Migration, Chemo-Attraction, and Co-Culture Assays for Human Stem Cell-Derived Endothelial Cells and GABAergic Neurons

Endothelial cells form the lining of blood vessels and mediate important functions that include maintenance of vessel wall permeability, regulation of blood flow, platelet aggregation, and formation of new blood vessels. In the brain, endothelial cells form part of a critical blood-brain-barrier that tightly controls exchange of materials between the brain and the bloodstream¹. Our studies in the past decade have identified novel neurogenic roles of brain endothelial cells that have significant implications for brain development and behavior^2,3,4,5. We have shown that the mouse embryonic forebrain is vascularized by two distinct subtypes of vessels, the pial vessels and the periventricular vessels, that differ in anatomy, origin, and developmental profile². Endothelial cells lining these two vessel subtypes show distinct differences in their gene expression profiles. While pial endothelial cells mostly express genes related to inflammation and immune response, periventricular endothelial cells are uniquely enriched in expression of genes commonly associated with neurogenesis, neuronal migration, chemotaxis, and axon guidance³. Periventricular endothelial cells also house a novel GABA signaling pathway that is distinct from the traditional neuronal GABA signaling pathway⁵. Concomitant with its gene expression, periventricular endothelial cells were found to regulate migration and distribution of GABAergic interneurons in the developing neocortex. During embryonic development, periventricular endothelial cells undergo long-distance migration along a ventral-dorsal gradient to establish the periventricular vascular network^2,3. This migratory route is mirrored a day later by interneurons. Migrating interneurons physically interact with the pre-formed periventricular vascular network and use it as a guiderail to reach their final destination in the neocortex. In addition to acting as a physical substrate, periventricular endothelial cells serve as the source of navigational cues for migrating neurons. Periventricular endothelial cell-secreted GABA guides interneuron migration and regulates their final distribution patterns⁴. Defects in interneuron migration and distribution are associated with neuropsychiatric disorders such as autism, epilepsy, schizophrenia and depression^6,7,8,9,10. Therefore, study of periventricular endothelial cell functions and their influence on interneuron migration in human context becomes critical for addressing the pathogenesis of these disorders.

We have generated human periventricular-like endothelial cells from human embryonic stem cells in our laboratory¹¹, using induced pluripotent stem cell (iPSC) technology^12,13. To validate whether human periventricular endothelial cells faithfully mimic mouse periventricular endothelial cells, and to quantitatively assess their influence on interneuron migration, we developed three in vitro assays: a long-distance migration assay, a co-culture migration assay, and a chemo-attraction assay. Here we describe protocols for these assays in detail. All three assays are based on the usage of silicone culture inserts to create a small rectangular patch of cells (of fixed dimensions) surrounded by cell-free space. Migration distance is evaluated by measuring the distance between the final positions of cells from the border of the rectangular patch that has been outlined on day 0. In the long-distance migration assay, human periventricular endothelial cells are seeded as a patch in the center of a 35 mm dish, and the distances traveled by the cells over a long range of time are calculated. In the co-culture migration assay, human periventricular endothelial cells are co-seeded with human interneurons as one patch in a 35 mm dish. This setup allows examination of the effect of direct physical interactions of these two cell types on the rate of migration of interneurons. The chemo-attraction assay measures the migration of interneurons in response to chemo-attractive cues secreted by human periventricular endothelial cells. Interneurons are seeded as a rectangular patch, with human periventricular endothelial cells and control non-periventricular endothelial cells seeded as similar sized patches on either side. Each of the cell patches are separated by a cell-free gap of 500 µm. Response of interneurons is assessed by quantifying the number of cells that have migrated towards periventricular endothelial cells compared to control non-periventricular endothelial cells.

These assays provide robust assessment of human periventricular endothelial cell functions and their influence on interneuron migration. The novel setup of long-distance assay and co-culture migration assay provides cell-free space in the range of centimeters (~1-1.5 cm) to allow detection of long-distance migration. A summary of the features of our assays compared to other popular assays is presented in Table 1. Collectively, the assays described here will serve as a platform for assessing "diseased" periventricular endothelial cells and interneurons generated from iPSCs of brain disorders like schizophrenia, autism or epilepsy. These assays can also be used to determine how different conditions (e.g. inhibitors, ligands, RNAi) affect cell migration. Finally, these assays can be optimized for other cell types to measure long-distance migration, chemo-attraction or cell-cell mediated migration.

1. Culture and Storage of Human Periventricular Endothelial Cells

1. Maintain human periventricular endothelial cells on basement membrane matrix-coated (see Table of Materials) 6-well plates in periventricular endothelial cell medium (E6 medium containing 50 ng/mL VEGF-A, 100 ng/mL FGF2 and 5 µM GABA) at 37 °C and 5% CO₂. Change medium every alternate day.
2. Thaw basement membrane matrix in 4 °C, and make a 1:100 solution by diluting it in cold DMEM/F12 medium. Coat each well of a 6-well plate with 1 mL of matrix solution. Incubate plates at 37 °C for at least 1 h before use.
3. Allow human periventricular endothelial cells to reach a confluency of 80%-90%. Aspirate medium from the well. Wash the wells once with 1 mL of sterile 1x PBS per well.
4. Detach cells by adding 1 mL of cell dissociation solution (see Table of Materials) per well. Incubate at 37 °C for 5 min. After 5 min, add 1 mL of periventricular endothelial cell medium. Transfer the cell solution into a 15 mL conical tube.
   NOTE: We use Accutase for cell dissociation here, as opposed to TrypLE in sections 3 and 4.
5. Centrifuge cells at 500 x g for 5 min at room temperature, aspirate the supernatant and resuspend the cell pellet in 1 mL of periventricular endothelial cell medium.
6. Count live cells using the trypan blue exclusion method. Seed cells in fresh matrix-coated plates at a density of 1.2 x 10⁵ cells/cm². Incubate at 37 °C and 5% CO₂.
7. Store human periventricular endothelial cells by cryopreserving in freezing medium (90% periventricular endothelial cell medium and 10% DMSO).
   1. Dissociate and collect cells following steps 1.3 and 1.4 above. Count cells in the solution by the trypan blue exclusion method.
   2. Centrifuge cells at 500 x g for 5 min at room temperature. Aspirate the supernatant and resuspend the cell pellet at 5 x 10⁶ cells/ mL of freezing medium.
   3. Dispense 1 mL of freezing medium plus cells per cryovial. Place the vials in isopropanol-filled chamber and cool overnight in -80 °C at 1 °C/min. Transfer vials to a liquid nitrogen tank on the next day for long term storage.

2. Preparation of Human Periventricular Endothelial Cells for Assay

1. Allow human periventricular endothelial cells to reach 70%-80% confluency.
2. Dissociate cells following steps 1.3 to 1.5 as described above. Count cells using the trypan blue exclusion method.

3. Preparation of Human GABAergic Interneurons for Assay

NOTE: Human induced pluripotent stem cell (iPSC)-derived GABAergic interneurons and the neuronal medium were commercially purchased (see Table of Materials). The neurons are generated by differentiating a human fibroblast-derived iPSC line following a protocol developed by the manufacturer. The cells were thawed and cultured according to manufacturer's protocol.

1. Thaw human GABAergic interneurons and culture them in 12-well plate for two weeks to a confluency of 70%-80%.
2. On the day of the assay, warm cell dissociation solution (see Table of Materials) and an aliquot of neuronal medium at 37 °C for 10 min before use.
3. Aspirate medium from each well containing the cells. Wash cells with 1 mL of sterile 1x PBS per well.
4. Detach cells by adding 0.5 mL of pre-warmed dissociation solution per well and incubate at 37 °C for 5 min. Add 1 mL of neuronal medium per well. Transfer cell solution into a 15 mL conical tube. Gently triturate to dissociate cell clumps.
5. Centrifuge cells at 380 x g for 5 min at room temperature, aspirate the supernatant and resuspend the cell pellet in 1 mL of neuronal medium. Count live cells using the trypan blue exclusion method.

4. Preparation of Control Human Endothelial Cells for Assay

NOTE: Control human iPSC-derived endothelial cells and endothelial cell medium were commercially purchased (Table of Materials). These endothelial cells are generated by differentiating a human fibroblast-derived iPSC line to endothelial fate following a protocol developed by the manufacturer. The cells were thawed and cultured on Fibronectin substrate according to manufacturer's protocol. Fibronectin-coated plates were prepared following manufacturer's protocol.

1. Thaw control human endothelial cells and culture them in 6-well plate to a confluency of 80%-90%.
2. On the day of the assay, warm cell dissociation solution (see Table of Materials) and an aliquot of endothelial medium at 37 °C for 10 min before use.
3. Aspirate the medium from each well containing the cells. Wash cells with 1 mL of sterile 1x PBS per well.
4. Detach cells by adding 0.5 mL of pre-warmed dissociation solution per well. Incubate at room temperature for 5 min. Add 1 mL of endothelial cell medium per well to neutralize the dissociation solution. Transfer cell solution into a 15 mL conical tube.
5. Centrifuge cells at 200 x g for 5 min at room temperature. Aspirate supernatant and resuspend the cell pellet in 1 mL of endothelial cell medium. Count live cells using the trypan blue exclusion method.

5. Preparation of One-well Culture Inserts

1. Thaw 1 mg/mL laminin solution at room temperature or overnight at 4 °C.
2. Coat an appropriate number of 35 mm dishes with 0.01% poly-L-ornithine solution (1 mL per dish). Incubate the dishes in room temperature for at least 1 h.
3. Dilute 1 mg/mL laminin solution 1:300 in sterile water to a final concentration of 3.3 µg/mL immediately before use.
4. Completely aspirate poly-L-ornithine from each dish. Rinse each dish thoroughly 3x with sterile water and aspirate completely to avoid poly-L-ornithine-induced cell toxicity.
5. Add 1 mL of 3.3 µg/mL laminin solution to each dish and incubate at 37 °C overnight or at least 1 h. Remove laminin solution from the dish immediately before use.
   NOTE: Alternatively, store the laminin-containing dishes in 4 °C. Equilibrate the dishes in a 37 °C cell culture incubator before use.
6. Cut three sides of one well of a two-well silicone culture insert (Figure 1A) using a sterile blade to generate a one-well insert (Figure 1B).
   NOTE: Keep the two-well insert firmly attached to the surface of the original packaging while cutting to ensure a smooth cut and protect the adhesiveness of the insert.
7. Aspirate laminin solution from the dishes.
   NOTE: Do not wash the dishes with sterile PBS or water after laminin incubation. Wet surfaces will prevent tight adhesion of the culture insert.
8. Remove one-well insert with sterile tweezers and place it in the center of the poly-L-ornithine/laminin coated dish. Press along the edges of the insert to fix it to the surface of the dish.
9. Carefully turn the dish upside down to verify that the insert is firmly adhered.
10. Keep the dish upside down and mark the boundary of the insert compartment using a permanent black marker with an ultra-fine tip (Figure 1C).

6. Long-distance Migration Assay

1. Suspend human GABAergic interneurons at a concentration of 3 x 10⁴ cells/70 µL of neuronal medium. Seed 70 µL of cell solution inside each one-well culture insert.
   NOTE: The seeding density of interneurons is as per manufacturer's recommendation.
2. Suspend human periventricular endothelial cells at a concentration of 3 x 10⁴ cells/70 µL of periventricular endothelial cell medium. Seed 70 µL of cell solution inside each one-well culture insert.
   NOTE: The number of human periventricular endothelial cells seeded at a 1:1 ratio with the number of seeded neurons.
3. Add 1 mL of neuronal medium in neuron dish to fill the area around the insert and prevent the coating from drying. Similarly, add 1 mL of periventricular endothelial cell medium in periventricular endothelial cell dish.
   NOTE: Add medium slowly along the edge of the dish so that the insert is not disturbed.
4. Check under a microscope to verify that cells are not leaking from the insert compartment.
5. Incubate cells for 24 h at 37 °C and 5% CO₂. After 24 h incubation, check under microscope to verify that cells have attached properly and there is no overnight leak.
6. After 48 h of seeding, gently remove the insert using a sterile tweezer. Check under the microscope to verify that the cell layer remains undisturbed (day 0).
7. Remove medium from the neuron dish and add 1 mL of fresh neuronal medium. Similarly, remove medium from the periventricular endothelial cell dish and add 1 mL of fresh periventricular endothelial cell medium.
   NOTE: Set aside a required number of dishes and fix with 4% PFA for day 0 images.
8. Incubate cells for 5 days at 37 °C and 5% CO₂. After 5 days, remove medium, fix cells with 4% PFA for 10 min, and wash 3x with 1x PBS.
9. Stain neurons with anti-human β-Tubulin or anti-human MAP2 antibody, and endothelial cells with anti-human CD31 antibody. At the end of immunostaining, add 1 mL of antifade mounting medium to each dish.

7. Co-culture Migration Assay

1. Co-suspend 3 x 10⁴ GABAergic interneurons and 3 x 10⁴ human periventricular endothelial cells in 70 µL of co-culture medium (50% periventricular maintenance medium without GABA and 50% neuronal medium). Seed this cell solution inside the one-well insert compartment. Prepare an appropriate number of such assay dishes.
   NOTE: GABA was not added in the co-culture medium to exclude the effect of exogenous GABA on migration.
2. Check under a microscope to verify that cells are not leaking from the insert compartment.
3. Slowly add 1 mL of co-culture medium along the side of the dish to prevent the coating from drying.
   NOTE: Add medium slowly along the edge of the dish so that the insert is not disturbed.
4. As a first control, seed 3 x 10⁴ human GABAergic interneurons only in 70 µL of co-culture medium per one-well insert. Prepare an appropriate number of such dishes.
5. As a second control, co-seed 3 x 10⁴ GABAergic human interneurons with 3 x 10⁴ control human endothelial cells in 70 µL of co-culture medium per one-well insert. Prepare an appropriate number of dishes.
6. Incubate the dishes for 24 h at 37 °C and 5% CO₂. After 24 h incubation, check under a microscope to verify that cells have attached properly and there is no leak.
7. After 48 h of seeding, gently remove the insert using a sterile tweezer. Check under the microscope to verify that the cell layer is not disturbed (day 0).
8. Remove medium and add 1 mL of fresh co-culture medium.
   NOTE: Set aside an appropriate number of dishes for acquiring day 0 images.
9. Incubate cells for 5 days at 37 °C and 5% CO₂.
10. After 5 days, remove medium, fix cells with 4% PFA for 10 min, and wash 3x with 1x PBS.
11. Stain with anti-human β-Tubulin or anti-human MAP2 antibody to label the neurons. At the end of immunostaining, add 1 mL of antifade mounting medium to each dish.

8. Chemo-attraction Assay

1. Place a three-well culture insert in the center of a poly-L-ornithine/laminin coated 35 mm dish using sterile tweezers.
2. Turn the dish upside down. Mark the boundary around the middle compartment of the insert using a permanent black marker with ultra-fine tip.
3. Seed 3 x 10⁴ human GABAergic interneurons in the middle compartment in 70 µL of neuronal medium (Figure 3A).
4. Seed 10⁴ human periventricular endothelial cells in 70 µL of periventricular endothelial cell medium and 10⁴ control endothelial cells in 70 µL of control endothelial cell medium in the two outer compartments respectively (Figure 3A).
5. Add 1 mL of co-culture medium (50% periventricular maintenance medium without GABA and 50% neuronal medium) along the side of the dish to prevent the coating on the dish from drying.
6. Check under microscope to verify that cells are not leaking from the insert compartment.
7. Incubate cells for 24 h at 37 °C and 5% CO₂. After 24 h incubation, check under a microscope to verify that cells have attached properly and there is no leak.
8. After 48 h of seeding, gently remove the insert using a sterile tweezer. Check under the microscope to verify that the cell layer is not disturbed (day 0).
9. Remove medium and add 1 mL of fresh co-culture medium.
   NOTE: Set aside required number of dishes for day 0 images.
10. Incubate cells for 36 h at 37 °C and 5% CO₂. After 36 h, aspirate medium, fix cells with 4% PFA for 10 min, and wash 3x with 1x PBS.
11. Stain human GABAergic interneurons with anti-human β-Tubulin or anti-human MAP2 antibody. At the end of the staining procedure, add 1 mL of antifade mounting medium in each dish.

9. Imaging and Data Analysis

1. Place the immuno-stained assay dish under a microscope at 4X magnification.
2. Keep one long-edge of the rectangular boundary (made in step 5.10 above) in the field of view. Take images of cells in the cell-free space adjacent to that boundary. Acquire images along the right long-edge and the left long-edge of the rectangular boundary (Figure 2B).
   NOTE: Cells positioned diagonally with respect to the rectangle are not considered due to ambiguity in selecting the short- or long-edge as starting mark. Also, numbers of cells migrating across the short-edge are often significantly fewer (possibly due to lesser number of starting cells along the short-edge) and are not considered.
3. Open the images in ImageJ. Calculate distance between each cell and the boundary mark (Figure 2D) using ImageJ.
4. To assess migration in terms of cell numbers, set a specific distance from the boundary in the acquired image in ImageJ. Count the number of cells that are present within this distance. Calculate average number, standard deviation, and statistical significance using appropriate software.

The steps to set up a one-well culture insert inside a 35 mm dish are shown in Figure 1. Long-distance migration assay and co-culture migration assay used a one-well insert to seed the desired number of cells in the center of a poly-L-ornithine/laminin coated 35 mm dish. On day 0, cells were present as a rectangular patch (Figure 2A,C). In day 0 images, the day 0 line could be easily identified by the sharp edge of the cell layer (white dotted line in Figure 2C). By 48 h, cells had migrated out into the cell-free space (Figure 2B,D). In post-day 0 images, the black border drawn around the insert (at the back of the dish) could be clearly observed as a black gap. The edge of the gap was assigned as the day 0 line (white dotted line in Figure 2D). As mentioned in Step 9.2, only those cells which fell in the area adjacent to the right and left long-edges of the cell layer (yellow area in Figure 2B) were considered for data analysis. The distance travelled by a cell was measured by calculating the distance between the cell (white arrow in Figure 2D) and the day 0 line. Immunocytochemical staining with anti-active Caspase 3 antibody, a marker of apoptosis, showed no apoptotic signal in the seeded cells (Figure 2E). In the co-culture migration assay, when interneurons were co-seeded with human periventricular endothelial cells, neurons travelled farther distances compared to when interneurons were seeded alone or when co-seeded with control endothelial cells (Figure 2F). Also, for the same distance range, a higher number of interneurons migrated out when co-seeded with periventricular endothelial cells in comparison to interneurons in the other two groups. This shows that, like mouse periventricular endothelial cells, human periventricular endothelial cells promote human interneuron migration.

In the chemo-attraction assay, using three-well culture inserts, human interneurons were seeded as a small rectangular patch in a 35 mm poly-L-ornithine/laminin coated culture dish. Periventricular endothelial cells and control non-periventricular endothelial cells were seeded as patches on either side of the neuronal patch, with the gap between each patch being 500 µm (Figure 3A). The number of interneurons that migrated towards periventricular endothelial cells versus control endothelial cells was quantified after 36 h. A significantly higher number of interneurons migrated towards periventricular endothelial cells compared to control endothelial cells (Figure 3B,C), confirming that GABAergic interneurons respond selectively to chemo-attractive cues secreted by human periventricular endothelial cells.

Figure 1: Preparation of the culture insert. (A) A two-well culture insert. (B) A one-well insert fixed in the center of a 35 mm dish. (C) The outline of the rectangular patch as observed after removing the insert. Please click here to view a larger version of this figure.

Figure 2: Schema and representative result of migration assay. (A) Schema of cell layer (red rectangle) on day 0. (B) Schema of cells migrating out into the cell-free space. Red dots indicate migrating cells. The yellow region marks the area that is imaged for data acquisition. The dotted box in A and B corresponds to the area shown in panels C and D. (C,D) Representative fluorescent images of anti-β-Tubulin antibody labeled interneurons on day 0 (C) and day 2 (D) of the migration assay. The white dotted line marks day 0. The yellow line in D indicates the distance travelled by a cell (marked by white arrow) in 48 h. (E) Neurons (on day 0) are co-labeled with anti-β-Tubulin antibodies (red) and anti-active Caspase 3 antibodies (green), which mark apoptotic cells. Nuclei are stained with DAPI (blue). Apoptotic cells were not detected in seeded cells. (F) Graph from day 5 of the co-culture assay, where the number of interneurons that have migrated is plotted against distance travelled. In comparison to interneurons that were seeded alone or co-seeded with control endothelial cells, interneurons co-seeded with periventricular endothelial cells migrated out in higher numbers, and also travelled farther distance. Data represents mean ± S.D (n = 5; **p<0.01, ***p< 0.001, Student's t test). Scale bars = 100 µm. IN = interneurons; PV EC = periventricular endothelial cells. Please click here to view a larger version of this figure.

Figure 3: Chemo-attraction assay. (A) Schema of the chemo-attraction assay. Using a three-well culture insert, interneurons (IN) were seeded in the middle (green dotted rectangle), while periventricular endothelial cells (PV ECs; orange dotted rectangle) and control endothelial cells (ECs; yellow dotted rectangle) were seeded on either side. (B) Images of β-Tubulin labeled interneurons showing robust migration towards periventricular endothelial cells but not towards control endothelial cells. (C) Quantification of the chemo-attractive response of interneurons. A significantly higher number of neurons migrated towards periventricular endothelial cells than towards control endothelial cells. Data represents mean ± S.D (n = 5; *p < 0.05, Student's t test). Scale bars = 100 µm. Please click here to view a larger version of this figure.

Table 1: Comparison of assay methods. (A) Comparison of common in vitro migration assays with the long-distance migration assay and co-culture assay. (B) Comparison of common chemotaxis assays with the chemo-attraction assay.

Here, we described three in vitro assays that together provide quantitative assessment of human periventricular endothelial cell-specific properties. These assays will be valuable in gaining mechanistic insights into the interaction of human periventricular endothelial cells with human interneurons. Experiments using ligands, inhibitors, or cells with gene-specific knockdown or overexpression will identify or validate molecular players that mediate endothelial cell-guided interneuron migration or long-distance migratory properties of periventricular endothelial cells. These assays can also be modified to perform live-cell time-lapse migration studies. Furthermore, there is evidence for interaction of endothelial cells with cells other than interneurons. Studies from our group and others have alluded to the influence of periventricular endothelial cells on patterning of projection neurons and proliferation of neural precursor cells^5,14,15. It would be of interest to test these possible interactions using our assay settings. Finally, these assays will serve as a platform for assessment of diseased periventricular endothelial cells. Our work has established novel autonomous links between the periventricular vascular network and the origin of neuropsychiatric disorders like schizophrenia, epilepsy, autism, and major depression^3,5. These assays will be invaluable in identifying potential defects in long distance migration, chemo-attraction, or juxtracrine signaling of diseased-periventricular endothelial cells in these neuropsychiatric disorder conditions.

These assays are simple, reproducible, and low cost, and they can be modified to measure cell migration and effects of co-culture or chemo-attractive cues on migration in various cell types, except for non-adherent cells. There are some critical steps that need to be followed to obtain accurate and reproducible results. First, it is critical to optimize seeding cell number for each assay. The number of cells to be seeded in a single compartment should depend on cell type, desired level of confluency, and assay-specific factors like co-culture ratio. Second, it is necessary to optimize the cell culture medium for each assay. In the co-culture migration assay and the chemo-attraction assay, where more than one cell type is seeded in a single dish, the assay medium should be conducive to all cell types. In pilot experiments, we examined the effect of co-culture medium on viability (using trypan blue exclusion method) and morphology (using immunocytochemistry) of each cell type. We cultured human GABAergic neurons with co-culture medium for one week and observed no significant difference in viability and morphology of neurons in co-culture medium compared to neurons cultured in neuronal medium. In a similar fashion, periventricular endothelial cells and control endothelial cells, cultured in co-culture medium for two passages, did not show any significant variation in cell survival and morphology. Third, since rate of migration varies among different cell types, it is important to determine the time frame for each assay for the cell type(s) being studied. Fourth, it is critical to handle the culture inserts carefully. Inserts should be fixed firmly on the dish by gently pressing with a fingertip. The dish should be turned upside down to verify that the insert is not moving. Care should also be taken while removing the insert so as not to disturb the cell layer. Finally, it is recommended to increase sample size to reduce experimental variability.

In conclusion, these assays will significantly expand our understanding of human periventricular endothelial cell biology and its role on brain development in normal and diseased conditions.