[CANCER RESEARCH 41, 3877-3880, October 1981]
0008-5472/81 /0041-OOOOS02.00
X-Ray Microanalysis of Electrolyte Content of Normal, Preneoplastic,
and Neoplastic Mouse Mammary Tissue
Nancy K. R. Smith,1 Sidra B. Stabler, Ivan L. Cameron, and Daniel Medina
Department of Anatomy, The University of Texas Health Science Center at San Antonio, San Antonio 78284 [N. K. R. S., S. B. S., I. L. C.J, and Department of Cell
Biology, Baylor College of Medicine, Houston, 77030 [D. M.], Texas
ABSTRACT
Intracellular sodium, chlorine, and potassium concentrations
(mmol/kg dry weight) were determined by electron probe Xray microanalysis of individual epithelial cells in freeze-dried 2jum sections of mouse mammary tissue which were cut at
-30°. A model system was utilized in order to compare ele
mental content of cells from normal pregnant mammary tissue
and preneoplastic and neoplastic mammary tissues from fe
male BALB/cCrlMed
mice. Animals were killed by cervical
dislocation, and tissue was rapidly frozen in liquid propane.
Normal mammary glands were obtained from primiparous mice
at 16 to 17 days of gestation. Tissue from the hyperplastic
alveolar nodule line D1 was removed from donor mice 12 to 16
weeks after transplantation into the cleared mammary fat pad.
All mammary adenocarcinomas,
D1T, were primary tumors
which developed in mice with transplants of nodule line D1.
Data were collected from five animals (10 cells/animal) in
each of the three groups. It was found that the electrolyte
content of cells of preneoplastic tissue was the same as that of
the normal mammary tissue but was significantly elevated in
neoplastic tissue (162, 130, and 48% increases for sodium,
chlorine, and potassium, respectively). Thus, an increase in
electrolyte content seems to be associated with the transfor
mation to a neoplastic state and not associated with conversion
to the preneoplastic state.
INTRODUCTION
Our laboratory is interested in the role of intracellular ions,
particularly sodium, and changes in their concentration with
regard to mitogenesis, growth control, and oncogenesis. It
appears that changes in the intracellular concentration of in
organic substances and small ions might cause a cascade of
physiological, biochemical, and morphological events leading
to a coordinated and appropriate response on the part of the
cell. Our findings have been presented in recent reviews (5, 8,
9, 28) which discuss the importance of the intracellular envi
ronment in cellular regulation. We have used electron probe Xray microanalysis to test the hypothesis of Cone (10, 11) and
Hazlewood (18) that the intracellular concentration of sodium
should be greater in transformed cells than in their normal
counterparts. In 1978, we reported that sodium and chlorine
were more than twice as high in transplantable H6 hepatoma
cells as compared to hepatocytes from the host mice or from
control mice (30). No significant differences were found be
tween nuclear and cytoplasmic concentrations for either ele
' Supported by Grant RR07187-01, Biomedicai Research Grant Program,
Division of Research Resources, NIH, and in part by Grant PCM-804084. National
Science Foundation. To whom requests for reprints should be addressed.
Received May 15, 1981 : accepted July 7, 1981.
OCTOBER
1981
ment. Several other normal and tumor populations were ex
amined and similar results were found (3, 5-9). Other elements
(potassium, magnesium, phosphorus, sulfur) did not show a
consistent pattern of elemental concentration differences be
tween tumor and nontumor populations.
The concept of multistage development of neoplasia is gen
erally accepted. The qualitative and quantitative changes un
dergone by a normal cell population exposed to a carcinogen
are called preneoplastic progression. The neoplasm formed
continues to evolve, undergoing "tumor progression" (15,16).
Having established that transformed cells have higher sodium
and chlorine concentrations than their normal counterparts, it
was of great interest to investigate the electrolyte content of
cells which might be at an intermediate stage of transformation.
Cameron et al. (6) performed an X-ray microanalytical study of
hepatocytes from animals to which a carcinogen, hydrazine
sulfate, had been administered and found that the sodium and
chlorine levels were intermediate between those for normal and
transformed hepatocytes. These data support our basic hy
pothesis but are subject to criticism in that the carcinogen
undoubtedly had toxic effects and those effects could not be
distinguished from the carcinogenic effects. In order to study
populations of cells which are between normal cells and fully
transformed cells and for which the acute toxic effects of a
carcinogen can be ruled out as a source of confusion, it is
important to utilize a model system in which the intermediate
populations can be visualized, defined, and manipulated. The
model system which was used in the present study is the HAN,2
which is the principal preneoplastic lesion in mouse mammary
glands. The HAN is a focus of hyperplastic lobuloalveolar
development in an area of nonstimulated mammary gland.
HAN's can be induced in mice by mammary tumor viruses,
chemical carcinogens, irradiation, or prolonged hormone stim
ulation (21). All of these agents can also induce the formation
of neoplastic tissue in the mammary gland. DeOme ef al. (14)
demonstrated the high-risk preneoplastic nature of HAN's,
showing that HAN's transplanted into the mammary gland-free
fat pads of isogenic mice gave rise to hyperplastic alveolar
outgrowths which produced mammary tumors sooner and in
greater frequency than did normal tissue transplanted into the
contralateral mammary fat pads.
Samples of the alveolar outgrowth tissues can be serially
transplanted from one fat pad to another indefinitely, always
giving rise to lobuloaveolar tissue which completely fills the fat
pad. By this technique, nodule outgrowth lines can be estab
lished. These lines can be considered the in vivo equivalent of
in vitro cell lines (21).
Numerous nodule outgrowth lines have been developed us2The abbreviations used are: ANOVA, analysis of variance statistical test;
HAN, hyperplastic alveolar nodule; NMR, nuclear magnetic resonance.
3877
Downloaded from cancerres.aacrjournals.org on June 16, 2017. © 1981 American Association for Cancer Research.
N. K. R. Smith et al.
ing the fat pad transplantation technique of DeOme ef al. (14).
These lines have been characterized with regard to numerous
parameters (Refs. 21, 22, and 24; see "Discussion")
in an
effort to identify a marker(s) for mammary preneoplasia and to
determine whether such a marker(s) would provide information
on the essential differences among normal, preneoplastic, and
neoplastic cells. For these reasons, it was of great interest to
us to utilize this model system to compare the electrolyte
content of the 3 types of mammary epithelial cells.
MATERIALS AND METHODS
Tissues were obtained from female BALB/cCrlMed
mice that were
bred and maintained in a closed conventional mouse colony in the
Department of Cell Biology, Baylor College of Medicine. Normal mam
mary glands were obtained from primiparous mice at 16 to 17 days of
gestation. The HAN line 01 originally arose in a hormonally treated
mouse and had been propagated by serial transplantation in cleared
mammary fat pads of syngeneic mice (26). Nodule line D1, although
originally characterized by low tumor-producing capabilities (21), has
recently progressed to a high tumor-producing line.3 The hyperplastic
tissue was removed from donor mice 12 to 16 weeks after transplan
tation. The mammary adenocarcinomas developed in mice with trans
plants of nodule line D1. All adenocarcinomas,
designated as D1T,
were primary tumors.
Animals were sacrificed by cervical dislocation, avoiding use of any
anesthesia which might affect elemental content of cells. Following the
rapid processing for X-ray microanalysis, additional tissue was fixed in
10% formalin with phosphate buffer and processed for light micros
copy.
For X-ray microanalysis, tissues were rapidly frozen and prepared
according to procedures which, with the following exception, we have
reported previously (3-8, 29). Frozen 2-jam-thick sections were
mounted on carbon-coated nylon support films over a 2-mm hole in a
1-inch-diameter spectrographically pure carbon disc (Ernest F. Fullam,
Inc., Latham, N. Y.). An aluminum foil retainer strip with a hole which
was covered with carbon-coated Formvar was placed over the section,
thus sandwiching the section and keeping it flat. The retainer was
attached to the carbon planchet at 2 points on opposite sides of the
hole by means of small amounts of colloidal graphite adhesive in
isopropanol (DAG-154; Ted Pella Co., Inc., Tustin, Calif.). Tissue was
freeze-dried and subsequently handled as reported previously. The top
retainer was popped off the planchet before the specimen was trans
ferred to the JSM-35 scanning electron microscope (JEOL USA, Peabody, Mass.) for anaysis under our standard conditions (accelerating
voltage, 15 kV; specimen current, 0.15 na; image mode, secondary
electron; raster size, 2.6 sq /im; analysis time, 100 sec; takeoff angle,
40°; specimen-to-detector
distance, 15 mm). A Si(Li) X-ray detector
(United Scientific. Mountain View, Calif.) and Tracor Northern (Middletown, Wis.) NS-880 X-ray analysis system were used.
For quantitation, the technique of Hall (17) of relating elemental
values for the characteristic X-ray peak (minus background)/continuum ratio to concentration was used. The values for the characteristic
X-ray peak (minus background)/continuum
ratio were converted to
absolute units of mmol/kg dry weight by applying conversion factors
which were derived from analysis of a series of standard solutions of
salts in bovine serum albumin which were processed in all ways
identical to the samples (7, 29).
Data were collected from 10 cells from each of 5 animals/experi
mental group and were analyzed statistically in our Computer Re
sources Center using the SPSS (Statistics Package for the Social
Sciences) (27) program. To determine whether mean elemental con
centration values were significantly different among the 3 groups, a
3 D. Medina, unpublished data.
3878
one-way ANOVA was performed for each element of interest (sodium,
chlorine, potassium). Since ANOVA F values simply indicate whether
or not significant differences exist, the following multiple range tests
were used for determining which means were significantly different
from each other at the a level (in parentheses): Student-Newman-Keuls
(0.05); Tukeys (0.05); and Scheffe (0.01 and 0.001).
RESULTS AND DISCUSSION
Examination of tissue prepared for light microscopy showed
the morphology of mammary tissue from the 3 groups of mice
to be as described by Medina (21). In the scanning electron
microscope, morphology was readily identified by means of the
secondary electron image. The 16- to 17-day pregnant mouse
was chosen as the control animal in our model system since,
morphologically, HAN's are indistinguishable from the lobuloalveolar development seen in normal prelactating mammary
cells (21 ). Histologically, alveoli in the preneoplastic outgrowth
are composed of a single row of epithelial cells surrounding a
small lumen. Fat and protein droplets are commonly seen in
the apical cytoplasm. Since the cytoplasm of the cells was
scant, making it difficult to locate an area of cytoplasm of
sufficient size to raster with certainty and not be picking up a
signal from excitation of extracellular contents, only data from
nuclei were analyzed. Previous studies from our laboratory (4,
7-9) have consistently not found a significant gradient between
nucleus and cytoplasm for any of the elements analyzed in any
of the tissue types analyzed (control, tumor, rapidly proliferat
ing normal, slowly proliferating normal) in rat or mouse. A
consistent criterion used for choice of cells for analysis was
that they be located near the blood supply in order to ensure
that the cells were viable and not necrotic. Only cells in the
outermost layer of a group of cells were analyzed. Areas of
obvious damage were avoided. Myoepithelial cells, which were
present in the normal and preneoplastic tissue but not in the
neoplastic tissue, were avoided. In nuclei which were analyzed,
excitation of the nucleolus was also avoided.
Results of a one-way ANOVA for each of the 3 elements
(sodium, chlorine, potassium), as well as the average concen
tration values (mmol/kg dry weight) and the standard errors of
the means, are given in Table 1. For each of the 3 elements,
concentration values for the control and preneoplastic cells are
not different from each other but are significantly different from
values for the tumor cells. On a dry weight basis, sodium is
Table 1
Mean intracellular (nuclear) electrolyte concentrations in normal, preneoplastic
(Dì),and neoplastic (D1T) mammary epithelial cells
Mammary tissue was taken in 2-mm cubes from female mice from 3 groups:
control, 16 to 17 days pregnant: preneoplastic. bearing nodule outgrowth line
01; and neoplastic, bearing tumor outgrowth line D1T. Tissues were frozen in
liquid propane, and 2-pm sections were cut at -30°, mounted on carbon-coated
nylon, and freeze dried. Energy-dispersive X-ray analysis of nuclei was carried
out at 15 kV, 0.15-na specimen current, for 100 sec.
mmol/kg dry wt
7a122
±
NormalPreneoplasticNeoplasticF
7304±
6156±
13253
±
5332±
9426±
±62480.0000Chlorine144
73020.0000Potassium287
±
10700.0000
±
ratioF
probabilitySodium116
" Each concentration value is the mean of 50 nuclei (5 animals, 10 cells/
animal) ±S.E.
CANCER
RESEARCH
Downloaded from cancerres.aacrjournals.org on June 16, 2017. © 1981 American Association for Cancer Research.
VOL. 41
Electrolyte
Content of Transformed Mammary Tissue
approximately 3 times higher in the tumor cells than in the for all 3 cell types, being intermediate in value for preneoplastic
preneoplastic and normal mammary cells and chlorine is twice cells, whereas the electrolyte content is the same for normal
as high. Potassium is elevated in the tumor cells but not to the and preneoplastic cells, which are both different from neoplas
tic cells.
same extent as sodium and chlorine.
Medina (21 -23, 25) has compiled data on the various prop
Elevation of sodium and chlorine in the mouse mammary
erties of preneoplastic HAN's as compared to normal and
tumor cells as compared to normal counterparts agrees with
our previous results for H6 hepatoma in the mouse (30), Morris neoplastic mammary tissue. Most of the properties of HAN's
are the same as for normal tissue. HAN's are different from
hepatoma 7777 in the rat (6), and mammary adenocarcinomas
C3H in the mouse and 13762 NF in the rat (8). Potassium was normal cells and similar to neoplastic cells only with regard to
also found to have increased for the H6 and 7777 hepatomas indefinite division potential and loss of steroid regulation of
but was not different from normal counterpart cells for the 2 DNA synthesis (24).
One interest in the model system reported herein derives
mammary adenocarcinomas.
It is of great interest to consider whether the increased from a quest for markers which will distinguish preneoplastic
electrolyte content, on a dry weight basis, of the mammary and/or malignant cells from their normal counterparts (2, 24).
tumor cells could be due to their taking on increased water and The search has proven to be lengthy and frustrating. Measure
hence not be a reflection of a true concentration change on a ment of NMR water proton relaxation times does have some
wet weight basis. If such were the case, one might expect each potential as a diagnostic tool. Most properties, however, are
of the major electrolytes to change to the same extent, which measured as averages for cell populations in tissue rather than
they do not do. It has been known for many years that the for individual cells. Electron probe X-ray microanalysis has the
water content of tumors can be higher than for the host tissues advantage of measuring intracellular elemental concentration
of origin (12). In 1971, Damadian (13) reported that the NMR values for individual cells, but with the model used, such
relaxation times of water protons in Walker sarcoma and in analysis has not revealed differences which would enable one
Novikoff hepatoma were significantly higher than in their normal to distinguish between normal and preneoplastic cells, al
tissues of origin in rats. The relaxation times were correlated though one can rather confidently identify a neoplastic cell by
its electrolyte content.
with water content and with state of transformation. The find
It would appear from the electrolyte content data reported
ings of Damadian (13) were confirmed by Hazlewood ef al. (19,
20) and were extended to mammary tissues. Hazlewood ef al. herein that an increase in electrolyte content occurs in asso
(19, 20) used NMR to analyze tissues from the same model ciation with some critical step undergone by a cell when it
system being reported herein, investigating several nodule transforms to the neoplastic state. Whatever this critical step
outgrowth lines and their derived tumors. Hazlewood et al. (19, is, a change in electrolyte content does not appear to be a
20) concluded that the relaxation times of water protons in progressive phenomenon with development of preneoplastic
HAN tissue.
tumors were almost doubled relative to normal and preneo
plastic mammary gland tissues and that there appeared to be
a progressive change in the relaxation times for the normal, REFERENCES
preneoplastic, and neoplastic mammary tissues so that the 3
1. Beali, P. T., Asch, B. B., Chang, D. C., Medina, D., and Hazlewood, C. F.
morphological stages could be distinguished by the average
Distinction of normal, preneoplastic, and neoplastic mouse mammary pri
mary cell cultures by water nuclear magnetic resonance relaxation times. J.
values of the relaxation times. Examination of the data showed
Nati. Cancer Inst., 64: 335-338, 1980.
a definite difference between normal and neoplastic tissue, but
2. Beali, P. T., Asch, B. B., Medina, D., and Hazlewood. C. Distinction of
the preneoplastic tissue showed high variability among different
normal, preneoplastic, and neoplastic mouse mammary cells and tissues by
nuclear magnetic resonance techniques. In: I. L. Cameron and T. B. Pool
outgrowth lines. The water content for the preneoplastic lines
(fids.). The Transformed Cell. pp. 294-326. New York: Academic Press,
also showed a high variability among outgrowth lines (31 to
Inc., 1981.
3. Cameron, I. L., Pool, T. B., and Smith, N. K. R. An X-ray microanalysis
58% water). Interpretation of the data was complicated by the
survey of the concentration of elements in the cytoplasm of different mam
fact that the analyses were carried out on whole tissue. In order
malian cell types. J. Cell. Physiol., 101: 493-502, 1979.
to circumvent this problem and to rule out the effects of
4. Cameron, I. L., Sheridan, P. J., and Smith, N. R. An X-ray microanalysis
study of the differences in concentration of elements in brain cells due to:
extracellular contents, connective tissue, fat, etc., primary cell
opiates, cell type, and subcellular location. J. Neurosci. Res., 3: 397-410,
cultures of the 3 tissue types were prepared and analyzed by
1978.
Beali et al. (1, 2). In the primary culture system, it was possible
5. Cameron, I. L., and Smith, N. K. R. Energy dispersive X-ray microanalysis of
the concentration of elements in relation to cell reproduction in normal and
to distinguish among normal, preneoplastic, and neoplastic
in cancer cells. Scan. Electron Microsc., 2; 463-474, 1980.
mammary epithelial cells with a high degree of accuracy (a <
6. Cameron, I. L.. Smith, N. K. R., and Pool, T. B. Increase in the intracellular
0.001) by NMR, with the relaxation times for preneoplastic
concentration of Na and CI as hepatocytes transform from the normal to the
neoplastic state: an electron probe microanalysis study. In J. M. Sturgess
cells being intermediate between those for normal and neo
(ed.), Electron Microscopy, Vol. 2, pp. 120-121. Baton Rouge. La.: Claitors
plastic cells. No significant differences were found in hydration
Publishing Division, 1978.
7. Cameron, I; L., Smith, N. K. R., and Pool, T. B. Element concentration
for the 3 cell types; each showed a water content of 90 to 91 %.
changes in mitotically active and postmitotic enterocytes: an X-ray microa
Assuming a similar equality of hydration for the 3 cell types in
nalysis study. J. Cell Biol., 80: 444-450, 1979.
vivo, it can be concluded that our observed differences in
8. Cameron, I. L.. Smith, N. K. R., and Pool, T. B. Intracellular concentration of
sodium and other elements as related to mitogenesis and to oncogenesis in
electrolyte content between normal, preneoplastic, and neo
vivo. Cancer Res., 40: 1493-1500,
1980.
plastic mammary epithelial cells were not due to differences in
9. Cameron, I. L., Smith. N. K. R., Pool, T. B., Grubbs, B. G.. Sparks, R. L,
intracellular water content. Comparison of our data and those
and Jeter, J. R., Jr. Regulation of cell reproduction in normal and cancer
cells: the role of Na, Mg, CI, K. and Ca. In: Gary L. Whitson (ed.), Nuclearof Hazlewood et al. (19, 20) and Beali ef al. (1, 2) indicates
Cytoplasmic Interactions in the Cell Cycle, pp. 249-270. New York: Aca
that there is not a direct correlation between elemental content
demic Press, Inc., 1980.
and freedom of the water since relaxation times are different 10. Cone, C. D., Jr. Unified theory on the basic mechanism of normal mitotic
OCTOBER
1981
3879
Downloaded from cancerres.aacrjournals.org on June 16, 2017. © 1981 American Association for Cancer Research.
N. K. R. Smith et al.
control and oncogenesis. J. Theor. Biol., 30. 151-181, 1971.
11. Cone, C. D . Jr. The role of the surface electrical transmembrane potential
in normal and malignant mitogenesis. Ann. N. Y. Acad. Sci., 23S: 420-435,
1974.
12. Cramer, W. On the biochemical mechanism of growth. J. Physio). (Lond.),
50. 322-334, 1916.
13. Damadian, R. Tumor detection by nuclear magnetic resonance. Science
(Wash. D. C.), »77:1151-1153, 1971.
14. DeOme, K. B., Faulkin. L. J.. Jr.. Bern, H. A., and Blair. P. B. Development
of mammary tumors from hyperplastic alveolar nodules transplanted into
gland-free mammary fat pads of female C3H mice. Cancer Res., 79: 515520, 1959.
15. Foulds. L. Tumor progression: a review. Cancer Res., 14: 327-339, 1954.
16. Foulds, L. Tumor progression and neoplastic development. In: P. Emmelot
and O. Mühlbock
(eds.). Cellular Control Mechanisms and Cancer, pp. 242258. Amsterdam: Elsevier. 1964.
17. Hall, T. The microprobe assay of chemical elements. In: Q. Oster (ed.).
Techniques in Biological Research. Vol. 1A, pp. 151-267. New York:
Academic Press, Inc., 1975.
18. Hazlewood, C. F. A role for water in the regulation of the electrical properties
and electrolyte composition of tissues during growth, differentiation and
tumorigenesis. In: R. Damadian(ed.). Nuclear Resonance Effects in Cancer.
Pierce City. Mo.: Pacific Press, in press. 1981.
19. Hazlewood, C. F., Chang, D. C., Medina, D , Cleveland, Q., and Nichols. B.
L. Distinction between the preneoplastic and neoplastic state of murine
mammary glands. Proc. Nati. Acad. Sei. U. S. A., 69:1478-1480, 1972.
20. Hazlewood, C. F., Cleveland, Q., and Medina. D. Relationship between
hydration and proton nuclear magnetic resonance relaxation times in tissues
of tumor bearing and non-tumor bearing mice: implications for cancer
detection. J. Nati. Cancer Inst , 52.' 1849-1853. 1974.
3880
21. Medina, D. Preneoplastic lesions in mouse mammary tumorigenesis. Meth
ods Cancer Res., 7: 3-53, 1973.
22. Medina, D. Tumor progression. In F. Becker (ed.). Cancer, A Comprehensive
Treatise. Vol. 3, pp. 99-119. New York: Plenum Publishing Corp.. 1975.
23. Medina, D. Preneoplasia in breast cancer, in William L. McGuire (ed.).
Breast Cancer, Vol. 2. pp. 47-102. New York: Plenum Publishing Corp.,
1978.
24. Medina, D., and Asch, B. B. Cell markers for mouse mammary preneoplasias.
In C. M. McQrath (ed.), Cell Biology of Breast Cancer, pp. 363-371. New
York: Academic Press, Inc., 1980.
25. Medina, D., Asch, B. B., Brinkley, B., and Mace, M. L. In vivo and in vitro
models for transformation of breast cells. In: M. J. Eennan, C. M. McQrath.
and M. A. Rich (eds.). Breast Cancer: New Concepts in Etiology and Control,
pp. 53-66. New York: Academic Press, Inc., 1980.
26. Medina. D., and DeOme,K. B. Effects of various oncogenic agents on tumorproducing capabilities of series D BALB/c mammarynodule outgrowth lines.
J. Nati. Cancer Inst., 45: 353-363, 1970.
27. Nie. N. H.. Hull, C., Jenkins. J. Q.. Steinbrenner, K., and Bent, D. H. SPSS
Statistical Package for the Social Sciences, pp. 398-430. New York:
McQraw Hill Book Co., 1975.
28. Pool, T. B., Cameron, I. L., Smith, N. K. R., and Sparks, R. L. Intracellular
sodium and growth control: a comparison of normal and transformed cells.
in-1. L. Cameron and T. B. Pool (eds.), The Transformed Cell, pp. 397-420.
New York: Academic Press. Inc., 1981.
29. Pool, T. B., Smith, N. K. R., Doyle, K. H., and Cameron, I. L. Evaluation of
a preparative method for X-ray microanalysls of soft tissues. Cytoblos, 28:
17-33, 1980.
30. Smith. N. R., Sparks, R. L., and Pool, T. B. Differences in the intracellular
concentration of elements in normal and cancerous liver cells as determined
by X-ray microanalysis. Cancer Res., 38: 1952-1959, 1978.
CANCER
RESEARCH
Downloaded from cancerres.aacrjournals.org on June 16, 2017. © 1981 American Association for Cancer Research.
VOL. 41
X-Ray Microanalysis of Electrolyte Content of Normal,
Preneoplastic, and Neoplastic Mouse Mammary Tissue
Nancy K. R. Smith, Sidra B. Stabler, Ivan L. Cameron, et al.
Cancer Res 1981;41:3877-3880.
Updated version
E-mail alerts
Reprints and
Subscriptions
Permissions
Access the most recent version of this article at:
http://cancerres.aacrjournals.org/content/41/10/3877
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. © 1981 American Association for Cancer Research.