A New Biomarker for Mitotic Cells
James W. Jacobberger,1* Phyllis S. Frisa,1R. Michael Sramkoski,1Tammy Stefan,1
Keith E. Shults,2Deena V. Soni3
Many epitopes are phosphorylated during mitosis. These epitopes are useful biomarkers
for mitotic cells. The most commonly used are MPM-2 and serine 10 of histone H3. Here
we investigated the use of an antibody generated against a phospho peptide matching
residues 774–788 of the human retinoblastoma protein 1 (Rb) to detect mitotic cells.
Human cell lines were stained with DNA dyes and antibodies reactive with epitopes
defined by antibody MPM-2, phospho-S10-histone-H3, and the phospho-serine peptide,
TRPPTLSPIPHIPRC (phospho-S780-Rb). Immunoreactivity and DNA content were
measured by flow and image cytometry. Correlation and pattern recognition analyses were
performed on list mode data. Western blots and immunoprecipitation were used to inves-
tigate the number of peptides reactive with phospho-S780-Rb and the relationship
between reactivity with this antibody and MPM-2. Costaining for bromodeoxyuridine
(BrdU) was used to determine acid resistance of the phospho-S780-Rb epitope. Cell cycle
related phospho-S780-Rb immunofluorescence correlated strongly with that of MPM-2.
Laser scanning cytometry showed that phospho-S780-Rb immunofluorescence is
expressed at high levels on all stages of mitotic cells. Western blotting and immunoprecipi-
tation showed that the epitope is expressed on several peptides including Rb protein. Cost-
aining of BrdU showed that the epitope is stable to acid. Kinetic experiments showed util-
ity in complex cell cycle analysis aimed at measuring cell cycle transition state timing. The
phospho-S780-Rb epitope is a robust marker of mitosis that allows cytometric detection
of mitotic cells beginning with chromatin condensation and ending after cytokinesis.
Costaining of cells with DNA dyes allows discrimination and counting of mitotic cells
and post-cytokinetic (‘‘newborn’’) cells. To facilitate use without confusion about specific-
ity, we suggest the trivial name, pS780 for this mitotic epitope.
' 2007 International Society for
? Key terms
cell cycle analysis; flow cytometry; image cytometry; immunofluorescence; retinoblas-
toma protein; phosphorylation; mitosis
CYTOMETRIC analysis of the cell cycle that includes mitosis has been reported inter-
mittently for many years (see references in 1,2). The most straightforward assays use
immunofluorescence of mitotic epitopes combined with DNA content. These epi-
topes are on single proteins (2,3), reside on multiple proteins (1,4), are degraded in
mitosis (2,5), or unmasked during mitosis (6). Although advantages of objectivity,
ability to count large numbers of events, and ability to correlate measurements are
obvious reasons to use flow or laser scanning cytometry (LSC) to count mitotic cells,
the approach has been used infrequently until recently. Now, it is more common to
see the use of cytometry to quantify mitotic cells in studies that are biological and
not methodological in theme. This may be due to the wide commercial availability of
primary conjugates of MPM-2 and phospho-S10-histone H3 antibodies.
An area of cytometry that has seen recent increased interest is the correlated
measurement of biochemical activities, especially phosphorylation (7–9). In this
work, the central theme, sometimes more implicit than explicit, has been analysis of
pathways and networks rather than assays of single activities. Perhaps for the first
1Case Comprehensive Cancer Center,
Case Western Reserve University,
Cleveland, Ohio 44106
2Esoterix Center for Innovation, 201
Summit View Drive, Brentwood,
3Department of Environmental Health,
Case Western Reserve University,
Cleveland, Ohio 44106
Received 21 August 2007; Accepted
7 November 2007
This article contains supplementary
material available via the Internet at
Grant sponsor: PHS; Grant number: R01-
CA473413; Grant sponsor: National
Cancer Institute; Grant number: P30-
CA43703; Grant sponsor: Advanced
Technology Center, Beckman Coulter,
Present address of Deena V. Soni:
Department of Molecular Pharmacology,
Stanford University, Palo Alto, CA.
*Correspondence to: James W.
Jacobberger, Case Comprehensive
Cancer Center, Case Western Reserve
University, Cleveland, OH 44106, USA.
Published online 4 December 2007 in
Wiley InterScience (www.interscience.
© 2007 International Society for
Cytometry Part A ? 73A: 5?15, 2008
time, this approach employs cytometry mainly for quantitative
correlation and analysis of complexity rather than the more
often cited reasons of convenience and objectivity. For analysis
of biochemical networks, cytometry is perhaps the most
powerful approach available when the questions are directed
and limited to ?20 variables or less. This is because of correla-
tion, precise quantification, and ability to examine all possible
states1within a single or limited number of samples.
Although most emphasis on biochemical networks within
the cytometry community is on signaling pathways, the cell
cycle is a biochemical network and highly multiparametric
approaches to cell cycle analyses can be equally compelling.
Indeed, any regulatory pathway can be analyzed by cytometry
with powerful results (e.g., 10). The reason for this is that bio-
chemical networks cannot be conceived except dynamically,
and cell populations exist in all possible biochemical states
with respect to time for a given environment. This is especially
easy to see for asynchronous populations of cycling cells in
which all biochemical states are extant at any given time. Anal-
ysis of the ‘‘regulatory’’ cell cycle within this context is more
easily digested when layered on a backbone of phases and
stages that we ‘‘understand’’ from kinetics and morphology. In
this sense, the backbone cytometry assay is DNA content
coupled with a mitotic marker and at least for the late phases
(S, G2, and M), cyclins A2 and B1 for somatic cells. For each
of these backbone markers, it is good to have alternative mar-
kers. For mitosis, there are two reasons not to be satisfied with
phospho-S10-histone H3 and MPM-2 as the only cytometric
mitotic markers. First, in studies of drug inhibitions and inter-
actions, phospho-S10-histone H3 phosphorylation in mitosis
is governed by aurora kinase B (11–13), and therefore, any
treatment that results in reduced aurora kinase B activity, may
result in a compromised or eliminated backbone. MPM-2 is a
good alternative in this case because the epitope is expressed
on many proteins and phosphorylated by several enzymes
(14,15). Therefore, in studies in which enzyme activity is
depleted, the epitope may not be so depleted so as to render
the backbone unusable.2Second, in multivariate studies
matching fluorescence colors to epitopes in terms of fluoro-
phore characteristics (wavelength, extinction coefficient, and
quantum yield) and antigen abundance is a continuous chal-
lenge (for a discussion of similar problems on surface immu-
nophenotype, see 17). While many phosphorylated epitopes,
which turn out to be the best mitotic markers, are abundant
in mitosis, the commercial availability of only two fluoro-
chromes is limiting. Therefore, it would be valuable to have
additional antibodies to robust markers available in forms that
can be integrated in assays of several mouse monoclonal anti-
bodies. The serine at residue 780 of Rb is a specific substrate
of cyclin D/Cdk4 (18,19) and not Cdk2. We discovered by
accident that antibody raised to the phospho-peptide,
RPPTLS780PIPHIPR stains mitotic cells intensely. Here, we
show that the epitope, defined by this antibody is a robust
mitotic marker that is equivalent to MPM-2.
MATERIALS AND METHODS
Cells and Cell Culture
The culture and sources of human cell lines: DU-145, PC3,
LNCaP, and 22Rv1 (prostate cancer); Molt4 and K562 (T cell
lymphoma and chronic myelogenous leukemia), and Hela
(cervical carcinoma) are conventional and have been published
(20–22). hTert-RPE-1 cells (Clontech; Mountain View, CA) were
cultured in DMEM/F-12 with 2 mM L-glutamine, 0.375% so-
dium bicarbonate, 50 lg/ml gentamicin sulfate, and 10% fetal
bovine serum (FBS) (Cambrex, Charles City, IA). SAOS-2 cells
were grown in McCoy’s modified media with 15% FBS and gen-
tamicin sulfate. Media and additives were from Thermo Fisher
Scientific (Waltham, MA) and Gibco (Carlsbad, CA).
Culture for Laser Scanning Confocal Microscopy
hTert-RPE-1 cells (30,000) were applied in 100 ll of cul-
ture medium to the microwell of collagen (rat tail type I, Sigma-
Aldrich, St. Louis, MO) coated 35 mm glass bottom microwell
dishes (MatTek, Ashland, MA). An additional 2 ml of culture
medium was added after 30 min, and the cells were incubated at
378C in a humidified 5% CO2atmosphere for 48 h.
Okadaic acid (OA) and nocodazole were obtained from
Sigma-Aldrich (St. Louis, MO).
Antibodies were obtained and used per 1–2 3 106cells as
follows: 0.125 lg MPM-2 and MPM-2-FSE (FSE 5 fluores-
cein-5-EX succinimidyl ester; Millipore/Upstate Biotechnolo-
gies, Lake Placid, NY); 0.125–0.5 lg Alexa Fluor 488 or 647
conjugated phospho-S10-histone H3 (Millipore/Upstate Bio-
technologies); Rb (BD Biosciences, San Jose, CA; clone G3-
245; cat 554136); 0.5–1.0 ll phospho-S780-Rb (Cell Signaling
Technology, Danvers, MA); 0.5 lg Alexa Fluor 647 conjugated
anti-BrdU (A647-BrdU, Phoenix Flow Systems, San Diego,
CA); 1 lg of phycoerythrin (PE) conjugated anti-cyclin A
(Beckman Coulter, Miami, FL). Secondary antibodies were
used at 2:1 ratio w/w and obtained from Molecular Probes-
Invitrogen (Carlsbad, CA). The conjugated fluorochromes
varied and are denoted in Figure legends. All staining reactions
were carried out in 50 ll volumes of PBS with 2% BSA.
DAPI or Hoechst 33342 (Molecular Probes, Invitrogen)
were used at 1 lg per 106cells. 7-Amino actinomycin D (7-
AAD, Molecular Probes, Invitrogen) was used at 25 lg/ml. For
7-AAD staining, cells were incubated with 10,000 Kunitz units
Rnase (Worthington Biochemical, Lakewood, NJ).
1Here, a biochemical state of the cell is defined as the conjunction of
biochemical activities that are extant for a long enough period of time
for asynchronous cells to ‘‘cluster’’ at those measurements.
2Indeed, we can knock down cyclin B1 and reduce cyclin B1/Cdk1
kinase activity by 99% and reduce MPM-2 reactivity by only 30%
(unpublished results) but the cytometric pattern for DNA versus a
mitotic marker remains intact. This robustness has been argued
previously as a positive factor for MPM-2 as a marker for mitosis (16).
6New Biomarker for Detecting Mitotic Cells
Preparation of attached cells as single cells suspensions
was by trypsinization and has been described in detail (23).
Fixation was with 90% methanol or a low concentration of
formaldehyde followed by 90% methanol and has been
described in detail (24). Hematopoietic cells are fixed directly
by addition of a low concentration of formaldehyde to the cul-
ture medium followed by washing and fixation/denaturation
with 90% methanol (21).
Flow Cytometry Staining
This has been described in detail (21–26).
Staining for Laser Scanning Confocal Microscopy
The cells were fixed on the plate with 0.125% formalde-
hyde added directly to the culture medium for 10 min at 378C
followed by cold 90% MeOH in PBS. Antibody staining was
carried out as for flow cytometry except that the reaction vol-
ume was 100 ll. DNAwas stained with 0.5 lg/ml DAPI in PBS
with 1% glycerol (v/v) and 2% n-propyl gallate (w/v).
Molt4 were pulsed with 20 lM BrdU (Sigma-Aldrich)
then chased with 100 lM thymidine (Sigma-Aldrich), sampled
periodically, fixed with 0.125% formaldehyde/90% methanol
(24) then processed for BrdU staining as described (22) with
the following changes: all staining and washes were done at 48C
and all staining and wash solutions (PBS with 2% BSA) con-
tained 0.5% Triton X-100 except the final running solution.
Flow cytometry was performed on BD Biosciences LSR I,
LSR II, and Beckman Coulter XL, and FC-500 instruments set
up with stock filters supplied by the manufacturers.
Laser Scanning Cytometry
LSC was performed with a Compucyte (Cambridge, MA)
iCyte using violet and red lasers to excite DAPI and Alexa 647
fluorochromes, respectively. Stock filters from the manufac-
turer were used and the contouring algorithm relied on a
composite gray scale image using the Max function for two
input images (DAPI and phospo-S780-Rb). This had the
advantage that anaphase through cytokinetic cells were con-
toured as a single event rather than two (as would happen if
the nuclear image alone were used).
Cells were imaged with a Zeiss LSM 510 inverted laser-
scanning confocal fluorescence microscope using 63X (NA 5
1.4) oil planochromat objective (Carl Zeiss, Thornwood, NY).
Each image was obtained by independent laser scanning. Laser
excitations were at 325, 488, and 633 nm. Images were written
in TIF format with supplied software. Composite images were
produced with Adobe Photoshop CS2 (Adobe Systems, San
SDS-PAGE and Western blotting
SDS-PAGE and immunoblotting techniques were conven-
tional and have been reported in detail (27). Cell lysates con-
tained phosphatase inhibitors (5 mM sodium orthovanadate
and 5 mM sodium fluoride in addition to protease inhibitors.
Blots were visualized with alkaline phosphatase conjugated
rabbit or mouse secondary antibody (Promega, Madison, WI)
using chemiluminescent detection with CDP-Star substrate
(Tropix, Bedford, MA). Images were developed on X-ray film
or the FluorS Imager (Biorad, Hercules, CA). Quantification
was performed using Quantity One 4.1.1 software (Bio-Rad).
Western blots were divided in two and treated with calf
intestinal mucosa alkaline phosphatase (CAP, New England
Biolabs, Ipswich, MA) or buffer alone before immunostaining
with phospho-S780-antibody. CAP was 500 U/ml in buffer
containing 100 mM NaCl, 50 mM Tris–HCl, 10 mM MgCl2,
and 1 mM dithiothreitol at pH 7.9 (manufacturer’s formula).
After 30 min incubation at room temperature the membrane
was washed three times in Tris buffered saline with 0.1%
Tween 20, pH 8.0, dried, and soaked in methanol for 5 min to
inactivate the CAP. The membrane was rehydrated and pro-
cessed as above.
Cells were lysed in nondenaturing lysis buffer (50 mM
Tris, 150 mM NaCl, 5 mM dithiothreitol, 1% NP40, pH 7.4
with protease and phosphatase inhibitors). The lysate (500 lg
protein) was cleared by nutating with 2 lg rabbit IgG and
50 ll of protein A agarose beads (Sigma-Aldrich) for 1 h then
centrifuging and removing and saving the supernatant. A
1:100 dilution of the phospoho-S780-Rb antibody was added
to the supernatant, nutated for 2.5 h and then 50 ll Protein A
agarose beads were added and nutated over night. The washed
pellet was dissolved in SDS sample buffer by heating, with a
final SDS concentration of 4.4%.
Data processing, display, gating, and region setting were
performed with WinList 6.0 (Verity Software House, Topsham,
ME). Pattern recognition was by eye and experience. Regions
were set on clusters using boxes, ovals, or amorphous contain-
ers or on contour levels. For kinetic analysis, data were fit by
nonlinear least square methods (GraphPad Prism 5.00, Graph-
Pad Software, San Diego, CA, www.graphpad.com). The ‘‘per-
centage of labeled cells’’ data versus time were fit with sigmoid
dose response functions. Except for the data from the regions
labeled R6/R6b and R12/R12b (see Fig. 9), the bottom and top
parameters were set to 0 and 100. For the R6/R6b data, the bot-
tom was set to the labeling index (54) to account for contami-
nation of S phase cells and movement of the label during the 30
min of labeling time. For the R12/R12b data, a five parameter
function was used to obtain an adequate fit. This is likely due to
contamination of true positive cells with negative cells.
Cytometry Part A ? 73A: 5?15, 20087
RESULTS AND DISCUSSION
Bivariate Analysis Shows that Reactivity with the
Phospho-S780-Rb Antibody Marks Mitosis
Because serine 780 of Rb is a substrate for Cdk4 (18,19), we
first stained cells with phospho-S780-Rb in a search for epitopes
that might prove interesting for analysis of the G1 phase of the
cell cycle. To our surprise, analysis of correlated cytometric data
for DNA and immunofluorescence was typical for a mitotic
marker (Fig. 1). Compared with an established marker, phos-
pho-S10-histone H3 (2), the ratio of the signal of M to G2 was
less but the fraction of cells clustering at peak levels was similar
(Fig. 1). We costained cells for a second established mitotic
marker, MPM-2 (4), and directly demonstrated in Molt4 cells
concordance between the phospho-S780-Rb epitope (pS780)3
and MPM-2 epitope (Fig. 2). Therefore, we showed that immu-
nostaining for pS780 marks mitotic cells by direct and indirect
methods with two established mitotic markers.
Figure 1. Mitotic patterns: Molt4 cells were fixed and stained for
the mitotic marker, phospo-S10-histone H3 (phospho-S10-H3), or
with antibody phospho-S780-Rb followed with Alexa 647 (A647)
conjugated secondary and DAPI. IF 5 immunofluorescence. A
and C: gated singlet 2C population with regions set on mitotic
cells (boxes). B and D: doublet discrimination (ungated events)
with mitotic cells color coded black (gated from regions set in A
and C). Some late mitotic cells [telophase and cytokinetic cells
that have not separated can register as doublets if the cytometer
is aligned well enough and the beam is narrow enough (39)].
Insets: Peak DNA signal histograms of mitotic cells; telophase and
cytokinetic clusters are denoted by arrows.
Figure 2. Phospho-S780-Rb and MPM-2 immunofluorescence are
coincident. Molt4 cells were fixed and stained with primary
conjugated MPM-2 and phospho-S780-Rb as above except a phy-
coerythrin (PE) conjugated secondary was used in conjunction with
7-AAD for DNA.The singlet2C stemlinewas gated bydoublet discri-
mination (not shown). MPM-2 versus DNA displayed a typical
mitotic cell cycle pattern. Mitotic cells were gated on MPM-2 reactiv-
ity (left panel) and color-coded black. The color coded mitotic cells
are mapped to the phospho-S780-Rb IF versus DNA pattern (right
Figure 3. LSC of phospho-S780-Rb immunofluorescence and
DNA content. hTert-RPE-1 cells were grown in dishes with cover-
slip inserts, fixed, then stained for pS780 followed with Alexa 647
secondary and DAPI. Top: immunofluorescence plotted versus
DNA content. Images in the left lower panel are composite signals
gated from the mitotic cluster (red box) plus the transition region
(green box). The lower right panel contains similar images from
the G2 gate (cyan box). Left lower panel green arrows: transition
cells (4 of 5 localized to green box); white arrows: prophase cells.
Lower right panel thin green arrow: prophase cell. X: mistakes in
3To avoid confusion about specificity, from this point, the antibody is
referred to as ‘‘phospho-S780-Rb’’ and the epitope will be referred to as
8 New Biomarker for Detecting Mitotic Cells
All Mitotic Stages are Positive for pS780 Epitope
Usually, when evaluating the use of a mitotic probe, we
have relied on flow cytometry and cell sorting coupled with
microscopy to determine whether all the mitotic stages are
detected or not (e.g., 25). LSC (28) is perhaps a better approach
in that uncertainties of cell preparation of attached cells are
eliminated as well as any error in sorting and slide capture.
Therefore, we used LSC to ask whether all stages of mitosis
were positive for pS780. The results for hTert-RPE-1 cells are
shown in Figure 3. hTert-RPE-1 cells are an immortal, euploid,
human epithelial cell line (29). In this experiment, the DNA/
immunofluorescence bivariate histogram is typical for a mitotic
marker and is similar to bivariate flow cytometry data for
pS780 (e.g., Figs. 1 and 2). All stages of mitosis are represented.
Although in this experiment, late mitotic stages are more preva-
lent, this was a result of biology rather than technology. All of
the mitotic cells on the evaluated coverslip are shown. There
were five cells that had higher than G2 level pS780 fluorescence
but G2 nuclear morphology in the larger mitotic region (Fig. 3,
red and green regions on histogram; left lower panel, green
arrows). These decreased to 1 with the gate set more tightly
around the M cluster (red region on histogram, Fig. 3), suggest-
ing that these cells might be classified as either transition cells
or early mitotic cells. The white arrows show obvious prophase
cells with condensed chromatin and circumscribed nuclei (evi-
dence for intact nuclear membranes). The G2 cells (lower right
panel) are randomly representative but all of the G2 cells are
not shown. There is one clear and obvious prometaphase cell in
this group (right side, thin green arrow). Since, this image
shows that the cell is distinctly more red than all the other
related cells, the inclusion here is likely to be algorithmic in na-
ture rather than biological. The total number of G2 cells that
were evaluated was 150 (100 are shown). Therefore, the false
positive rate for mitotic cells should be less than 1% [depending
Figure 4. LSC and confocal images of mitotic cells. The cells and experiment are the same as Figure 3. The panel on the top left is the same as
in Figure 3 and represents the composite image of all mitotic cells analyzed. The panel on the top right is the pS780 signal only. Green arrows:
prophase; white arrows, top right: prometaphase. Lower panel: laser scanning confocal microscopy of hTert-RPE-1 mitotic cells. Red images
were obtained by excitation of Alexa 647 secondary for phospho-S780-Rb staining. Green images were obtained by excitation of MPM-2-FSE.
The cells were also stained with DAPI for DNA (not shown). Green and red confocal images were processed to reduce the visual impact of
low level intensity and increase high level intensity (S shaped input/output curve). This improved S/N and contrast but did not materially
change the images. The merged image (bottom row) has not been processed other than merging. Arrows, bottom row: prophase through
anaphase----centrosomes; telophase/CK----contractile ring area.
Cytometry Part A ? 73A: 5?15, 20089
on how tightly the region is set and the classification scheme/
criteria (i.e., an arbitrary decision on the nature of the transi-
tion cells between the G2 and M clusters)]. For G2, the false
positive rate is likely to be less than 1%.
The images from LSC suggested that the pS780 fluores-
cence was predominantly nuclear in prophase (Fig. 4 top right
panel, green arrows); uniform in prometaphase (Fig. 4 top
right panel, white arrows), and appeared to be reduced or
excluded from the chromatin from metaphase onward. To
explore this further, we performed confocal microscopy on
samples that were prepared identically to those presented in
Figure 4 (top panels) except that MPM-2 staining was also
performed. Representative cells are shown in Figure 4 (bottom
panel). It is clear that both pS780 and MPM-2 staining were
excluded or very reduced on the chromatin during all stages.
Both pS780 and MPM-2 were nuclear and cytoplasmic during
prophase and both become distributed throughout the cell af-
ter nuclear envelope breakdown (inferred). Both epitopes are
present on centrosomes (Fig. 4 bottom panel, white arrows).
MPM-2 dimly lights up the mitotic spindle. The pS780 epi-
tope is expressed strongly on the cleavage furrow (Fig. 4 bot-
tom panel, green arrow; also cytokinetic cells in top right
panel). Overall, there is some coincident staining but largely,
the green and red signals are separate and not coincident.
Cell Lines that are Deficient for Rb Expression are
Efficiently Stained for Mitotic Cells by
The data of Figure 2 suggest that there is a high correla-
tion between expression of the MPM-2 and p-S780-pRb epi-
topes. The sequence surrounding S780 of Rb (TRPPTLSPI-
PHIPRSC) contains a putative MPM-2 site of XSPXX where
the first and fourth amino acids are hydrophobic and the last
uncharged or basic (30). To both evaluate the relationship
between coexpression of the MPM-2 and pS780 epitopes and
the level of contribution of Rb phosphorylation to the staining
intensity, we analyzed samples of three human prostate tumor
lines, one (DU-145) of which does not express appreciable
levels of Rb (31) and compared these to Molt4. The cells were
stained for DNA, MPM-2, pS780, and cyclin A. The results are
presented in Figure 5. Bivariate plots of cyclin A versus MPM-
2 were used to gate G2 and M cells that have the highest levels
of cyclin A. These are late G2 and prophase cells, since cyclin
A is degraded after nuclear membrane breakdown (32). Gated
bivariate plots of MPM-2 versus pS780 were used to make the
following observations and calculate Table 1. The following
statements can be made.
First, it is clear that in Molt4 cells, there is a high correla-
tion between MPM-2 and pS780 and that result has been
observed several times and can be observed visually and by the
correlation coefficient for G21M cells (Table 1).
Second, there is a high but inexact correlation between
the two epitopes in each of the prostate cell lines. In each case,
Figure 5. Cyclin A, MPM-2, and pS780 for 4 cell lines. The left hand panel for each named pair shows the cyclin A versus MPM-2 distribu-
tion and the right hand panel shows the correlation plots (MPM-2 versus pS780) for only the ‘‘late’’ G2, transition phase, and mitotic cells
gated from the left display (gray boxes). Gray arrows indicate that the ‘‘pointed to’’ bivariate plot was gated from the gray boxed region.
The black arrows in the Molt4 plot shows the 2D data vector----i.e., the age of the cells increases as a function of position within a distribu-
tion about the centroid and orthogonal to the arrows.
Table 1. Quantitative properties of epitope expression
RATIO M:G2 CORRELATION
MPM-2 pS780G2M G2 1 M
10 New Biomarker for Detecting Mitotic Cells
since we know that cyclin A increases in G2 then decreases
after prophase, we know the direction of the vector that
describes coexpression of cyclin A and MPM-2 through the
cell cycle (black arrows in the Molt4 plot). Plots of cyclin A
versus pS780 are similar, and the same inference can be drawn.
The lack of perfect correlation between the two epitopes dur-
ing the period of time covering the G2 to M transition4can be
accounted for by different rates of expression. The difference
in rates appears to occur during the period of G2 to M transi-
tion when the epitopes are phosphorylated above the inter-
phase level but not yet to the stable, high mitotic level. During
this period of time, for the prostate cell lines, the rate of net
phosphorylation of pS780 is less per MPM-2 epitope than it is
either before (interphase) or after (prophase). Therefore, if
mitotic cells are defined from the first onset of increased phos-
phorylation, the mitotic count will differ by a small amount
dependent on the marker used. These same statements can be
made about the end of mitosis when decreasing phosphoryla-
tion is occuring (with regard to rates and differences).
Third, the ratio of intensities, M to G2, for each of the
lines differs and the intensity of pS780 and MPM-2 does not
correlate between cell lines. Therefore, the antigen density
and/or the net activity of the responsible enzyme systems var-
ies between cell lines. This variation is not enough to eliminate
the use of the pS780 as a mitotic marker.
Fourth, the detection of mitotic cells in DU-145 cells that
do not express appreciable Rb was robust and efficient using
the phospho-S780-Rb antibody. Therefore, pS780 resides on
proteins in addition to Rb.
Multiple Proteins are Recognized by Antibody to
To examine specificity of the phospoho-S780 antibody at
the peptide level, we performed western blots on lysates of sev-
eral human cell lines that were untreated or treated with noco-
dazole (Fig. 6). In these experiments, a band at 140 kDa and a
cluster of bands from 57 to 63 kDa increased upon nocodazole
treatment. A band that migrated at 103 kDa that was consistent
with Rb (confirmed by immunoblotting with antibody to Rb)
increased in several experiments (e.g., Fig. 6A) but not all (e.g.,
Fig. 6B). Further, bands were detected that migrated at the same
mobility as Rb but were clearly not Rb (Fig. 6B, SAOS-2) and
also did not increase with nocodazole treatment. Flow cytometry
of SAOS-2 cells stained with phospho-S780-Rb and MPM2
showed robust detection of mitotic cells (Fig. 6C). SAOS-2 cells,
like DU-145, poorly express a truncated unstable protein (33).
These data further support the idea that the pS780 epitope
resides on several proteins; that it is present in interphase cells as
well as mitotic cells, and that the reactivity for some proteins
increased in mitosis.5Further, these data suggest that Rb is phos-
phorylated in mitosis at this site but not necessarily at an
increased fraction of total Rb. Detection of Rb phosphorylation
in mitosis is affected by the presence of reactive bands that
appear not to undergo increased mitotic phosphorylation but
migrate the same as Rb. Also, the level of Rb protein increases
throughout the cell cycle and the increase in phospho-S780-Rb
reactivity is similar to the increase in total Rb reactivity (calcu-
lated for some experiments from immunoblots of untreated and
nocodazole treated samples). We also immunoprecipitated pep-
tides from SAOS-2 and K562 cells with phospho-S780-Rb.
Rb was immunoprecipitated from K562 but not SAOS-2 cells
(Supplemental Fig. 1). We also performed a western blot on the
immunoprecipitates with MPM-2. Very few MPM-2 positive
peptides were immunoprecipitated by phospho-S780-Rb; how-
Figure 6. Phospho-S780-Rb recognizes peptides in addition to Rb.
(A) Immunoblots for several cell lines including DU-145, which
poorly express Rb with an exon 21 deletion. NDZ 5 nocodazole
treatment for 17 h. (B) Immunoblot of SAOS-2 cells that poorly
express a truncated Rb. K562 cells are an Rb positive control. The
a-pS780 lanes are in duplicate. (C) Bivariate histograms of pS780
and MPM-2 for Rb-positive Molt4 (left) and Rb-negative SAOS-2
(right) cells showing mitotic cell detection by immunoreactivity of
phospho-S780-Rb antibody in Rb positive and negative cell lines.
(Antibodies (a) for each immunoblot are labeled on the blots.)
4M to G1 transition cells that are not resolved by the pS780-MPM-2
plots in Figure 5 but are readily apparent in the MPM-2/Cyclin A plots
are also present.
5These data do not distinguish between increased phosphorylation,
increased protein content, or both. However, increased phosphorylation
is more likely.
Cytometry Part A ? 73A: 5?15, 2008 11
ever, an MPM-2 positive peptide that migrated at the molecular
mass of Rb was detected in K562 and not in SAOS-2 cells (Sup-
plemental Fig. 1).
Bands at 57 and 140 kDa are Phosphorylated Peptides
A membrane with transferred peptides from nocodazole
treated cells was divided in two and treated with calf alkaline
phosphatase (CAP) or phosphatase buffer before immunostain-
ing for pS780. Figure 7 shows that the bands at 140 kDa and 57
kDa disappeared, and therefore, are phosphorylated epitopes.
The prominent bands at 63 kDa decreased upon CAP treatment
and, therefore, are likely to be phosphopeptides. The case for
phosphorylation of all the bands that migrate with Rb is less
clear. This result was obtained twice. To confirm further, we
treated Molt4 and K562 cells with okadaic acid at a concentra-
tion that inhibits protein phosphatase 2A (34,35) and subjected
lysates to western blotting for pS780. The immunoreactivity of
both the 140 kDa and 57–63 kDa bands increased significantly
in the OA treated samples (Supplemental Fig. 2).
The pS780 Epitope Reactivity Remains after Acid
Denaturation of DNA
Beyond simple counting of mitotic cells and phase fraction
analysis, cytometry can be used for complex cell cycle analysis
(36). An analysis that we use often is BrdU labeling to obtain
timing information (37,38). Combining a mitotic marker, DNA
content, and cyclin A can be a powerful combination for mea-
suring the G2 ? M ? G1 transitions (22). Although there are
other approaches, in our hands, acid denaturation of DNA is
the most reliable and general of the methods for detecting BrdU
incorporation. To combine two or more immunofluorescence
probes (e.g., cyclin A, BrdU, and a mitotic marker) in which all
probes are mouse monoclonal antibodies requires primary la-
beled antibodies or antibodies of different heavy chain classes
and isotype specific secondary antibodies. We have found it con-
venient to use different combinations of primary and secondary
labeling. In this regard, the use of rabbit antibodies to phospho-
S10-histone H3 as a mitotic marker and secondary staining has
been especially useful because the secondary antibody fluoro-
chrome could be chosen as needed. For example, we have used
secondaries labeled with Pacific Blue, Pacific Orange, and Alexa
700 to enable analyses in which FITC, phycoerythrin, and Alexa
647 have been used on other probes. A problem with phospho-
S10-histone H3 as a mitotic marker in BrdU studies is that his-
tones are extracted by acid treatment. Therefore, we tested the
pS780 epitope for stability to acid denaturation. Figure 8 shows
bivariate plots of pS780 versus DNA with and without BrdU
labeling (top panels). Bivariate histograms of BrdU immunoflu-
orescence versus DNA for the same samples are shown in the
bottom panels. Detection of mitotic cells is as robust in these
samples as in samples that are not treated with acid (e.g., Fig. 1).
To demonstrate further the utility in BrdU labeling, we
performed a kinetic analysis of exponentially growing Molt4 af-
ter 30 min labeling with BrdU. The cells were labeled, then
chased with thymidine, and thereafter sampled at time intervals
and fixed for later staining with antibodies to cyclin A, pS780,
and BrdU, and DAPI. The primary gating strategy is shown in
supplemental data Figure 3. The secondary gating strategy is
shown in Figure 9A. After gating on singlets and live cells,6mi-
totic cells (color coded blue), and G2 cells (color coded green),
the bivariate distribution of cyclin A and pS780 was used to set
up regions of interest (Fig. 9A). In this plot, the regions R6 and
R7 are sequential (early ? late) windows on G2.7R8 is a win-
dow in the intermediate region that can be classified as early pro-
phase, late G2, or in between based on available information. It
is the period during which the enzymes that result in phospho-
rylation of the pS780 epitope (net activity of kinases and phos-
phatases) are changing such that phosphorylation is between
bistable states. R9 through R11 are sequential regions through
mitosis. R9 is consistent with prophase cells (high levels of cyclin
Figure 8. The pS780 epitope is resistant to acid treatment. Bivari-
ate histograms of pS780 versus DNA content or anti-BrdU fluores-
cence for Molt4 cells are shown. Left panels are from the control
(no BrdU incorporation, acid treatment). Right panels are from
cells labeled for 30 min. Compare top two panels to Figure 1
(bottom left panel).
Figure 7. The pS780 epitope is a phospho-epitope. Western blots
with phospho-S780-Rb antibody with and without treatment of
the membrane with calf alkaline phosphatase are shown.
6That is, cells that were alive at the time of fixation.
7See supplemental Figure 3 and its legend for explanation of G2
12New Biomarker for Detecting Mitotic Cells
A). R10 is consistent with prometaphase (cyclin A levels are in-
termediate and in the process of decreasing). R11 is consistent
with metaphase through telophase and cytokinesis (cyclin A is
depleted to the levels associated with absence of mitotic cyclins
(the anti-mitotic window governed by Cdh1)). R12 is a region
set on a nub of hyper reactivity of pS780 in G1 (defined by ab-
sence of cyclin A and 2C DNA content). R13 is set on G1 cells
that have levels of pS780 below the median reactivity for all G1.
The precedence that we expected was R6 ? R12. Figures 9B and
9C show the resulting single parameter BrdU immunofluores-
cence histograms at t 5 0.5 and 3.5 h with each panel and:
‘‘Rxb’’ corresponding to Rx gates/regions in 9A with additional
Figure 9. Gating scheme for kinetic analysis of the G2 ? M ? G1 transition. The top panel (A) shows a pS780 versus cyclin A bivariate histo-
gram for BrdU labeled Molt4 cells at t 5 0.5 h (immediately after labeling for 30 min). The cells represented in this plot were from the 2C stem
line only (i.e., the 4C endoreduplicated cells have been gated out). Color coding: G2 5 green, M 5 blue, gray 5 all the rest. The complete gat-
ing scheme is described in detail in Supplemental Figure 3. Gates were set on the upper panel histogram to delimit the approximate first and
second tertiles of the bivariate G2 distribution (R6, R7); the transition-phase (between G2 and M) (R8); prophase cells (R9); prometaphase cells
(R10); and late mitosis (R11). Additional gates were set on a pS780 positive G1 ‘‘nub’’ (R12) and the lower half of the main cluster of G1 cells
(R13) (identified additionally by DNA content). The second and third panels (B and C) are the BrdU distributions at t 5 0.5h (B) and t 5 3.5h (C)
with each panel (moving left to right) representing the cells gated from R6 through R13 (R6b maps to R6, etc.). Note the change in positive cell
numbers in the first 6 panels from 0.5 to 3.5h. D and E show the fraction of labeled cells plotted versus time for each of the panels depicted in
B and C (R6b through R13b are the positive gates mapping to R6 through R13). Panel D shows the relevant curves that are used to calculate
transit times for the G2 to M transitions. Panel E shows the same curves covering the M to G1 transition.
Cytometry Part A ? 73A: 5?15, 200813
gates to ensure that only the 2C stemline was analyzed (as shown
in Supplemental data).
The percentage of positive labeled cells at each time point
for each region (R6b through R13b) are plotted in Figures 9D
and 9E. As expected, the labeled cells ‘‘pass through’’ the win-
dows R6 through R12 sequentially. Note that the slope of the
curves used to fit the data, are sharper for the mitotic windows
than for the G2 or G1 windows, as expected for the reduction
of variation as the cell cycle progresses. Also, note that the
time delay at the half-max points between the ‘‘newborn’’ nub
(R12) and the region set on the bottom half of the pS780 epi-
tope G1 distribution (R13) is long enough to guess that the
pS780 distribution in G1 is not informative with respect to
time—that is, it is not likely that G1 cells are appreciably
younger or older at different levels pS780 (with the exception
of the ‘‘nub’’). Therefore, there are two states in G1 with
respect to the pS780 epitope—‘‘newborn’’ cells that are post-
cytokinetic and all the rest. This means that cells start out in
the ‘‘newborn’’ nub then descend (lose the mitotic phospho-
rylation) and then start the interphase cycle again from some
random point within the large G1 cluster. This may not apply
in the situation where a true G0 state exists, since activated G1
lymphocytes can easily be distinguished from resting lympho-
cytes (Supplemental Fig. 4). A final note to make about these
data is that very short time periods along the cell cycle can be
measured effectively. Subtracting the halftimes of the curves
for R9b–R8b, R10b–R9b, and R11b–R10b gives values of 21.2,
13.6, and 8.2 min, respectively. Adding these together, the me-
dian early mitosis length was 42 min (from the onset of pS780
mitotic phosphorylation to metaphase).
The pS780 Mitotic Pattern is Ubiquitous
Every cell line or cell type that we have tested has ren-
dered an immunofluorescence versus DNA pattern that identi-
fies mitotic cells. Table 2 lists the cell lines we have tested.
Most are human solid tumor or hematopoietic cancer cell
lines. The hTert lines are immortal, non-tumorigenic, euploid
human cell lines. We have also tested stimulated human T
cells. Data demonstrating this for T cells are presented in
Supplemental Figure 4.
Nature of the pS780 Epitope
We have presented significant evidence that the antibody
phospho-S780-Rb recognizes an epitope or closely related
epitopes on several peptides with high affinity. These are phos-
phopeptides and are highly expressed in mitosis. This expres-
sion is likely to be due to elevated kinase activity since it is
highly correlated with MPM-2 epitope expression. Although
reactivity with the two antibodies is highly correlated on a cel-
lular basis, there is little overlap in terms of the peptides on
which the epitopes are expressed. Like MPM-2, pS780 is resist-
ant to the acid treatment used to stain cells for BrdU. Thus,
pS780 is a robust marker for mitosis and useful as an alterna-
tive to MPM-2 in complex cytometric analysis of the cell cycle.
We searched the human genome for translated sequences that
might constitute the core of the epitope. As a start toward
identifying the epitope, we limited the search to peptides of
predicted mass of 130–150 kDa; at least one serine residue,
and at least four contiguous amino acids contained within the
sequence PTLSPIPH. That search produced 18 peptides (due
to isoforms) and 13 genes that code TLSP; 8 peptides that
code SPIP; 6 peptides (5 genes) that code PTLS, and 3 pep-
tides (2 genes) that code LSPI. This suggests that using infor-
matics as a guide and synthetic peptide inhibition, we may be
able to identify the epitope.
We wish to thank Minh Lam for assistance with confocal
microscopy. We thank Randy Wetzel and Brad Smith (Cell
Signaling Technology, Danvers, MA) for generous gifts of
phospho-S780-Rb antibody and SOAS-2 cells; T. Vincent
Shankey for the generous gift of PE conjugated cyclin A anti-
body, and Ed Luther (Compucyte Inc., Cambridge, MA) for
the algorithm to identify mitotic cells by LSC.
1. Anderson HJ, de Jong G, Vincent I, Roberge M. Flow cytometry of mitotic cells. Exp
Cell Res 1998;238:498–502.
2. Juan G, Traganos F, James WM, Ray JM, Roberge M, Sauve DM, Anderson H,
Darzynkiewicz Z. Histone H3 phosphorylation and expression of cyclins A and
B1 measured in individual cells during their progression through G2 and mitosis.
3. Clevenger CV, Epstein AL, Bauer KD. Modulation of the nuclear antigen p105 as a
function of cell-cycle progression. J Cell Physiol 1987;130:336–343.
4. Andreassen PR, Margolis RL. Microtubule dependency of p34cdc2 inactivation and
mitotic exit in mammalian cells. J Cell Biol 1994;127:789–802.
5. Gong J, Traganos F, Darzynkiewicz Z. Discrimination of G2 and mitotic cells by flow
cytometry based on different expression of cyclins A and B1. Exp Cell Res 1995;
6. Di Vinci A, Geido E, Pfeffer U, Vidali G, Giaretti W. Quantitative analysis of mitotic
and early-G1 cells using monoclonal antibodies against the AF-2 protein. Cytometry
7. Nolan GP. Deeper insights into hematological oncology disorders via single-cell
phospho-signaling analysis. Hematology Am Soc Hematol Educ Program 2006;
8. Perez OD, Nolan GP. Phospho-proteomic immune analysis by flow cytometry: From
mechanism to translational medicine at the single-cell level. Immunol Rev 2006;
9. Hedley DW, Chow S, Goolsby CL, Shankey TV. Pharmacodynamic monitoring of
molecular-targeted agents in the peripheral blood of leukemia patients using flow
cytometry. Toxicologic Pathology (in press).
10. Jacobberger JW, Sramkoski RM, Zhang D, Zumstein LA, Doerksen LD, Merritt JA,
Wright SA, Shults KE. Bivariate analysis of the p53 pathway to evaluate Ad-p53 gene
therapy efficacy. Cytometry 1999;38:201–213.
11. Monier K, Mouradian S, Sullivan KF. DNA methylation promotes Aurora-B-driven
phosphorylation of histone H3 in chromosomal subdomains. J Cell Sci 2007;120:
12. Scrittori L, Skoufias DA, Hans F, Gerson V, Sassone-Corsi P, Dimitrov S, Margolis
RL. A small C-terminal sequence of Aurora B is responsible for localization and func-
tion. Mol Biol Cell 2005;16:292–305.
13. Crosio C, Fimia GM, Loury R, Kimura M, Okano Y, Zhou H, Sen S, Allis CD, Sas-
sone-Corsi P. Mitotic phosphorylation of histone H3: spatio-temporal regulation by
mammalian Aurora kinases. Mol Cell Biol 2002;22:874–885.
14. Wang R, He G, Nelman-Gonzalez M, Ashorn CL, Gallick GE, Stukenberg PT, Kirsch-
ner MW, Kuang J. Regulation of Cdc25C by ERK-MAP kinases during the G2/M
transition. Cell 2007;128:1119–1132.
Table 2. Cells tested
14New Biomarker for Detecting Mitotic Cells
15. Escargueil AE, Larsen AK. Mitosis-specific MPM-2 phosphorylation of DNA topo-
isomerase IIalpha is regulated directly by protein phosphatase 2A. Biochem J
16. Andreassen PR, Skoufias DA, Margolis RL. Analysis of the spindle-assembly check-
point in HeLa cells. Methods Mol Biol 2004;281:213–225.
17. Mahnke YD, Roederer M. Optimizing a multicolor immunophenotyping assay. Clin
Lab Med 2007;27:469–485.
18. Geng Y, Yu Q, Sicinska E, Das M, Bronson RT, Sicinski P. Deletion of the p27Kip1
gene restores normal development in cyclin D1-deficient mice. Proc Natl Acad Sci
19. Kitagawa M, Higashi H, Jung HK, Suzuki-Takahashi I, Ikeda M, Tamai K, Kato J,
Segawa K, Yoshida E, Nishimura S, Taya Y. The consensus motif for phosphorylation
by cyclin D1-Cdk4 is different from that for phosphorylation by cyclin A/E-Cdk2.
Embo J 1996;15:7060–7069.
20. Sramkoski RM, Pretlow TG II, Giaconia JM, Pretlow TP, Schwartz S, Sy MS, Marengo
SR, Rhim JS, Zhang D, Jacobberger JW. A new human prostate carcinoma cell line,
22Rv1. In Vitro Cell Dev Biol Anim 1999;35:403–409.
21. Jacobberger JW, Sramkoski RM, Frisa PS, Ye PP, Gottlieb MA, Hedley DW, Shankey
TV, Smith BL, Paniagua M, Goolsby CL. Immunoreactivity of Stat5 phosphorylated
on tyrosine as a cell-based measure of Bcr/Abl kinase activity. Cytometry A
22. Soni DV, Jacobberger JW. Inhibition of cdk1 by alsterpaullone and thioflavopiridol
correlates with increased transit time from mid G2 through prophase. Cell Cycle
23. Jacobberger JW, Fogleman D, Lehman JM. Analysis of intracellular antigens by flow
cytometry. Cytometry 1986;7:356–364.
24. Schimenti KJ, Jacobberger JW. Fixation of mammalian cells for flow cytometric eva-
luation of DNA content and nuclear immunofluorescence. Cytometry 1992;13:48–59.
25. Sramkoski RM, Wormsley SW, Bolton WE, Crumpler DC, Jacobberger JW. Simulta-
neous detection of cyclin B1, p105, and DNA content provides complete cell cycle
phase fraction analysis of cells that endoreduplicate. Cytometry 1999;35:274–283.
26. Jacobberger JW. Stoichiometry of immunocytochemical staining reactions. Methods
Cell Biol 2001;63:271–298.
27. Frisa PS, Lanford RE, Jacobberger JW. Molecular quantification of cell cycle-related
gene expression at the protein level. Cytometry 2000;39:79–89.
28. Kamentsky LA, Kamentsky LD. Microscope-based multiparameter laser scanning cyto-
meter yielding data comparable to flow cytometry data. Cytometry 1991;12:381–387.
29. Bodnar AG, Ouellette M, Frolkis M, Holt SE, Chiu CP, Morin GB, Harley CB, Shay
JW, Lichtsteiner S, Wright WE. Extension of life-span by introduction of telomerase
into normal human cells. Science 1998;279:349–352.
30. Che S, Weil MM, Nelman-Gonzalez M, Ashorn CL, Kuang J. MPM-2 epitope
sequence is not sufficient for recognition and phosphorylation by ME kinase-H.
FEBS Lett 1997;413:417–423.
31. Wang X, Deng H, Basu I, Zhu L. Induction of androgen receptor-dependent apopto-
sis in prostate cancer cells by the retinoblastoma protein. Cancer Res 2004;64:1377–
32. den Elzen N, Pines J. Cyclin A is destroyed in prometaphase and can delay chromo-
some alignment and anaphase. J Cell Biol 2001;153:121–136.
33. Shew JY, Lin BT, Chen PL, Tseng BY, Yang-Feng TL, Lee WH. C-terminal truncation
of the retinoblastoma gene product leads to functional inactivation. Proc Natl Acad
Sci USA 1990;87:6–10.
34. Favre B, Turowski P, Hemmings BA. Differential inhibition and posttranslational
modification of protein phosphatase 1 and 2A in MCF7 cells treated with calyculin-
A, okadaic acid, and tautomycin. J Biol Chem 1997;272:13856–13863.
35. Millward TA, Zolnierowicz S, Hemmings BA. Regulation of protein kinase cascades
by protein phosphatase 2A. Trends Biochem Sci 1999;24:186–191.
36. Darzynkiewicz Z, Crissman H, Jacobberger JW. Cytometry of the cell cycle: Cycling
through history. Cytometry A 2004;58A:21–32.
37. Begg AC, McNally NJ, Shrieve DC, Karcher H. A method to measure the duration of
DNA synthesis and the potential doubling time from a single sample. Cytometry
38. White RA, Terry NH. A quantitative method for evaluating bivariate flow cytometric
data obtained using monoclonal antibodies to bromodeoxyuridine. Cytometry
39. Gasnereau I, Ganier O, Bourgain F, de Gramont A, Gendron MC, Sobczak-Thepot J.
Flow cytometry to sort mammalian cells in cytokinesis. Cytometry A 2007;71A:1–7.
Cytometry Part A ? 73A: 5?15, 2008 15