[CANCER RESEARCH 38. 1893-1898,
0008-5472/78/0038-0000$02.00
July 1978]
A Comparison of Acridine Orange and Feulgen Cytochemistry of Human
Tumor Cell Nuclei1
James E. Gill, Leon L. Wheeless, Jr., Carol Manna-Madden,- Richard J. Marisa, and Paul K. Horan
Analytical
Cytology
Division,
Department
of Pathology,
University of Rochester
ABSTRACT
Specimens of cells derived from tumors of the human
female genital tract plus normal cells as standards have
been divided into aliquots and stained according to acridine orange or pararosanilin:Feulgen procedures. Acri
dine orange-stained cells were slit-scanned for 535 nm
nuclear fluorescence; Feulgen-stained cells were combscanned for 580 nm nuclear absorbance. For each speci
men examined, the tumor celhnormal cell ratio of mean
nuclear fluorescence following acridine orange staining
was greater than the tumor celhnormal cell ratio of mean
nuclear absorbance following Feulgen staining. The tumor
celhnormal cell ratio of mean nuclear fluorescence
ranged from 2.3 for a nonkeratinizing squamous cell
carcinoma to 3.9 for a keratinizing squamous cell carci
noma. The tumor celhnormal cell ratio of mean nuclear
absorbance ranged from 1.4 for a mixed mesoderma!
sarcoma to 2.3 for a small cell squamous cell carcinoma.
These results indicate that the elevated nuclear fluores
cence intensity from acridine orange-stained tumor cells
cannot be explained solely on the basis of elevated
Feulgen:DNA content. An alternative hypothesis, consist
ent with these results, is that DNA is the principal binding
substrate for intranuclear acridine orange and that the
DNA of certain tumor cells is more accessible to acridine
orange than is the DNA of normal cells.
INTRODUCTION
Cells derived from many human malignant tissues have
altered chromatin content and structure. The cellular Feulgen:DNA content of malignant tissues is generally elevated
(1), and the chromatin patterns as seen under the light
microscope are different from those associated with normal
tissues (9). Both of these generalizations hold true for
cancers of the uterine cervix in particular (13).
Research efforts directed toward cytology automation
have attempted to use elevated DNA content (14) and
altered chromatin patterns (12) as markers for abnormal
cells. Fluorescent dyes believed to bind to DNA within cells
have been tested for their ability to discriminate between
normal and abnormal cells. In particular, AO3 has been
shown to provide effective discrimination between normal
and abnormal cells when 535 nm nuclear fluorescence of
AO-stained cells has been measured by a slit-scan tech
nique (19).
1These studies were performed under National Cancer Institute Contract
N01-CB-53930.
2 Present address: Doctors' Laboratory, Cuyahoga Falls, Ohio.
3 The abbreviation used is: AO, acridine orange.
Received October 6. 1977; accepted March 22, 1978.
School of Medicine
and Denistry,
Rochester,
New York 14642
The purpose of the research was to determine the cytochemical basis for this discrimination. Although it has been
reported that intracellular AO:DNA complexes fluoresce
green, while AO:RNA complexes fluoresce red, the actual
cytochemical behavior of AO is complex (11, 21). The
staining protocol used for the experiments reported here
was derived (20) from the protocol of Von Bertalanffy (17).
It differs significantly from those (15) for which DNA and
RNA specificity has been claimed. However, since AO binds
to DNA and since the DNA content of abnormal cells is
elevated, it appeared plausible that the elevated nuclear
fluorescence from AO-stained abnormal cells might reflect
elevated DNA content (6). To test this hypothesis, we have
examined the AO staining of DNase I- and RNase-treated
cells and have compared Feulgen:DNA values and AO
nuclear fluorescence values of aliquots of the same clinical
specimens. The results of these experiments indicate that
DNA is the primary substrate for AO binding within cell
nuclei but contradict the hypothesis that the integrated 535
nm (green) fluorescence of AO-stained cells provides a
measure of their relative Feulgen:DNA content.
MATERIALS
AND METHODS
Collection and Preparation of Specimens. Tumor speci
mens were collected from surgery, surgical pathology, and
autopsy services. Specimens and their sources are sum
marized in Table 1. Cells derived from solid tumors were
rinsed from these tissues into clear Mucosol (Lerner Labo
ratories, Stamford, Conn.), syringed to break up clumps of
cells, and mixed with normal material if none was present
in the specimen. Cells taken by cervicovaginal scrape were
suspended immediately in clear Mucosol and combined
with a suspension of normal cervical material. Normal
specimens were collected from an outpatient clinic.
Nuclease Treatment. Cells suspended in Mucosol were
centrifuged and resuspended in 20 mM citrate:phosphate,
pH 3.0, with 0.1% Triton X-100 for 1 min to facilitate
nuclease penetration (5). Cells were then rinsed in pH 7.5
buffer, resuspended in buffer with appropriate salts, and
brought to 4 x 103 units/ml with DNase I (Sigma Chemical
Company, St. Louis, Mo.) or 103 units/ml with RNase
(Sigma). Cell suspensions were held at 37°Cfor 30 min. The
cells were then centrifuged and stained.
AO Staining. An aliquot of the cell suspension was
centrifuged, and the pellet was resuspended in 2.5 x 10~5
M AO:0.2 M phosphate buffer, pH 7.O. After 10 min, the cells
were centrifuged, resuspended in 3.8%glutaraldehyde:0.14
M phosphate buffer at pH 7, and stored in this suspension
until needed (3).
Pararosanilin:Feulgen
Staining. A second aliquot of the
JULY 1978
Downloaded from cancerres.aacrjournals.org on June 16, 2017. © 1978 American Association for Cancer Research.
1893
J. E. Gill et al.
Table 1
Description of specimens analyzed
in series; the reflector transition wavelength was 510 nm,
and the barrier filters were an LP 528 and a 525 to 566 nm
band pass (full width at half-maximum);
green fluorescence
materialDiagnosis"Mixed
Tumor
of nor
was selected for measurement. The objective was x50 with
mal ma
a numerical aperture of 0.95; effective slit width was 5 /nm.
terialbClinicClinicClinicft
Epi-illumination
intensity was reduced to approximately 5%
pathol
andfluidSolidCervicalscrapeSolidSolidVaginalscrapeSourceSurgical
mesodermalsarcomaOvarian
for location and positioning of cells (to avoid bleaching the
ogyAutopsyOperating
specimen) and then increased to the maximum for slit
Adeno-carcinomaSmall
scanning. Scanning of the stage and collection
of data
roomAutopsySurgical
cellsquamous
were controlled by a POP 11/40 computer (Digital Equip
cellcarcinomaKeratinizingsquamous
ment Corp., Maynard, Mass.). The fluorescence
intensity
profile was collected by the computer and analyzed for cell
diameter, nucleus diameter, ratio of nucleus diameter to
cellcarcinoma.metastatic
cell diameter, integrated nuclear fluorescence,
and inte
toliverNonkeratinizingsquamous
grated cell fluorescence (18).
Comb-Scan Measurements
of Integrated Nuclear Abpathol
sorbance. Absorbance measurements of Feulgen-stained
ogyOperating
cellcarcinoma
nuclei were obtained with the Axiomat system. For trans
ofcervixSarcoma;metastatic
mitted illumination, the light source was a tungsten:halogen
roomSource
lamp operated at 12 Vd.c. A linear wedge interference filter
froma
was used to select a band of light centered at 580 nm. The
mixedmesodermalsarcomaTissueSolid
field stop diameter was 0.16 mm, giving a 34-/j,m-diameter
circle of illumination
in the specimen plane. The objective
was x50 with a numerical aperture of 0.95 to provide an
" For diagnostic, see Ref. 13.
A No additional normal cells required.
effective photometer spot aperture with a diameter of 0.5
t¿min the specimen plane. The image was comb-scanned
cell suspension was centrifugea, 1 drop of the concentrate
was placed on 1 slide, and "pull-apart"
smears were gen
erated with a second slide. These slides were immersed in
methanol:formalin:glacial
acetic acid (85:10:5, v/v) for at
least 1 hr and then taken through a Feulgen staining
protocol (2).
Schiffs reagent was prepared according to the method
of Graumann (10) with the following modifications.
Basic
fuchsin (C.I. 42,500), 0.1 g, was dissolved in 15 ml of 1 N
HCI and then combined with 85 ml of 2.6 x 10~2 M potas
sium metabisulfite in distilled water and stored overnight in
the dark in a capped Teflon bottle. The solution was shaken
with activated charcoal (Darco G-60) for 2 min and then
filtered through a Millipore prefilter. The resulting staining
solution was clear and nonfluorescent.
Analyses of Stains. Absorption spectra of stains were
obtained with a Beckman Acta V spectrophotometer
and
compared with published spectra (22) to verify identity
and purity. Fluorescence
analysis of stains were made
with an Aminco-Bowman
spectrophotofluorometer
modi
fied for photon counting (8). Thin-layer chromatography
of "purified"
AO (Polyscience,
Inc., Warrington,
Pa.)
on silica or on cellulose
with a solvent mixture
of
butanol:ethanol:H,O:NH4OH
(32:10:10:0.3, v/v) revealed 1
minor impurity. Other commercial samples of AO had sub
stantial contamination.
The "purified" AO was used without
further purification.
Slit-Scan Measurements of Integrated Nuclear Fluores
cence. Slit-scans of individual AO-stained cells were re
corded with a Zeiss Axiomat microscope-photometer
sys
tem, equipped for both epi- and transmitted illumination.
For epi-illumination,
the light source was an Osram XBO150 xenon arc lamp. Exciter filters were KP 490 and KP 500
1894
across this spot aperture by stepping the scanning stage in
0.5-/¿mincrements at approximately
150 Hz under control
of the computer. Light passing through the aperture was
detected by a photomultiplier
tube, and the photomultiplier
output was converted to a voltage. This voltage was re
corded by the computer after each increment of the stage.
The average value from the first line of the comb scan was
calculated and considered as background intensity, /„.For
each succeeding
increment, the computer recorded the
measured intensity, /,/, and calculated the specimen absorbance of that position as A' = log (V/.«1')- The total
absorbance of the scanned nucleus was calculated as the
sum of all A'. Duplicate comb-scan measurements
were
made of specimen nuclear absorbance,
averaged if the
values agreed within 20%; otherwise, they were discarded,
which was rare. Single slit-scan measurements were made
of nuclear fluorescence,
since each measurement caused
partial fading of fluorescence.
Before specimens were measured for either nuclear ab
sorbance or fluorescence,
the absorbance
of a marked
"standard"
Feulgen-stained
nucleus was measured repeat
edly, and the mean and standard deviation was checked
against previous measurements
of this nucleus to verify
performance of the system. (Standard deviations were typi
cally 5 to 6% of the mean.)
Statistical Analysis and Protocol. Analysis of the data
was based on the following assumptions: (a) nuclear fluo
rescence from slit-scan measurements and nuclear absorb
ance from comb-scan
measurements
constitute
random
variables; (b) each of these random variables is associated
with a probability
density function such that mean and
standard deviations exist; (c) if A is the mean of the
integrated absorbance of abnormal cell nuclei and 8 is the
mean of the integrated absorbance of normal cell nuclei,
CANCER
RESEARCH
VOL. 38
Downloaded from cancerres.aacrjournals.org on June 16, 2017. © 1978 American Association for Cancer Research.
Cytochemistry of Human Tumor Cell Nuclei
Obtain
Tumor
Form Suspension
Seed with Normal
Divide Suspension
Aliquot
Tissue
of Single
Cells
Cells if Required
into Two Aliquots
Aliquot
1
I
i
Apply
Stain with AO
Cells to Slides
I
I
Fix in Glutaraldehyde
Fix in MFG
i
I
Feulgen
2
Apply
Stain
Cells to Slides
I
Slit-Scan for Nuclear
Fluorescence
I
Comb Scan for Nuclear
Absorbance
i
I
Computer Analysis and
Histogram of Nuclear
Fluorescence Distribution
Computer Analysis and
Histogram of FeulgenDNA Distribution
Chart 1. A flow diagram of the basic experimental protocol. See "Materi
als and Methods" for details. MFG, methanol:formalin:glacial acetic acid
(85:10:5, v/v).
then the ratio of the means, R = Ä/B, is also a random
variable; (d) repeated measurements of R would be nor
mally distributed; (e) if nuclear fluorescence from AOstained cells represents Feulgen:DNA content, then the
ratio of the means of abnormal and normal nuclear fluores
cence will be the same, statistically, as the ratio of the
means of abnormal and normal nuclear absorbance; (f) the
appropriate test for significant difference in the ratios of the
means is the f test (23).
The basic protocol for the experiments with human ma
terial is given in Chart 1.
Isolation and Staining of Rat Liver Nuclei. Isolation
followed the procedure of Umana and Dounce (16) with the
following modifications. Solutions of 0.44 M sucrose con
tained 1 mw MgCI2; the centrifugation through the 2.2 M
sucrose was carried out in a Beckman J21C centrifuge with
an SW-13 rotor spun at 13,000 rpm for 1 hr. The final
nuclear pellets were resuspended in 10 ml of 0.44 M su
crose, and 10 ml of 20% formalin in 0.44 M sucrose were
added and mixed by vortexing. The nuclei were fixed in this
medium overnight and then stained according to a pararosanilin:Feulgen procedure for cell suspensions (7).
Flow Cytometry of Rat Liver Nuclei. Suspensions of
Feulgen-stained nuclei in distilled water were diluted with
distilled water to a concentration of about 106 nuclei/ml.
The analysis and sorting were carried out with a Coulter
Model TPS-1 sorter (Coulter Electronics, Hialeah, Fla.).
Table 2
Mean nuclear fluorescence of normal and tumor cells after AO staining
of
means.72
SpecimenMixed
mesodermal sarcoma
Ovarian adenocarcinoma
Small cell squamous
cellcarcinomaKeratinizing
cellcarcinomaNonkeratinizing
squamous
90±
161 (52)
100 ±34 (30)
330
3 .30 1 .95
±24±
100100100100±24"
115± (54)(55)(32)(33)2
±
(27)(21)(24)(30)272
274386229346Tumor±
.74.86.29.461±±±1
.32.42.02.26
2323Ratio
1212.11
26±
219±
squamous
12±
98±
carcinomaSarcoma,
cell
metastaticNormal"100
14(22)c
220(149)
" The mean of the normal cell nuclear fluorescence is normalized to 100 for each
specimen. The mean of the tumor cell nuclear fluorescence is normalized by the same
factor.
b Mean ±S.D.
c Numbers in parentheses, number of cells measured.
Table 3
Mean nuclear absorbance of normal and tumor cells after pararosanilin:Feulgen staining
SpecimenMixed
means1
of
±29.46 (28)°
±59.2(25)
.41 ±0.72
mesodermal sarcoma
100 ±21.3 (45)
172 ±83.3(50)
1.73 ±0.91
Ovarian adenocarcinoma
Small cell squamous cell
(28)100
100 ± 14.7
(63)225
227 ±59.8
0.682.25
2.27 ±
carcinoma
±19.2 (33)
±101 (31)
± 1.10
Keratinizing squamous cell
carcinoma
184 ±62.4(30)
1 .84 ±0.62
Nonkeratinizing squamous
100 ±26.0 (25)
cell carcinoma
1 .52 ±0.83
Sarcoma, metastaticNormal"100
100 ±22.0
(29)Tumor141 151 ±76.4(29)Ratio
" The mean of the normal cell nuclear absorbance is normalized to 100 for each specimen. The
mean of the tumor cell nuclear fluorescence is normalized by the same factor.
* Mean ±S.D.
c Numbers in parentheses, number of cells measured.
JULY 1978
Downloaded from cancerres.aacrjournals.org on June 16, 2017. © 1978 American Association for Cancer Research.
1895
J. E. Gill et al.
Histograms of frequency versus integrated fluorescence
intensity were obtained (4). The light scatter channel was
used to gate out fluorescence
signals from small bits of
fluorescent debris. Sorted nuclei were examined to verify
that peaks in the histogram corresponded
to fluorescence
from intact single nuclei.
Quantitative Microscopy of Rat Liver Nuclei. For meas
urements of nuclear fluorescence or absorbance with the
Axiomat photometer system, 1 drop of the stained rat liver
nuclei suspension was smeared on a glass slide and al
lowed to air dry. The smear was covered with 1 drop of
Cargille's type A immersion oil and coverslipped. Individual
nuclei were slit-scanned
for integrated fluorescence
or
comb-scanned
for integrated
absorbance
as described
above. Statistical analysis of the ratio of tetraploid to diploid
fluorescence or absorbance was as described above.
RESULTS
AND DISCUSSION
Mean values for normal and tumor cell absorbance or
fluorescence
and the tumor celhnormal
cell ratio of the
means have been calculated for each of the 6 specimens
tested. Nuclear fluorescence data are summarized in Table
2; nuclear absorbance data are summarized in Table 3. If
nuclear fluorescence
from AO-stained cells represented
nuclear Feulgen:DNA content, then the ratio of the means
of nuclear fluorescence
would be indistinguishable
from
the ratio of the means of nuclear absorbance. Statistical
analyses of the distributions,
summarized in Table 4, indi
cate that in every case this "null" hypothesis, namely, that
the ratios are statistically
indistinguishable,
can be rejected
with a confidence level of at least 95%.
Histograms of data for 2 of the specimens tested are
shown in Charts 2 and 3. In each chart A represents the
nuclear fluorescence distributions
of normal and abnormal
cells, and B represents the Feulgen absorbance distribu
tions. Chart 2 represents the mixed mesodermal sarcoma
(Tables 1 to 4, Line 1) while Chart 3 represents the small
cell squamous cell carcinoma (Tables 1 to 4, Line 3).
Nuclease treatments of normal clinical specimens were
performed to test whether AO staining of nuclei was specific
for DMA. DNase treatment of cells reduced the nuclear
fluorescence
below the level of detection by the slit-scan
technique. From observation of slit-scan contours, residual
nuclear fluorescence
of the DNase treatment is estimated
to be less than 10% of the value for untreated cells.
Fluorescence microscopy of treated cells revealed a lack of
nuclear chromatin-like
structures. Any remaining nuclear
fluorescence was associated with the nuclear membrane.
RNase treatment
of cells left the nuclear fluorescence
statistically indistinguishable
from that of untreated cells.
These data indicate that AO staining of normal cell nuclei
is 90% or more specific for DNA and does not involve
nuclear RNA. However, these data do not indicate whether
the amount of nuclear fluorescence from AO-stained cells
represents the amount of DNA present. The variability of
nuclear fluorescence
from normal AO-stained cells (19)
indicates that factors other than DNA content must play a
role. In short, specificity does not indicate stoichiometry.
Experiments involving nuclease treatments of tumor cells
are pending.
It was necessary to test the possibility that the Axiomat
system was providing fluorescence
or absorbance meas-
30
25
20
15ID
ID
S'
ì
O
u.
l
s
500
0J
O
NUCLEAR FLUORESCENCE
20
500
NUCLEAR FLUORESCENCE
30-
15-
25-
IO-
20I5
5
IO
O
500
NUCLEAR
ABSORBANCE
Chart 2. Nuclear AO fluorescence and nuclear Feulgen absorbance distri
butions for normal epithelial cells (open bar) and cells derived from a mixed
mesodermal sarcoma (stippled bar). Both fluorescence and absorbance are
in arbitrary units, normalized so that the mean value for normal cells is 100.
Each bar is 40 units wide: both axes are linear.
1896
5
O
500
NUCLEAR ABSORBANCE
Chart 3. Nuclear AO fluorescence and nuclear Feulgen absorbance distri
butions for normal polymorphonuclear leukocytes (open oar) and cells
derived from a small cell squamous cell carcinoma (stippled bar).
CANCER
RESEARCH
VOL. 38
Downloaded from cancerres.aacrjournals.org on June 16, 2017. © 1978 American Association for Cancer Research.
Cytochemistry of Human Tumor Cell Nuclei
Table 4
Confidence levels for rejection of hypothesis that ratios of means
are indistinguishable
1000
800
level0.99950.99950.950.9950.950.9995
SpecimenMixed
sarcomaOvarian
mesodermal
adenocarcinomaSmall
cellcarcinomaKeratinizing
cell squamous
600
400
cellcarcinomaNonkeratinizing
squamous
§
200
cellcarcinomaSarcoma,
squamous
metastaticf6.464.651.973.511.984.47Confidence
O,
100
40
60
80
120
CHANNEL NUMBER
(RELATIVEFLUORESCENCEINTENSITY)
Chart 4. A flow cytometry histogram of frequency versus relative fluores
cence for pararosanilin:Feulgen-stained rat liver nuclei. Relative fluores
cence is the putative relative DMA content.
20
Table 5
Mean absorbance and fluorescence of Feulgen-stained rat liver nuclei
Property mea
suredAbsorbance"
means1
of
±12.5* (9)'
±19.3(11)
.97 ±0.31
Fluorescence0Diploid100
214 ±18.6(26)Ratio 2.14 ±0.43f1.28
100 ±18.0 (20)Tetraploid197
" The mean of the diploid nuclei absorbance is normalized to 100, and the mean of the
tetraploid nuclei is normalized by the same factor.
6 Mean ±S.D.
' Numbers in parentheses, number of nuclei measured.
dThe mean of the diploid nuclei fluorescence is normalized to 100, and the mean of
the tetraploid nuclei is normalized by the same factor.
urements that were not proportional to the fluorescence or
absorbance of the stained nuclei. For example, the meas
urements might have indicated that 1 nucleus had 1.5 times
the absorbance of a second nucleus when in fact the first
nucleus had twice the absorbance of the second. To make
such a test, we analyzed Feulgen-stained rat liver nuclei for
both fluorescence and absorbance. Additionally, the
stained rat liver nuclei were analyzed and sorted on the
basis of fluorescence intensity with a flow cytometer.
Pararosanilin:Feulgen staining of nuclei or cells gener
ates both fluorescence and absorbance with the integrated
intensities proportional to nuclear DMA content (2, 7). Rat
livers contain both diploid and tetraploid cells; hence,
fluorescence or absorbance measurements of Feulgenstained rat liver nuclei can be expected to generate bimodal
distributions, with the mean of the high-intensity peak equal
to twice the mean of the low-intensity peak. Flow cytometer
measurements of fluorescence from the rat liver nuclei
generated the histogram shown in Chart 4. The ratio of the
means of the 2 peaks, high intensityilow intensity, is 2.06
±0.45 (S.D.).
Measurements of absorbance or fluorescence from an
other aliquot of these rat liver nuclei, made with the Axiomat
system, are summarized in Table 5. The ratios of the means
of absorbance and fluorescence data are not significantly
different from 2.0. This result shows that the Axiomat
system does provide measurements that are proportional to
the fluorescence or absorbance of stained nuclei over the
intensity ranges of interest. The differences between the
fluorescence and absorbance ratios of means in Tables 2
and 3 are not due to measurement artifact.
JULY
Thus, for each clinical specimen tested, the data indicate
that nuclear fluorescence from AO-stained cells is not
proportional to Feulgen:DNA content but does provide a
measure of some other nuclear property associated with
abnormality. If DNA is assumed to be the substrate for AO
binding within the tumor as well as normal cell nuclei, it
appears that this other property may be increased DNA
accessibility within the chromatin of tumor-derived cells.
In work to be published elsewhere,4 it is shown that
propidium iodide as well as AO provides discrimination
between normal and tumor cells, based on measurements
of nuclear fluorescence. The data indicate that intercalative
binding of these fluorochromes occurs in both normal and
tumor cells, providing further evidence that DNA is the
binding substrate for AO in normal and tumor cells.
An important practical result of these findings, relevant
to cytology automation, is that integrated nuclear fluores
cence at 535 nm from AO-stained tumor cells provides a
better indication of abnormality than does Feulgen absorb
ance (or any other measure of Feulgen:DNA content).
In summary, the increase in 535 nm nuclear fluorescence
from AO-stained tumor cells is greater than the increase in
580 nm absorbance of Feulgen-stained tumor cells, relative
to values for normal cells, with a confidence level of at least
95%. This result can be accounted for on the basis of
increased accessibility of DNA to AO within the chromatin
of tumor cells.
4 J. E. Gill, L. L. Wheeless, Jr., C. Hanna-Madden. and R. J. Marisa. Cytofluorometric and Cytochemical Comparisons of Normal and Abnormal
Human Cells, manuscript in preparation.
1978
Downloaded from cancerres.aacrjournals.org on June 16, 2017. © 1978 American Association for Cancer Research.
1897
J. E. Gill et al.
ACKNOWLEDGMENTS
We thank Margaret Gambier and Marcia Zach for collecting the specimens
used in these studies and Thomas Schmitt for supplying the rat liver.
REFERENCES
1. Bohm, N.. and Sandritter, W. DMA in Human Tumors: A Cytophotometric
Study. Current Topics Pathol., 60. 151-214, 1975.
2 Bohm, N., and Springer, E. Fluorescence Cytophotometry: A Valuable
Method for the Quantitative Determination of Nuclear Feulgen-DNA.
Histochemistry, 76: 100-118, 1968.
3. Cambier, M. A.. Wheeless, L. L., Jr., and Patten. S. F., Jr. A New PostStaining Fixation Technique for Acridine Orange. Acta Cytol , 21: 477480, 1977.
4. Crissman. H. A., Mullaney, P. F., and Steinkamp, J. A. Methods and
Applications of Flow Systems for Analysis and Sorting of Mammalian
Cells. Methods Cell Biol.,9: 180-242, 1975.
5. Darzynkiewicz, Z.. Tráganos, F., Sharpless, T., and Melamed, M. R.
Lymphocyte Stimulation: A Rapid Multiparameter Analysis. Proc. Nati.
Acad. Sei. U. S., 73: 2881-2884, 1976.
6. Gill, J. E., Moran, P. K., and Wheeless. L. L., Jr. Cylofluorometric and
Cytochemical Comparisons of Normal and Abnormal Human Cells. J
Cell Biol.,70: 100a, 1976.
7. Gill, J. E., and Jotz, M. M. Further Observations on the Chemistry of
Pararosaniline-Feulgen Staining. Histochemistry, 46. 147-160, 1976.
8. Gill, J. E., Jotz, M. M., Young, S. G., Modest, E. J.. and Sengupta. S. K.
7-Amino-Actinomycin D as a Cytochemical Probe. I. Spectral Properties.
J. Histochem. Cytochem., 23. 793-799, 1975.
9. Graham. R. M. The Cytologie Diagnosis of Cancer, Ed. 2. Philadelphia:
W. B. SaundersCo.. 1963.
10. Graumann. W. Zur Standardisierung des Schiffschen Reagens. Z. wiss.
Mikroskopie, 67: 225-226, 1953.
11. Kasten, F. H. Cytochemical Studies with Acridine Orange and the
Influence of Dye Contaminants in the Staining of Nucleic Acids. Intern.
Rev. Cytol.. 2T: 141-198.1967.
1898
12. Kopp, R. E., Lisa. J.. Mendelsohn, J., Permick, B.. Stone. H., and
Wohlers, R. Coherent Optical Processing of Cervical Cytology Samples
J. Histochem. Cytochem., 24. 122-137, 1976.
13. Patten, S. F., Jr. Diagnostic Cytology of the Uterine Cervix. Baltimore:
The Williams & Wilkins Co., 1969.
14. Sprenger, E., Sandritter. W., Naujoks. H., Hilgarth. M., Wagner. D., and
Vogt-Schaden, M. Routine Use of Flow-Through Photometric Prescreen
ing in the Detection of Cervical Carcinoma. Acta Cytol., 21: 435-440,
1977.
15. Tráganos, F., Darzynkiewicz, Z., Sharpless, T., and Melamed, M. R.
Simultaneous Staining of Ribonucleic and Deoxyribonucleic Acids in
Unfixed Cells Using Acridine Orange in a Flow Cytofluorometric System.
J. Histochem. Cytochem., 25: 46-56, 1977.
16. Umana, R., and Dounce, A. L. Relationship of Rat Liver Cathepsins to
Autolytic Degradation of Proteins of the Liver Cell Nucleus during
Isolation. Exptl Cell Res.,35. 277-292, 1964.
17. Von Bertalanffy, L., Masin, M.. and Masin, F. A New and Rapid Method
for Diagnosis of Vaginal and Cervical Cancer by Fluorescence Micros
copy. Cancer, 77: 873-887, 1958
18. Wheeless, L. L., Jr., and Patten, S. F., Jr. Slit-Scan Cytofluorometry.
Acta Cytol.. 77: 333-339, 1973.
19. Wheeless, L. L., Jr., Patten, S. F., Jr.. and Cambier, M. A. Slit-Scan
Cytofluorometry: Data Base for Automated Cytopathology Acta Cytol.,
79. 460-464. 1975.
20 Wheeless, L. L., Jr., Patten. S. F., Jr., and Onderdonk, M. A. Quantitative
Comparison of Slide and Suspension Technique for Acridine Orange
Staining of Human Cells. Acta Cytol., 78: 8-12, 1974.
21 Wied, G. L.. Messina, A. M., Meier. P., and Rosenthal, E. Deoxyribonucleic Acid Assessments on Feulgen-Stained Endometrial Cells, and
Comparison with Fluorometric Values on Acridine Orange Stained
Material. Lab. Invest.. 14: 1494-1499. 1965.
22 Zanker, V. Ãœberden Nachweis definierter reversibler Assoziate (revers
ible Polymerisate") des Acridinorange durch Absorptions- und Fluoroeszenzmessungen in WässrigerLösung.Z. Physik. Chem., 799: 225258,1952.
23. Zar, J. H. Biostatistical Analysis, pp. 105-107. Englewood Cliffs. N. J.:
Prentice Hall. Inc., 1974.
CANCER
RESEARCH
VOL. 38
Downloaded from cancerres.aacrjournals.org on June 16, 2017. © 1978 American Association for Cancer Research.
A Comparison of Acridine Orange and Feulgen Cytochemistry
of Human Tumor Cell Nuclei
James E. Gill, Leon L. Wheeless, Jr., Carol Hanna-Madden, et al.
Cancer Res 1978;38:1893-1898.
Updated version
E-mail alerts
Reprints and
Subscriptions
Permissions
Access the most recent version of this article at:
http://cancerres.aacrjournals.org/content/38/7/1893
Sign up to receive free email-alerts related to this article or journal.
To order reprints of this article or to subscribe to the journal, contact the AACR Publications
Department at [email protected].
To request permission to re-use all or part of this article, contact the AACR Publications
Department at [email protected].
Downloaded from cancerres.aacrjournals.org on June 16, 2017. © 1978 American Association for Cancer Research.