Separation of dead cells from cultures - (Jul/04/2005 )
Dear all:
Our lab recently had recently acquired a rare cell line from some lab, after numerous email requests and hassals to say the least. This cell line is a non-adherent line and shipped in a complete RPMI-filled T25 flask. However, when we spun and resuspended the cells, we only found very very few live cells around (less than 1% by trypan blue staining).
Since there's no way for this person to resend us this line again, I'm wondering if there's any feasible technique to successfully separate dead cells from live ones so that we could try to see if somehow we could enrich the live population and ressurect this cell line? Some one had suggested that we may try using lower centrifugation speed so that dead cells will be spun to accumulate at bottom of the tube, while the live ones are still floating in the media. However, I've never read about this protocol anywhere else, so was dare not to try. For now, we had only kept all cells in its original flask, awaiting solution to take care of this. But, we're concened that the presence of way too many dead cells might comprimise the surviving possibility of live ones.
Any suggestions are gratefully appreciated since we're desparated to death... 
I think the technique you are referring to is a little more involved than this. When isolating lymphocytes from whole blood the sample is layered over Ficoll and spun. The live cells will float on top of the interface and the dead cells and red blood cells will accumalate at the bottom. I have seen this being done with unhealthy tissue cultures but the technique does take a little practice to cleanly isolate the cells off the interface and I wouldn't recommend risking your one and only culture on a first time technique??
A second alternative is to grow the cells in a small volume and regularly exchange half the media (every two days) to dilute out toxins released by the dying cells. The way you would select clones from a drug selection.
Either way once you rescue this culture remember that you have passed it through a harsh selection pressure and effectively isolated a subpopulation from the original culture. The characteristics of this new line may vary from the original.
I hope this helps.
A fellow labmate has used this technique with a B-cell line quite successfully.
Spin cells ~500-600rpm 2-3 minutes. In his case the live cells pelleted and most dead cells / fragments floated. Note: at this speed there will still be some live cells floating.
Resuspend in low volume, culture, and repeat.
You can always plate the super' in a separate dish to verify where the majority of your live cells are going.
I have been told that Ficolling could help, but I have never tried that myself.
I also agree with centifugation protocol.
My suggestion is to spin down the culture suspension at about 125-200g for 5 min, the live cells will pellet down and dead cells will remain in the supernantant. Resuspend cell pellet in a small volume and grow in smaller culture plates (say 35mm in your case), also culture supernatant in separate dish.
I hope this helps.
I also agree with the centrifugation procedure as it seems safer.
Another option is cell sorting if you could label the cells with an antibody and then use FACS to sort them out, probably tough thought considering you have so few cells!
Too bad they didn't send you a frozen vial!
Hi. I will follow the majority on this. According to my experience with Jurkat cells, low-speed centrifugation and pellet plating in small volumes works fine in this cases. Good luck.
I'm not sure how you would use a facs to physically remove dead from live cells...but in my experience ficolling works with lymphocytes, though you have to follow the protocol exactly. If these are attached cells (EC's, SMC's) you can just wash the flask with media, as the dead will not be attached.
Hi,
I've experienced a similar problem with a suspension cell line from breast tumor. I tried the centrifugation for shorter time/lower speed as it made sense that the live cells would be heavier and therefore pellet quicker. To my horror, the dead cells always pelleted first. In the case of this cell line, the cells clumped up when dead to form a dark mass which was visible under a normal light microscope. Because of the clumping, we were never able to successfully separate the dead from live ones (spinning at lower speed and retaining the supernatant with live cells did work initially, but there were always dead cells as well).
Eventually, we decided to try inducing the live cells to adhere onto laminin. It worked. Dead cells wouldn't adhere and we could wash them off after the live ones were adherent. But I don't like this technique for the following reasons:
1) forcing cell adherance may alter cell properties?
2) the live cells stick for as long as the laminin lasts, and they seem to be able to proliferate to a certain extent while adhering (but not to the extent observed with common cell lines)...but that means any manipulation would have to be done on this laminin background
3) Eventually, the cells float and I could never ever achieve confluence.
Anyway have any clue as to why dead cells would clump? is there any benefit in continuing with the laminin based system? Thanks.
3)