Isolation and Characterization of Tumor-initiating Cells from Sarcoma Patient-derived Xenografts

We discovered that HLA-I is epigenetically down regulated in a subpopulation of cancer cells that exhibit stem cell properties and tumor-initiating capacity. HLA-I downregulation is directly related to immune escape, providing survival advantages for cancer stem cells, therefore making HLA-I negativity an accurate functional marker for these cells. Demonstrating the procedure will be Sudeh Izadmehr, a postdoc from our laboratory.

Upon acquisition of the sample, place the tissue in a 100-millimeter Petri dish in a sterilized bio safety cabinet, and add 500 microliters of sterile PBS to the dish. Use a sterile scalpel to mechanically triturate the tissue into small pieces until no fragment is larger than 0.1 millimeter. Transfer the entire 500-microliter volume of PBS through a 35-micrometer pore cell strainer into a 50-milliliter conical tube.

Add another 500 microliters of PBS to the tissue fragments, and mince the tissues again. Then filter the supernatant through the cell strainer into the same collection tube, and continue to mince the tissue pieces until the sample is completely dissociated. When the last volume of cells has been collected, spin down the tumor cells by centrifugation, and re-suspend the pellet in five milliliters of hemolysis buffer.

After five minutes at room temperature, collect the cells with another centrifugation, and wash the pellet with another five milliliters of PBS. Re-suspend the pellet in one milliliter of PBS, and count the viable cells. Dilute the cells to a one times 10 to the seventh cells per 200 microliters of PBS concentration on ice, and add 200 microliters of basement membrane matrix to the cells with gentle mixing.

Then inject the cell suspension-basement membrane matrix subcutaneously into the flank of a NOD scid gamma mouse, and check the growth of the tumor at the injection site two times a week. When the tumor reaches one centimeter in diameter, cut the tumor in half, and fix half of the xenograft with 4%paraformaldehyde overnight for histological analysis. Process the second half of the tumor as just demonstrated to achieve a single cell suspension, and dilute the cells to a two times 10 to the six cells per milliliter of PBS supplemented with 5%FBS concentration.

Maintain the cells on ice for 30 minutes before dividing the cell suspension between one isotype control and one antibody tube. Note the number of cells in each tube, and mix the cells in the antibody tube with anti-HLA antibody and the cells in the isotype control tube with an appropriate anti-isotype control antibody for a 90-minute incubation on ice. At the end of the incubation, collect both cell populations by centrifugation, followed by two 10-milliliter PBS washes by centrifugation.

After the second wash, re-suspend the pellets in 10 micrograms per milliliter of DAPI to a one times 10 to the seven cells per milliliter concentration. Then filter each cell suspension through a 35-micrometer pore strainer into individual 12-by-75-millimeter polystyrene tubes. After fluorescence-activated cell sorting, dilute the HLA-I-negative and positive cell populations to individual one times 10 to the fifth cell concentrations in 15 milliliters of the sarcosphere growth medium per subset.

Serially dilute the cells to one times 10 to the fourth, 10 to the third, and 10 to the second concentrations in 15 milliliters of fresh sarcosphere growth medium per dilution, and add 100 microliters of cells from each dilution to each well of one 96-well ultra-low attachment cell culture plate per dilution. Place the plates into a 37-degree Celsius, 5%carbon dioxide cell culture incubator, monitoring the sarcosphere formation daily by light microscopy and adding fresh basic fibroblast growth factor and epidermal growth factor to each well without changing the medium every three days. After three weeks, count the number of sarcosphere-positive and sarcosphere-negative wells for each cell dilution of both HLA-I-negative and HLA-I-positive cells to allow calculation of the sphere-forming cell frequency based on a Poisson probability distribution.

Typically, human patient-derived sarcoma xenografts consist of two distinct HLA-I-positive and negative populations, with a similar histology to that demonstrated by the parental primary tumor. Fluorescence-activated cell sorting of sarcoma patient-derived xenograft tumor-initiating cells as demonstrated facilitates the enrichment of a high number of HLA-I-negative cells from the parental cell population. HLA-I-negative cells are able to form spheres with an initial input of as little as 10 cells and exhibit a higher tumor formation ability than HLA-I-positive sarcoma patient arrived xenograft tumor-initiating cells.

Injection of the same number of HLA-I negative and positive cells subcutaneously into opposing flanks of the same mouse demonstrates that HLA-I-negative cells possess a significantly higher tumor formation ability. While xenografts formed by both HLA-I negative and positive subpopulations are cellularly heterogeneous tumors. Further, gene expression analysis of the tumor-initiating cells reveals that HLA-I-negative cells express stem cell differentiation markers and can be induced to differentiate along both lipogenic and osteogenic pathways, demonstrating strong Oil-Red-O and Alizarin-Red-S staining and culture.

This method can be used to isolate cancer stem cells in various human cancers. Molecular carbonization, for example, by next generation sequencing may reveal common markers for targeting the cells therapeutically. Our discovery demonstrates that the restoration of HLA-I is a critical step for the success of functional unit immune cell augmentation strategies.