Research Technologies Branch

Factors Affecting Sorting

Cell Staining for Sorting

It is best to stain cells in media without phenol red (because it is fluorescent and increases background). You may stain large numbers of cells at the same concentration normally used for FACS analysis, i.e., 2 x 107/ml. Addition of the antibody can be performed in two ways:

1. Add the correct volume of the antibody at the pre-diluted concentration. If you normally add 10 ml of diluted antibody to 106 cells, add 100 ml to 10 x 106 cells.
2. Add the correct volume of stock antibody. If you normally add 10 ml of antibody diluted 1:20 from stock to 106 cells, add 5 ml of stock antibody to 10 x 106 cells. This probably works the best, since possible dilution errors are minimized when scaling up from small to large volumes.

After washing to remove unbound antibody cells for sorting should be resuspended at 5 x 106/ml.

Controls

Two types of control cells must be provided for each experiment:

1. Unstained cells of the type to be sorted. This will allow us to adjust the instrument according to the inherent autofluorescence of unstained cells excited by the laser wavelengths that will be used.
2. Single color controls of each antibody/dye combination to be used. This allows us to compensate for minor color "spill over" of one antibody/dye combination into other color detectors. Compensation should be performed using the brightest antibody/dye combination for each color to be used that day. (If your antibody populations do not stain brightly or are less than 10% positive, it is advisable to stain with a brighter antibody conjugated to the same fluorochrome for compensation purposes.)

Coincidence

Sorting involves a trade-off between maximizing cell purity and maximizing cell recovery. Processing cells at higher flow rates increases the yield but the flow rate is limited by the requirement for a minimum time interval between cells due to measurement time, electronic processing, sort decision time, and droplet deflection. Cells that enter the laser beam before the minimum time interval after the preceding cells are termed coincident. There are two types of coincidence:

1. Cells that are detected as individuals but are too close together for the electronics to make individual sorting decisions for each cell. In this case, no sort decision is made and both cells are discarded as waste.


2. Cells so closely spaced that the detection system fails to identify them independently. In this case the cells are treated as a single object and sorted if they meet the sort criteria. This type of coincidence affects purity, as in the example of a cell doublet with one of the cells expressing the desired fluorochrome and the other cell being negative. Only upon post-sort analysis will the negative cell become evident.

Sort Purity

Sort purity is influenced by these factors:

1. Signal to noise ratio—the precision of sorting depends upon how well the populations of interest can be resolved from the background autofluorescence. Populations that may be resolved on the FACScan may not be as well resolved on a large cytometer, due to the optical restrictions of a "stream in air" sorting instrument compared to the cuvette of an analysis instrument. There are several procedures that may improve the signal-to-noise ratio:
1. Amplify the weakest antibody by using avidin or a secondary reagent (e.g., @Hamster Ig).
2. Less autofluorescence is produced by a dye laser (590nm) than by an argon laser (488nm) so use fluorochromes such as APC and Texas Red that are excited by a dye laser.
3. With careful consideration of the dyes, the dual laser systems can be used to reduce or eliminate the need for color compensation between the various antibody/dye combinations.


2. Undetected coincidence—two cells so close together that they are detected as one cell:
1. The smaller the frequency of cells of interest, the more likely undetected coincidence will occur with an unwanted cell. Pre-enrichment can and should be used to increase purity and yield of sorted populations.
2. Positive staining of unwanted cells with subsequent sorting for a negative population will virtually eliminate undetected coincidence since an undetected doublet would consist of two negative cells. Positive staining can be accomplished by staining unwanted cells with multiple antibodies conjugated with the same fluorochrome or by increasing the number of antibody/color combinations used to positively and/or negatively select for unwanted cells.
3. Low flow rates and/or dilute cell concentrations will increase the time interval between cells and therefore reduce undetected coincidence. Purity will improve but the time required to sort will increase.

Sort Recovery

Sort recovery (yield) depends on these factors:

1. The percentage of the cells of interest. The time required to sort increases dramatically when the frequency of the population of interest falls below 10% or 15%. Pre-enrichment techniques should then be used when possible:
1. Magnetic cell sorting (MACS) can be used as a positive enrichment since the small size (<200nm) of MACS beads make them undetectable by the flow cytometer. Positive and/or negative selections with MACS can be based on
1. direct staining
2. secondary reagent staining
3. streptavidin staining
4. anti-FITC staining (if you can stain your cells with FITC, then they can be enriched)


2. Complement mediated cytotoxicity to specifically remove unwanted cells.


3. FACS sorting. Instruments can sort in enrich mode first (increases cell frequency approximately 10-fold), then re-sort in normal mode. This technique works best when high purity is preferred over cell recovery.


2. Detected coincidence. Cells that are detected as individuals but are too close together for the electronics to make individual sorting decisions for each cell. In this case, no sort decision is made and both cells are discarded as waste, reducing the net sorting rate. This can be minimized by:
1. Slightly lower flow rates and/or more dilute cell concentrations will increase the time interval between cells slightly and therefore reduce detected coincidence. At 2000-5000 cells/sec we are able to achieve our best combination of number of cells recovered and high purity. A concentration of 10 – 20 x 106 cells/ml allows sorting at this optimal rate.
   
3. Sort time. Time is needed to prepare the instrument for sorting and for shutting it down at the end of the day. Therefore, time for sorting is limited to approximately 6.5 hours maximum unless special arrangements are made in advance.
4. Beginning number of cells. In order to estimate how long your sort will take and how many cells to bring, you can use the following formulas:
   
   Time required for sort (hrs) = [Number of recovered cells desired/fraction of desired cells in starting population)/(Flow rate)]
   where:

   Fraction of desired cells in starting population = percent/100
   Flow rate = cells/hr or (cells/sec)*3600.
   Number of cells to bring = Time required * Flow rate * 2

back to top