Summary - Viable by LS; dead by PI (?)

Below are the responses I received to my question about cells that appear
viable by light scatter but nevertheless take up propidium iodide.  I also
asked if 1 ug/ml PI was too low a concentration for this purpose.  Some
responses addressed one issue and not the other.

What I didn't mention was the fact that the cells in question are murine
bone marrow cells, cultured in vitro, and 'selected' by some form of 'drug'
treatment, in vitro.

Responses from the group suggested they could be granulocyte-like cells that
are non-viable because they said grans aren't as susceptible to light
scatter changes when going from live to dead; others suggested apoptotic
cells - which could be saying the same thing, I suppose.

Thanks for everyone who responded (we really should sort them, I suppose).

Ray Hester
Univ. of South Alabama
Mobile, AL
rhester@jaguar1.usouthal.edu


.................................


I suspect that the decrease in forward scatter [often seen in non-viable
cells]
 is due to swelling due to loss of
osmotic regulation. I have seen decreases in FS signal in lymphoid cells
that correlate almost exactly with increases in Coulter cell volume. In
contrast, the granulocytes do not have these cell surface villi. If the
theory is correct, then grans will not show the death associated decrease in
FS that is seen in lymphocytes.  In addition, much of the light scatter
signal in grans is from the cytoplasmic granules, unless the cell
degranulated this signal should be relatively unaffected.

......................
Is it possible that these cells are gran precursors that are late
enough in development to have significant granules present? This would
make
them look like grans in the scatter plot. If the correct growth factors
are
not present to support them, they may well have died by the time you have
the cells for analysis. This could be checked by looking at the expression
of the species appropriate markers in the gran differentiation sequence. I
normally avoid the S-word but this is one place where a small sort onto a
microscope slide could answer the question more directly.

.......................
This is not unusual. In fact, when true viability is required, I
always tell my users to use PI (or some other DNA dye) rather than
light scatter. The larger cells can also be early apoptotic cells
which do swell and would also be PI permeant...they could also be
doublets...
the concentration is correct..we usually use 2.5 ug/ml

........................

I for one used to  use scatter for live dead cell discrimination on cultured
cell lines.

When people started mutating / transfecting the cells, we found that cells
that would be assumed dead by changes in light scatter were not in fact all
dead.  So we switched to PI in all un fixed samples, a really wide scatter
gate and then a PI negative gate Complexing the two gates to yield
fluorescence from live cells only.

I suspect that your last paragraph is what is going on here, i.e.,

[Can certain "drug" treatments cause cell membranes to become permeable to
propidium iodide without changing their light scatter properties, i.e., they
appear viable by LS but non-viable based on the PI uptake?]


..........................

PI uptake (or at least PI uptake associated with bright nuclear staining)
is generally considered to indicate membrane damage, i.e., a "hole" in the
membrane big enough to let water and electrolytes cross, changing the
refractive index of the cell (hence the change in scatter signal) and
dropping membrane potential to zero.

In bacteria, at least, it has been noted that some drug treatments (e.g., a
sublethal dose of a beta-lactam antibiotic) will result in PI uptake in
cells that can be shown by ratiometric measurements to retain membrane
potential. If we look at the effects of pore-forming antibiotics on the
bacteria, we find that gramicidin, which forms a pore said to be about 5
Angstroms across, reduces membrane potential to zero (it is known that
sodium, potassium, hydrogen, and chloride ions can pass through a
gramicidin pore), but does not render bacteria permeable to PI. Nisin,
which forms an 8 Angstrom pore, does render bacteria permeable to PI, as
well as reducing membrane potential to zero. Since the PI uptake in
bacteria that possess a membrane potential cannot be based on passage
through a nonspecific pore or hole, it is likely to result from activation
of an influx pump and/or inhibition of an efflux pump. Dyes that,like PI,
have two positive charges, one of which comes from a quaternary ammonium
group, behave like PI in this system; when examined in a flow cytometer
with both 488 nm and 633 nm sources, bacteria exposed to both PI and
TO-PRO3, which has similar charge characteristics, take up either both dyes
or neither dye.

I have heard that at least some eukaryotic cells have a transport mechanism
that can process PI and other doubly charged molecules, but I haven't seen
the references. However, such a mechanism could explain the PI uptake in
the absence of membrane damage significant enough to change scatter
characteristics. The bottom line is that PI and similar compounds are
better described as "permeability" indicators than as "nonviability"
indicators.

........................

You asked:  Can certain "drug" treatments cause cell membranes to become
permeable to
propidium iodide without changing their light scatter properties, i.e., they
appear viable by LS but non-viable based on the PI uptake?

Absolutely.  You don't say what the drug does but there are lots of
mechanisms for this.  Affecting membrane potential and/or perturbing ion
transport is one way.  Inducing apoptosis, whether intentional or not, can
also make the cells a little leaky before the light scatter changes much.
And this can vary A LOT depending on the cell type.  Some cells actually
swell during apoptosis.  I typically don't consider cells "dead" by PI
uptake unless they are very bright.  That is, you start to see a DNA
distribution.  That being said, I wouldn't want to sort the low PI cells if
I am trying to grow them, because they probably are going to die anyway.  PI
leaks in a little if they are left in it for too long, especially if not
kept on ice.  Too high a concentration of DMSO (or other solvents typically
used in stock solutions of drug treatments) can cause problems.  Limit DMSO
to 0.1%! .  DMSO can even differentiate certain cell types.  Make sure they
aren't seeing some low level of detergent along the way as well.  If they
are in fact lymphocytes, activation can lead to a large amount of
apoptosis/leaky membranes as well.  In case you're wondering, I would think
that all these things are documented phenomena and not just personal
anecdotes since I am certain that I'm not the first to observe them.

..........................
Have you already tried to refresh these cells back in culture medium?

...........................
I've seen this, when phagocytosis of the dead has taken place.
We have IntraPrep treated the cells and looked at phenotype.
Surface markers not seen externally on the larger cells are found internally
and
correspond to the phenotype of the dead cells.

...........................
i typically see about 10% which are pi positive and fall within a
reasonably conservative fsc vs ssc gate. it is for this reason that i
ask all my users to run a viability test (via exclusion dye) with their
experiments. when i see data presented i always ask if they have run
such a test. people tend to remember my questions so that when in the
spotlight they don't have an answer.

..........................
Funny that you mentioned this. We routinely use PI to screen for
specimen viability and it sometimes comes as less than expected.  I
especially noted that bone marrow specimens with lots of NRBCs have
strong PI staining, supposedly coming from NRBCs

...........................

I'm struggling with the same problem with PI staining fpr fluorescence
microscopy.
So far I have 6 different protocol on that subject. The concentrations
are reaching from 0,5 to 5,0 uM.
I don't have the molar weight at hand now, but 1,5 uM is 1mg/mL.
The staining times vary from 1 to 30 minutes, the recommended
temperature is either room temperature or 4°C. I haven't tried tthe
different protocol yet (I'm trying to optimise my FLICA-staining first),
but 10 minutes at 1,5 uM at 37°C doesn't seem to work with my cells.

...........................
That should work if you put it on a log scale.	The concentration should not
be too high as PI will cause cell death in itself.  By eye, the samples
should have  a slight pink tinge and 1 ug/ml should give you that.  For
live/dead discrimination the concentration of PI is not all that critical.

............................

well it is impossible to answer to a question like that. When you try a
dye on a sample, you have to test several concentrations, and also make
a time series to see what is the best incubation time. Very often the
dye concentration has to be adapted to the cell concentration. You have
to find out which are the best conditions. What I tell you is not only
good for PI, but works also with any other dye you use for the first time.






This attachment - 'Cell_Viability.doc' -  35.84 KBytes - can be viewed at
http://www.cyto.purdue.edu/MD-parts/42bb2c7f4e078049193e763a0d89e1776f587a57.doc