- Email: [email protected]
Amplified expression of p21 ras protein in hormone-dependent mammary carcinomas of humans and rodents
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
Vol. 127, No. 2, 1985
Pages 699-706
March 15, 1985
AMPLIFIED EXPRESSION OF p21 ras PROTEIN IN HORMONE-DEPENDENT MAMMARY CARCINOMAS OF HUMANS AND RODENTS Michele E. DeBortoli*, Hussein Abou-lssa+, Boyd E. Haley#, and Yoon Sang Cho-Chung* *Cellular Biochemistry Section, Laboratory of Pathophysiology, National Cancer Institute, Building I0, Room 5B38, Bethesda, Maryland 20205 +The Ohio State University, Breast Cancer Research Laboratories, Department of Surgery, Columbus, Ohio 43210 #The University
of Wyoming, Department of Biochemistry, Laramie, Wyoming 82071
Received January 25, 1985
SUMMARY: Western blotting analysis was utilized to determine the amount of a ras oncogene product, p21 present in mammary carcinomas of humans and rats. The levels of p21 in hormone-dependent rat tumors was about 7-fold that of hormone-independent tumors. The majority of human breast carcinomas examined had high p21 levels, about 10-fold that of the normal breast tissue; 70% of these tumors were estrogen and progesterone receptor positive, p21 levels in the remaining tumors were 3-fold that of the normal breast tissue, regardless of the receptor status. Fibroadenomas and fibrocystic disease showed p21 levels similar to that of the normal mammary glands. Moreover, the high p21 levels in the mammary carcinomas correlated directly with high GTPase activity, as revealed by the photo-incorporation of 8-N3-[~-32p]GTP into the tumor lysates. The results suggest that hormone-dependency of mammary carcinomas may correlate with quantitative change in 'normal' p21 protein. © 1 9 8 5 A c a d ~ i c Press, Inc.
INTRODUCTION:
A fraction (~30%) of human mammary carcinomas have been
observed to regress after ovariectomy (I).
This hormone-dependence
mammary carcinomas (~)anthracene
of human carcinomas
of the rat (2, 3).
(DMBA)-induced
such an experimental
tumor,
of the host
is mimicked by some
Primary 7,12-dimethylbenz-
mammary carcinoma in the rat represents since it grows only when stimulated by estro-
gen and regresses upon ovariectomy The hormone-dependent
and/or hypophysectomy
of the host (2).
growth has been linked to the presence of an
estrogen receptor in target tissues, the nucleus has been considered
and the action of estrogen receptor in
to be of fundamental
importance
in the
0006-291X785 $1.50 699
VoI. 127, NO. 2, 1985
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
response of the tissues to hormonal stimulation regulated expression hormone-dependent
(4).
Furthermore,
estrogen-
of specific genes have been shown to correlate with the
growth (5).
More recently,
hormone-dependent
growth of DMBA-induced mammary carcinoma
was shown (6) to be associated with an enhanced expression of a cellular protooncogene ras (c-rasH),
the cellular homolog of the transforming
gene of Harvey
J
sarcoma virus (v-ras H ) (7).
The 21,000-dalton
(8) of c-ras H was a prominent
transforming
gene product
(p21)
in vitro translation product of mRNAs from the
growing tumors and the p21 translated protein sharply decreased in the regressing
tumors upon hormone-withdrawal
We now show that hormone-dependent rats contain high levels of 'normal'
(6).
mammary carcinomas
of both humans and
p21 protein.
MATERIALS %ND METHODS: DMBA-induced mammary carcinomas (2) in Sprague-Dawley random-bred female rats (200-250g) were used as a model of hormone-dependent tumors. N-nitrosomethylurea (NMU)-induced mammary carcinomas (3) and transplanted DMBA #i mammary carcinomas (tumor received from W.F. Dunning and carried on in the Laboratory of Pathophysiology) in Fischer 344 female rats (150-200g) were used as models of hormone-independent tumors. In addition to these rat tumor models, primary, estrogen and progesterone receptor defined human breast carcinomas as well as benign tumors and normal breast tissues (from The Ohio State University, Breast Cancer Research Laboratory, Columbus, Ohio) were used in the present study. Western blotting analysis was used to determine the amounts of p21 present in both rat and human mammary tumors. Cellular proteins present in the tumor extracts were separated by 12% NaDodSO4-PAGE and transferred to nitrocellulose. Nitrocellulose sheets were first incubated with 3% bovine serum albumin solution (37°C, 3h) to remove bound NaDodSO 4 and to block the residual sites on the cellulose, then sequentially exposed to p21 monoclonal antibody #Y13-259 (9), directed against the Harvey murine sarcoma virus-encoded p21 (kindly provided by D.R. Lowy), rabbit anti-rat IgG, and 125I-labeled protein A and were finally exposed for autoradiography. Comparison of the resulting autoradiograms with others in which normal rat serum was substituted for p21 monoclonal antibody, permitted identification of p21. To provide a reference p21, we analyzed a cell lysate from NIH 3T3 clone 13-3B-4 which had been transfected with Harvey murine sarcoma virus DNA (kindly provided by D.R. Lowy).
RESULTS AND DISCUSSION:
Fig I shows a Western blotting analysis of p21
in rat mammary carcinomas.
Extensively
labeled doublet bands, co-migrating
with the lower band of p21 from 13-3B-4 cells (Fig la, Lane I), were detected in the extracts of hormone-dependent la, Lanes 2 and 3).
The corresponding
DMBA-induced
tumors (Fig
bands detected in extracts
?00
of
Vol. 127, No. 2, 1985
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
a
I
I i
b.
1
43K
25.7K p21
:
14.3K 1
2
3
4
5
6
M
1
2
3
4
5
6
Fig I. Western blotting analysis of p21 in rat mammary carcinomas, rat mammary glands, and 13-3B-4 cells. Tumors excised and cleared of necrotic areas, and mammary glands removed, were immediately dipped into liquid nitrogen and pulverized, and homogenized in a Teflon-glass homogenizer with four volumes of ice-cold Buffer Ten (0.1M NaCI - 5mM MgCI 2- 1% Nonidet P-40 - 0.5% Na deoxycholate - 2 KIU/ml bovine aprotinin - 20mM Tris-HCl, pH 7.4). The homogenates were centrifuged at 750 x g for 20 min at 4 ° , and the resulting supernatants were used as the lysates. 13-3B-4 cell pellets were suspended in Buffer Ten (5 x 106 cells/ml), vortexed and allowed to stand for 20 min at 4 ° , passed through a 22-gauge needle i0 times, centrifuged at 750 x g for 20 min at 4 ° , and the resulting supernatants were used as cell lysates. Tumor and cell lysates were used either immediately after their preparations or after being stored at -20 ° up to I week. Cellular proteins were separated by 12% NaDodS04-PAGE (ii) and transferred to nitrocellulose sheets (12) (0.2 ~m, Schleicher & Schuell). Nitrocellulose sheets were washed and first incubated with 3% bovine serum albumin in NTE-NP40 (50 mM Tris-HCl, pH 7.5 150mM NaCI - 2mM EDTA - 0.1% Nonldet P-40) for 3 hr at 37 ° , and then sequentially incubated with media containing p21 monoclonal antiserum #Y13-259 (or normal rat serum) for 16 hr at 4 ° , rabbit anti-rat IgG (Cappel Lab.) for 2.5 hr in ice water bath, and 5 x 105 cpm/ml 1251-proteln A (Amersham Corp., specific activity ~ 30 mCi/mg) for I hr in ice water bath. The nitrocellulose sheets were air-dryed and exposed to Kodak XAR films for 12-36 hr at -20 ° . a, with monoclonal antibody #Y13-259; b, with normal rat serum; Lane I, 13-3B-4; Lane 2 & 3, DMBA-induced tumors; Lane 4, DMBA #i tumor; Lane 5, NMU-induced tumor; Lane 6, normal rat mammary gland; M, prestained molecular weight markers (Bethesda Research Laboratories, Inc.) 43K, ovalbumin; 25.7K, a-chymotrypslnogen; 14.3K, Lysozyme. Each Lane contained 150 pg of proteins for NaDodSO4-PAGE.
hormone-independent of n o r m a l (Fig
mammary
la, L a n e s
serum was used
4,
tumors glands 5 and
in p l a c e
(DMBA
were 6).
#i and N M U - i n d u c e d ) ,
of a p p r o x i m a t e l y The p21 b a n d s
of the Y 1 3 - 2 5 9
were
antibody
701
as w e l l
the same absent (Fig
as
those
low intensity when
Ib).
control
rat
Approximately
Vol. 127, No. 2, 1985
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
12% of DMBA-induced
tumors are known to be hormone-independent
To accurately correlate p21 levels with hormone-dependency we measured p21 in biopsy specimens
from DMBA-induced
to monitor tumor growth following host ovariectomy.
(I0).
of tumors,
tumors and continued It was found that
all tumors (5 out of 5) that continued to grow after ovariectomy contained low levels of p21 comparable to those in DMBA #I and NMU-induced
tumors,
whereas every tumor (17 out of 17) that regressed following ovariectomy contained high levels of p21. of autoradiograms
Quantitation
by densitometric
showed that p21 in the hormone-dependent
tracings
DMBA-induced
tumors was present in amounts roughly 7-fold greater than p21 levels in the hormone-independent
tumors.
To rule out the possibility the hormone-independent
that the diminished labeling of p21 in
tumors was due to a higher proteolytic
in the tumors, we tested various mixtures -independent
tumor extracts;
the amount of detectable independent
of hormone-dependent
activity and
these mixtures showed an additive effect on
p21 (data not shown),
suggesting that hormone-
tumors do not have higher proteolytic activity toward p21.
Thus, elevated levels of p21 in the hormone-dependent carcinomas appear to be specfically
rat mammary
related to the hormone-dependency
of
of these tumors. Fig 2 shows a Western blotting analysis carcinomas.
of p21 in primary human breast
Strongly labeled p21 doublets were detected in extracts
6 out of 7 estrogen and progesterone
receptor-positive
In the extracts of estrogen and progesterone ever, p21 bands were either undetectable exception of one tumor (Lanes 11-16). the receptor-negative
from
tumors (Lanes 4-10).
receptor-negative
tumors, how-
or only weakly labeled, with the
Interestingly,
tumors showed a preferential
the p21 doublet in
labeling of either the
lower band (Lanes 12, 13 and 15) or the upper band (Lane 14 and 16), whereas
in the receptor-positive
tumors, all doublets
labeling of each band (Lanes 4-10).
showed an equal
A low level of p21 was found in
normal breast tissues (Lane 1-3) and benign tumors (see Fig 3).
702
The p21
Vol. 127, No. 2, 1985
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
-
-
68K
- - 43K
- - 25.7K p21 --,'--18.4K
1 2 3 4 5 6 7 8 910111213141516M I
I
I
NB
I I R+
I R-
Fig
2. Western blotting analysis of p21 in estrogen and progesterone receptorpositive and-negative human breast carcinomas and normal breast tissues. The lysates were prepared from frozen-pulverized breast tumors and normal breast tissues, and the cellular proteins were separated by NaDodSO4-PAGE , transferred to nitrocellulose sheets, immunoblotted with monoclonal antibody #Y13-259, and then exposed for autoradiography as described in the legend to Fig i. Each lane was loaded with I00 ~g protein for NaDodS04-PAGE. NB, normal breast; R+, estrogen (ER) and progesterone (PR) receptor-positlve tumors; R-, the receptornegative tumors. Positive ER = > i0 fmole/mg protein; positive PR = > i0 fmole/mg protein. Cytoplasmic estrogen and progesterone receptor levels were determined by Scatchard analysis (13) of the data obtained from the dextran-coated charcoal assay (14, 15). Cytosol protein concentrations were measured by the method of Lowry et al (16). M, prestained molecular weight markers (Bethesda Research Laboratories, Inc.)
doublet bands from human breast tumors co-migrated
in NaDodSO4-PAGE
with
the lower band of p21 from 13-3B-4 cells, and the bands were absent when control rat serum was used in place of the #Y13-259 m o n o c l o n a l
antibody
(data not shown). The levels of p21 in a series of human breast tumors were estimated quantitatively
after compiling the results from several experiments
the type shown in Fig. 2. densitometric films.
tracings
The concentrations
of the band intensities
of p21 were compared from from the autoradiographic
It can be seen in Fig. 3 that 15 of 22 (68%) primary,
progesterone
receptor-positive
breast carcinomas
fold greater than those of normal breast tissues, the r e c e p t o r - n e g a t i v e
breast carcinomas
that of normal breast tissues.
of
estrogen and
contained p21 levels ~ lOwhereas
13 of 21 (62%)
contained p21 levels only ~ 3-fold
Levels of p21 in benign tumors of fibre-
703
Vol. 127, No. 2, 1985
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
13
t 00
•= 11
•
•00 00o0
&&
03
9 _J W
>
W _J
o•
7
e~
uJ >
5
w
• •
A && && &AA&
• 0
&&&
00
< 3
~^t~
R+
CARCINOMAS
I IT'I
BENIGN TUMORS
Fig 3. Relative p21 levels in estrogen and progesterone receptor positive ~-~nd negative (R-) breast carcinomas and benign lesions of flbroadenoma (~) and fibrocystic disease ( [] ). Each symbol represents a single tumor from a different patient. The p21 levels were quantified from densitometrLc tracings of autoradiographs of the Western blotting analysis. The levels of p21 in the tumors are expressed relative to the level of p21 in normal breast tissues which is set equal to 1 arbitrary unit.
adenoma and fibrocystie disease were not elevated above the levels present in normal breast tissues, whereas practically all the breast carcinomas (41 out of 43) contained elevated p21 levels at least 2-fold that of normal breast tissues (Fig 3). About 70% of estrogen and progesterone receptor-positive human breast carcinomas are responsive to endocrine therapy, i.e., hormone-dependent
(17).
Since 68% of the receptor-positive tumors examined in this study exhibited high p21 levels, a correlation may exist between hormone-dependency of breast tumors and p21 elevation.
It is not known, however, whether the
receptor-positive tumors containing high p21 levels reported here are all responsive to endocrine therapy.
These results with human breast tumors
are, however compatible with the data obtained with experimental rat mammary tumors of well defined hormone dependency (Fig i). GTP (or GDP)-bindlng is the only known biological function common to both v-ras p21 and c-ras p21 (18).
We examined whether the p21 species
identified in the human and rat mammary tumors would exhibit guanine nucleo704
Vol. 127, No. 2, 1985
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
tide binding activity by the use of the photoaffinity ligand, 8-N3-[y-32p] GTP (19).
Tumor lysates containing either high or low levels of p21 were
incubated with 8-N3-[y-32p]GTP for various lengths of time.
The incubation
mixtures were then exposed to UV light, quenched, and subjected to NaDodSO 4PAGE.
Only in these lysates containing low levels of p21, was there a time-
dependent appearance of a GTP-labeled protein band of MW 21K, presumably p21, as well as other labeled protein bands (data not shown).
These protein bands
were present only after photolysis with UV light, indicating a covalent attachment of the labeled nucleotide to the proteins.
That the tumor lysates
containing high levels of p21 did not exhibit labeling of p21 as well as other GTP binding proteins suggested the presence of high GTPase activities in the lysates.
In fact, 8-N3-[y-32p ] GTP photo-incorporation into mixtures of
lysates containing high and low levels of p21 resulted in a complete loss of the p21 band and marked decrease of other GTP binding protein bands (data not shown).
The low photo-incorporation of 8-N3-[y-32p]GTP into the tumor lysates
containing high levels of p21 was not due to low GTP binding activity: direct binding of [a-32p]GTP or [y-32p]GT P to p21 transferred to nitrocellulose sheets was found to be proportional to the amount of p21 present in the tumors (data not shown). Recently, quantitative enhancement of p21 ras expression in human mammary carcinomas has been reported (20).
In the present study, we showed a correla-
tion between hormone-dependency of mammary carcinoma and elevation of p21.
Estrogen and progesterone-receptor positive human mammary carcinomas
contained p21 at levels as much as lO-fold higher than those found in normal mammary tissues.
The elevated p21 in the mammary carcinomas was
accompanied by a high GTPase activity.
In the light of a recent report
(21) that 'normal' p21 possesses GTPase activity, while in mutant p21 this function of GTPase is impaired, the elevated p21 in the hormonedependent mammary carcinomas may be 'normal' p21 rather than structurally altered protein.
Thus, amplified, hormone-stimulated expression of
'normal' p21 may be involved in the hormone-dependent growth of mammary
705
Vol. 127, No. 2, 1985
carcinomas.
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
This conclusion has further support from our previous finding
(6) that over-expression of c-ras gene in hormone-dependent rat mammary carcinomas ceases in the regressing tumors following hormone-withdrawal. Western blotting analysis and photoactivated binding of 8-N3-[~-32p]GTP now make possible both quantitative and qualitative assessments of p21 expression in tissues.
The precise role of ras p21 in the processes of mammary
carcinogenesis is currently under study. ACKNOWLEDGEMENTS: We thank D.R. Lowy for his generosity in providing p21 monoclonal antibody #Y13-259 and clone 13-3B-4, the Harvey murine sarcoma virus transfectant of NIH 3T3 cells; M. Barbacid, R.H. Bassin, P.M. Gullino and D.R. Lowy for valuable comments on the manuscript; H. Perno for preparation of manuscript and S. Oremland for typing the manuscript. REFERENCES i. 2. 3. 4. 5. 6. 7. 8. 9. i0. ii. 12. 13. 14. 15. 16. 17.
18. 19. 20. 21.
Dao, T.L., (1972) Ann. Rev. Med. 23, 1-18. Huggin§, C., Briziarelli, G., Sutton, H., Jr. (1959) J. Exp. Med. 109, 25-41. Gullino, P.M., Pettigrew, H.M., and Grantham, F.H. (1975) J. Natl. Cancer Inst. 54, 401-404. Jensen, E.V. and DeSombre, E.R. (1972) Ann. Rev. Biochem. 789, 203-230. Huang, F.L. and Cho-Chung, Y.S. (1982) Biochem. Biophys. Res. Commun. 107, 411-415. Huang, F.L. and Cho-Chung, Y.S. (1984) Biochem. Biophys. Res. Commun. 123, 141-147. Ellis, R.W., DeFeo, D., Maryak, J.M., Young, H.A., Shih, T.Y., Chang, E.H., Lowy, D.R. and Scolnick, E.M. (1980) J. Virol. 36, 408-420. Shih, T.Y., Weeks, M.O., Young, H.A. and Scolnick, E.M. (1979) Virology 96, 64-79. Furth, M.E., Davis, L.J., Fleurdelys, B. and Scolnick, E.M. (1982) J. Virol. 43, 294-304. Cho-Chung, Y.S., Bodwin, J.S. and Clair, T. (1978) Eur. J. Biochem. 86, 51-60. Laemmli, U.K. (1970) Nature 227, 680-685. Towbin, H., Stahelin, T. and Gordon, J. (1979) Proc. Natl. Acad. Sci. USA 76, 4350-4354. Scatchard, G. (1949) Ann. N.Y. Acad. Sci. 51, 660-672. Korenman, S.G. and Dukes, B.A. (1970) J. Clln. Endocrinol. Metab. 30, 639-645. Powell, B., Garola, R.E., Chamness, G.C. and McGuire, W.L. (1979) Cancer Res. 39, 1678-1682. Lowry, O.H., Rosebrough, N.J., Farr, A.L. and Randall, R.J. (1951) J. Biol. Chem. 193, 265-275. McGuire, W.L., Carbone, P.O., Sears, M.E. and Eschert, G.C. (1975) In: Estrogen Receptors in Human Breast Cancer, (McGuire, W.L. Eds.) Raven, New York, pp. 1-7. Shih, T.Y., Papageorge, A.G., Stokes, P.E., Weeks, M.O. and Scolnick, E.M. (1980) Nature 287, 686-691 Geahlen, R.L. and Haley, B.E. (1979) J. Biol. Chem. 254, 11982-11987. Hand, P.H., Thor, A., Wunderlich, D., Muraro, R., Caruso, A. and Schlom, J. (1984) Proc. Natl. Acad. Sci. USA 81, 5227-5231. McGrath, J.P., Capon, D.J., Goeddel, D.V. and Levinson, A.D. (1984) Nature 310, 644-649.
706
Vol. 127, No. 2, 1985
Pages 699-706
March 15, 1985
AMPLIFIED EXPRESSION OF p21 ras PROTEIN IN HORMONE-DEPENDENT MAMMARY CARCINOMAS OF HUMANS AND RODENTS Michele E. DeBortoli*, Hussein Abou-lssa+, Boyd E. Haley#, and Yoon Sang Cho-Chung* *Cellular Biochemistry Section, Laboratory of Pathophysiology, National Cancer Institute, Building I0, Room 5B38, Bethesda, Maryland 20205 +The Ohio State University, Breast Cancer Research Laboratories, Department of Surgery, Columbus, Ohio 43210 #The University
of Wyoming, Department of Biochemistry, Laramie, Wyoming 82071
Received January 25, 1985
SUMMARY: Western blotting analysis was utilized to determine the amount of a ras oncogene product, p21 present in mammary carcinomas of humans and rats. The levels of p21 in hormone-dependent rat tumors was about 7-fold that of hormone-independent tumors. The majority of human breast carcinomas examined had high p21 levels, about 10-fold that of the normal breast tissue; 70% of these tumors were estrogen and progesterone receptor positive, p21 levels in the remaining tumors were 3-fold that of the normal breast tissue, regardless of the receptor status. Fibroadenomas and fibrocystic disease showed p21 levels similar to that of the normal mammary glands. Moreover, the high p21 levels in the mammary carcinomas correlated directly with high GTPase activity, as revealed by the photo-incorporation of 8-N3-[~-32p]GTP into the tumor lysates. The results suggest that hormone-dependency of mammary carcinomas may correlate with quantitative change in 'normal' p21 protein. © 1 9 8 5 A c a d ~ i c Press, Inc.
INTRODUCTION:
A fraction (~30%) of human mammary carcinomas have been
observed to regress after ovariectomy (I).
This hormone-dependence
mammary carcinomas (~)anthracene
of human carcinomas
of the rat (2, 3).
(DMBA)-induced
such an experimental
tumor,
of the host
is mimicked by some
Primary 7,12-dimethylbenz-
mammary carcinoma in the rat represents since it grows only when stimulated by estro-
gen and regresses upon ovariectomy The hormone-dependent
and/or hypophysectomy
of the host (2).
growth has been linked to the presence of an
estrogen receptor in target tissues, the nucleus has been considered
and the action of estrogen receptor in
to be of fundamental
importance
in the
0006-291X785 $1.50 699
VoI. 127, NO. 2, 1985
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
response of the tissues to hormonal stimulation regulated expression hormone-dependent
(4).
Furthermore,
estrogen-
of specific genes have been shown to correlate with the
growth (5).
More recently,
hormone-dependent
growth of DMBA-induced mammary carcinoma
was shown (6) to be associated with an enhanced expression of a cellular protooncogene ras (c-rasH),
the cellular homolog of the transforming
gene of Harvey
J
sarcoma virus (v-ras H ) (7).
The 21,000-dalton
(8) of c-ras H was a prominent
transforming
gene product
(p21)
in vitro translation product of mRNAs from the
growing tumors and the p21 translated protein sharply decreased in the regressing
tumors upon hormone-withdrawal
We now show that hormone-dependent rats contain high levels of 'normal'
(6).
mammary carcinomas
of both humans and
p21 protein.
MATERIALS %ND METHODS: DMBA-induced mammary carcinomas (2) in Sprague-Dawley random-bred female rats (200-250g) were used as a model of hormone-dependent tumors. N-nitrosomethylurea (NMU)-induced mammary carcinomas (3) and transplanted DMBA #i mammary carcinomas (tumor received from W.F. Dunning and carried on in the Laboratory of Pathophysiology) in Fischer 344 female rats (150-200g) were used as models of hormone-independent tumors. In addition to these rat tumor models, primary, estrogen and progesterone receptor defined human breast carcinomas as well as benign tumors and normal breast tissues (from The Ohio State University, Breast Cancer Research Laboratory, Columbus, Ohio) were used in the present study. Western blotting analysis was used to determine the amounts of p21 present in both rat and human mammary tumors. Cellular proteins present in the tumor extracts were separated by 12% NaDodSO4-PAGE and transferred to nitrocellulose. Nitrocellulose sheets were first incubated with 3% bovine serum albumin solution (37°C, 3h) to remove bound NaDodSO 4 and to block the residual sites on the cellulose, then sequentially exposed to p21 monoclonal antibody #Y13-259 (9), directed against the Harvey murine sarcoma virus-encoded p21 (kindly provided by D.R. Lowy), rabbit anti-rat IgG, and 125I-labeled protein A and were finally exposed for autoradiography. Comparison of the resulting autoradiograms with others in which normal rat serum was substituted for p21 monoclonal antibody, permitted identification of p21. To provide a reference p21, we analyzed a cell lysate from NIH 3T3 clone 13-3B-4 which had been transfected with Harvey murine sarcoma virus DNA (kindly provided by D.R. Lowy).
RESULTS AND DISCUSSION:
Fig I shows a Western blotting analysis of p21
in rat mammary carcinomas.
Extensively
labeled doublet bands, co-migrating
with the lower band of p21 from 13-3B-4 cells (Fig la, Lane I), were detected in the extracts of hormone-dependent la, Lanes 2 and 3).
The corresponding
DMBA-induced
tumors (Fig
bands detected in extracts
?00
of
Vol. 127, No. 2, 1985
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
a
I
I i
b.
1
43K
25.7K p21
:
14.3K 1
2
3
4
5
6
M
1
2
3
4
5
6
Fig I. Western blotting analysis of p21 in rat mammary carcinomas, rat mammary glands, and 13-3B-4 cells. Tumors excised and cleared of necrotic areas, and mammary glands removed, were immediately dipped into liquid nitrogen and pulverized, and homogenized in a Teflon-glass homogenizer with four volumes of ice-cold Buffer Ten (0.1M NaCI - 5mM MgCI 2- 1% Nonidet P-40 - 0.5% Na deoxycholate - 2 KIU/ml bovine aprotinin - 20mM Tris-HCl, pH 7.4). The homogenates were centrifuged at 750 x g for 20 min at 4 ° , and the resulting supernatants were used as the lysates. 13-3B-4 cell pellets were suspended in Buffer Ten (5 x 106 cells/ml), vortexed and allowed to stand for 20 min at 4 ° , passed through a 22-gauge needle i0 times, centrifuged at 750 x g for 20 min at 4 ° , and the resulting supernatants were used as cell lysates. Tumor and cell lysates were used either immediately after their preparations or after being stored at -20 ° up to I week. Cellular proteins were separated by 12% NaDodS04-PAGE (ii) and transferred to nitrocellulose sheets (12) (0.2 ~m, Schleicher & Schuell). Nitrocellulose sheets were washed and first incubated with 3% bovine serum albumin in NTE-NP40 (50 mM Tris-HCl, pH 7.5 150mM NaCI - 2mM EDTA - 0.1% Nonldet P-40) for 3 hr at 37 ° , and then sequentially incubated with media containing p21 monoclonal antiserum #Y13-259 (or normal rat serum) for 16 hr at 4 ° , rabbit anti-rat IgG (Cappel Lab.) for 2.5 hr in ice water bath, and 5 x 105 cpm/ml 1251-proteln A (Amersham Corp., specific activity ~ 30 mCi/mg) for I hr in ice water bath. The nitrocellulose sheets were air-dryed and exposed to Kodak XAR films for 12-36 hr at -20 ° . a, with monoclonal antibody #Y13-259; b, with normal rat serum; Lane I, 13-3B-4; Lane 2 & 3, DMBA-induced tumors; Lane 4, DMBA #i tumor; Lane 5, NMU-induced tumor; Lane 6, normal rat mammary gland; M, prestained molecular weight markers (Bethesda Research Laboratories, Inc.) 43K, ovalbumin; 25.7K, a-chymotrypslnogen; 14.3K, Lysozyme. Each Lane contained 150 pg of proteins for NaDodSO4-PAGE.
hormone-independent of n o r m a l (Fig
mammary
la, L a n e s
serum was used
4,
tumors glands 5 and
in p l a c e
(DMBA
were 6).
#i and N M U - i n d u c e d ) ,
of a p p r o x i m a t e l y The p21 b a n d s
of the Y 1 3 - 2 5 9
were
antibody
701
as w e l l
the same absent (Fig
as
those
low intensity when
Ib).
control
rat
Approximately
Vol. 127, No. 2, 1985
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
12% of DMBA-induced
tumors are known to be hormone-independent
To accurately correlate p21 levels with hormone-dependency we measured p21 in biopsy specimens
from DMBA-induced
to monitor tumor growth following host ovariectomy.
(I0).
of tumors,
tumors and continued It was found that
all tumors (5 out of 5) that continued to grow after ovariectomy contained low levels of p21 comparable to those in DMBA #I and NMU-induced
tumors,
whereas every tumor (17 out of 17) that regressed following ovariectomy contained high levels of p21. of autoradiograms
Quantitation
by densitometric
showed that p21 in the hormone-dependent
tracings
DMBA-induced
tumors was present in amounts roughly 7-fold greater than p21 levels in the hormone-independent
tumors.
To rule out the possibility the hormone-independent
that the diminished labeling of p21 in
tumors was due to a higher proteolytic
in the tumors, we tested various mixtures -independent
tumor extracts;
the amount of detectable independent
of hormone-dependent
activity and
these mixtures showed an additive effect on
p21 (data not shown),
suggesting that hormone-
tumors do not have higher proteolytic activity toward p21.
Thus, elevated levels of p21 in the hormone-dependent carcinomas appear to be specfically
rat mammary
related to the hormone-dependency
of
of these tumors. Fig 2 shows a Western blotting analysis carcinomas.
of p21 in primary human breast
Strongly labeled p21 doublets were detected in extracts
6 out of 7 estrogen and progesterone
receptor-positive
In the extracts of estrogen and progesterone ever, p21 bands were either undetectable exception of one tumor (Lanes 11-16). the receptor-negative
from
tumors (Lanes 4-10).
receptor-negative
tumors, how-
or only weakly labeled, with the
Interestingly,
tumors showed a preferential
the p21 doublet in
labeling of either the
lower band (Lanes 12, 13 and 15) or the upper band (Lane 14 and 16), whereas
in the receptor-positive
tumors, all doublets
labeling of each band (Lanes 4-10).
showed an equal
A low level of p21 was found in
normal breast tissues (Lane 1-3) and benign tumors (see Fig 3).
702
The p21
Vol. 127, No. 2, 1985
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
-
-
68K
- - 43K
- - 25.7K p21 --,'--18.4K
1 2 3 4 5 6 7 8 910111213141516M I
I
I
NB
I I R+
I R-
Fig
2. Western blotting analysis of p21 in estrogen and progesterone receptorpositive and-negative human breast carcinomas and normal breast tissues. The lysates were prepared from frozen-pulverized breast tumors and normal breast tissues, and the cellular proteins were separated by NaDodSO4-PAGE , transferred to nitrocellulose sheets, immunoblotted with monoclonal antibody #Y13-259, and then exposed for autoradiography as described in the legend to Fig i. Each lane was loaded with I00 ~g protein for NaDodS04-PAGE. NB, normal breast; R+, estrogen (ER) and progesterone (PR) receptor-positlve tumors; R-, the receptornegative tumors. Positive ER = > i0 fmole/mg protein; positive PR = > i0 fmole/mg protein. Cytoplasmic estrogen and progesterone receptor levels were determined by Scatchard analysis (13) of the data obtained from the dextran-coated charcoal assay (14, 15). Cytosol protein concentrations were measured by the method of Lowry et al (16). M, prestained molecular weight markers (Bethesda Research Laboratories, Inc.)
doublet bands from human breast tumors co-migrated
in NaDodSO4-PAGE
with
the lower band of p21 from 13-3B-4 cells, and the bands were absent when control rat serum was used in place of the #Y13-259 m o n o c l o n a l
antibody
(data not shown). The levels of p21 in a series of human breast tumors were estimated quantitatively
after compiling the results from several experiments
the type shown in Fig. 2. densitometric films.
tracings
The concentrations
of the band intensities
of p21 were compared from from the autoradiographic
It can be seen in Fig. 3 that 15 of 22 (68%) primary,
progesterone
receptor-positive
breast carcinomas
fold greater than those of normal breast tissues, the r e c e p t o r - n e g a t i v e
breast carcinomas
that of normal breast tissues.
of
estrogen and
contained p21 levels ~ lOwhereas
13 of 21 (62%)
contained p21 levels only ~ 3-fold
Levels of p21 in benign tumors of fibre-
703
Vol. 127, No. 2, 1985
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
13
t 00
•= 11
•
•00 00o0
&&
03
9 _J W
>
W _J
o•
7
e~
uJ >
5
w
• •
A && && &AA&
• 0
&&&
00
< 3
~^t~
R+
CARCINOMAS
I IT'I
BENIGN TUMORS
Fig 3. Relative p21 levels in estrogen and progesterone receptor positive ~-~nd negative (R-) breast carcinomas and benign lesions of flbroadenoma (~) and fibrocystic disease ( [] ). Each symbol represents a single tumor from a different patient. The p21 levels were quantified from densitometrLc tracings of autoradiographs of the Western blotting analysis. The levels of p21 in the tumors are expressed relative to the level of p21 in normal breast tissues which is set equal to 1 arbitrary unit.
adenoma and fibrocystie disease were not elevated above the levels present in normal breast tissues, whereas practically all the breast carcinomas (41 out of 43) contained elevated p21 levels at least 2-fold that of normal breast tissues (Fig 3). About 70% of estrogen and progesterone receptor-positive human breast carcinomas are responsive to endocrine therapy, i.e., hormone-dependent
(17).
Since 68% of the receptor-positive tumors examined in this study exhibited high p21 levels, a correlation may exist between hormone-dependency of breast tumors and p21 elevation.
It is not known, however, whether the
receptor-positive tumors containing high p21 levels reported here are all responsive to endocrine therapy.
These results with human breast tumors
are, however compatible with the data obtained with experimental rat mammary tumors of well defined hormone dependency (Fig i). GTP (or GDP)-bindlng is the only known biological function common to both v-ras p21 and c-ras p21 (18).
We examined whether the p21 species
identified in the human and rat mammary tumors would exhibit guanine nucleo704
Vol. 127, No. 2, 1985
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
tide binding activity by the use of the photoaffinity ligand, 8-N3-[y-32p] GTP (19).
Tumor lysates containing either high or low levels of p21 were
incubated with 8-N3-[y-32p]GTP for various lengths of time.
The incubation
mixtures were then exposed to UV light, quenched, and subjected to NaDodSO 4PAGE.
Only in these lysates containing low levels of p21, was there a time-
dependent appearance of a GTP-labeled protein band of MW 21K, presumably p21, as well as other labeled protein bands (data not shown).
These protein bands
were present only after photolysis with UV light, indicating a covalent attachment of the labeled nucleotide to the proteins.
That the tumor lysates
containing high levels of p21 did not exhibit labeling of p21 as well as other GTP binding proteins suggested the presence of high GTPase activities in the lysates.
In fact, 8-N3-[y-32p ] GTP photo-incorporation into mixtures of
lysates containing high and low levels of p21 resulted in a complete loss of the p21 band and marked decrease of other GTP binding protein bands (data not shown).
The low photo-incorporation of 8-N3-[y-32p]GTP into the tumor lysates
containing high levels of p21 was not due to low GTP binding activity: direct binding of [a-32p]GTP or [y-32p]GT P to p21 transferred to nitrocellulose sheets was found to be proportional to the amount of p21 present in the tumors (data not shown). Recently, quantitative enhancement of p21 ras expression in human mammary carcinomas has been reported (20).
In the present study, we showed a correla-
tion between hormone-dependency of mammary carcinoma and elevation of p21.
Estrogen and progesterone-receptor positive human mammary carcinomas
contained p21 at levels as much as lO-fold higher than those found in normal mammary tissues.
The elevated p21 in the mammary carcinomas was
accompanied by a high GTPase activity.
In the light of a recent report
(21) that 'normal' p21 possesses GTPase activity, while in mutant p21 this function of GTPase is impaired, the elevated p21 in the hormonedependent mammary carcinomas may be 'normal' p21 rather than structurally altered protein.
Thus, amplified, hormone-stimulated expression of
'normal' p21 may be involved in the hormone-dependent growth of mammary
705
Vol. 127, No. 2, 1985
carcinomas.
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
This conclusion has further support from our previous finding
(6) that over-expression of c-ras gene in hormone-dependent rat mammary carcinomas ceases in the regressing tumors following hormone-withdrawal. Western blotting analysis and photoactivated binding of 8-N3-[~-32p]GTP now make possible both quantitative and qualitative assessments of p21 expression in tissues.
The precise role of ras p21 in the processes of mammary
carcinogenesis is currently under study. ACKNOWLEDGEMENTS: We thank D.R. Lowy for his generosity in providing p21 monoclonal antibody #Y13-259 and clone 13-3B-4, the Harvey murine sarcoma virus transfectant of NIH 3T3 cells; M. Barbacid, R.H. Bassin, P.M. Gullino and D.R. Lowy for valuable comments on the manuscript; H. Perno for preparation of manuscript and S. Oremland for typing the manuscript. REFERENCES i. 2. 3. 4. 5. 6. 7. 8. 9. i0. ii. 12. 13. 14. 15. 16. 17.
18. 19. 20. 21.
Dao, T.L., (1972) Ann. Rev. Med. 23, 1-18. Huggin§, C., Briziarelli, G., Sutton, H., Jr. (1959) J. Exp. Med. 109, 25-41. Gullino, P.M., Pettigrew, H.M., and Grantham, F.H. (1975) J. Natl. Cancer Inst. 54, 401-404. Jensen, E.V. and DeSombre, E.R. (1972) Ann. Rev. Biochem. 789, 203-230. Huang, F.L. and Cho-Chung, Y.S. (1982) Biochem. Biophys. Res. Commun. 107, 411-415. Huang, F.L. and Cho-Chung, Y.S. (1984) Biochem. Biophys. Res. Commun. 123, 141-147. Ellis, R.W., DeFeo, D., Maryak, J.M., Young, H.A., Shih, T.Y., Chang, E.H., Lowy, D.R. and Scolnick, E.M. (1980) J. Virol. 36, 408-420. Shih, T.Y., Weeks, M.O., Young, H.A. and Scolnick, E.M. (1979) Virology 96, 64-79. Furth, M.E., Davis, L.J., Fleurdelys, B. and Scolnick, E.M. (1982) J. Virol. 43, 294-304. Cho-Chung, Y.S., Bodwin, J.S. and Clair, T. (1978) Eur. J. Biochem. 86, 51-60. Laemmli, U.K. (1970) Nature 227, 680-685. Towbin, H., Stahelin, T. and Gordon, J. (1979) Proc. Natl. Acad. Sci. USA 76, 4350-4354. Scatchard, G. (1949) Ann. N.Y. Acad. Sci. 51, 660-672. Korenman, S.G. and Dukes, B.A. (1970) J. Clln. Endocrinol. Metab. 30, 639-645. Powell, B., Garola, R.E., Chamness, G.C. and McGuire, W.L. (1979) Cancer Res. 39, 1678-1682. Lowry, O.H., Rosebrough, N.J., Farr, A.L. and Randall, R.J. (1951) J. Biol. Chem. 193, 265-275. McGuire, W.L., Carbone, P.O., Sears, M.E. and Eschert, G.C. (1975) In: Estrogen Receptors in Human Breast Cancer, (McGuire, W.L. Eds.) Raven, New York, pp. 1-7. Shih, T.Y., Papageorge, A.G., Stokes, P.E., Weeks, M.O. and Scolnick, E.M. (1980) Nature 287, 686-691 Geahlen, R.L. and Haley, B.E. (1979) J. Biol. Chem. 254, 11982-11987. Hand, P.H., Thor, A., Wunderlich, D., Muraro, R., Caruso, A. and Schlom, J. (1984) Proc. Natl. Acad. Sci. USA 81, 5227-5231. McGrath, J.P., Capon, D.J., Goeddel, D.V. and Levinson, A.D. (1984) Nature 310, 644-649.
706