[CANCER RESEARCH 49,4298-4304, August 1. 1989]

Study of Estrogen Receptor, Progesterone Receptor, and the Estrogen-regulated Mr24,000 Protein in Patients with Carcinomas of the Endometrium and Cervix1

Daniel R. Ciocca,2 Libertad A. Puy,3 and Liliana C. Fasoli

Unidad de Histo-Patologia Endocrina, Area de Biologíade la Reproducción y Biotecnología(LAKLAC), Centro Regional de investigaciones Científicasy Tecnológicas(CRICYT), Casilla de Correo 855, Mendoza 5500, Argentina

ABSTRACT

The presence of an estrogen-regulated protein with 24,000 molecular

weight has been studied in 47 patients with endometrial carcinomas andin 29 patients with cervical carcinomas in order to correlate its presencewith that of estrogen receptors (ERs) and progesterone receptors (PgRs).In the cytosol tumor samples the M, 24,000 protein was detected by theWestern blot technique using a monoclonal antibody (CM), while thepresence of ER and PgR was studied by the one-point dextran-coated

charcoal assay. In the tumor tissue sections immunohistochemistry wasapplied to detect M, 24,000 protein, ER, and PgR; in these casesmonoclonal antireceptor antibodies (H222 and ml'KI) were used to

localize the receptor proteins. In endometrial and endocervical adenocar-

cinomas the presence of M, 24,000 protein correlated significantly withthat of ER (P < 0.05) in the cytosol samples; when the evaluation was

performed in the tumor sections, the presence of M, 24,000 proteincorrelated with that of ER (P < 0.005) and PgR (P < 0.05) as well. Thestudy also showed that almost 70% of the well-differentiated adenocar-

cinomas had ER, PgR, and M, 24,000 protein. In 25% of the endometrialadenocarcinomas examined the tumors were associated with normal,proliferativi*, and hyperplastic endometrium; in these cases the presence

of ER, PgR, and M, 24,000 protein was evaluated by immunohistochemistry in the malignant and nonmalignant endometrium.

On the other hand, there was a lack of correlation between M, 24,000protein, ER, and PgR in the squamous carcinomas of the uterine cervixand in the endometrial adenocarcinomas with squamous cells. In most ofthese cases the tumors lacked ER and PgR although 80% of themcontained the M, 24,000 protein to a variable degree. It is suggested thatM, 24,000 protein is involved in growth and differentiation (the M,24,000 protein is a heat shock protein) and that the gene coding of thisprotein is under hormonal control only in those tissues where growth anddifferentiation are strongly hormonally controlled (breast and endometrium).

INTRODUCTION

It is generally accepted that the presence of ER" in human

breast cancer is useful for selecting patients for endocrinetherapy and to predict the prognosis of the disease (1, 2).However, the interaction between the hormone and the receptoris one of the first steps in explaining the mechanism of estrogenaction, and the presence of ER alone does not assure completehormonal responsiveness of the tumor (3). There are manybiochemical and physiological events after the binding of thehormone to its receptor, and for the exploration of these eventsit may be useful to know if the tumor remains responsive to thesteroid at postreceptor level. One way to explore functional ER

Received 8/11/88; revised 2/9/89; accepted 4/17/89.The costs of publication of this article were defrayed in part by the payment

of page charges. This article must therefore be hereby marked advertisement inaccordance with 18 U.S.C. Section 1734 solely to indicate this fact.

1Supported by Grant PID 3-903604/85 from the Consejo Nacional de Inves

tigaciones Científicasy Técnicas(CONICET) of Argentina.1To whom requests for reprints should be addressed, at LARLAC. Casilla de

Correo 855, Mendoza 5500, Argentina.3 Postdoctoral Research Fellow of the Consejo Nacional de Investigaciones

Científicasy Técnicas(CONICET) of Argentina.4The abbreviations used are: ER, estrogen receptor; PgR, progesterone recep

tor; SDS-PAGE, sodium dodecyl sulfate-polyacrylamide gel electrophoresis;ABC, avidin-biotin-peroxidase complex.

is achieved by measuring in the tumor the presence of PgRsince they are generally up-regulated by estradici (4), but PgRmay also be constitutively present in human breast tumor cells(5). Another possibility is to measure the amount of PgR aftera short period of tamoxifen treatment, since this anliestrogenmay induce PgR synthesis (6). Recently, a nuclear binding assayhas been introduced to assess biologically active receptors ableto bind to nuclear acceptor sites (7). Additional markers ofhormone action at postreceptor level are some estrogen-regulated proteins and a tamoxifen-induced protein (8, 9). One ofthe estrogen-regulated proteins under study is that found in thecytoplasm of the human breast cancer cell line MCF-7, characterized by its molecular weight of 24,000 (10). This proteinhas also been termed 28K in previous publications since thishas been the reported molecular weight after its purification bymonoclonal antibody affinity chromatography (11). The function of the M, 24,000 protein is unknown but it has beenidentified as a heat shock protein (12). Immunohistochemicallocalization of the M, 24,000 protein using monoclonal antibodies showed that it is mainly present in the female reproductive tract, and in breast tumor cell lines with ER and PgR (13).Elevated presence of M, 24,000 protein was found to correlatewell with presence of ER and PgR in human breast tumorbiopsy samples (14).

We have reported that in the human normal endometriumM, 24,000 estrogen-regulated protein undergoes quantitativedifferences in the various phases of the normal menstrual cycle(IS). In women with abnormal endometrium, this protein appeared with maximum immunostaining in active forms of persistent proliferative endometria and in cystic glandular hyper-plasia and decreased in adenomatous and atypical hyperplasia(16), and in malignant endometrium its presence was correlatedwith the degree of tumor differentiation (16, 17). In addition,the M, 24,000 protein has also been found in normal andabnormal samples from uterine cervix ( 18,19). In the ectocervixthe protein was detected in the parabasal and intermediate celllayers, with no prominent changes during the normal menstrualcycle, while in the ectoendocervix it was highly present duringthe process of indirect squamous metaplasia. In order to continue the analysis of the ,V/,24,000 estrogen-regulated protein,we have now performed a study correlating the presence of theprotein with the presence of ER and PgR in patients withadenocarcinomas and squamous carcinomas of the uterus.

MATERIALS AND METHODS

Patients. The tumors were obtained from patients seen at Mendoza's

Hospitals by the Cooperative Group for the Study of Endocrine Dependent Tumors during the period of February 1986 to May 1988.These patients had operable endometrial or cervical carcinomas, andthe diagnosis was done by preoperative endometrial curettage or cervical punch biopsy. Of the 86 cases entering into the study, 10 cases wereexcluded due to inadequate tissue for complete analysis or because noresidual tumor was seen on histological examination; the 76 remainingcases comprised 47 primary endometrial carcinomas and 29 primarycervical carcinomas. Pre- and postmenopausal patients were included

4298

Research. on October 25, 2020. © 1989 American Association for Cancercancerres.aacrjournals.org Downloaded from

RECEPTORS AND M, 24,000 PROTEIN IN UTERINE TUMORS

in the study; they had no clinical metastasis, had received no radiotherapy before surgery, and only three patients with endometrial adenocar-cinoma received endocrine therapy (medroxyprogesterone acetate) before surgery.

Tissue Handling. The tumor samples were obtained immediatelyafter surgery, trying to avoid as much as possible the incorporation ofnecrotic and hémorragieareas and of myometrium. In order to obtaina homogeneous representation of the tumor samples for the study, thetumor tissues were divided into small pieces which were mixed beforeseparation into three fractions: (a) for ER and PgR analysis by thedextran-coated charcoal assay and for immunoblot study of the M,24,000 protein; (/>)for ER and PgR analysis by an immunohistochem-ical procedure; and (c) for histológica!examination and immunohisto-chemical study of the M, 24,000 protein. Fractions a and b wereimmediately frozen and stored at —¿�70"Cuntil assays, while the tissue

of the fraction c for histológica! examination was fixed in a neutral-buffered formalin and processed routinely for paraffin embedding.

Biochemical Studies. The frozen tissues were pulverized, thawed inthree volumes of homogenizing buffer (5 imi molybdic acid/1.5 HIMEDTA/500 mM dithiothreitol/10 HIMTris-HCl, pH 7.4), and homogenized in an Ultra-Turrax homogenizer at 4°C.Cytosol was prepared

from the homogenate by centrifugation at 140,000 x g for 50 min at4°C.The protein concentration of the supernatant was determined bythe method of Lowry (20). The cytosols were diluted to 1.5-2 mgprotein/ml, and the content of ER and PgR was analyzed using a one-point dextran-coated charcoal assay as described elsewhere (13). Ali-quots of 200 fi\ of cytosol were incubated overnight at 4°Cwith 0.25pmol of 3H-exalabeled estradici or with 1 pmol of 3H-R5020 (NewEngland Nuclear, Boston, MA). A 100-fold excess of diet hyIstiIbest rol(Sigma Chemical Co., MO) or of R5020 (NEN) was used in the coldassay tubes of ER and PgR determinations, respectively. Endogenousunbound steroid was removed by 0.25% (w/v) Norit A/0.0025% dextranin 10 mM Tris, pH 8.0. Radioactivity was determined in a KontrónMR300 liquid scintillation counter using 5 ml solution of 0.4% (w/v)Omnifluor (NEN) in toluene.

SDS-PAGE analysis was performed on the cytosol samples by themethod of Laemmli (21). The samples were dissolved into 1.9% SDSsolution containing 9.5% glycerol/4.76% mercaptoethanol/10 HIMTris, pH 6.8. A 100 Mgof sample protein was loaded on the stackinggel prepared with 3.2% of acrylamide and separated in the resolvinggel with 13.4% of acrylamide, under conditions of constant current andrunning at 35 mA/gel (slab gels of 2.7 x 140 mm). In each gel, onelane was loaded with molecular weight standards: ribonuclease A (M,13,700), chymotrypsinogen A (M, 25,000), ovalbumin (M, 43,000), andbovine serum albumin (M, 67,000) (Pharmacia Fine Chemicals, Pisca-taway, NJ). The proteins were then transferred from SDS-PAGE to0.22-pm nitrocellulose membranes (Schleicher & Schull, Keene, NH)using electroblotting techniques (22). The electroblot buffer contained150 mM glycine/20 mM Tris in methanol, pH 7.5-8.0. Followingtransfer, the standard-containing nitrocellulose lanes were separatedand stained in 0.1% (w/v) Amido black/45% methanol, and 10% aceticacid for 10 min, and destaining was effected in 45% methanol and 10%acetic acid. The cytosol-containing nitrocellulose lanes were immuno-stained with the Cll monoclonal antibody against the M, 24,000estrogen-regulated protein and treated with the ABC system to detectthe proteins immunoenzymatically as reported previously (23). Thedilution and time of incubation with the antiserum were the same asthose for immunohistochemistry.

Immunohistochemical Studies. The presence of ER and PgR on thetumors was detected by monoclonal antireceptor antibodies utilizingfrozen tissue sections. The H222 antibody (rat monoclonal origin)against ER was kindly provided by Dr. E. De Sombre (University ofChicago), while the antibody mPRI against PgR (mouse monoclonalorigin) was purchased from Transbio SARL (Paris, France). The specificity and use of these antibodies in immunohistochemistry has beenpreviously reported (24-27). Briefly, frozen 6-i/m thick tissue sectionswere thaw-mounted onto 0.1 % poly-L-lysine (w/v) (Sigma)-coated slidesand immediately fixed in a buffered picric acid-formaldehyde solution(28) for 5 min at room temperature. After rinsing in phosphate buffer,the sections were incubated with 3% normal goat or sheep serum for

30 min. Then, the sections were incubated overnight at 4"C with H222

(10 Mg/ml), mPRI (10 Mg/ml) or with normal rat and mouse serum ina moist incubation chamber. Biotinylated anti-rat or anti-mouse antibodies (Amersham International, UK) diluted 1:50 were used as secondantibodies (incubated for 45 min at room temperature), while ABCcomplex (Vector Laboratories, Burlingame, CA) diluted 1:100 was usedas the enzyme reagent (incubated for 45 min). Hydrogen-peroxide anddiaminobenzidine tetrahydrochloride (Sigma) prepared as reported previously (29) was used for 15 min to give the brown color; after washingthe sections were dehydrated in serial alcohols to xylene and mountedwith Permount without counterstaining.

The paraffin-embedded sections (5-6 urn thickness) were mountedon gelatinized slides and stained with hematoxylin & eosin or immu-nostained. Briefly, the histológica!criteria used to diagnose endometrialand endocervical carcinomas were the presence of glandular structuresclosely packed together lined by a pseudosi ratified or stratified epithelium, with frequent papillary projections. The epithelium was composedby large nuclei, at times pleomorphic, with prominent nucleoli, andsparse cytoplasm. Mitosis were frequently seen (30-32). Well-differentiated tumors had a glandular or papillary pattern, moderately differentiated tumors had a glandular or papillary pattern mixed withsolid cores, and poorly differentiated tumors presented solid sheetswith absence of glandular or papillary formations. The term adenoa-canthomas was used for those adenocarcinomas with histologicalllybenign-appearing squamous elements, while the term mixed-adeno-squamous carcinoma was used for those adenocarcinomas with histo-logically malignant foci of squamous cells. The M, 24,000 proteinimmunoreactivities in endometrial carcinomas according to the histo-logical, mitotic, nuclear, and nucleolar grades have been presentedpreviously (17). On the other hand, invasive squamous cell carcinomasof the cervix comprised large cell keratinizing carcinomas, large cellnonkeratinizing carcinomas, and small cell nonkeratinizing carcinomas. Discrepancies in some interpretations were resolved by consensus.

The immunostaining of the paraffin sections was done as reportedelsewhere (29) using the ABC system (Vector). The M, 24,000 estrogen-regulated protein was detected with the Cll mouse monoclonal antibody (16) kindly provided by Dr. W. L. McGuire (Texas). In sometumors keratin was detected by a rabbit antibody raised against purifiedhuman epidermal keratin (Polyscience Inc., Warrington, PA). Theantibody against the M, 24,000 protein was used at a concentration of5 Mg/ml, while the antibody against keratin was used at 1:100 dilution;and the samples were incubated for 24 h in humidity chambers at 4°C.

In the negative control slides the primary antibody was omitted andsometimes we also used the primary antibody preabsorbed. After immunostaining, one of the slides from each sample was counterstainedwith hematoxylin, and all of them were mounted with Permount.

The tissue samples were observed and photographed with a ZeissIM35 microscope, with the use of differential interference contrastoptics whenever required. A quantitative analysis of the immunon-active

cells was carried out by two observers who counted the percentage ofreacting nuclei (for ER and PgR immunostained cells) and the percentage of reacting cells containing nuclei (for M, 24,000 protein immu-noreactive cells). Since in many of the tumors the distribution of theimmunostaining was heterogeneous we used a systematic random sampling method similar to that employed in the selection of tissue blocksfor electron microscopy (33). In this way sample fields representativeof the different areas of the tumor were examined under the x40objective.

Test of incidence in contingency tables, two-way x% was applied toevaluate the biochemical data of ER versus M, 24,000 (Tables 1 and 3)and PgR versus M, 24,000 protein (Tables 2 and 3); and for theimmunohistochemical data of ER verm? M, 24,000 (Table 1) and PgRversus M, 24,000 protein (Table 2). The two-way x2 test was also used

to evaluate the data on the presence of ER, PgR, and M, 24,000 proteinin carcinomas with squamous cells (Table 5). A regression analysis wasperformed when the percentage of ER-positive cells was compared withthe percentage of M, 24,000-positive cells in endometrial adenocarcinomas (Fig. 2). In this case the straight line is the result of the equation:log (%ERPICA + 1) = log(a) + /3(%ERICA); where log(a) = 0.52 andß= 0.04.

4299

Research. on October 25, 2020. © 1989 American Association for Cancercancerres.aacrjournals.org Downloaded from

RECEPTORS AND M, 24,000 PROTEIN IN UTERINE TUMORS

RESULTS

Adenocarcinomas. The correlation between the presence ofthe MT 24,000 estrogen-regulated protein and ER in adenocar-cinomas of the endornetrium and endocervix is shown in Table1. The biochemical studies of the cytosol tumor samples revealed that M, 24,000 protein and ER are not independent (P< 0.05). When these two parameters were studied by immuno-histochemistry, a higher significant positive correlation (P <

Table Comparison between expression ofM, 24,000 protein and ER in uterineadenocarcinomas

ERrPositive

NegativeM,

24.000 proteinimmunoblol"Positive

Negative21(78)'

2(7)

2 (7) 2 (7)ER"PositiveNegativeM,

24,000 proteinimmunocyto-

chemistry*Positive

Negative30(68)

6(14)0 8(18)

" Positive or negative according to the presence or not of the specific band.'Positive, when M, 24,000 protein appeared in >10% of the tumor cell

population.c Positive, ER >10 fmol/mg cytosol protein.d Positive, when ER appeared in >10% of the tumor cell population.' Numbers in parentheses, percentage.

0.005) was found. This was not unexpected since in the samplestaken for the biochemical studies there were neoplastic cellsmixed with stromal and sometimes with myometrial cells, cellsthat also contain ER but that have variable amount of M,24,000 protein; therefore, the discrimination between the different cell populations was possible in the tissue sections evaluated by immunohistochemistry. Fig. 1, A-D, shows sectionsfrom a patient with endometrial adenocarcinoma reacting withthe monoclonal antibodies against M, 24,000 protein, ER, andPgR. On the other hand, Fig. 1, E-H, shows the presence ofM, 24,000 protein, ER, and PgR in a patient with an endocer-vical adenocarcinoma. In addition, the immunohistochemicalstudy allowed us to evaluate the heterogeneity of the immuno-staining. Fig. 2 shows the correlation obtained by counting thenumber of positive cells immunostained for M, 24,000 proteinand for ER.

A significant positive correlation between M, 24,000 proteinand PgR (P < 0.05) was found in the adenocarcinomas studiedby immunohistochemistry (Table 2), but not when these parameters were studied biochemically in the cytosol tumor samples.In the biochemical assay we decided to use a cutoff point of190 fmol/mg cytosol protein for PgR because this has been thereported value for those endometrial adenocarcinomas that are

•¿�Vi' ~ *^^^H.r^P

Fig. 1. Immunohistochemical detection ofM, 24,000 protein, ER, and PgR in patientswith endometrial (A-D) and endocervical (E-11}carcinoma. A, in this hematoxylin & cosilistained section the characteristic of the adenocarcinoma examined may be observed (x120); lì,tissue obtained from the same patientshown in A, displaying variable M, 24,000protein immunostaining in the tumor cells (X120); C, tissue section from the same patientshown in A and B, displaying ER in most ofthe nuclei of the tumor cells (x 150); D, sectionfrom the same patient shown in A-C, showingimmunoreactivity for PgR (x 150); E and F,serial sections from a patient with endocervicalcarcinoma, note the lack of cytoplasmic immunostaining for M, 24,000 protein when thetissue was incubated with a negative controlantibody (/,. with hematoxylin counterstain-ing); compare with the strong cytoplasmic immunostaining when the tissue was incubatedwith the antibody against M, 24,000 protein(F, with hematoxylin counterstaining). N, normal gland; '/', tumoral gland (x 120): G, tissueobtained from the same patient shown in /.'and /•'.ER immunostaining can be seen in the

nuclei of the tumor cells (x 130); H, tissuesection from the same patient shown in E-G,low immunostaining intensity for PgR can beseen in the nuclei of the tumor cells (x 130).

H

4300

Research. on October 25, 2020. © 1989 American Association for Cancercancerres.aacrjournals.org Downloaded from

RECEPTORS AND M, 24.000 PROTEIN IN UTERINE TUMORS

-ÃŽMOO

20

10

£5Q-tt

20 40 60 80 100

% ERICA Positive CellsFig. 2. Correlation between percentage of M, 24,000 estrogen-regulated pro

tein-positive cells in immunocytochemistry (ERPICA) and percentage of ER-positive cells in immunocytochemistry (ERICA) in 43 patients with endometrialadenocarcinomas. For statistical analysis see "Materials and Methods" (n = 44,a = 1.68, 0 = 0.04, R2 = 0.73).

Table 2 Comparison between expression ofM, 24,000 protein and PgR in uterineadenocarcinomas

M, 24,000 proteinimmunoblot"PgRf

PositiveNegativePositive12(46)'11 (42)Negative1(4)2(8)PgR*

PositiveNegativeM,

24,000 proteinimmunocyto

chemistry*Positive24

(55)6(13)Negative2(5)12(27)

°Positive or negative according to the presence or not of the specific band.* Positive, when M, 24,000 protein appeared in >10% of the tumor cell

population.* Positive, PgR >190 fmol/mg cytosol protein.**Positive, when PgR appeared in >10% of the tumor cell population.' Numbers in parentheses, percentage.

Table 3 Presence ofM, 24,000 protein by immunoblot (Western blot) accordingto the amount of ER and PgR in uterine adenocarcinomas

ER*<1010-100>100PgR"<1010-190>190M,n°43130124,000

proteinnegativePercentage12.59.33.110.73.5M,n01683101124,000proteinpositivePercentage50.025.010.735.739.2

" Number of patients.* Expressed in fmol/mg of cytosol protein.

likely to be under the influence of estrogen (34). Even higherthreshold values of PgR have been reported (35). As can beseen in Table 3, the presence of the M, 24,000 protein detectedin the cytosol samples correlates significantly with the amountof ER (P < 0.005) and with the amount of PgR (P < 0.05).

On the other hand, when the degree of tumor differentiationwas evaluated, we found that 69% of the well differentiatedadenocarcinomas were positive for M, 24,000 protein, for ER,and for PgR, while most of the moderately and poorly differentiated tumors were positive only for ER (40%), or lacked ER,PgR, and M, 24,000 protein (20%) (Fig. 3). It is important tomention here that we excluded from the group of adenocarcinomas those presenting squamous elements (36), because in thesix cases examined there was M, 24,000 protein immunodetection while ER was detected in only two cases and PgR in one(Fig. 4, A and B). Although not shown here, the presence of thesquamous cells was corroborated by immunostaining with theantikeratin antibody.

In 25% of the endometrial adenocarcinomas examined we

80-70-0)

60-UC

50-0)2

40-U£

30-•S20'

gç10§ER

+24K+PR

+niER

+24K+PR

-nER24KPRli

Mod.andni+•

ER-24K-+

PR -poorly

cER

+24K-PR

-

Fig. 3. Relation between percentage of incidence of ER. PgR. and M, 24,000protein in well differentiated versus moderately and poorly differentiated endometrial adenocarcinomas.

found varying amounts of normal, proliferative, and hyperplas-tic endometrium. In these cases we were able to compare, byimmunohistochemistry, the expression of ER, M, 24,000 protein, and PgR in the malignant endometrium and in the associated nonmalignant endometrium (Table 4). In these postmenopausal patients normal endometrium was rich in ER, whilethe content of PgR was moderate, and M, 24,000 proteincontent was low to absent. Proliferative endometrium showedfrom 50% to more than 90% of the epithelial cells with moderate to strong immunostaining intensity for ER; M, 24,000protein was also high in these samples, while PgR was negativein one case and high in the others. Cystic and adenomatoushyperplasias showed variable degree of ER, M, 24,000 protein,and PgR; some of them were rich in the three parametersstudied while some others showed low content of ER, PgR, andalmost absence of M, 24,000 protein. It is interesting to notethe patient (Case 4, Table 4) with an adenocarcinoma negativefor ER, M, 24,000 protein, and PgR but whose associatedproliferative endometrium and cystic hyperplasia showed considerable amount of ER, M, 24,000 protein, and PgR.

Squamous Carcinomas and Adenocarcinomas with SquamousCells. The presence of ER, PgR, and M, 24,000 protein phe-notype in the squamous carcinomas of the uterine cervix andin the adenoacanthomas and mixed adenosquamous carcinomas of the corpus uteri is shown in Table 5. The statisticalstudy revealed a lack of correlation among the three parametersevaluated; in most of the cases these tumors lacked ER andPgR. However, 82% of these receptor-negative tumors expressed the M, 24,000 protein in variable degree. Fig. 4, C andD, shows a squamous carcinoma of the uterine cervix withimmunoreactive M, 24,000 protein but lacking ER. The identityof the protein was corroborated in the immunoblot studies,where it was compared with the M, 24,000 protein present inendometrial adenocarcinomas and from a positive control patient with a breast carcinoma (Fig. 5).

DISCUSSION

The results obtained show that in endometrial and endocer-vical adenocarcinomas the presence of M, 24,000 protein iscorrelated with that of ER and PgR. We have used two differentimmunodetection systems for M, 24,000 protein, Western blotand immunohistochemistry, and when the evaluation was donein the tumor cytosols the presence of M, 24,000 protein correlated significantly with that of ER. However, when the evaluation was performed in the tumor sections by immunohistochemistry the presence of M, 24,000 protein correlated significantly with that of ER and PgR as well. In a previous study

4301

Research. on October 25, 2020. © 1989 American Association for Cancercancerres.aacrjournals.org Downloaded from

RECEPTORS AND M, 24,000 PROTEIN IN UTERINE TUMORS

Fig. 4. Immunohistochemical study of ERand Mr 24,000 protein in a patient with anendometrial carcinoma (mixed adenosqua-mous carcinoma) (A, B) and in a patient witha squamous carcinoma of the ectocervix (C,D). A, variable M, 24,000 protein immuno-staining intensity can be seen in this area withsquamous tumor cells. Hematoxylin counter-staining (x 175); B, tissue obtained from thesame patient shown in .•(,note the lack of ERin the squamous tumor cells (T) (x 175); C,cytoplasmic M, 24,000 protein immunostain-ing of variable intensity can be seen in thetumor cells (x 175); D, tissue obtained fromthe same patient shown in C, no ER-positivecells are seen in the tumor area ('/'). the dotted

line is the limit with the stroma (S), notepositive staining of cervical stroma (x 175).

D

Table 4 Immunohistochemical evaluation of ER, M, 24,000 protein, and PgR in nonmalignant endometrium associated to endometrial adenocarcinomas

AdenocarcinomaCase

Age1

602

563

654

505

736697488639

541060115012

46Histology"WWWWWWWMWWWWER*++,80++,83+++,100Neg+.78+,35++,86++,70++,45Neg+++,60++,81M,

24,000protein*++,20++,75++,62Neg++,35+.24+,64++,68+,60Neg++,68++,28PgR*+,15+,55++,98Neg+,53+,11+,40++,73++,48Neg++,30++,53Associated

nonmalignantendometriumHistology"PPECHPPECHNECHPPECHPPEAHNEAHCHNECHNECHNEER*++,50++,48+++,95+++,98++,93++,95+++,92+++,97++,80+,10+,70++,93++,82+++,90+,32++,60+,45++,82M,

24,000protein*+,30+,40++,80++,87++,70++,61++,80++,83+,68+,15NegNeg+,30Neg+,11++,58++,37+,68PgR*+,50+,47NegcNeg++,95++,70+,52++,57+,79+,8+,32++,85++,40+,28+,19+,63++,42++,50

" W, well-differentiated adenocarcinoma; M, moderately differentiated adenocarcinoma; NE, normal endometrium; PPE, persistent proliferative endometrium;

CH, cystic hyperplasia; AH, adenomatous hyperplasia.* Immunostaining intensity: +, light; ++, moderate; +++, strong. The numbers are percentages of immunostained cells.' Neg, negative (no immunostaining).

performed in human breast tumors, it was also found thatelevated expression of A/r 24,000 protein correlated well withpresence of both ER and PgR, with the correlation comingprimarily through ER (14). We believe that the immunohisto-chemical study is better than the biochemical study becausewith the former we were able to analyze the presence anddistribution of ER, PgR, and M, 24,000 protein exclusively inthe tumor cells, avoiding stroma, myometrium, and nontu-morous areas which are sometimes present in the homogenate.This is consistent with the heterogeneity of ER and PgR inhuman endometrial adenocarcinomas reported previously (25,37).

The present study also shows that almost 70% of the well-

differentiated adenocarcinomas had ER, PgR, and M, 24,000

protein, and most authors agree that well-differentiated endometrial adenocarcinomas frequently have ER and PgR (38-40).However, we cannot predict the hormone dependence by morphological examination, it is necessary to study the receptorstatus of these tumors (40, 41). Today the immunohistochem-ical study of ER and PgR is increasingly more accepted becauseof its advantages (lower cost, exploration of tumor heterogeneity). Nevertheless, at present the monoclonal antireceptorantibodies are recommended for use in frozen sections, and ourstudy shows that it is possible as an alternative and/or complementary method to explore the presence of the M, 24,000protein in tumors fixed and processed for routine paraffinsections. We are now following these patients to analyze if thepresence of A/r 24,000 protein is useful in predicting their

4302

Research. on October 25, 2020. © 1989 American Association for Cancercancerres.aacrjournals.org Downloaded from

RECEPTORS AND M, 24,000 PROTEIN IN UTERINE TUMORS

Table 5 Expression of ER, PgR, and M, 24,000 protein pnenotype in carcinomaswith squamous celts of the uterine corpus and cervix

M, 24,000Case ER" PgR* proteincHistology1

+ —¿� +Adenoacanthoma2- -+3—¿� —¿� + Mixed adenosquamouscarcinoma4- -+5- -+6+ ++7—¿� —¿� +Squamou8—¿� —¿�+9++10++1112

+13- NDrf+14- -+15- -+16- -+1718

- -+1920

- -+21- -+22+ -+23- -+24-H ++25- -+2627

- -+28- -+29+ND30- ND +s

carcinoma

" Positive, with >10 fmol/mg protein and/or -10', of cells immunostained.* Positive, with >100 fmol/mg protein and/or >IO% of cells immunostained.' Positive, with detectable band in immunoblot and >10% of cells immuno

stained.d ND, not done.

67 K>»

13.7 K>-

Fig. 5. Immunoenzymatic study of M, 24,000 protein by Western blot analysisin cytosol samples. /, patient with endometrial adenocarcinoma; 2, positivecontrol from a patient with breast carcinoma; 3, patient with endometrial adenocarcinoma lacking M, 24,000 protein; 4, patient with endocervical adenocarcinoma; and 5, patient with squamous carcinoma of the ectocervix. Standardmolecular weight markers are on the left side. Note that the specific band appearedat M, 25,000 using chymotrypsinogen A as standard, which may explain theslight difference with the M, 24,000 reported previously (using another standard).

response to endocrine therapy and the clinical outcome.Although we are not presenting a high number of endometrial

cancer patients with associated nonmalignant endometrium,our study allows us to draw some observations. In these post-menopausal patients, the normal endometrium was usually richin ER, while the expression of PgR and M, 24,000 protein wasvariable. If estrogens in circulation are low, it is reasonable tofind a moderate to low amount of PgR and M, 24,000 protein,because they are regulated by estradiol. In a previous study,

normal postmenopausal endometria showed strong immuno-staining for ER in epithelial and straniai cells (42). On theother hand, proliferative endometrium showed a moderate torich ER, PgR, and M, 24,000 protein content (except in onecase without PgR). Cystic and adenomatous hyperplasia showeda variable expression of receptors and M, 24,000 protein, sometimes they were abundant and sometimes they were rather poor.However, those cystic formations with numerous and stronglystained ER usually expressed numerous M, 24,000 protein andPgR immunostained cells. Finally, our results suggest that someendometrial tumors lacking ER, PgR, and M, 24,000 proteinmay be associated with proliferative endometrium and hyper-plastic endometrium containing ER, PgR, and M, 24,000 protein.

On the other hand, the present results show a lack of correlation between M, 24,000, ER, and PgR in squamous carcinomas of the uterus. We know that the normal squamous epithelium of the cervix has ER, PgR, and M, 24,000 protein (18,43,44), and our results show that M, 24,000 protein may beexpressed in squamous cervical carcinomas without requirement of ER and PgR. At this point it is important to rememberthat in transformed human cervical tissue, sex steroid receptorsmay show considerable biochemical alterations (45); that thepresence of ER and PgR in cervical carcinomas is being evaluated as prognostic parameters (46); and that patients withcervical cancer are generally refractory to hormonal therapy(47). In addition, M, 24,000 protein has also been found innormal skin biopsy samples from males and females, in normalmerocrine sweat glands, and in skin and bladder carcinomas(13). Therefore, it is reasonable to suggest that M, 24,000protein is under hormonal control in certain tissues and undercertain conditions, but that in another tissue and/or conditionsits expression is independent of hormonal regulation. For instance, in the breast M, 24,000 protein has been described asan estrogen-regulated protein (10), in the endometrium itspresence is modulated by both estrogen and progesterone (13,15), while in neoplastic squamous epithelia from the cervix itseems that M, 24,000 protein is not modulated by a mechanismdependent on estrogen-ER and/or progesterone-PgR interaction. We believe that an explanation for this may come fromthe fact that M, 24,000 protein has been identified as a heatshock protein (12) and that in eukaryotes these proteins seemto be involved in growth and differentiation processes (48, 49).If M, 24,000 protein is effectively involved in growth anddifferentiation, then it is not surprising that the gene coding ofthis protein is under hormonal control only in those tissueswhere growth and differentiation are strongly hormonally controlled (breast and endometrium). We can compare this situation with another model involved in growth processes, the c-myc oncogene whose expression is stimulated in fibroblasts byplatelet-derived growth factor, in T-lymphocytes by concana-valin A, while in ER-positive human breast cells by estrogenadministration (50, 51). More evidence that M, 24,000 proteinis involved in differentiation processes was found in the presentstudy, since the protein was always found in squamous cells inpatients with adenoacanthomas and with mixed adenosqua-mous carcinomas. This is consistent with the finding of A/,24,000 protein in patients with cervical intraepithelial neoplasms rising from indirect squamous metaplasia.5

REFERENCES1. McGuire, W. L., Carbone, P. P., and Vollmer, E. P. (eds.) Estrogen Receptors

in Human Breast Cancers. New York: Raven Press, 1975.

' Personal communication.

4303

Research. on October 25, 2020. © 1989 American Association for Cancercancerres.aacrjournals.org Downloaded from

RECEPTORS AND M, 24,000 PROTEIN IN UTERINE TUMORS

2. Allegra, J. C, Lippman, M. E., Thompson, E. B., Simon, R., Bablock, A., 27.Green, L., Hoff, K. K., Do, H. M. T., latken, S. C., and Warren, R. Estrogenreceptor status: an important variable in predicting response to endocrinetherapy in metastatic breast cancer. Eur. J. Cancer, 16: 323-331, 1980.

3. Jensen, E. V., and De Sombre, E. R. The diagnostic implications of steroid 28.binding in malignant tissues. Adv. Clin. Chem., 19: 57-89, 1977.

4. McGuire, W. L., and Clark, G. M. The prognostic role of progesteronereceptors in human breast cancer. Semin. Oncol., 10: 2-6, 1983. 29.

5. Wei, L. L., Krett, N. L., Francis, M. D., Gordon, D. F., Wood, W. M.,O'Malley, B. W., and Horwitz, K. B. Multiple human progesterone receptor

messenger ribonucleic acids and their autoregulation by progestin agonistsand antagonists in breast cancer cells. Mol. Endocrino!., 2: 62-72, 1988. 30.

6. Howell, A., Harland, R. N. L., Barnes, D. M., Baildam, A. D., Wilkinson,M. J. S., Hayward, E., Swindell, R., and Seilwood, R. A. Endocrine therapyfor advanced carcinoma of the breast: relationship between the effect of 31.tamoxifen upon concentrations of progesterone receptor and subsequentresponse to treatment. Cancer Res., 47:300-304, 1987. 32.

7. Spelsberg, T. C., Graham II, M. L., Berg, N. J., Umehara, T., Riehl, E.,Coulam, C. B., and Ingle, J. N. A nuclear binding assay to assess the biologicalactivity of steroid receptors in isolated animal and human tissues. Endocri- 33.nology, 121: 631-644, 1987.

8. Ciocca, D. R. Tumores hormonodependientes: conceptos actuales. Medicina 34.(Bs.As.), 4S: 87-94, 1988.

9. Bronzer!, D., Silverman, S., and Lippman, M. Estrogen inhibition of a M, 35.39,000 glycoprotein secreted by human breast cancer cells. Cancer Res., 47:1234-1238, 1987.

10. Edwards, D. P., Adams, D. J., and McGuire, W. L. Estradici stimulates 36.synthesis of a major intracellular protein in a human breast cancer cell line(MCF-7). Breast Cancer Res. Treat., 1: 209-223, 1981.

11. Adams, D. J., Hajj, H., Bitar, K. G., Edwards, D. P., and McGuire, W. L. 37.Purification of an estrogen-regulated breast cancer protein by monoclonalantibody affinity chromatography. Endocrinology, 113:415-417, 1983.

12. Fuqua, S. A. W., Moretti-Rojas, I., Blum, M., and McGuire, W. L. The 24K 38.estrogen-regulated protein is a heat shock protein. Breast Cancer Res. Treat.,/0.-A150, 1987.

13. Ciocca, D. R., Adams, D. J., Edwards, D. P., Bjercke, R. J., and McGuire,W. L. Distribution of an estrogen-induced protein with a molecular weight 39.of 24,000 in normal and malignant human tissues and cells. Cancer Res., 43:1204-1210, 1983.

14. Adams, D. J., and McGuire, W. L. Quantitative enzyme-linked immunosor-bent assay for the estrogen-regulated M, 24,000 protein in human breasttumors: correlation with estrogen and progesterone receptors. Cancer Res., 40.45: 2445-2449, 1985.

15. Ciocca, D. R., Asch, R. H., Adams, D. J., and McGuire, W. L. Evidence formodulation of a 24K protein in human endometrium during the menstrualcycle. J. Clin. Endocrino!. Metab., 57:496-499, 1983. 41.

16. Ciocca, D. R., Puy, L. A., Edwards, D. P., Adams, D. J., and McGuire, W.L. The presence of an estrogen-regulated protein detected by monoclonalantibody in abnormal human endometrium. J. Clin. Endocrino!. Metab., 60:137-143, 1985. 42.

17. McGuire, W. L., Dressier, L. G., Sledge, G. W., Ramzy, I., and Ciocca, D.R. An estrogen-regulated protein in normal and malignant endometrium. J.Steroid Biochem., 24: 155-159, 1986.

18. Ciocca, D. R., Puy, L. A., and Lo Castro, G. Localization of an estrogen- 43.responsive protein in the human cervix during menstrual cycle, pregnancy,and menopause and in abnormal cervical epithelia without atypia. Am. J.Obstet. Gynecol., 155: 1090-1096, 1986.

19. Dressier, L. G.. Ramzy, I., Sledge, G. W., and McGuire, W. L. A new marker 44.of maturation in the cervix: the estrogen-regulated 24K protein. Obstet.Gyneco!., 68: 825-831, 1986.

20. Lowry, O. H., Rosebrough, N. J., Fair, A. L., and Randall, R. J. Protein 45.measurement with the Folin phenol reagent. J. Biol. Chem., 193: 265-275,1951.

21. Laemmly, U. K. Cleavage of structural proteins during the assembly of the 46.head of bacteriophage T4. Nature (Lond.), 227: 680-685, 1970.

22. Towbin, H., Staehelin, T., and Gordon, J. Electrophoretic transfer of proteinfrom polyacrylamide gels to nitrocellulose sheets: procedure and some applications. Proc. Nati. Acad. Sci. USA, 76:4350-4354, 1979. 47.

23. Ciocca, D. R.. Winters, C. A., and Dufau, M. L. Expression of an estrogen-regulated protein in rat testis Leydig cells. J. Steroid Biochem., 24:119-229, 48.1986.

24. Press, M. F., and Greene, G. L. An immunocytochemical method for demonstrating estrogen receptor in human uterus using monoclonal antibodies 49.to human estrophilin. Lab. Invest., 50: 480-486, 1984.

25. Budwit-Novotny, D. A., McCarty, K. S., Cox, E. B., Soper, J. T., Mutch, D.G., Creasman, W. T., Flowers, J. L., and McCarty, K. S., Jr. Immunohisto- 50.chemical analysis of estrogen receptor in endometrial adenocarcinoma usinga monoclonal antibody. Cancer Res., 46: 5419-5425, 1986.

26. De Sombre, E. R., Thorpe, S. M., Rose, C., Blough, R. R., Andersen, K. W.,Rasmussen, B. B., and King, W. J. Prognostic usefulness of estrogen receptor 51.immunocytochemical assays for human breast cancer. Cancer Res., 46(Suppl.): 4256s-4264s, 1986.

Perrot-Applanat, M., Groyer-Picard, M. T., Lorenzo, F., Jolivet, A., Hai,M. T. V., Pallud, C., Spyratos, F., and Milgrom, E. Immunocytochemicalstudy with monoclonal antibodies to progesterone receptor in human breasttumors. Cancer Res., 47: 2652-2661, 1987.Freeman, A. E., Vanderpool, E. A., Black, P. H., Turner, H. C., and Huebner,R. J. Fixation of ejaculated spermatozoa for electron microscopy. Nature(Lond.), 216: 173-174, 1967.Ciocca, D. R., and Bjercke, R. J. Immunohistochemical techniques usingmonoclonal antibodies. ///: J. J. Langone and H. Van Vunakis (eds.), Methods in Enzymology, Vol. 121, pp. 652-679. New York: Academic Press,Inc., 1986.Marquez, M. Carcinoma invasor del cuello uterino. In: J. González-Merlo(ed.), El Cuello Uterino. AnatomíaPatológica, Citologia y Colposcopia, pp.223-257. España:Salvai, 1974.Ashley, D. J. B. Evans' Histológica! Appearances of Tumours, pp. 705-737.

Edinburgh: Longman Group Limited, 1978.Kurman, R. J., and Norris, H. J. Endometrial neoplasia, Inpcrplasia andcarcinoma. In: A. Blaustein (ed.). Pathology of the Female Genital Tract,Ed. 2, pp. 311-351. New York: Springer-Verlag, 1982.Weibel, E. R. Stereological principles for morphometry in electron microscopic cytology. Int. Rev. Cytol., 26: 235-302, 1969.Deligdisch, L., and Holinka, C. F. Progesterone receptors in two groups ofendometrial carcinoma. Cancer (Phila.), 57:1385-1388, 1986.Holinka, C. F., Deligdisch, L., and Gurpide, E. Histológica! evaluation of invitro responses of endometrial adenocarcinoma to progestins and their relation to progesterone receptor levels. Cancer Res., 44: 293-296, 1984.Ng, A. B. P., Reagan, J. W., Storaasli, P. J., and Wentz, W. B. Mixedadenosquamous carcinoma of the endometrium. Am. J. Pathol., 59: 765-767, 1973.Zaino, R. J., Clarke, C. L., Mörtel,R., and Satyaswaroop, P. G. Heterogeneity of progesterone receptor distribution in human endometrial adenocarcinoma. Cancer Res., 48:1889-1895, 1988.Spona, J., Ulm, R., Bieglmayer, C., and Husslein, P. Hormone serum levelsand hormone receptor contents of endometria in women with normal menstrual cycles and patients bearing endometrial carcinoma. Gynecol. Obstet.Invest., 70:71-80, 1979..Ialine. O., Kauppila, A., Kontulak, K., Syrjälä,P., and Vihko, R. Female sexsteroid receptors in normal, hyperplastic and carcinomatous endometrium:the relationship to serum steroid hormones and gonadotrophins and changesduring medroxyprogesterone acetate administration. Int. J. Cancer, 24:545-554, 1979.Geisinger, K. R., Marshal!, R. B., Kute, T. E., and Homesley, H. D. Correlation of female sex steroid hormone receptors with histologie and ultrastructural differentiation in adenocarcinoma of the endometrium. Cancer (Phila.),58: 1506-1517, 1986.Kauppila, A. J. I., Isolalo, H. E., Kivinen, S. T., and Vihko, R. K. Predictionof clinical outcome with estrogen and progestin receptor concentrations andtheir relationships to clinical and histopathological variables in endometrialcancer. Cancer Res., 46: 5380-5384, 1986.Press, M. F., Nousek-Goebl, N., King, W. J., Herbst, A. L., and Greene, G.L. Immunohistochemical assessment of estrogen receptor distribution in thehuman endometrium throughout the menstrual cycle. Lab. Invest., 51:495-503, 1984.Sanborn, B. M., Kuo, H. S., and Held, B. Estrogen and progesterone bindingsite concentrations in human endometrium and cervix throughout the menstrual cycle and in tissue from women taking oral contraceptives. J. SteroidBiochem., 9: 951-955, 1978.Press, M. F., Nousek-Goebl, N. A., Bur, M., and Greene, G. L. Estrogenreceptor localization in the female genital tract. Am. J. Pathol., 123: 280-292, 1986.Sulovic, V., Milasin, J., Ribarac-Stepic, N., Brocic, M., and Kanazir, D. Sexsteroid receptors in transformed human cervical tissue. Eur. J. Gynecol.Oncol., 7:30-33, 1986.Twiggs, L. B., Polish, R. A., Leung, B. S., Carson, L. F., Adcock, L. L.,Savage, J. E., and Prem, K. A. Cytosolic estrogen and progesterone receptorsas prognostic parameters in stage IB cervical carcinoma. Gynecol. Oncol.,28: 156-160, 1987.Malkasian, G. D., Decker, D. G., and Jorgensen, E. O. Chemotherapy ofcarcinoma of the cervix. Gynecol. Oncol., 5: 109-120, 1977.Morange, M., Diu, A., Bensaude, O., and Babinet, C. Altered expression ofheat shock protein in embryonal carcinoma and mouse early embryonic cells.Mol. Cell Biol., 4: 730-735, 1984.Carper, S. W., Duffy, J. J., and Gerner, E. W. Heat shock proteins inthermotolerance and other cellular processes. Cancer Res., 47: 5249-5255,1987.Blanchard, J. M., Piechaczyk, M., Dani, C., Chambard, J. C., Franchi, A.,Pouyssegur, J., and Jeanteur, P. C-myc gene is transcribed at high rate inGo-arrested fibroblasts and is post-transcriptionally regulated in response togrowth factors. Nature (Lond.), 317: 443-445, 1985.Dubik, D., Dembinski, T. C., and Shiu, R. P. Stimulation of c-myc oncogeneexpression associated with estrogen-induced proliferation of human breastcancer cells. Cancer Res., 47: 6517-6521, 1987.

4304

Research. on October 25, 2020. © 1989 American Association for Cancercancerres.aacrjournals.org Downloaded from

1989;49:4298-4304. Cancer Res   Daniel R. Ciocca, Libertad A. Puy and Liliana C. Fasoli  Carcinomas of the Endometrium and Cervix

24,000 Protein in Patients withrMEstrogen-regulated Study of Estrogen Receptor, Progesterone Receptor, and the

  Updated version

  http://cancerres.aacrjournals.org/content/49/15/4298

Access the most recent version of this article at:

   

   

   

  E-mail alerts related to this article or journal.Sign up to receive free email-alerts

  Subscriptions

Reprints and

  .pubs@aacr.orgDepartment at

To order reprints of this article or to subscribe to the journal, contact the AACR Publications

  Permissions

  Rightslink site. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC)

.http://cancerres.aacrjournals.org/content/49/15/4298To request permission to re-use all or part of this article, use this link

Research. on October 25, 2020. © 1989 American Association for Cancercancerres.aacrjournals.org Downloaded from