RE: Proliferating lymphocytes in peripheral blood and CD25 expression

Zbigniew Darzynkiewicz wrote:

>Let me add a note of caution regarding the use of pyronin Y as a marker
>of proliferating cells or RNA. Both Howard [Shapiro] and Edward 
>[Srour] are right that a
>bivariate analysis of Hoechst vs pyronin Y fluorescence allows one to
>discriminate between Go vs G1 cells. Edward is also correct stating that
>pyronin Y is toxic to cells. I would argue, however, that when used
>supravitally pyronin Y preferentially stains mitochondria and not RNA
>(see e.g. "Cytostatic and cytotoxic properties of pyronin Y: Relation to
>mitochondrial localization of the dye and its interaction with RNA",
>Cancer Res., 46: 5760-5766, 1986). In fact, the cytotoxic and cytostatic
>properties of pyronin Y remind very much rhodamine 123 and relate to its
>interaction with mitochondria. The cytotoxicity is enhanced when cells
>stained with pyronin Y are illuminated with white light - e.g. under
>light microscopy one can easily see that mitochondria containing pyronin
>Y swell and rupture within seconds after turning on light.
>Noncycling (Go) cells, have much fewer mitochondria and their
>stainability with Rhodamine 123, similar as with pyronin Y, is many-fold
>lower that that of G1 -S- G2M lymphocytes ("Increased mitochondrial
>uptake of Rhodamine 123 during lymphocyte stimulation" Proc. Natl Acad.
>Sci. USA 78: 2383-2387, 1981)
>  The photosensitizing effects of pyronin Y through mitochondria was
>described in another publication ("Photosensitizing effect of tricyclic
>heteroaromatic cationic dyes pyronin Y and toluidine blue O (Tolonium
>chloride" Cancer Res., 48: 1295-1299, 1988)"). Perhaps to lower the
>cytotoxicity of pyronin Y one has to carefully screen cells from the
>ambient light. Rhodamine 123 may be less toxic than pyronin Y and both
>discriminate Go from G1 cells.
>Of course, pyronin Y and Hoechst can be used to stain differentially RNA
>and DNA in fixed cells. Pyronin Y can also be used to discriminate
>between ds and ssRNA ("Application of pyronin Y in cytochemistry of
>nucleic acids". Cytometry, 8:138-145,1987).

Zbigniew, Frank Traganos, and Tom Sharpless were kind enough to run 
acridine orange stains on my CCRF-CEM cells when I was generating the 
data for my 1981 paper on staining with Hoechst 33342 and pyronin Y, 
and I am mindful of both the papers on pyronin Y (PY) localization 
and toxicity and of the 1981 paper on rhodamine 123 (R123) uptake in 
stimulated lymphocytes. I would, however, opine that the question of 
how much of the PY fluorescence from supravitally stained cells comes 
from RNA and how much comes from mitochondria remains to be resolved.

The R123 studies with stimulated lymphocytes were done by incubating 
cells with high (8-25 uM) concentrations of dye and subsequently 
washing the cells, more or less following the procedure originally 
described by Lan Bo Chen's group, which first reported mitochondrial 
localization of R123 (Proc Natl Acad Sci USA 77:990, 1980). At the 
concentrations used for cell loading, R123 (and most other 
fluorescent dyes) can be expected to be quenched substantially. Even 
if R123 uptake is Nernstian, the combination of an interior negative 
cytoplasmic membrane potential of tens of millivolts and a 
mitochondrial potential of over 100 mV will produce an 
intramitochondrial dye concentration in the millimolar range before 
the cells are washed. When washed cells are observed under the 
fluorescence microscope, most of the dye appears in mitochondria, 
which retain a much higher concentration of dye than is left 
elsewhere in the cell, and the results of fluorescence microscopy 
correlate with those obtained from flow cytometry - i.e., signals 
from cells with a large mass or volume of energized mitochondria are 
brighter than those from cells with a smaller mitochondrial mass or 
volume or with deenergized mitochondria, in which membrane potentials 
is low or absent.

At lower concentrations (nanomoles to tens of nanomoles/L), permeant 
cationic dyes, a class which includes PY as well as R123 and the 
cyanines, are used for cytometric estimation of cytoplasmic membrane 
potential. This is most commonly done using cyanine dyes such as 
DiOC6(3) and DiIC1(5), but, as I showed by swapping dyes with Lan Bo 
Chen about 25 years ago, R123 at low concentrations, left in 
equilibrium with cells, can be used to study cytoplasmic membrane 
potential, and cyanines can be loaded into cells at high 
concentrations and, after washing, be used to estimate mitochondrial 
potential. I also established that there was not a substantial 
difference in DiOC6(3) fluorescence, measured by flow cytometry, 
between lymphoblastoid cell controls and cells treated with various 
agents that deenergize mitochondria. This suggests that the bulk of 
intramitochondrial dye is quenched, even at an external dye 
concentration of 50 nM, and, since one would expect the 
intramitochondrial concentration under this circumstances to be tens 
of micromoles/L, this explanation is not unreasonable. However, if 
one were to look at the cells under the microscope, the cytoplasmic 
fluorescence would bleach out in a few seconds or less, leaving 
fluorescence from quenched dye in mitochondria visible.

The Hoechst/pyronin Y staining procedure as I orignally described it 
left intact, presumably viable, cells in equilibrium with 5 uM PY; 
Edward uses a lower concentration, 500 nM. Either would produce a 
sufficiently high concentration in energized mitochondria to result 
in substantial quenching. In the brief exposure associated with flow 
cytometry, one would expect a greater contribution from 
extramitochondrial PY, presumably including dye bound to RNA, and 
relatively little from intramitochondrial dye. Extramitochondrial dye 
would be bleached very rapidly from cells placed under the 
fluorescence microscope, and the impression would therefore be that 
dye was primarily associated with mitochondria. Under such 
conditions, what you see is not necessarily what you get.

PY fluorescence signals from ethanol-fixed and intact cells are not 
vastly different in magnitude, and much (at least 75%) of PY 
fluorescence in fixed cells disappears after RNAse treatment. This 
also suggests that PY fluorescence in intact cells reflects RNA 
content. It would be informative to compare PY fluorescence from 
cells in various phases of the cell cycle (e.g., stimulated 
lymphocytes) in the presence and absence of mitochondrial 
depolarizing agents. It should also be informative to examine cells 
stained simultaneously with Hoechst 33342 and with Molecular Probes's 
newly described SYTO RNASelect. The nucleolar localization of the 
fluorescence of this dye in both live and fixed cells and its 
susceptibility to RNAse but not DNAse in fixed cells provide evidence 
for selective staining of intracellular RNA; weak fluorescence is 
also observed from mitochondria. Again, comparison of cells in the 
presence and absence of mitochondrial inhibitors would be informative.

Hoechst 33342, R123, and PY are all sensitive to some extent to the 
action of efflux pumps, which affect staining of both stem cells and 
lymphocytes in different activation states. And I should mention that 
PY staining best reflects RNA content only when PY is used in 
combination with Hoechst 33342, 7-AAD, or another DNA-selective dye. 
This adds a few more worms to the can, consideration of which I will 
defer until next year.

Happy New Year!!!

-Howard



-- End --