Glioblastomas are devastating brain tumors with a high invasive profile. This co-culture system mimics glioblastoma cells migration on neurons to recapitulate one of the invasive routes observed in patients. The geometry and composition of our co-culture model is precisely controlled.
It does facilitate a better reproducibility and also a straightforward quantification of complex biological processes. This method can be adapted to clinical diagnosis by co-culturing freshly dissociated glioblastoma cells and neurons. It defines a clinical index, which is very important as clinical outcome.
Our method can also be used to quantify other migrating cells, such as fibroblasts or immune cells. Demonstrating the procedure will be Joris Guyon, a PhD student from my team. To make a substrate for the micropatterning, treat 18 millimeter circular glass coverslips by air or plasma activation for five minutes before placing the coverslips in a desiccator with 100 microliters of triethoxysilane for one hour, then incubate a solution of PEG-SVA at 100 milligrams per milliliter for one hour.
For gel deposition, at the end of the incubation, add three microliters of PLPP and 50 microliters of absolute ethanol onto the center of the slide and wait until it dries completely. For glass slide micropatterning, mount the coverslip in a Ludin chamber and place the chamber onto the stage of a microscope equipped with an auto-focus system. After imaging, load the micropattern images into the software.
After automatic UV illumination sequencing, use a pipette to wash the PLPP from the coverslip extensively with PBS. Then incubate the coverslip with 50 micrograms per milliliter of laminin for 30 minutes, followed by another wash with PBS as demonstrated. To set up an embryonic rat hippocampal neuron culture on the micropattern coverslips, after the last wash, rehydrate the glass slides with neuronal cell culture medium.
Seed five times 10 to the fourth rat hippocampal neurons suspended in Neurobasal medium enriched with 3%horse serum per square centimeter onto each micropatterned coverslip for a 24-hour incubation in a 5%carbon dioxide incubator at 37 degrees Celsius. Centrifuge dissociated glioblastoma cells for five minutes at 1, 000 RPM and resuspend the pellet in glioblastoma cell culture medium, then deposit one times 10 to the third GBM cells over the micropatterned neurons. For live cell imaging of the cells, place the co-culture onto the stage of an inverted microscope equipped with a 37 degree Celsius thermostat chamber and select the 20X objective.
Then use the multi-dimensional acquisitions toolbox in the microscope software to acquire live Brightfield and epifluorescence GFP tomato images every two minutes for 12 hours in 16 different positions based on the number of patterns with neurons. For neuronal network analysis, after imaging, select one image from the stack. Right-click on the network tool to open the corresponding options dialog box and adjust the settings to produce a precise segmentation of the images.
Click OK.Left-click on the network tool to duplicate the selected image and split the image into the red, gray, and green color channels. Select the gray channel and perform contrast stretch enhancement to improve the separation between the different areas. Use the Sobel edge detector to perform the 2D signal processing convolution as grouped under the find edge command.
For double filtering, apply Gaussian blur and a median filter to reduce the noise and to smooth the object signal. To convert to mask, execute adapted threshold algorithms to obtain a binary picture with black and white pixels. Next, skeletonize the cell area into a simple network and use filter particles to remove small non-networked particles in the results.
Network filter particles in a network image. To obtain red and green channels, perform double filtering and convert to mask as demonstrated using the adapted thresholding method. Use analyze particles to determine the cell morphology in the binary green image.
Use the or operator to merge all of the channels using their regions of interest and readjust their initial color into a simple RGB image. To perform a single-cell motility analysis, right-click on the single cell tracking tool to open the corresponding options dialog box and adjust the settings to produce a precise segmentation of images. Click OK and left-click on single-cell tracking to remove the gray channel.
To generate an image corresponding to an image stack according to the time, apply Z projection and double filter and convert to mask the trails left by the cells. Remove the small particles from the binary red and green image as demonstrated. Use the region of interest tool to select each contour of the cell trace and check the skip edge detection box in the options dialog window.
Isolate the red channel on the original stack and select one region of interest. Double filter all of the images and convert to mask to allow the centroid XY position of each binarized nucleus to be determined. The XY positions can be used to calculate the mean square displacement, directionality ratio, and average speed for the cell.
For multiple cells tracking analysis, right-click on the tracking tool to open the corresponding options dialog box and adjust the settings to produce a precise segmentation of images. Left-click on the tracking tool to remove the gray channel. Split the red and green channels, double filter, and convert to mask.
Then use the image calculator command with the and operator to merge the channels leaving only the nucleus signal located within the membranes and calculate the trajectory plot, mean square displacement, directionality ratio, and average speed for the cells as demonstrated. Fluorescent GBM co-cultured with patterned neurons quickly modify their shape and show migration along neuronal extensions in a random motion. GBM cells seeded onto neurons display an elongated shape with multiple protrusions that follow the neuron tracks while the cells retain their rounded shape when cultured on laminin.
At later stages of culture, thin protrusions linking to cells can be observed in GBM neuronal co-cultures. GBM cells seeded onto neurons demonstrate a greater migratory capacity than GBM cells seeded onto laminin as observed in these trajectory plots. Analysis of the fluorescent confluence of the cells demonstrates that over a 500-minute observation period, more cell migration is observed when the spheroids are co-cultured with neurons than when the cells are cultured on laminin alone.
Indeed by the end of the analysis, nearly half of the pattern is covered with GBM cells while the spheroids cultured on laminin remain unadhered to the coverslip. Depending on the biological questions you want to answer, one should take care that the pattern design and also the cell density are representative of the in vivo conditions. The cells can be fixed and imaged by confocal microscopy.
Live imaging is also possible since our method does not impair the imaging capabilities. This technique is well-suited for studying molecular interaction, such as metabolic exchanges between glioblastoma cells and neurons. It allows the exploration of function biological processes.
High-throughput experiments can be also done for clinical purposes.
