[CANCER RESEARCH 33,786-790,
April 1973]
Microfluorometric Comparisons of Chromatin Thermal Stability
in Situ between Normal and Neoplastic Cells1
Marvin R. Alvarez
Department of Biology, Cell Biology-Physiology Section, University of South Florida, Tampa, Florida 33620
SUMMARY
Microfluorometric
comparisons
of chromatin
thermal
stability in situ were made among nuclei isolated from various
types of normal and neoplastia cells of the mouse. Normal
tissue cell nuclei were isolated from liver, kidney, and spleen.
Neoplastic cell nuclei were derived from transplanted murine
lymphoma, sarcoma, mammary adenocarcinoma, and Ehrlich
ascites tumor.
The isolated nuclei were seeded on glass slides in equal
densities, fixed, and heated to various temperatures in 0.15 M
NaCl-0.015 M sodium citrate containing formaldehyde in
order to induce denaturation of the chromatin. The nuclei
were subsequently stained with acridine orange. Chromatin
thermal stability was compared by measuring the fluorescence
emission ratios of the chromatin-ligand complex at 530 and
590 nm for each temperature.
The fluorescence emission profiles obtained were charac
teristic of each tissue type. However, the curves from all
normal tissue cells showed a gradual shift in the emission ratio
with temperature while those of the neoplastic cells, in
general, showed a rapid shift in the fluorescence emission ratio
and achieved a consistently higher maximum ratio. The data
indicate that the chromatin of these neoplastic cells is more
thermolabile than that of normal tissue cells and suggest the
possible application of the method to cytodiagnosis.
INTRODUCTION
Since genetic regulation in eukaryotic cells may be related
to configurational changes in nuclear chromatin (8), it would
appear profitable to compare the physicochemical state of
chromatin in normal and neoplastic cells. This is particularly
true if the comparison could be done at the level of the light
microscope because it could serve as a means of distinguishing
incipient neoplastic cells from normal cells, which they may
resemble morphologically.
Acridine orange, a fluorescent nucleic acid ligand, has been
used as a probe to measure changes in the structural order of
chromatin in situ (4, 5, 10, 12, 16, 20). The application of this
microfluorometric
procedure permits one to monitor subtle
changes in the structural order of nuclear chromatin by
measuring changes in the fluorescence emission ratio at 590
nm (F59o) and 530 nm (FS30) after having heated to various
temperatures cells attached to glass slides.
1This investigation was supported by Research Grant F 71SF-2 from
the American Cancer Society, Florida Division, Inc.
Received September 27, 1972; accepted December 29, 1972.
786
This paper reports on microfluorometric
comparisons of
chromatin thermal stability between nuclei isolated from 3
normal tissues and 4 transplanted murine neoplasms. The
purpose of the work was to attempt to distinguish tumor cells
objectively on the basis of their acridine orange staining
properties after thermal denaturation of chromatin.
MATERIALS
AND METHODS
Nuclei of normal liver, spleen, and kidney cells were
obtained from adult BALB/c mice of either sex. The animals
were maintained in an air-conditioned room and provided with
food and water ad libitum. Ehrlich ascites tumor was
maintained by weekly transplant into BALB/c mice. Methylcholanthrene-induced
sarcomas (MC-8) were obtained from
C3H male mice after 174 passages, and lymphomas (M5183)
were obtained from BALB/c females after 213 passages.
Mammary adenocarcinomas (M4833) were taken from DBA/2
females after 157 passages. Mice bearing the latter 3 types of
tumors were obtained from Dr. W. Dunning at the Papanicolaou Cancer Research Institute and were sacrificed imme
diately upon arrival at our laboratory.
Animals were killed by etherization, and the entire organ or
tumor was quickly removed. In the cases of the solid
transplanted tumors, any obvious necrotic peripheral tissue
was dissected away upon excision of the tumors. The tissues
were washed briefly in ice-cold 0.25 M sucrose containing 2
mM MgCl2 (9). The minced pieces were gently homogenized in
9 volumes of the same solution for 2 to 4 min with a hand
glass homogenizer. The homogenates were filtered through a
single layer of cheesecloth into conical glass tubes and
centrifuged at 900 X g for 15 min at 4°.The supernatants
were discarded, and the pellets were resuspended in a solution
of 0.5% (w/v) Triton X-100,0.25 M sucrose, and 1 mM MgCl2
by repeated expulsion through a transfer pipet. The suspen
sions were centrifuged again at 900 X g for 15 min. This step
was repeated until there was no evidence of erythrocytes in
the suspensions. The final pellets were suspended in 0.25 M
sucrose containing 1 mM MgCl2. The isolation procedure was
carried out at 4°.
Nuclei were isolated from Ehrlich ascites tumor cells in the
following manner. Intact cells suspended in ascites fluid were
aspirated from the peritoneal cavities of mice killed by
etherization between 6 and 8 days after inoculation. The cells
in ascites fluid were packed by centrifugation, the supernatant
was decanted, and the pellet was resuspended in 0.9% NaCl
solution and packed again by centrifugation. At least 2 such
washes were repeated to remove the ascites fluid. The cells,
suspended in a solution of 0.5% Triton X-100,0.25 M sucrose,
CANCER RESEARCH
VOL. 33
Downloaded from cancerres.aacrjournals.org on June 14, 2017. © 1973 American Association for Cancer Research.
Chromatin Stability in Normal and Neoplastic Cells
and 1 mM MgCl2, were homogenized for 5 min in a glass
homogenizer with a Teflon pestle driven at slow motor speed.
The homogenate was filtered through a single layer of
cheesecloth into glass centrifuge tubes and centrifuged at 900
X g for 15 min. The supernatant was discarded, and the final
pellet was resuspended in 0.25 M sucrose containing 1 mM
MgCl2.
Drops of the nuclear suspensions were placed on chemically
clean thin glass slides and allowed to settle onto the surface.
The densities were adjusted so that the average number of
nuclei per unit area was the same for each cell type. The
preparations were immediately fixed in 10% formalin, pH 7.0,
at 26°for 30 min. The fixed, isolated nuclei attached to glass
slides were immersed for 1 hr in vessels containing 100 ml of
0.15 M NaCl and 0.015 M sodium citrate made in 10%
formalin adjusted to pH 7.O. The temperatures of the various
baths ranged from 25°to 90°.All slides used to construct a
given denaturation profile were treated simultaneously. After
denaturation,
the slides were quenched in ice-cold NaClsodium citrate containing formaldehyde for 5 min, dehydrated
in a graded ethanol series, and processed for staining without
storage (15).
The 3 phases of the acridine orange procedure, i.e.,
acetylation, staining, and diffusion, were carried out at
25 ±0.5° (15). Reactions were timed precisely so that
treatment of slide sets was as uniform as possible. Replicates
were made of all runs and microfluorometric measurements in
all cases were made immediately after staining.
Slide preparations were acetylated for 15 min in 40% acetic
anhydride in pyridine following a 5-min equilibration in
water-free pyridine. The nuclei were subsequently rinsed in
100% ethanol and brought to water through a graded ethanol
series. Slides were equilibrated for 5 min in NaHPO,»-citrate
buffer (pH 4.1, T/2 0.6) and stained for 15 min in IO-4 M
acridine orange prepared in the buffer at pH 4.1. Excess dye
was removed by diffusion in two 2.5-min changes of buffer,
and coverslips were mounted over buffer and sealed with clear
fingernail polish.
The fluorescence of the ligand bound to chromatin in
individual nuclei was measured at 530 nm (^530) and at 590
nm (FSQO) with a previously described microfluorometer (14).
Exciting radiation was generated by a xenon arc (XBO-150W)
operating from a time-stabilized power supply. The exciting
wavelength (X = 400 nm) was selected with a BG-12 filter and
focused on the specimen through a 0.90-N.A. phase-bright
field condenser with transillumination. Emitted and exciting
light were collected by a Phaco 40 X, 0.65-N.A. objective and
separated by a 510 nm cutoff filter. The image of the single
nucleus being measured was delimited by a variable circular
diaphram and focused by an optical train onto the entrance
slit of a Bausch and Lomb monochromator and projected to
fill the grating. The monochromator
was set up with a
dispersion of 6.4 nm/mm with exit and entrance slits of 1.50
and 2.68 mm, respectively, thus giving a band width of 9.6
nm. The photodetector was a RCA 1P21 photomultiplier tube
positioned over the exit slit of the monochromator. The tube
was operated at an input of 940 V from a solid-state
zener-regulated power supply. The output of the phototube,
which is proportional to the fluorescence intensity of the
stained object, was amplified through the vertical amplifier of
APRIL
an oscilloscope and read as a vertical deflection on the screen
in V/cm. For avoidance of photodecomposition
which results
from prolonged exposure to high-intensity exciting light, the
nuclei were aligned and focused in low-intensity light effected
by partial closure of the substage diaphragm.
RESULTS
Chromatin thermal denaturation profiles of normal spleen
cell nuclei are shown in Chart 1. The curves obtained with this
tissue are typical of those obtained with other normal tissue
cells, and these data are presented to show the levels and
sources of variation encountered with the procedure in the 3
normal tissues examined. The degree of denaturation at a given
temperature was determined by the ratio of fluorescence
emission of the dye-chromatin complex at 590 and 530 nm.
Each point represents the mean ±S.E. from at least 10 nuclei
selected from different locations on the slides. The 3 curves
shown in this illustration were derived from 3 different
animals of approximately the same age and weight. The curves
were run at different times, but the slides within a given run
were treated simultaneously.
Internuclear variation within a slide heated to a given
temperature was consistently small, although the standard
errors of the means generally increased with temperature.
Although the curves differed somewhat from run to run in
spite of the fact that controllable parameters of the procedure
0.6-
in
0*-
in
Ü.
0
0.2 -
O
¡75
t/>
LU
40
TEMPERATURE
to
to
100
(«C)
Chart 1. In situ thermal denaturation profiles of chromatin of
normal isolated mouse spleen cell nuclei. Each curve was from a
different animal and was run independently. Note small intracurve
variation in contrast with variation between curves.
1973
Downloaded from cancerres.aacrjournals.org on June 14, 2017. © 1973 American Association for Cancer Research.
787
Marvin R. Alvarez
were held constant, the profiles resemble each other and are
distinguishable from that of other normal tissues.
Three chromatin thermal denaturation profiles of Ehrlich
ascites tumor cell nuclei are shown in Chart 2. Again, the 3
curves were derived from tumors from 3 different animals and
each curve was run independently of the others. As in the case
of the spleen, intraslide variability was minimal, but there was
a notable difference between runs.
However, the fluorescence emission ratio of the tumor cells
is higher than that of the spleen at 25°by a factor of nearly 2.
Note also the steepness of the ascites tumor curves compared
to those of the spleen, indicating that deoxyribonucleoprotein
denaturation occurs more rapidly in the ascites tumor cell
nuclei than in the spleen cell nuclei.
The differences in chromatin thermal stability observed
between several normal and neoplastic cell types are
summarized in Chart 3. This composite graph shows the mean
profiles obtained from normal liver, kidney, and spleen and
from Ehrlich ascites tumor, adenocarcinoma, lymphoma, and
sarcoma cell nuclei. The curves from the neoplasms are shown
in dashed lines and those of the normal tissues in dark, solid
lines. In general, the curves of the neoplasms are steeper than
those of the normal tissues. These data indicate that at 90°the
degree of denaturation achieved is markedly and consistently
greater in all of the neoplastic cells than in the normal cells.
Thus, after heating at 90°, the neoplastic cells could be
distinguished from normal cells solely on the objective basis of
their fluorescence color.
thought to be weaker than in the highly condensed chromatin
(8).
The present results indicate that binding of the ligand shifts
from the intercalated to aggregate mode at a lower tempera
ture in neoplastic cells than in cells from normal tissues. This
appears to be generally true of cells that are suddenly
stimulated into increased RNA synthesis prior to extensive
proliferation activity. Such changes have been observed during
lymphocyte activation by phytohemagglutinin
(10), in nu
cleated erythrocytes activated by cell hybridization (5), in
activated lymphocytes in infectious mononucleosis (4), and in
plant cells stimulated by kinetin (12). The present results
indicate that proliferating neoplastic cells, rich in cytoplasmic
RNA, are no exception and that this property may prove
useful in recognizing malignant cells in the absence of mitotic
figures.
While each of the transplantable tumors used in this
investigation are composed of a homogenous cell population,
the normal tissues with which they were compared contain
several cell types. However, variations in the 590/530 ratios
were small for each point throughout the thermal denaturation
range in both cases. This unexpected observation results from
the fact that only nuclei of 1 or 2 cell types for each tissue was
measured. Thus, in the liver, only large round nuclei
corresponding to hepatocytes were measured while smaller
elongated nuclei corresponding to Kupffer cells or fibroblasts
were ignored. The same was true in the kidney, in which
parenchyma cell nuclei were selected in favor of fibroblast
DISCUSSION
Acridine orange binds to DNA in either of 2 modes (13,
14). The physical mode in which the dye binds appears to be
determined by the degree of structural order of the polymer if
other parameters such as dye concentration, pH, and ionic
strength remain constant. Thus, double-stranded, helical DNA
bound the dye by intercalation of acridine orange molecules
between adjacent base pairs within the interior of the DNA O
ro
helix. Single-stranded DNA regions bind the dye in an if)
04aggregate form through an electrostatic interaction between
the negatively charged groups of the ligand molecule and the
positively charged phosphate groups on the exterior of the
DNA helix (13, 14). The DNA-dye complex fluoresces in the
green region of the spectrum (530 nm peak) when the dye is CD
Ul
bound in the intercalated form, but the emission peaks shift
CC
0.21/5
toward the red as single strandedness and thus electrostatic or
aggregate dye binding increase (590 nm peak). Therefore, by O
measuring the ratio of emission at these 2 wavelengths, it is
possible to obtain information about the degree of structural
LJ
order of the DNA in situ (15).
The presence of polycationic histones decreases the amount
of ligand bound by the DNA by neutralizing the phosphate
groups on the exterior of the DNA helix and also stabilizes the
20
40
helix against strand separation induced by heat, thus decreas
ing the binding of the ligand in aggregate form (7). It has been
TEMPERATURE
(°C)
shown recently that acridine orange binds almost exclusively
Chart 2. In situ thermal denaturation profiles of chromatin of
to diffuse chromatin in mammalian cells (8); this is also the isolated Ehrlich ascites tumor cell nuclei. Each curve was from a
phase that supports the bulk of RNA synthesis (7). The different animal and was run independently. Note sharp rise and higher
DNA-histone interaction in this diffuse chromatin phase is finalF590/F530 in contrast with normal spleen in Chart 1.
788
CANCER
RESEARCH
VOL. 33
Downloaded from cancerres.aacrjournals.org on June 14, 2017. © 1973 American Association for Cancer Research.
Chromatin Stability in Normal and Neoplastic Cells
differences in fluorescence existed under several conditions of
pathology, individual evaluations showed marked overlap.
A part of the problem in attempting to recognize malignant
cells on the basis of the amount of fluorescent ligand bound
by the nucleus lies in the fact that many, if not most,
neoplastic cells are hyperdiploid. Since acridine orange binds
semiquantitatively to DNA, the amount of dye bound is a
reflection of the ploidy level and thus leads to scattered,
overlapping values. The method described in this paper is
based on differential thermal stability of the chromatin, which
appears to be independent of ploidy (2). However, technical
difficulties such as the effects of cell density (3) and the
effects of mixiing cell types (1) on chromatin-dye emission
ratios must be further studied. If, however, the results of these
preliminary investigations can be generalized to include many
types of cancers, it appears possible that this procedure could
be applied in cytodiagnosis to decrease subjectivity and might
prove to be of value in the diagnosis of precancerous lesions.
0.6-
o
n
in
Ü.
04-
Ifl
u_
O
CE
W
z
ai
at
ACKNOWLEDGMENTS
l/l
5
LU
I thank Dr. W. Dunning and the Papanicolaou Cancer Research
Institute for providing the solid murine tumors and Judith Harvey,
Elizabeth Ingalls, and Anne Truitt for their technical assistance.
20
too
REFERENCES
TEMPERATURE (°C)
Chart 3. Composite graph showing mean chromatin thermal
denaturation profile of nuclei isolated from normal spleen (X), kidney
(o), and liver (•)and from transplanted sarcoma (+), lymphoma (A),
Ehrlich ascites tumor (o), and mammary adenocarcinoma (O) of the
mouse. Each curve is the mean of the means of 3 separate runs. In
general, the neoplastic cells show a more rapid rise and a higher
ultimate emission ratio than the normal tissue cells.
nuclei, and in the spleen, in which lymphocyte and white pulp
parenchyma predominate numerically.
The cytological recognition of malignant cells in a clinical
situation is based primarily on nuclear characteristics such as
variations in size, shape, structure, and hyperchromasia (11).
However, the certainty of identification increases with the
number of malignant characteristics recognized. The presence
of 2 or more characteristics is necessary to make a safe
diagnosis, and accuracy is increased when based on changes in
many cells. For these reasons, additional characters, partic
ularly objective ones, are most desirable in aiding cytodiagnosis.
Acridine orange has been used previously in clinical
cytology to recognize malignant cells. The method of von
Bertalanffy et al. (17) and its modifications are based on
fluorescence metachromasia in cells resulting from the binding
of the ligand to both DNA and RNA. Since malignant cells
frequently exhibit more cytoplasmic RNA than normal cells, it
is sometimes possible to distinguish them in smears on this
basis. More recently, Wied et al. (18, 19) studied acridine
orange binding in smears of glandular epithelium of uterine
endometrium and endocervix in various pathological condi
tions. These authors measured microfluorometrically
the
amount of ligand bound in the intercalated mode by the cell
nuclei. Although they found that statistically significant
1. Alvarez, M. R. Alterations of Deoxyribonucleoprotein Thermal
Stability Induced by Lymphocyte-Ehrlich Ascites Tumor Cell
Interaction. Exptl. Cell Res., 69: 72-80, 1971.
2. Alvarez, M. R., and Reyniers, J. Microspectrofluorometric Com
parison of Acridine Orange Dye Binding in Meristematic and
Parenchymal Nuclei of the Orchid Embryo. Exptl. Cell Res., 67:
326-332,1970.
3. Bolund, L., Darzynkiewicz, Z., and Ringertz, N. Cell Concentration
and the Staining Properties of Nuclear Deoxyribonucleoprotein.
Exptl. Cell Res., 62: 76-89, 1970.
4. Bolund, L., Gahrton, G., Killander, D., Rigler, R., and Wahren, B.
Structural Changes in the Deoxyribonucleoprotein Complex of
Leukocytes from Patients with Infectious Mononucleosis. Blood,
25: 322-332, 1970.
5. Bolund, L., Ringertz, N., and Harris, H. Changes in the
Cytochemical Properties of Erythrocyte Nuclei Reactivated by Cell
Fusion. J. Cell Sci., 4: 71-87, 1969.
6. Frenster, J. Nuclear Polyanions as De-repressors of Synthesis of
Ribonucleic Acid. Nature, 206; 680-683, 1965.
7. Frenster, J. A Model of Specific De-repression within Interphasc
Chromatin. Nature, 206: 1269-1270, 1965.
8. Frenster, J. Electron Microscopic Localization of Acridine Orange
Binding to DNA within Human Leukemic Bone Marrow Cells.
Cancer Res., 31: 1128-1133,1971.
9. Hymer, W., and Kuff, E. Isolation of Nuclei from Mammalian
Tissues through the Use of Triton X-100. J. Histochem.
Cytochem., 12: 359-363, 1964.
10. Killander, D., and Rigler, R. Activation of Deoxyribonucleoprotein
in Human Leukocytes Stimulated by Phytohemagglutinin. Exptl.
Cell Res., 54: 163-170,1969.
11. Koss, L. Diagnostic Cytology. Philadelphia: J. B. Lippincott Co.,
1972.
12. Piesco, N., and Alvarez, M. R. Nuclear Cytochemical Changes in
Onion Roots Stimulated by Kinetin. Exptl. Cell Res., 73:
129-139, 1972.
13. Rigler, R. Microfluorometric Characterization of Intracellular
APRIL 1973
Downloaded from cancerres.aacrjournals.org on June 14, 2017. © 1973 American Association for Cancer Research.
789
Marvin R. Alvarez
Nucleic Acids and Nucleoproteins by Acridine Orange. Acta
Physiol. Scand. Suppl., 67: 3-123, 1966.
14. Rigler, R. Acridine Orange in Nucleic Acid Analysis. Ann. N. Y.
Acad. Sci., 151: 211-224, 1969.
15. Rigler, R., Killander, D., Bolund, L., and Ringertz, N. Cytochemical Characterization of Deoxyribonucleoprotein in Individual Cell
Nuclei. Techniques for Obtaining Heat Denaturation Curves with
the Aid of Acridine Orange Microfluorometry and Ultraviolet
Microspectrophotometry. Exptl. Cell Res., 55: 215-224, 1969.
16. Ringertz, N., Gledhill, B., and Darzynkiewicz, Z. Changes in
Deoxyribonucleoprotein during Spermiogenesis in the Bull. Exptl.
Cell Res., 62:204-218,
1970.
790
17. von Bertalanffy, L., Masin, F., and Masin, M. Use of AcridineOrange Fluorescence Technique in Exfoliative Cytology. Science,
124: 1024-1025, 1956.
18. Weid, G., Messina, A., Mangliano, J., Meier, P., and Blough, R.
Fluorospectrophotometric
Analysis of Endometrial Endocervical
Epithelial Cells. Acta Cytol., 8: 408-415, 1964.
19. Weid, G., Messina, A., Meier, P., and Blough, R. Fluorospectro
photometric Analysis on Cervical Epithelial Cells. Acta Cytol., 8:
61-67, 1964.
20. Zetterburg, A., and Auer, G. Proliferative Activity and Cytochemical Properties of Nuclear Chromatin Related to Local Density of
Epithelial Cells. Exptl. Cell. Res., 62: 262-270, 1970.
CANCER RESEARCH
Downloaded from cancerres.aacrjournals.org on June 14, 2017. © 1973 American Association for Cancer Research.
VOL. 33
Microfluorometric Comparisons of Chromatin Thermal Stability
in Situ between Normal and Neoplastic Cells
Marvin R. Alvarez
Cancer Res 1973;33:786-790.
Updated version
E-mail alerts
Reprints and
Subscriptions
Permissions
Access the most recent version of this article at:
http://cancerres.aacrjournals.org/content/33/4/786
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 14, 2017. © 1973 American Association for Cancer Research.