Todd Belanger wrote- I am new to sorting (but I have ten years of flow experience) and we just purchased a FACSVantage/DiVa. Some of our projects require sorting rare cells at levels of 1 in a million or ten million. Some of the researchers say it could be one in 100 million (which seems quite impossible to me). One in 100 million is about ten times as good as the current state of the art under the best of circumstances. Does anyone have any pointers or particularly good papers that would help me in this task? How low can you go (in terms of rare events) and still be relatively confident in what you sorted? Currently the researchers I will be doing the sort for has two markers- PI to discriminate live cells and a FITC conjugated marker. I know more markers would be better for discriminating rare events but their doesn't seem to be any for this particular experiment. If your company bought the sorter solely or primarily for these projects, tell them to send it back and get a refund. Pertinent references are: Gross H-J, Verwer B, Houck D, Recktenwald D: Detection of rare cells at a frequency of one per million by flow cytometry. Cytometry 14:519-26, 1993 Gross HJ, Verwer B, Houck D, Hoffman RA, Recktenwald D: Model study detecting breast cancer cells in peripheral blood mononuclear cells at frequencies as low as 10(-7). Proc Natl Acad Sci U S A 92:537-41, 1995 Ger van den Engh, who knows whereof he speaks when he talks about high speed sorting, says 10,000 cells/second is about as fast as one can generally count on being able to analyze cells; an informal poll of Boston area high speed sorting experts finds that they rarely run more than 25,000 cells/sec. That's 100,000 cells in 4 sec, or 1,000,000 cells in 40 sec. So, if you were 100% accurate and had 100% recovery in selecting cells present at a frequency of 1/1,000,000, it would take you a little over an hour to collect 100 cells. It would take over ten hours to collect 100 cells if they represented 1/10,000,000 of the population, and over a hundred hours if the cells were present at a frequency of 1 in 100,000,000. However, the kicker is not how many cells/second you can run through the machine, but how well you can discriminate the cells you want from the cells you don't want. From the frequencies you quote, and the name of your company, I am guessing that the project involves the use of tetramer technology to detect and isolate lymphocytes with reactivity to a specific antigen. Using PI to discriminate dead cells from live ones is not going to be a big help; your main problem is discriminating the specifically labeled cells from unlabeled cells, essentially using single-parameter data (fluorescein fluorescence), and a single-parameter fluorescence measurement rarely is able to discriminate cells present at levels below a substantial fraction of 1 per cent. The one cell in a million and one cell in ten million papers use multiparameter analysis to identify the cells of interest; most successful rare event sorting uses a cocktail of stains that react with the unwanted cells as well as two or more markers specific for the wanted ones. If the researchers behind this project want to isolate and grow reactive cells, 100 cells (single-cell sorted into a multiwell plate) may be enough to work with. If their notion is to sort and analyze thousands, or tens of thousands, of cells, forget it. Cells present at 1/100,000,000 will be found once an hour; sorting 10,000 of them will take five years if you work single shifts, and well over a year if you sort round the clock. And, of course, it will be necessary to kill a lot of animals, or take a lot of blood from a lot of humans, to get the input material. You can improve your odds somewhat by using a pre-enrichment step, the most popular being immunomagnetic separation using technology from Dynal, Immunicon, Miltenyi, etc. If you can enrich the cells you are looking for so they represent more than 1/1,000 cells in the input population to the sorter, you have a fighting chance. But do the math to find out whether or not this project will run past retirement age for all involved. -Howard
This archive was generated by hypermail 2b29 : Sun Jan 05 2003 - 19:26:18 EST