CAR T cells are being constantly improved to achieve better responses in human clinical trials. We are identifying ways to evaluate this in our lab using an unbiased high throughput method, such as using fluorescent imaging to screen combinations that work best. There are several combinations of constructs that can be designed for any single-chain variable fragment by modifying the geometry of the extracellular and hinge domains.
We have designed this protocol to identify the combinations with the best efficacy in eliminating target cells using Jurkat cells. This is a rapid high-throughput protocol to screen CAR constructs based on its target cell cytotoxicity using Jurkat cells, which are, when validated with conventional T cells, a dozen different CAR constructs can be tested in a 96-well plate simultaneously using this technique. Here in Akhavan lab, we quantify in absolute numbers the direct cytotoxicities of target cells by the CAR-expressing Jurkat cells in a high-throughput manner to screen multiple CAR constructs.
Our laboratory's primary aim is to improve the clinical translation of CAR T cell therapy against high-grade brain tumors. We leverage high-throughput drug screens as one modality to improve CAR T cell cytotoxicity as well as targeting the tumor microenvironment. To begin, using a serological pipette, remove the growth media from the T75 flask containing 70%confluent MDA-MB-231 culture.
Add three milliliters of trypsin to the flask and incubate at 37 degrees with 5%carbon dioxide for three to five minutes to detach the cells from the flask. Then to neutralize the trypsin, add three milliliters of DMEM into the flask. Collect the cell suspension in a centrifuge tube and spin at 400 G for four minutes.
Discard the supernatant using a pipette and resuspend the pellet in two milliliters of PBS. Using a cell counter, determine the cell concentration. Transfer eight times 10 to the fifth cells into a 15 milliliter tube and add PBS to make the volume to one milliliter.
Then add two microliters of CFSE into the tube and mix the cell suspension thoroughly using a one milliliter pipette. Place the tube at 37 degrees Celsius and 5%carbon dioxide incubator for 20 minutes. Next, add five milliliters of DMEM to the tube and centrifuge to pellet down the CFSE-labeled cells.
Discard the supernatant and resuspend the pellet in one milliliter of DMEM. After reevaluating the cell concentration, transfer four times 10 to the fifth cells into a 25 milliliter reagent reservoir. Add DMEM to make the volume eight milliliters and mix the cell suspension with a five milliliter serological pipette.
Using a multichannel pipette, transfer 100 microliters of cell suspension into each row on the left half of a black 96-well plate. Similarly, add EGFR knockout cell suspension on the right half of the plate. To ensure uniform cell distribution, move the plate back and forth and side to side on the platform.
Incubate the plate for four hours at 37 degrees Celsius and 5%carbon dioxide for the tumor cells to attach. Based on counts of untransduced and CAR-expressing Jurkat cells, transfer four times 10 to the fifth CAR-J cells into a 25 milliliter reservoir. Add DMEM to the reservoir to a final volume of two milliliters.
For an effector-to-tumor ratio of four-to-one, add two times 10 to the fourth CAR-J cells per well in 100 microliters of medium along the side of the well without disturbing the attached tumor cells. Then add 100 microliters of DMEM to the side of the wells containing tumor and Jurkat cells to get 300 microliters of volume per well. Move the plate as demonstrated to ensure uniform distribution of the Jurkat cells on the tumor cells.
Allow the co-culture to establish at 37 degrees Celsius with 5%carbon dioxide for 48 hours. After co-culturing CAR-expressing Jurkats with CFSE-labeled tumor cells, gently invert the plate on a paper towel and tap to discard the supernatant containing CAR-J. Using a multichannel pipette, add 100 microliters of FluoroBrite DMEM media containing propidium iodine or PI into the wells without disturbing the adhered tumor cells.
Then add 20 microliters of 20%Triton X to the first well of each tumor type serving as a total dead control. Incubate the plate at 37 degrees Celsius and 5%carbon dioxide for 20 minutes before imaging. After fluorescent imaging, use one of the tumor-only wells of cells to calibrate the green fluorescent channel.
To exclude cells at the edge of the wells, decrease the well mask to 98%Identify CFSE-labeled tumor cells on the green fluorescent channel and adjust the fluorescence intensity threshold value to capture all the cells in the well. Set the minimum cell diameter to 25 micrometers to exclude debris detected in the CFSE channel. Then enable separate touching objects to identify individual cells.
Next, set gates to define live versus dead cell populations. Select the CFSE-labeled cells and generate a histogram of counts on the y-axis versus mean PI intensity on the x-axis. Adjust the x-axis value to better visualize the cells and draw a splitter to differentiate between low PI-stained cells and high PI-stained cells.
Label the low PI-stained cells as live cells. Next, set another histogram on live cells with area on the x-axis and counts on the y-axis. Label the non-debris cells as big live cells.
Then draw another splitter using the tumor-only well to capture cells and filter out debris. After running the analysis on the entire plate, export the spreadsheet containing the numbers. Plot the counts of big cells to determine the number of live CFSE-labeled tumor cells remaining in the well after exposure to CAR-J.
Finally, perform statistical analysis using one-way ANOVA. The cytotoxicity of CAR-J1 increased with a higher effector-to-tumor ratio and more than 50%killing was observed at an effector-to-tumor ratio of four-to-one. EGFR targeting CAR constructs with modified hinge domains showed antigen-specific killing against EGFR expressing MDA-MB-231 cells and no significant killing was observed against EGFR knockout cells.
CAR constructs were also expressed on PBMC-derived CD3 T cells. However, there was no expression of CD8 hinge-containing EGFR CAR. The IgG short showed the least cytotoxic potency against MDA-MB-231 cells compared to the IgG long and IgG medium.
