neurons.11) Using this system, we successfully screened >16000 mouse genes and identified approximately 400 that could potentially regulate axonal regeneration.12) In the present study, we performed an in vitro regeneration assay using hiPSC (line 1231A313))-derived neurons instead of rodent neurons. A floating metal pin tool was used for axotomizing hiPSC-derived neurons that were cultured in 96-well plates. We examined several conditions, such as coating reagents, culture duration, and cell number, and determined the optimal settings. Enhanced axonal regeneration of hiPSC-derived neurons was observed by treating with the Rho-associated kinase (ROCK) inhibitor Y-27632 in this in vitro regeneration assay. Therefore, we developed a novel axonal regeneration assay using hiPSC-derived neurons.
The following reagents were used in this study: A-83-01, Fibronectin, non-essential amino acids, and recombinant human brain-derived neurotrophic factor (BDNF) (FUJIFILM Wako Pure Chemical Corporation, Osaka, Japan); N-[2S-(3,5-difluorophenyl) acetyl]-L-alanyl-2-phenyl-1, 1-dimethylethyl ester-glycine (DAPT) , LDN-193189, PD0325901, SU5402, Y-27632, and XAV-939 (Selleckchem, Houston, TX, U.S.A.); B-27 supplement without vitamin A, Essential 6, Essential 8, GlutaMax, Neurobasal-A, Rhodamine-phalloidin, and TrypLE (Thermo Fisher Scientific, Waltham, MA, U.S.A.); Laminin from Engelbreth-Holm-Swarm murine sarcoma basement membrane, and L-ascorbic acid (Sigma-Aldrich, St. Louis, MO, U.S.A.), iMatrix-511 silk (Matrixome, Osaka, Japan).
Generation of Cortical Neurons from hiPSCsThe hiPSC line 1231A313) was provided by the RIKEN BRC through the National BioResource Project of the Ministry of Education, Culture, Sports, Science and Technology (MEXT)/Japan Agency for Medical Research and Development (AMED), Japan. Studies of hiPSCs were conducted with approval of the Ethical Review Committee for Medical and Health Research Involving Human Subjects of the Kyoto Pharmaceutical University. hiPSCs were maintained as described previously.14,15) Cortical neurons were differentiated from P20-P22 hiPSCs using feeder-free 2D monolayer culture system as described previously.16)
Axon Regeneration AssayThe axon regeneration assay with mouse primary neurons performed previously was adapted to this hiPSCs derived neuron study. Day 11 neurons (1 × 105cells) differentiated from hiPSCs were plated on 96-well tissue culture plates coated with the indicated reagents in 200 µL of neurobasal-A containing B27, GlutaMax, supplemented with ascorbic acid (200 µM, days 11–46), BDNF (20 ng/mL, days 11–46), DAPT (10 µM, days 11–46), PD0325901 (1 µM, days 11–18), SU5402 (5 µM, days 11–46) and Y-27632 (10 µM, day 11). Then, half the volume of media was replaced every day. At day 16, 96-well cultures were scraped using a floating pin tool and allowed to regenerate for the indicated periods. The scraping pin tools scratch pins (TK-SP01) modified with 350 µm diameter were attached to scratch plate tool (TK-SP05) were obtained from Tokken (Chiba, Japan).
Then neurons were fixed with 4% paraformaldehyde. Regenerating axons in the scrape zone were visualized using an antibody against βIII-tubulin (1 : 2000, Promega, Madison, WI, U.S.A.). Images were taken on a 10× objective with Operetta (PerkinElmer, Inc., Waltham, MA, U.S.A.) or All-in-One Fluorescence Microscope (Keyence, Osaka, Japan). The scraped area of the stained image was cropped with a 500–600 µm width rectangle, the image was thresholded and neurite length was quantitated using ImageJ to determine the extent of axon regeneration. The ImageJ script was; run (“8-bit”); setThreshold (XX, XX); setOption (“BlackBackground,” false); run (“Convert to Mask”). Measurements from different 6 to 10 wells for the same condition in any one experiment were averaged together for one n value.
Indirect Immunofluorescence Confocal MicroscopyNeurons (1 × 106 cells per well) differentiated from hiPSCs at day 11 were re-seeded on a poly-D-lysine (PDL)/laminin (L)-coated cover glasses in a 12 well plate. At 5 d after re-seeding, neurons were fixed with a solution containing 4% paraformaldehyde for 15 min and then permeabilized with 0.1% Triton X-100 in phosphate-buffered saline (PBS) for 15 min. The samples were incubated overnight at 4 °C with anti-βIII-tubulin (1 : 100; Promega), MAP2 (1 : 100; Cell signaling Technology, Danvers, MA, U.S.A.), VGLUT1 (1 : 200; Proteintech, Rosemont, IL, U.S.A.), glial fibrillary acidic protein (GFAP) (1 : 100; Sigma-Aldrich) and Iba1 (1 : 100; FUJIFILM Wako Pure Chemical Corporation) antibodies. Then, either Alexa 488-conjugated donkey anti-mouse immunoglobulin G (IgG) and Alexa 594-conjugated donkey anti-rabbit IgG (1 : 100; all from Invitrogen, Waltham, MA, U.S.A.) were used to detect primary antibodies. Samples were mounted with mounting solution with dapi (Nacalai Tesque, Kyoto, Japan) and observed using a Zeiss LSM800 laser scanning confocal microscope (Carl Zeiss, Oberkochen, Germany).
RT-PCR and Quantitative PCRTotal RNA was prepared using RNAiso Plus (TaKaRa Bio; Shiga, Japan) and subjected to reverse transcriptase (RT)-PCR using ReverTra Ace qPCR RT Kit (Toyobo; Osaka, Japan). Quantitative real-time PCR analyses was conducted using a THUNDERBIRD Next SYBR qPCR Mix (Toyobo) on a CFX Connect Real-Time System (Bio-Rad, Hercules, CA, U.S.A.) with standard cycles. Following primers were used; hPAX6 (NM_001368930.2), F-GACTGCATTTGAAGGCCTGG, R-GGCCTCAATTTGCTCTTGGG; hSOX2 (NM_003106.4), F-AACCAGCGCATGGACAGTTA, R-CGAGCTGGTCATGGAGTTGT; hVGLUT1(NM_020309.4), F-AGCTGGGATCCAGAGACTGT, R-CGAAAACTCTGTTGGCTGC;hGAPDH, F-GAAATCCCATCACCATCTTCCAGG, R-CAGTAGAGGCAGGGATGATGTTC.
Cell Viability AssayCell viability was determined by Cell Count Reagent SF (Nacalai Tesque) according to the manufacturer’s instructions. The absorbance was measured at a test wavelength of 450 nm (OD450) and a reference wavelength of 590 nm (OD590) using a microplate reader (Tecan, Zürich, Switzerland).
Statistical AnalysisThe significance of differences between group means was determined by Student’s t-test as specified in the figure legend using Excel software.
Recent advances in stem cell research have succeeded in producing neurons from hiPSC, and several studies have attempted to differentiate between several neuronal types. We established an in vitro regeneration assay using hiPSC-derived neurons by choosing cortical neurons because previous primary neuron-based regeneration assays have used cortical neurons from embryonic mice or rats.11) Neurons differentiated from hiPSCs were cultured on a poly-L-ornithine (100 µg/mL)/fibronectin (2 µg/mL)/laminin (10 µg/mL) (OFL)-coated 96-well plate according to the methods followed in previous studies.14) At differentiated day 11, more than 90% cells were PAX6 positive (data not shown). We developed an in vitro regeneration assay using scraping for axotomy, we scraped hiPSC-derived neurons on day 15 after differentiation (day 5 after reseeding on the OFL). Most neurons aggregated and formed colonies without scraping, and scraping resulted in further aggregation, resulting in the neurons easily detaching from the culture plate during immunostaining (Fig. 1A). Since mouse or rat primary neurons for in vitro regeneration assay were cultured on poly-D-lysine (100 µg/mL) (PDL)-coated well, we sought to culture the hiPSC-derived neurons on PDL-coated well for scraping. Neurons on the PDL-coated wells spread and diffused on the well surface as a monolayer, and they did not detach during immunostaining (Fig. 1B). We evaluated other coating conditions using a series of different reagent combinations to determine the suitable conditions for scraping. At differentiated day 11, the neurons were re-seeded on OFL, poly-L-ornithine (100 µg/mL)/laminin (10 µg/mL) (OL), fibronectin (2 µg/mL)/laminin (10 µg/mL) (FL), laminin (10 µg/mL) (L), PDL, or PDL/laminin (10 µg/mL) (PDL/L). Five days after culturing on the differently coated wells, the neurons were scraped and cultured for another 2 d. The neurons were subsequently fixed and stained with βIII-tubulin antibody for neuronal detection. Consequently, the neurons in the OFL-, OL-, FL-, and L-coated wells were detached or highly aggregated. Neurons in the PDL-or PDL/L-coated wells diffused and firmly attached to the well surface (Fig. 1C). Additionally, we evaluated different laminin concentrations (0.5, 1, 2, and 10 µg/mL) because PDL/laminin (10 µg/mL) coating caused the regenerating neurites to get entangled. The lowest laminin concentration resulted in no entangled neurite regeneration (data not shown). Therefore, we decided to use 0.5 µg/mL laminin coating for further experiments.
(A, B) Day 11 neurons derived from hiPSCs (1 × 105 cells per well) were re-seeded on a poly-L-ornithine/fibronectin/laminin (OFL) (A), or poly-D-lysine (PDL) (B)-coated 96 well plate. At 5 d after re-seeding, neurons were scraped with pipet tip (injured). The microphotographs show βIII-tubulin (green), phalloidin (red) and dapi (blue). Dashed lines indicate the scraped area. Scale bars represent 200 µm. (C) HiPSCs derived neurons (1 × 105 cells per well) seeded on OFL, poly-L-ornithine/laminin (OL), fibronectin/laminin (FL), laminin (L), PDL, or PDL/laminin (PDL/L) were scraped and observed with microscope. The morphology of neurons in each condition were summarized as a table.
Since hiPSC-derived neurons on PDL and PDL/L coatings could be useful for the scraping experiments, we evaluated regeneration under these two conditions. At differentiated day 11, the neurons were reseeded on a PDL (100 µg/mL) or PDL (100 µg/mL)/L (0.5 µg/mL)-coated 96-well plate. Five days after culture, the neurons were scraped and allowed to regenerate for another 2 d. While the neurons on the PDL did not regenerate, those on the PDL/L-coated wells regenerated into the scraped area (Fig. 2A). Thus, we defined the laminin and PDL coating combination as the best condition for the scraping assay using hiPSC-derived neurons. Further, we compared whether the regenerating neurites are βIII-tubulin or MAP2 positive (Supplementary Figs. 1A, B). The vast majority of neurites regrowing into the scraped area were βIII-tubulin positive (>95%), and about 5% of neurites were MAP2-positive. Therefore, we assumed that most of the regenerating neurites were axons. We also checked the neural population of the hiPSC-derived cells on a PDL/L coating plate using anti-βIII-tubulin, Microtubule-associated protein 2 (MAP2), GFAP, and Iba1 antibodies. Most cells were βIII-tubulin and MAP2 positive; however, neither GFAP nor Iba1 positive cells were detected (Fig. 2B). Moreover, the vast majority of cells were VGLUT1-positive neurons (Supplementary Fig. 1C).
(A) Day 11 neurons derived from hiPSCs were re-seeded on a PDL or PDL/L-coated 96 well plate. At 5 d after re-seeding, neurons were scraped and allowed to regenerate for 2 d. Then neurons were fixed and stained with anti-βIII-tubulin antibody (green). Scale bars represent 300 µm. (B) Day 11 neurons (1 × 106 cells per well) derived from hiPSCs were re-seeded on a PDL/L-coated cover glasses in a 12 well plate. At 5 d after re-seeding, neurons were fixed and stained with anti-βIII-tubulin, MAP2, GFAP and Iba1 antibodies. Scale bars represent 20 µm.
Next, the optimal cell number for the regeneration assay was determined. Neurons differentiated from hiPSCs on day 11 were seeded on a PDL/L coated 96-well plate at a density of 8 × 104, 1 × 105, or 1.2 × 105 cells per well. The neurons were axotomized 5 d after culture using a pin tool and allowed to regenerate for another 2 d. Subsequently, the neurons were stained with βIII-tubulin antibody to detect the regenerating axons under the microscope (Fig. 3A). Lower and higher neuronal numbers indicated less and robust regeneration, respectively. The edge of the scraped area should be straight and not wound to quantify the regenerating axons (Fig. 3B). Thus, we determined 1 × 105 cells per well as the optimal concentration for the regeneration assay (Fig. 3C). Additionally, we compared the duration of regeneration after scraping. Day 11 hiPSC-derived neurons were seeded on a PDL/L coated 96-well plate at a density of 1 × 105 cells per well. The neurons were axotomized 5 d after culture and allowed to regenerate for 1, 3, 5, or 7 d (Fig. 3D). Neuronal regeneration began 1 d after scraping and reached the center of the scraped area on day 3. On day 7 after scraping, the regenerating axons filled the scraped area (Fig. 3E). Since axonal regeneration was examined using relatively immature neurons (days 16–23 after differentiation), we investigated whether neurons cultured for longer periods could regrow using this scraping method. Day 11 hiPSC-derived neurons were reseeded on a PDL/L coated 96-well plate at a density of 1 × 105 cells per well and cultured for 4 weeks (Fig. 3F). At 28 d (day 39 after differentiation) after replating, the neurons were axotomized and cultured for a further 7 d. The neurons were subsequently fixed and stained with βIII-tubulin antibody for visualization (Fig. 3G). Although older neurons (Fig. 3G) exhibited lesser regeneration than younger neurons (Fig. 3E), several neurons exhibited robust regeneration.
(A) Day 11 neurons (at the indicated number of cells) derived from hiPSCs were re-seeded on a PDL/L-coated 96 well plate. At 5 d after re-seeding, neurons were scraped and allowed to regenerate for 2 d. Then neurons were fixed and stained with anti-βIII-tubulin antibody (green). Scale bars represent 200 µm. (B) Scheme of in vitro axon regeneration assay. Neurons were cultured in a 96-well plate as a monolayer and scraped with a pin tool as axotomy. After scraping, neurons were fixed and stained with βIII-tubulin antibody and taken photos. Scraped area was cropped (blue rectangle) and the fluorescent intensity was measured as regenerating axons. (C) Regenerating pattern in each cell number was summarized in a table. (D) Timeline for this experiment. (E) Day 11 neurons (1 × 105 cells per well) derived from hiPSCs were re-seeded on a PDL/L-coated 96 well plate. At 5 d after re-seeding, neurons were scraped and allowed to regenerate for the indicated periods. Then neurons were fixed and stained with anti-βIII-tubulin antibody (green). Scale bars represent 300 µm. (F) Timeline for this experiment. (G) Day 11 neurons (1 × 105 cells per well) derived from hiPSCs were re-seeded on a PDL/L-coated 96 well plate. At 28 d after re-seeding, neurons were scraped and allowed to regenerate for 7 d. Then neurons were fixed and stained with anti-βIII-tubulin antibody (green). Scale bars represent 300 µm.
This study aimed to establish a high-throughput screening for drugs that regulate axon regeneration. Therefore, we sought to determine whether treatment with the drug affected the regeneration of hiPSC-derived neurons. RhoA and ROCK signaling inhibit axonal regeneration by altering the F-actin cytoskeleton, and ROCK inhibition facilitates axonal growth.17) We have previously reported that the ROCK inhibitor Y-27632 enhances axonal regeneration in mouse cortical neurons.18,19) Thus, we used Y-27632 in an in vitro regeneration assay using hiPSC-derived neurons. After 5 d of culture on a PDL/L-coated plate, the neurons were scraped and allowed to regenerate for 3 d with or without Y-27632 treatment. The neurons were subsequently stained with βIII-tubulin antibody for visualization (Fig. 4A). The intensity of regenerating axons positive for βIII-tubulin was measured for quantification (Fig. 3B). Compared with the dimethyl sulfoxide (DMSO)-treated control, Y-27362 treatment resulted in the significant enhancement of axonal regeneration (Fig. 4B). Y-27362 treatment had no effect on cell viability (p = 0.178) (Fig. 4C), suggesting that the enhancement of regeneration was not affected by cell viability. We also analyzed the gene expression patterns of PAX6, SOX2 and VGLUT1 in neurons after Y-27632 treatment (Fig. 4D). Compared to DMSO treated control, Y-27632 had no effect on the expression of these genes.
(A) Day 11 neurons (1 × 105 cells per well) derived from hiPSCs were re-seeded on a PDL/L-coated 96 well plate. At 5 d after re-seeding, neurons were scraped and allowed to regenerate for 3 d without (DMSO) or with Y-27632 (5 µM). Then neurons were fixed and stained with anti-βIII-tubulin antibody (green). Scale bars represent 300 µm. (B) The graph shows quantification of axonal regeneration relative to control (DMSO). Mean ± standard error of the mean (S.E.M.), n = 8 biological replicates. *** p < 0.005, Student’s two-tailed t test. (C) Day 11 neurons (1 × 105 cells per well) derived from hiPSCs were re-seeded on a PDL/L-coated 96 well plate. At 5 d after re-seeding, neurons were scraped and allowed to regenerate for 3 d without (DMSO) or with Y-27632 (5 µM). Cell viability was measured with WST8 assay. Data show mean ± S.E.M. n = 3 independent experiments. n.s., not significant. Student’s two-tailed t test. (D) Day 11 neurons (3 × 105 cells per well) derived from hiPSCs were re-seeded on a PDL/L-coated 24 well plate. At 5 d after re-seeding, neurons were treated with DMSO or Y-27632 (5 µM) for 3 d. Then, total RNA samples were subjected to RT-PCR. Data are PAX6, SOX2 and VGLUT1 mRNA levels normalized to those of a GAPDH internal control with ± S.E.M. n = 4, n.s., not significant. Student’s two-tailed t test.
Herein, we developed a human cell-based in vitro axonal regeneration assay for an unbiased, high-throughput screening. This is the first known model of scraping injury used to quantify axonal regrowth after axotomy in hiPSC (line 1231A313))-derived neurons in a high-throughput study. An advantage of our scraping assay is the simultaneous creation of reproducible axonal injuries in 96-well plates. In a previous study, we successfully screened >16000 genes that affected axonal regeneration after neuronal axotomy of mouse primary cortical neurons using the scraping method.12) We have currently established a regeneration assay system for hiPSC-derived neurons; therefore, performing similar genome-wide screening using human neurons in the future is feasible. In this study, we investigated a single iPS cell line 1231A3,13) further studied should be performed to generalize this assay to any iPS cell line.
A neuron and a hiPSC can be differentiated in a monolayer culture using a proper combination of biochemical adhesion molecules for optimal neuronal differentiation to develop functional properties.20) We evaluated the different patterns of coating reagents to optimize the scraping assay and determined that the PDL/L coating was the best condition. Additionally, the duration of culture days affected the functional maturation of hiPSC-derived neurons as shown in previous studies.20,21) In the present study, day 16 hiPSC-derived neurons for axotomy were used for most experiments. Since robust axonal regeneration was observed even with prolonged culture time on day 39 after differentiation, this assay might have extended the culture duration to preferable time points beyond day 39. However, we have not evaluated older hiPSC-derived neurons beyond day 39. Future experiments will assess whether the older neurons can regenerate in this system. Additionally, the neuronal number and culture days after scraping affected the axonal regeneration ratio, and we could modify both parameters for the experiments, that is, prolonging the duration after scraping for inhibitory molecule screening or shortening it for promoting drugs. Because the scraped widths are not completely identical among the wells in the same 96-well plate (499.31 µm ± 25.56), multiple replicates may help for getting convincing results.
Although cell-based assays do not reflect much of the complex in vivo environment, exploring pharmacological drugs that increase axonal growth and in vitro axonal regeneration systems is beneficial for screening a compound library before studying animal models. Since primary neurons mimic the gene expression and phenotypic profiles of native neurons, using neuronal cells derived directly from primary tissues is more relevant to neurons in vivo than using immortalized cell lines.10) Several in vitro models have been used to analyze neuronal regeneration.10) Neurite outgrowth assays are commonly performed using cell-based in vitro models to evaluate axonal growth. Neurite growth assays have been used for high-throughput screening of hiPSC-derived neurons to identify neurotoxic environmental chemicals.22) Compared to the neurite outgrowth assays, the scraping method is beneficial for quantifying regrowth from injured preexisting axons. Stretch injury with silicone membranes attached to a 96-well plate was developed as an in vitro neuronal injury model and adjusted for hiPSC-derived neurons.23) However, this model is used to evaluate cell viability and neurite morphology but not axonal regeneration. Although laser-induced axotomy has been established as a model for in vitro injury,24) it has not been designed for high-throughput experiments. Therefore, we developed a scraping-assay-based screening system to evaluate axonal regeneration in hiPSC-derived neurons.
This work was supported by Grants from JSPS KAKENHI (Grant Number: 22H03544), and Takeda Science Foundation to Y.S.
The authors declare no conflict of interest.
This article contains supplementary materials.
