Re: Cell cycle analysis

From: Derek Davies (daviesd2@cancer.org.uk)
Date: Tue Dec 10 2002 - 06:19:18 EST


Hi Ken,

There will nearly always be a discrepancy between the machine count and
a haemocyometer count due to the fact that not all sorted cells land in
the tube, inaccuracies in counting by eye etc but I normally expect to
be +/- 20% of the machine count (generally - it has to be said).

However your yield is of more concern. With exponentially growing cells
you may have say 60% of cells in G1, allowing for a narrow sort gate you
could be sorting 40% of the population, then allowing fro other cell
losses you may get 75% of that figure so from 30 million cells you
should potentially be able to get about 10 million back. Whne you say
you have 30 million cells, do you mean at the point they go on the
sorter? I have users who claim to have had x million cells after
harvesting but then choose to ignore the y centrifugation steps where
they have lost a significant number.

I have done a lot of the type of sorting you describe. U2OS cells are
pretty big - you dont say what size nozzle you are using but I would use
a 100um nozzle. Do you have a particularly high hardware abort rate? It
shouldnt realy be more than 10% of the event rate. You may be able to
reduce it by diluting the sample or using a larger nozzle. You should
also ensure that the drop delay is accurate - when using PI stained
cells, especially large ones, I find that it is useful to set the delay
with the sample itself rather than beads.

Good luck!
Derek



On Fri, 6 Dec 2002, Kenichi Fujise wrote:
> I have been evaluating a cell cycle specific expression of various
> proteins.
> In stead of synchronizing cells using double thymidine block, etc, we
> are attempting to sort cells and perform Western on the sorted cells to
> evaluate the expression levels of proteins of interest.  We are using
> U2OS cells.
> We propagate U2OS cells so that next day, cells will be 70-80%
> confluent. We harvest them by trypsinization and fix them overnight with
> 70% MetOH.  Next day, we stain DNA with propidium iodine and take the
> cells to a core facility.
> Our problem has been that we only get 0.5 million G0/G1 cells out of 30
> million cells we start with.  The counter on the machine says 1.0
> million but when we count them by hemocytometer, the number is always
> lower. Because of this low yield, our Western blot has not been able to
> detect consistent signals of the proteins.
> Has anybody tried a similar approach?  How would you increase the yield
> of sorting?  What do you think is causing the discrepancy between the
> machine read-out and actual cells sorted (as assessed by hemocytometric
> counting).

************************************************************************
Derek Davies                       Voice: (44) 020 7269 3394
FACS Laboratory,                   FAX: (44) 020 7269 3100
Cancer Research UK,                e_mail:derek.davies@cancer.org.uk
London Research Institute,	   mobile: 07790 604112
44 Lincolns Inn Fields,
London, UK.

Web Page: http://sci.cancerresearchuk.org/axp/facs/davies/index.html

In tenebris lux
*************************************************************************



This archive was generated by hypermail 2b29 : Sun Jan 05 2003 - 19:26:31 EST