0021-972X/90/7001-0115$02.00/0 Journal of Clinical Endocrinology and Metabolism Copyright © 1990 by The Endocrine Society
Vol. 70, No. 1 Printed in U.S.A.
Cathepsin-D in Human Encjometrium: Induction by Progesterone and Potential Value as a Tumor Marker* THIERRY MAUDELONDE, PATRICIA MARTINEZ, JEAN-PAUL BROUILLET, FRANQOIS LAFFARGUE, ANDRE PAGES, AND HENRI ROCHEFORT Institut National de la Sante et de la Recherche Medicale, Unite Hormones et Cancer (U148) (T.M., P.M., H.R.), 34090 Montpellier; and Laboratoire de Biologie Cellulaire et Hormonale (T.M., J-P.B., H.R.) and Service de Gynecologie Obstetrique (F.L.), Maternite, and Laboratoire d'Anatomie Pathologique Centre Gui de Chauliac (A.P.), CHR, 34059 Montpellier, France
cathepsin-D was higher in 19 endometrial carcinoma than in 20 normal endometrium, but was not correlated with steroid receptor status. However, using 15 pmol/mg protein as a cut-off level, the cathepsin-D status (high or low) was correlated with the degree of myometrial invasion (>one third) by adenocarcinoma cells, whereas steroid receptor status was not. We conclude that cathepsin-D is induced by progesterone in human endometrium, as it is in normal rat uterus, and we suggest that a low concentration of cathepsin-D in the cytosol of endometrial adenocarcinoma may indicate a favorable prognosis, since it is correlated with low myometrial invasion. (J Clin Endocrinol Metab 70: 115, 1990)
ABSTRACT. Using an immunoenzymatic assay, cathepsin-D concentrations were measured in the cytosol of human endometrium biopsies. The level of cathepsin-D was higher in the luteal phase than in the follicular phase (P < 0.01), suggesting increased accumulation by progesterone. Induction by progestin was confirmed by immunoprecipitation of cathepsin-D from a lysate of epithelial endometrial cells previously treated in primary culture with R5020 (10 nM); estradiol (10 nM) had no effect. Immunohistochemistry showed that cathepsin-D is mainly localized in the epithelium and that its level is higher in the luteal phase. The plasma level of cathepsin-D was stable during the menstrual cycle, ranging between 2.5-10 pmol/mL, but increased slightly during pregnancy. The mean level of
P
ROTEASES may play an important role in the process of nidation in the human uterus (1) and in the cyclic destructive autophagy of endometrium (2). Cathepsin-D is a lysosomal protease regulated by progesterone in the rat uterus (3) and by estrogen (but not progesterone) in human breast cancer cells (4). This cellular marker appears to be correlated with invasive and/or proliferative breast cancer (5, 6, 21). In the present study we measured cathepsin-D in human plasma and endometrium at different phases of the menstrual cycle to determine which steroid regulates its production in humans. We also assayed its concentration in endometrial cancers to test its potential value as a marker of transformation or invasiveness.
yr (November 1986 to December 1987). All hormonal treatments were stopped at least 3 months before biopsies. Five patients were excluded because samples were too small for analysis or histological study. Endometrial samples were obtained from hysterectomy or by biopsy using a pipelle endometrial suction curette (7). Two biopsies were performed on each patient; one sample was immediately frozen in liquid nitrogen for cathepsin-D assay, and the other was examined histologically to establish the menstrual cycle phase according to the classification of Noyes et al. (8). Based on clinical and histological information, samples were assigned to either follicular or luteal phases.
Materials and Methods Study in the menstrual cycle Twenty normal women from the Department of Obstetrics and Gynecology (Maternite, Montpellier, France) studied for 1
We selected adenocarcinomas at early clinical stages (1 and 2; classification of the International Federation of Gynecology and Obstetrics) (9) not requiring radiation therapy before surgery. Nineteen endometrial adenocarcinomas were obtained from the Department of Gynecology and the Cancer Center of Montpellier. A sample of each adenocarcinoma was immediately frozen in liquid nitrogen for cytosolic determination of steroid receptors and cathepsin-D. The remaining tumor was analyzed by the pathologist for histological diagnosis, determination of histological grade (I, II, or III, depending on tumor differentiation), and the degree of myometrial (< or >one third) (10) and cervical invasion.
Endometrium adenocarcinoma
Received February 6, 1989. Address all correspondence and requests for reprints to: Dr. Henri Rochefort, Institut National de la Sante et de la Recherche Medicale, Unite Hormones et Cancer (U148), 60 rue de Navacelles, 34090 Montpellier, France. * This work was supported by the Institut National de la Sante et de la Recherche Medicale, the University of Montpellier I, the Association pour la Recherche sur le Cancer, and the Federation Nationale des Centres de Lutte contre le Cancer.
115
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 12 November 2015. at 22:53 For personal use only. No other uses without permission. . All rights reserved.
MAUDELONDE ET AL.
116
Cytosol preparation Frozen tissues were homogenized with a Dounce homogenizer (Kontes Co., Vineland, NJ) in 10 mM Tris, pH 7.4, containing 1.5 mM EDTA, 10 mM monothioglycerol, and 10 mM sodium molybdate (TET molybdate buffer). They were then centrifuged at high speed (105,000 X g for 1 h), and the supernatant was defined as the cytosol. Immunoenzymatic assay of cathepsin-D in cytosol Cathepsin-D was determined by a double determinant solid phase assay (11). Microtiter plates were coated with the monoclonal antibody D7E3. After saturation of the remaining free adsorption sites, plates were washed three times with 0.1 M phosphate buffer (pH 7.4). Cytosols and second antibody (M1G8) conjugated with alkaline phosphatase were incubated simultaneously for 16 h at 4 C. The plates were then washed five times. They were stained with a 0.1-M diethanolamine (pH 9.8) solution containing 1 mg/mL paranitrophenyl phosphate for 1 h at room temperature. The cathepsin-D level was measured by optical density (OD) at 405 nm using a Multiskan spectrophotometer (Titertek, Flow Laboratories, Puteaux, France) coupled to an Apple He microcomputer. The cytosolic concentration of cathepsin-D was established from the absorbance of four duplicate dilutions (1:10 to 1:80). For each microtiter plate, a standard curve was determined from five duplicate dilutions of a pool of breast cancer cytosols containing 78 pmol/ mL cathepsin-D with reference to a known concentration of pure cathepsin-D (10). The sensitivity limit was 0.03 pmol/ well. If interassay variability exceeded 15%, plate results were excluded. Results are expressed as picomoles per mg proteins. Protein concentration was measured by a Bio-Rad (Munich, West Germany) assay, using BSA as standard. Immunoenzymatic assay of cathepsin-D in plasma Blood samples were collected on sodium heparinate. Plasma was obtained by centrifugation at 3000 rpm for 10 min and then frozen at —20 C until the assay. Using the previously described assay for cytosol1 with modifications, the total cathepsin-D concentration in the plasma was obtained by the absorbance of three dilutions (1:30 to 1:120) and compared to a linear curve, determined from five dilutions of a standard medium of cultured MCF7 cells. The limit of sensitivity of the assay was 2.5 fmol/well; the standard curve was linear between 0-40 fmol/well. Steroid receptor assay Estrogen (RE) and progesterone receptors (RP) concentrations in cytosol were assayed with the Abbott enzyme immunoassay (Abbott Diagnostic, Rungis, France) using monoclonal antibodies (12, 13). Ten femtomoles per mg protein was taken as the limit of positivity for RE and RP. 1 Brouillet, J. P., Hanslik, B., Maudelonde, T., Vaucher, E., Piva, M.T., Blanc, F., Rochefort, H., submitted for publication.
JCE & M • 1990 Vol70«Nol
Immunoprecipitation of 3bS-labeled proteins synthetized by cultured endometrial cells Primary cultures of epithelial and stromal cells were prepared from human endometrium as previously described (14). Tissues collected from surgery were immediately put into medium A [Dulbecco's Modified Eagle's Medium (DMEM) mixed with Ham's F-12] containing antibiotics. After tissue mincing, epithelial glands and stromal cells were dissociated using a solution of collagenase (250 U/mL) and hyaluronidase (95 U/ mL) for 2 h at 37 C. The cell suspension was filtered through polyester screen filters (successively 300 and 50 jim). Glandular structures were retained by a 50-/um screen filter, but stromal and blood cells were not retained. Cells were cultured in medium A containing 10% fetal calf serum for 1 day and then 10% fetal calf serum treated with dextran-coated charcoal for 4 days. Duplicate steroid treatment (estradiol and/or R5020) was then carried out for 4 days. Cells were labeled for 6 h with [35S]methionine (60 MCi/well) in 300 fiL DMEM containing a low concentration (10 nM) of methionine. Immunoprecipitation of cathepsin-D with the monoclonal antibody D8F5 in cellular lysate was performed as previously described (15). Nonspecific immunoprecipitation was determined with a nonrelevant monoclonal antibody (IgGl MOPC 21, Bionetics, Kensington, MD). 35S-Labeled immunoprecipitates were further analyzed by sodium dodecyl sulfate-polyacrylamide gel electrophoresis as described previously (14). Fluorograms were scanned with a densitometer (Vernon, Paris, France) connected to a Delsi integrator (Elvetec, Marseille, France) to determine the percentage of total absorbance attributable to 52K and 34K proteins. Nonspecific immunoprecipitates were subtracted for each treatment. Results are shown as percentages of the untreated control corrected per /xg total DNA. Two luteal phase samples were dissociated in stromal and glandular cells as described above. Red blood cells were then destroyed with a lysis buffer ( 8 mM KHCO3, 155 mM NH4C1, and 1 mM EDTA). The 600 X g supernatant was discarded, and the pellet sonicated in a TET buffer. The 12,000 x g (30 min) supernatant was used for cathepsin-D assay. Immunohistochemical staining of cathepsin-D Six-micron thick frozen sections, fixed in cold acetone, were stained by the Steinberger peroxidase-antiperoxidase method as previously described (16), using 100 ^L/slide of a 10 Mg/mL M1G8 antibody solution. Monoclonal antibody MlG8 was more sensitive than the previously tested antibody D7E3 (16) when comparisons were made at different concentrations (1,3 and 7 Mg/mL) on the same breast cancer tissue (not shown). Each staining was associated with a negative control performed with a nonrelevant monoclonal antibody (IgGl, MOPC 21, Bionetics) of the same subclass. Statistical methods Statistical differences within the population were determined by Kruskall-Wallis or Mann-Whitney nonparametric tests for quantitative parameters. Fisher's exact test was used for qualitative parameters. Linear regression was calculated by Pear-
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 12 November 2015. at 22:53 For personal use only. No other uses without permission. . All rights reserved.
CATHEPSIN-D IN HUMAN ENDOMETRIUM
117
normal
son's least squares method. Pearson's correlation coefficients were analyzed by Student's t test.
Results Cathepsin-D concentration in normal endometrium according to the menstrual cycle
E
Clinical and histological dating of the phase of menstrual cycle concurred for each of the 20 patients studied. Patient ages were similar in the follicular and luteal phases (37.7 ± 7.5 and 33 ± 5 yr old, respectively). In the follicular phase, all cathepsin-D values were low and grouped (Fig. 1). Cathepsin-D levels varied more in the luteal phase (Fig. 1), and its mean value was significantly higher than that for the follicular phase, according to the Mann-Whitney test (P < 0.01).
o o
Cathepsin-D concentration in endometrial adenocarcinoma Nineteen adenocarcinomas were analyzed. Sixteen were well differentiated (grade I), and the other three were undifferentiated (grade III) or intermediate (grade II; Table 1). The cathepsin-D concentrations in endometrial carcinoma were dispersed (Fig. 1). The mean TABLE 1. Clinical and biological parameters in endometrial adenocarcinoma Patient Age Grade no. (yr)
Invasion
fmol/mg Cathepsin-D P (pmol/mg
Myometrium Others RE RP 1 2
3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19
53 55 57 69 60 76 67 51 61 64 69 48 73 48 59 77 43 66 58
I I I I I I I III I I
I I I II I I I III I
0 0 0
+ + + + + ++ ++ ++ ++ ++ ++ ++ ++ ++ ++ ++
0 0 0
0 0 0 0 Cervix Ovaries 0 0 0 0 0 0 0 0 Nodes 0
16 187 317 244 39 9 9 2 163 20 611 2108
0 ND 37 1 185 58 78 627 957 1 61 7 0 4 57 116 912 2310 106 9 296 37 336 829 144 5
r)
7 9 61 10 10 52 69 15 19 20 23 23 25 26 27 31 31 49 53
The myometrial invasion status was used to distinguish three groups: no invasion (0), less than one third myometrial invasion (+), and one third or more invasion (++). Histological grades were determined according to the International Federation of Gynecology and Obstetrics (6). The cytosolic concentrations of RE, RP, and cathepsin-D were measured in tumor cytosol as described in Materials and Methods. Patient 7 had inflammatory adenocarcinoma with macrophage invasion. ND, not determined. P, Progesterone.
o
10
FOL.
LUT.
,30
*/• invasiveness
FIG. 1. Cytosolic cathepsin-D concentrations in normal and cancerous endometrium. Cytosolic cathepsin-D was assayed by a double determinant immunoenzymatic assay as described in Materials and Methods. Normal endometria were classified as follicular (FOL.) or luteal (LUT.) phase according to clinical and histological information. Two groups of adenocarcinomas were identified according to the status (< or >30%) of myometrial invasion (percent invasiveness). The numbers of subjects are in parentheses, and the transverse lines are the means of each group. The broken transverse line is the cut-off level of 15 pmol/mg protein. Statistical differences between the groups were analyzed by the Kruskall-Wallis and Mann-Whitney nonparametric tests.
value is not statistically different from that for normal endometrium in the luteal phase (P = 0.39), but is significantly higher than normal endometrium values in the follicular phase (P < 0.01). Attempts have been made to correlate cathepsin-D concentration with myometrial invasion. The mean value [29.2 ± 26.6 (±SD) pmol/mg protein] in the group without myometrial invasion did not differ from that in tumors with high myometrial invasion (29.7 ± 11.2 pmol/ mg protein). However, using a cut-off level of 15 pmol/ mg protein, the highest level of cathepsin-D was significantly associated with high myometrial invasion (Table 2). None of the low cathepsin-D tumors was invasive, but three of the highest cathepsin-D levels were found in apparently noninvasive tumors (no. 3, 6, and 7; Table 1). One (no. 7) contained an inflammatory stroma with many lymphocytes, plasmocytes, and macrophages, which may have artificially increased the level of cathep-
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 12 November 2015. at 22:53 For personal use only. No other uses without permission. . All rights reserved.
MAUDELONDE ET AL.
118
JCE & M • 1990 Vol70«Nol
TABLE 2. Correlation between myometrial invasion and cellular markers
Markers
Status n°
No. of patients according to invasion
P
Low (0 or +) High (++) RE
+
19 0.37*
6 2
10 1
RP
+
18 0.64"
4 3
6 6
52K cathepsin-D
+
19 0.007
3 5
10 0
c
Qualitative correlation between myometrial invasion (0, +, or ++) and the status of three biochemical parameters. The cut-off level was 10 fmol/mg protein for steroid receptors and 15 pmol/mg protein for cathepsin-D. " Number of endometrial adenocarcinomas studied. *NS. c Statistically significant using Fisher's exact test.
sin-D in the sample (16). Only three patients showed invasion of other tissues (ovary, cervix, and lymph nodes), and no correlation was found with the metastasizing ability of these endometrial cancers. The cathepsin-D concentration was not correlated with RE (r = 0.002; P = 0.98) or RP (r = 0.28; P = 0.25) concentrations. Immunohistochemistry of cathepsin-D Using the monoclonal antibody D7E3, cathepsin-D was not detected by immunohistochemistry in most samples of human endometrium, as previously described (15). However, using a more sensitive monoclonal antibody (M1G8), cathepsin-D was weakly detected in proliferative phase endometrium and was highly concentrated in glandular epithelial cells of luteal phase endometrium (Fig. 2). The staining intensity was lower in stromal cells than in epithelial cells (Fig. 2). To confirm the difference in cathepsin-D accumulation in epithelial and stromal cells, we compared cathepsin-D concentrations in dissociated glandular epithelial cells and in fibroblasts from two luteal phase patients. In both cases cathepsin-D concentration was higher in glandular epithelial cells (mean, 72 pmol/mg protein) than in fibroblasts (mean, 30 pmol/mg protein). The higher values for dissociated endometrial cells than for the total endometrial cytosol (Fig. 1) may simply result from the removal of plasma protein during the digestion step.
was not significantly different in follicular (mean ± SD, 5.11 ± 1.9) or luteal phases (4.75 ± 2.2; Fig. 3). After ovarian stimulation by human menopausal gonadotropin-hCG, progesterone treatment (600 mg/day) did not alter the mean level of plasma cathepsin-D (4.55 ± 1.9). Plasma levels of cathepsin D in the 17th to 18th weeks of pregnancy (6.60 ± 2.1) were slightly but statistically higher than those in patients treated with 600-mg daily doses of progesterone (P < 0.01) and those in the normal follicular phase (P < 0.04), but not those in the normal luteal phase (P < 0.06).
Cathepsin-D concentration in plasma
Hormonal regulation of cathepsin-D in primary culture
The cathepsin-D plasma concentration was stable (from 2.5 to 10 pmol/mL or 0.03 to 0.14 pmol/mg protein) and lower than the endometrial cytosol concentration. It
Cathepsin-D was immunoprecipitated from the cellular lysate of cultured epithelial endometrial cells and analyzed by sodium dodecyl sulfate-polyacrylamide gel
FIG. 2. Immunoperoxidase staining of cathepsin-D in follicular (A) and luteal (B) endometrium. Using the M1G8 monoclonal antibody, frozen sections from endometrial biopsies were fixed, permeabilized, and analyzed using the indirect peroxidase-antiperoxidase technique described in Materials and Methods. In this patient the cytosolic cathepsin-D concentrations were 8 and 17 pmol/mg protein in the follicular and luteal phases, respectively.
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 12 November 2015. at 22:53 For personal use only. No other uses without permission. . All rights reserved.
CATHEPSIN-D IN HUMAN ENDOMETRIUM (9)
(10)
(101
119
MW
(11)
10
« o
7.5
E
'*
*
5.0-
J_
t
•
52 K^
•
V •
+
•
48K 34 K—
•
FOL.
LUT.
Prog.
PREG
FIG. 3. Plasma levels of cathepsin-D in women according to hormonal status. Assay of plasmatic cathepsin-D was performed by a double determinant immunoenzymatic assay as described in Materials and Methods. The follicular (FOL.) and luteal (LUT.) phases of the menstrual cycle were defined according to clinical and histological data. PROG., The group of patients treated for sterility with progesterone (600 mg/day) after the 14th or 15th day of an artificial cycle following successive treatment with the LHRH analogs human menopausal gonadotropin and hCG. Only the group of normal pregnancies (17-18 weeks gestation; PREG) had statistically higher values than the other groups (*, P = 0.01 us. PROG., P = 0.04 us. FOL., P = 0.06 vs. LUT.). The numbers of subjects are in parentheses.
electrophoresis. Two bands were only specifically precipitated by anticathepsin-D antibody. They correspond to the precursor (52K) and the large chain (34K) of mature cathepsin-D. Generally, other bands were nonspecific, as shown by the migration of the nonspecific control. Cell treatment by progestin R5020 increased the 34K and 52K bands (3 times that of the control), and estradiol had no effect. The stimulation induced by R5020 was not inhibited by estradiol (Fig. 4). The 48K nonspecific band was also increased by R5020, suggesting that this band contains at least two proteins, one that is nonspecific and precipitated by the nonrelevant antibody and the other a 48K specific protein corresponding to the one-chain mature form of cathepsin-D (5) which is also increased by progestin. The amount of cathepsin-D secreted into the medium by endometrial cells was not detectable, in contrast with amounts secreted by the MCF7 breast cancer cells analyzed in parallel (not shown).
Discussion There are three major findings in this study. First, the hormonal regulation of cathepsin-D in human normal
NS
7o(52K + 34K) -
C
E2 R5020 R5020
100 102
320
300
FlG. 4. Fluorography of immunoprecipitated cellular cathepsin-D. Epithelial endometrial cells obtained in the luteal phase were cultured for 11 days and treated for 4 days without [control (C)] or with estradiol (E2), R5020, or E2 plus R5020, each at 10 nM. Cathepsin-D was immunoprecipitated using monoclonal antibody D7E3. The same amount of trichloroacetic acid-precipitable 36S-labeled protein was used for controls and the different treatments. Mol wt (MW) were determined from the migration of the following standard proteins: lactoglobin (18,000), carbonic anhydrase (30,000), ovalbumin (46,000), BSA (69,000), and phosphorylase-/? (92,000). 52K, 48K, and 34K correspond to three forms of the immunoprecipitated cathepsin-D. They are also expressed as the percentage of the untreated control sample after quantification as described in Materials and Methods. NS corresponds to the immunoprecipitates obtained with the nonrelevant antibody MOPC-21.
endometrium is mostly due to progesterone, not to estradiol. Second, the cathepsin-D level appears to be higher in adenocarcinoma cells than in normal cells. Third, the low level of cathepsin-D in endometrial cancer appears to be correlated with low myometrial invasion. In normal endometrium, the level of cathepsin-D was higher in the luteal phase. It would, therefore, appear that in normal human endometrium, as reported in the rat uterus (3), cathepsin-D is induced by progesterone, but not by estrogens. This suggests tissue specificity in the regulation of cathepsin-D, since it was previously shown that this lysosomal enzyme is estrogen regulated
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 12 November 2015. at 22:53 For personal use only. No other uses without permission. . All rights reserved.
120
MAUDELONDE ET AL.
in human breast cancer cells (4, 5). Similar cathepsin-D unresponsiveness to estrogens has been demonstrated in the estrogen-responsive Ishikawa endometrial cell line (22). However, it is not yet clear if this hormonal regulation difference is due to tissue specificity (uterine us. mammary cells), the transformation process (normal vs. cancer cells), or both. This question could be resolved by examining the regulation in normal mammary cells containing estrogen receptors and uterine adenocarcinoma cells responding to both progestins and estrogens (22). Examples of tissue specificity include the synthesis of PRL, which is regulated by estrogen in the pituitary but by progesterone in endometrium (17); Moulton and Koenig (18) have also shown that progesterone regulation of cathepsin-D is specific for luminal epithelial cells in
the rat uterus. This cellular specificity of hormonal regulation suggests that cathepsin-D might have a role in the biological and morphological modifications that occur during the luteal phase. Its plasma level is very stable during the menstrual cycle, and in nonpregnant women the contribution of cathepsin-D secreted by steroid-regulated tissue such as endometrium is negligible. The cathepsin-D function in endometrium is unknown. Lysosomal activity increases during the luteal phase, and its role in cyclic menstrual bleeding has been suggested (2). Cathepsin-D is a lysosomal protease and may play a role in the cyclic destructive autophagy and remodeling of endometrium in the absence of nidation. The second finding, based on a limited number of patients, suggests that this marker could be an indicator of endometrial transformation. Since progesterone is absent in all postmenopausal cancer patients, the comparison with normal endometrium in the follicular phase indicates a 3-fold increase in the level of cathepsin-D in adenocarcinoma. Further studies are required, however, since we have not excluded the possibility that the higher levels in adenocarcinoma, compared with levels in normal endometrium in the follicular phase, may result from a higher density of epithelial cells. Moreover, the cathepsin-D level is generally lower in endometrial adenocarcinoma than in breast cancer, and the proportion of procathepsin-D secretion is much lower in Ishikawa adenocarcinoma cell lines (Touitou et al., unpublished) than in breast cancer cell lines. The third finding suggests that the cathepsin-D concentration in endometrial adenocarcinoma may be an additional marker of invasiveness. Cathepsin-D values in the group with low myometrial invasion were more dispersed, and five of the eight samples had low levels, corresponding to follicular phase values. At the 15 pmol/ mg protein cut-off level, which refers to the highest value obtained in the follicular phase, there was a significant correlation between myometrial invasion and cathepsin-
JCE & M • 1990 Vol 70 • No 1
D concentrations. This cut-off level may be adjusted after more patients are studied. Macrophages produce cathepsin-D (19), and the high concentrations in adenocarcinoma without significant myometrial invasiveness may have been due to the high proportion of histiocytes. This was highly probable in the sample with abundant inflammatory cells. Myometrial invasiveness is considered to be the major prognostic parameter of endometrial carcinoma (20), and the cellular concentration of cathepsin-D may be of interest as an associated biochemical marker of invasion. This result is consistent with prospective (6) and retrospective studies (21) of breast cancer patients, indicating that cytosolic cathepsin-D concentrations have a prognostic value independent of other prognostic parameters. We could not draw conclusions on the value of cyto-
solic cathepsin-D as a tumor and prognostic marker because of the low number of patients studied. However, the results are in agreement with the hypothesis that an increased level of cathepsin-D (5) is associated with cell transformation and invasion of surrounding tissues by cancer cells, at least in hormone-dependent tissues.
Acknowledgments We are grateful to Prof. H. Pujol and Dr. D. J. Domergue (Centre Paul Lamarque, Montpellier) for providing endometrial adenocarcinoma samples and Prof. B. Hedon for providing normal endometrial samples. We thank C. Esnault and G. Salazar-Retana for technical help, and E. Barrie and M. Egea for typing the manuscript.
References 1. Schlafke S, Enders AC. Cellular basis of interaction between trophoblast and uterus at implantation. Biol Reprod. 1975;12:41-65. 2. Henzl MR, Smith RE, Boost G, Tyler ET. Lysosomal concept of menstrual bleeding in humans. J Clin Endocrinol Metab. 1972; 34:860-75. 3. Elangovan S, Moulton BC. Progesterone and estrogen control of rates of synthesis of uterine cathepsin D. J Biol Chem. 1980; 255:7474-9. 4. Westley B, Rochefort H. A secreted glycoprotein induced by estrogen in human breast cancer cell lines. Cell. 1980;20:352-62. 5. Rochefort H, Capony F, Garcia M, et al. Estrogen-induced lysosomal proteases secreted by breast cancer cells. A role in carcinogenesis? J Cell Biochem. 1987;35:17-29. 6. Maudelonde T, Khalaf S, Garcia M, et al. Immunoenzymatic assay of Mr 52,000 cathepsin D in 182 breast cancer cytosols. Low correlation with other prognostic parameters. Cancer Res. 1988;48:462-6. 7. Hill GA, Herbert III CM, Parker RA, Wentz AC. Comparison of late luteal phase endometrial biopsies using the novak curette or pipelle endometrial suction curette. Obstet Gynecol. 1989;73:4435. 8. Noyes RV, Herng AT, Rock J. Dating the endometrial biopsy. Fertil Steril. 1950;l:3-25. 9. International Federation of Gynecologists and Obstetricians. Classification and staging of malignant tumors in the female pelvis. Acta Obstet Gynecol Scand. 1971;50:l-12. 10. Morrow CP. Endometrial carcinoma stages I and II: is surgery adequate? Int J Radiation Oncol Biol Physiol. 1980;6:365-6. 11. Rogier H, Freiss G, Besse MG, et al. Two-site immunoenzymometric assay for the 52-kDa cathepsin D in cytosols of breast cancer
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 12 November 2015. at 22:53 For personal use only. No other uses without permission. . All rights reserved.
CATHEPSIN-D IN HUMAN ENDOMETRIUM tissues. Clin Chem. 1989;35:81-5. 12. Nolan C, Przywara LW, Miller LS, Suduikis VBS, Tomita JT. A sensitive solid-phase enzyme immunoassay for human estrogen receptor. In: Ames FC, Blumenschein GR, Montague ED, eds. Current controversies in breast cancer. Austin: University of Texas Press; 1984;433-41. 13. Greene GL, Harris K, Bovo R, Nolan C. Preparation and characterization of monoclonal antibodies to human progesterone receptor. Mol Endocrinol. 1988;8:714-26. 14. Maudelonde T, Rochefort H. A 51K progestin-regulated secreted by human endometrial cells in primary culture. J Clin Endocrinol Metab 1987;64:1294-1302. 15. Morisset M, Capony F, Rochefort H. Processing and estrogen regulation of the 52-kDa protein inside MCF7 breast cancer cells. Endocrinology. 1986;119:2773-83. 16. Garcia M, Salazar-Retana G, Richer G, et al. Immunohistochemical detection of the estrogen-regulated Mr 52,000 protein in primary breast cancers but not in normal breast and uterus. J Clin Endocrinol Metab. 1984;59:564-6.
121
17. Huang JR, Tseng L, Bischof P, Janne OA. Regulation of prolactin by progestin, estrogen and relaxin in human endometrial stromal cells. Endocrinology. 1987;121:2011-7. 18. Moulton BC, Koenig BB. Progestin increases cathepsin D synthesis in uterus luminal epithelial cells. Am J Physiol. 1983;244:E442-6. 19. Imort M, Ziihlsdorf M, Feige U, Hasilik H, Von Figura K. Biosynthesis and transport of lysosomal enzymes in human monocytes and macrophages. Biochem J. 1983;214:671-8. 20. Hendrickson M, Ross J, Eifel RJ, et al. Adenocarcinoma of the endometrium. Analysis of 256 cases with carcinoma limited to the uterine corpus. Gynecol Oncol. 1982;13:373-7. 21. Thorpe SM, Rochefort H, Garcia M, Freiss G, Christensen J, Khala FS, Paolucci F, Pau B, Rasmussen BB, Rose C. Association between high concentrations of 52K cathepsin-D and poor prognosis in primary breast cancer. Cancer Res. Nov. 1989, in press 22. Touitou I, Cavailles V, Garcia M, DeFrenne A, Rochefort H. Differential regulation of 52K cathepsin D by estradiol and progesterone in mammary cancer and uterine cells Mol Cell Endocrinol. 1989, in press.
International Symposium on Thyroperoxidase Fundamental and Clinical Aspects Marseille, France" June 28-30, 1990 Scientific committee: G.F. Bottazzo (London), L. Braverman (Worcester), P. Carayon (Marseille), B. Czarnocka (Warsaw), J. De Vijlder (Amsterdam), R. Mornex (Lyon), A. Pinchera (Pisa), B. Rousset (Lyon), J. Ruf (Marseille), P. Scriba (Liibeck), G. Vassart (Brussels) An international symposium on recent advances in the implication of thyroperoxydase (TPO) in autoimmune disease of the thyroid gland. Topics include: structure and expression of the TPO gene, TPO structure, catalytic mechanisms, structure-activity relationship, hormonal regulation of TPO, traffic of TPO, congenital and acquired deficiencies of iodide organification, cellular and humoral mechanisms of autoimmune disease, anti-TPO antibodies in various thyroid and autoimmune diseases, TPO and antiTPO assays, immunohistochemistry, in-situ hybridization. The program will include lectures by invited speakers, selected oral communications and poster presentations. For further information please contact: Dr. Pierre Carayon, Laboratoire des Hormones Proteiques, Faculte de Medecine 27 bd Jean Moulin 13385 Marseille Cedex 5, France Tel: 91.78.68.55, Fax: 91.25.64.07
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 12 November 2015. at 22:53 For personal use only. No other uses without permission. . All rights reserved.
Vol. 70, No. 1 Printed in U.S.A.
Cathepsin-D in Human Encjometrium: Induction by Progesterone and Potential Value as a Tumor Marker* THIERRY MAUDELONDE, PATRICIA MARTINEZ, JEAN-PAUL BROUILLET, FRANQOIS LAFFARGUE, ANDRE PAGES, AND HENRI ROCHEFORT Institut National de la Sante et de la Recherche Medicale, Unite Hormones et Cancer (U148) (T.M., P.M., H.R.), 34090 Montpellier; and Laboratoire de Biologie Cellulaire et Hormonale (T.M., J-P.B., H.R.) and Service de Gynecologie Obstetrique (F.L.), Maternite, and Laboratoire d'Anatomie Pathologique Centre Gui de Chauliac (A.P.), CHR, 34059 Montpellier, France
cathepsin-D was higher in 19 endometrial carcinoma than in 20 normal endometrium, but was not correlated with steroid receptor status. However, using 15 pmol/mg protein as a cut-off level, the cathepsin-D status (high or low) was correlated with the degree of myometrial invasion (>one third) by adenocarcinoma cells, whereas steroid receptor status was not. We conclude that cathepsin-D is induced by progesterone in human endometrium, as it is in normal rat uterus, and we suggest that a low concentration of cathepsin-D in the cytosol of endometrial adenocarcinoma may indicate a favorable prognosis, since it is correlated with low myometrial invasion. (J Clin Endocrinol Metab 70: 115, 1990)
ABSTRACT. Using an immunoenzymatic assay, cathepsin-D concentrations were measured in the cytosol of human endometrium biopsies. The level of cathepsin-D was higher in the luteal phase than in the follicular phase (P < 0.01), suggesting increased accumulation by progesterone. Induction by progestin was confirmed by immunoprecipitation of cathepsin-D from a lysate of epithelial endometrial cells previously treated in primary culture with R5020 (10 nM); estradiol (10 nM) had no effect. Immunohistochemistry showed that cathepsin-D is mainly localized in the epithelium and that its level is higher in the luteal phase. The plasma level of cathepsin-D was stable during the menstrual cycle, ranging between 2.5-10 pmol/mL, but increased slightly during pregnancy. The mean level of
P
ROTEASES may play an important role in the process of nidation in the human uterus (1) and in the cyclic destructive autophagy of endometrium (2). Cathepsin-D is a lysosomal protease regulated by progesterone in the rat uterus (3) and by estrogen (but not progesterone) in human breast cancer cells (4). This cellular marker appears to be correlated with invasive and/or proliferative breast cancer (5, 6, 21). In the present study we measured cathepsin-D in human plasma and endometrium at different phases of the menstrual cycle to determine which steroid regulates its production in humans. We also assayed its concentration in endometrial cancers to test its potential value as a marker of transformation or invasiveness.
yr (November 1986 to December 1987). All hormonal treatments were stopped at least 3 months before biopsies. Five patients were excluded because samples were too small for analysis or histological study. Endometrial samples were obtained from hysterectomy or by biopsy using a pipelle endometrial suction curette (7). Two biopsies were performed on each patient; one sample was immediately frozen in liquid nitrogen for cathepsin-D assay, and the other was examined histologically to establish the menstrual cycle phase according to the classification of Noyes et al. (8). Based on clinical and histological information, samples were assigned to either follicular or luteal phases.
Materials and Methods Study in the menstrual cycle Twenty normal women from the Department of Obstetrics and Gynecology (Maternite, Montpellier, France) studied for 1
We selected adenocarcinomas at early clinical stages (1 and 2; classification of the International Federation of Gynecology and Obstetrics) (9) not requiring radiation therapy before surgery. Nineteen endometrial adenocarcinomas were obtained from the Department of Gynecology and the Cancer Center of Montpellier. A sample of each adenocarcinoma was immediately frozen in liquid nitrogen for cytosolic determination of steroid receptors and cathepsin-D. The remaining tumor was analyzed by the pathologist for histological diagnosis, determination of histological grade (I, II, or III, depending on tumor differentiation), and the degree of myometrial (< or >one third) (10) and cervical invasion.
Endometrium adenocarcinoma
Received February 6, 1989. Address all correspondence and requests for reprints to: Dr. Henri Rochefort, Institut National de la Sante et de la Recherche Medicale, Unite Hormones et Cancer (U148), 60 rue de Navacelles, 34090 Montpellier, France. * This work was supported by the Institut National de la Sante et de la Recherche Medicale, the University of Montpellier I, the Association pour la Recherche sur le Cancer, and the Federation Nationale des Centres de Lutte contre le Cancer.
115
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 12 November 2015. at 22:53 For personal use only. No other uses without permission. . All rights reserved.
MAUDELONDE ET AL.
116
Cytosol preparation Frozen tissues were homogenized with a Dounce homogenizer (Kontes Co., Vineland, NJ) in 10 mM Tris, pH 7.4, containing 1.5 mM EDTA, 10 mM monothioglycerol, and 10 mM sodium molybdate (TET molybdate buffer). They were then centrifuged at high speed (105,000 X g for 1 h), and the supernatant was defined as the cytosol. Immunoenzymatic assay of cathepsin-D in cytosol Cathepsin-D was determined by a double determinant solid phase assay (11). Microtiter plates were coated with the monoclonal antibody D7E3. After saturation of the remaining free adsorption sites, plates were washed three times with 0.1 M phosphate buffer (pH 7.4). Cytosols and second antibody (M1G8) conjugated with alkaline phosphatase were incubated simultaneously for 16 h at 4 C. The plates were then washed five times. They were stained with a 0.1-M diethanolamine (pH 9.8) solution containing 1 mg/mL paranitrophenyl phosphate for 1 h at room temperature. The cathepsin-D level was measured by optical density (OD) at 405 nm using a Multiskan spectrophotometer (Titertek, Flow Laboratories, Puteaux, France) coupled to an Apple He microcomputer. The cytosolic concentration of cathepsin-D was established from the absorbance of four duplicate dilutions (1:10 to 1:80). For each microtiter plate, a standard curve was determined from five duplicate dilutions of a pool of breast cancer cytosols containing 78 pmol/ mL cathepsin-D with reference to a known concentration of pure cathepsin-D (10). The sensitivity limit was 0.03 pmol/ well. If interassay variability exceeded 15%, plate results were excluded. Results are expressed as picomoles per mg proteins. Protein concentration was measured by a Bio-Rad (Munich, West Germany) assay, using BSA as standard. Immunoenzymatic assay of cathepsin-D in plasma Blood samples were collected on sodium heparinate. Plasma was obtained by centrifugation at 3000 rpm for 10 min and then frozen at —20 C until the assay. Using the previously described assay for cytosol1 with modifications, the total cathepsin-D concentration in the plasma was obtained by the absorbance of three dilutions (1:30 to 1:120) and compared to a linear curve, determined from five dilutions of a standard medium of cultured MCF7 cells. The limit of sensitivity of the assay was 2.5 fmol/well; the standard curve was linear between 0-40 fmol/well. Steroid receptor assay Estrogen (RE) and progesterone receptors (RP) concentrations in cytosol were assayed with the Abbott enzyme immunoassay (Abbott Diagnostic, Rungis, France) using monoclonal antibodies (12, 13). Ten femtomoles per mg protein was taken as the limit of positivity for RE and RP. 1 Brouillet, J. P., Hanslik, B., Maudelonde, T., Vaucher, E., Piva, M.T., Blanc, F., Rochefort, H., submitted for publication.
JCE & M • 1990 Vol70«Nol
Immunoprecipitation of 3bS-labeled proteins synthetized by cultured endometrial cells Primary cultures of epithelial and stromal cells were prepared from human endometrium as previously described (14). Tissues collected from surgery were immediately put into medium A [Dulbecco's Modified Eagle's Medium (DMEM) mixed with Ham's F-12] containing antibiotics. After tissue mincing, epithelial glands and stromal cells were dissociated using a solution of collagenase (250 U/mL) and hyaluronidase (95 U/ mL) for 2 h at 37 C. The cell suspension was filtered through polyester screen filters (successively 300 and 50 jim). Glandular structures were retained by a 50-/um screen filter, but stromal and blood cells were not retained. Cells were cultured in medium A containing 10% fetal calf serum for 1 day and then 10% fetal calf serum treated with dextran-coated charcoal for 4 days. Duplicate steroid treatment (estradiol and/or R5020) was then carried out for 4 days. Cells were labeled for 6 h with [35S]methionine (60 MCi/well) in 300 fiL DMEM containing a low concentration (10 nM) of methionine. Immunoprecipitation of cathepsin-D with the monoclonal antibody D8F5 in cellular lysate was performed as previously described (15). Nonspecific immunoprecipitation was determined with a nonrelevant monoclonal antibody (IgGl MOPC 21, Bionetics, Kensington, MD). 35S-Labeled immunoprecipitates were further analyzed by sodium dodecyl sulfate-polyacrylamide gel electrophoresis as described previously (14). Fluorograms were scanned with a densitometer (Vernon, Paris, France) connected to a Delsi integrator (Elvetec, Marseille, France) to determine the percentage of total absorbance attributable to 52K and 34K proteins. Nonspecific immunoprecipitates were subtracted for each treatment. Results are shown as percentages of the untreated control corrected per /xg total DNA. Two luteal phase samples were dissociated in stromal and glandular cells as described above. Red blood cells were then destroyed with a lysis buffer ( 8 mM KHCO3, 155 mM NH4C1, and 1 mM EDTA). The 600 X g supernatant was discarded, and the pellet sonicated in a TET buffer. The 12,000 x g (30 min) supernatant was used for cathepsin-D assay. Immunohistochemical staining of cathepsin-D Six-micron thick frozen sections, fixed in cold acetone, were stained by the Steinberger peroxidase-antiperoxidase method as previously described (16), using 100 ^L/slide of a 10 Mg/mL M1G8 antibody solution. Monoclonal antibody MlG8 was more sensitive than the previously tested antibody D7E3 (16) when comparisons were made at different concentrations (1,3 and 7 Mg/mL) on the same breast cancer tissue (not shown). Each staining was associated with a negative control performed with a nonrelevant monoclonal antibody (IgGl, MOPC 21, Bionetics) of the same subclass. Statistical methods Statistical differences within the population were determined by Kruskall-Wallis or Mann-Whitney nonparametric tests for quantitative parameters. Fisher's exact test was used for qualitative parameters. Linear regression was calculated by Pear-
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 12 November 2015. at 22:53 For personal use only. No other uses without permission. . All rights reserved.
CATHEPSIN-D IN HUMAN ENDOMETRIUM
117
normal
son's least squares method. Pearson's correlation coefficients were analyzed by Student's t test.
Results Cathepsin-D concentration in normal endometrium according to the menstrual cycle
E
Clinical and histological dating of the phase of menstrual cycle concurred for each of the 20 patients studied. Patient ages were similar in the follicular and luteal phases (37.7 ± 7.5 and 33 ± 5 yr old, respectively). In the follicular phase, all cathepsin-D values were low and grouped (Fig. 1). Cathepsin-D levels varied more in the luteal phase (Fig. 1), and its mean value was significantly higher than that for the follicular phase, according to the Mann-Whitney test (P < 0.01).
o o
Cathepsin-D concentration in endometrial adenocarcinoma Nineteen adenocarcinomas were analyzed. Sixteen were well differentiated (grade I), and the other three were undifferentiated (grade III) or intermediate (grade II; Table 1). The cathepsin-D concentrations in endometrial carcinoma were dispersed (Fig. 1). The mean TABLE 1. Clinical and biological parameters in endometrial adenocarcinoma Patient Age Grade no. (yr)
Invasion
fmol/mg Cathepsin-D P (pmol/mg
Myometrium Others RE RP 1 2
3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19
53 55 57 69 60 76 67 51 61 64 69 48 73 48 59 77 43 66 58
I I I I I I I III I I
I I I II I I I III I
0 0 0
+ + + + + ++ ++ ++ ++ ++ ++ ++ ++ ++ ++ ++
0 0 0
0 0 0 0 Cervix Ovaries 0 0 0 0 0 0 0 0 Nodes 0
16 187 317 244 39 9 9 2 163 20 611 2108
0 ND 37 1 185 58 78 627 957 1 61 7 0 4 57 116 912 2310 106 9 296 37 336 829 144 5
r)
7 9 61 10 10 52 69 15 19 20 23 23 25 26 27 31 31 49 53
The myometrial invasion status was used to distinguish three groups: no invasion (0), less than one third myometrial invasion (+), and one third or more invasion (++). Histological grades were determined according to the International Federation of Gynecology and Obstetrics (6). The cytosolic concentrations of RE, RP, and cathepsin-D were measured in tumor cytosol as described in Materials and Methods. Patient 7 had inflammatory adenocarcinoma with macrophage invasion. ND, not determined. P, Progesterone.
o
10
FOL.
LUT.
,30
*/• invasiveness
FIG. 1. Cytosolic cathepsin-D concentrations in normal and cancerous endometrium. Cytosolic cathepsin-D was assayed by a double determinant immunoenzymatic assay as described in Materials and Methods. Normal endometria were classified as follicular (FOL.) or luteal (LUT.) phase according to clinical and histological information. Two groups of adenocarcinomas were identified according to the status (< or >30%) of myometrial invasion (percent invasiveness). The numbers of subjects are in parentheses, and the transverse lines are the means of each group. The broken transverse line is the cut-off level of 15 pmol/mg protein. Statistical differences between the groups were analyzed by the Kruskall-Wallis and Mann-Whitney nonparametric tests.
value is not statistically different from that for normal endometrium in the luteal phase (P = 0.39), but is significantly higher than normal endometrium values in the follicular phase (P < 0.01). Attempts have been made to correlate cathepsin-D concentration with myometrial invasion. The mean value [29.2 ± 26.6 (±SD) pmol/mg protein] in the group without myometrial invasion did not differ from that in tumors with high myometrial invasion (29.7 ± 11.2 pmol/ mg protein). However, using a cut-off level of 15 pmol/ mg protein, the highest level of cathepsin-D was significantly associated with high myometrial invasion (Table 2). None of the low cathepsin-D tumors was invasive, but three of the highest cathepsin-D levels were found in apparently noninvasive tumors (no. 3, 6, and 7; Table 1). One (no. 7) contained an inflammatory stroma with many lymphocytes, plasmocytes, and macrophages, which may have artificially increased the level of cathep-
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 12 November 2015. at 22:53 For personal use only. No other uses without permission. . All rights reserved.
MAUDELONDE ET AL.
118
JCE & M • 1990 Vol70«Nol
TABLE 2. Correlation between myometrial invasion and cellular markers
Markers
Status n°
No. of patients according to invasion
P
Low (0 or +) High (++) RE
+
19 0.37*
6 2
10 1
RP
+
18 0.64"
4 3
6 6
52K cathepsin-D
+
19 0.007
3 5
10 0
c
Qualitative correlation between myometrial invasion (0, +, or ++) and the status of three biochemical parameters. The cut-off level was 10 fmol/mg protein for steroid receptors and 15 pmol/mg protein for cathepsin-D. " Number of endometrial adenocarcinomas studied. *NS. c Statistically significant using Fisher's exact test.
sin-D in the sample (16). Only three patients showed invasion of other tissues (ovary, cervix, and lymph nodes), and no correlation was found with the metastasizing ability of these endometrial cancers. The cathepsin-D concentration was not correlated with RE (r = 0.002; P = 0.98) or RP (r = 0.28; P = 0.25) concentrations. Immunohistochemistry of cathepsin-D Using the monoclonal antibody D7E3, cathepsin-D was not detected by immunohistochemistry in most samples of human endometrium, as previously described (15). However, using a more sensitive monoclonal antibody (M1G8), cathepsin-D was weakly detected in proliferative phase endometrium and was highly concentrated in glandular epithelial cells of luteal phase endometrium (Fig. 2). The staining intensity was lower in stromal cells than in epithelial cells (Fig. 2). To confirm the difference in cathepsin-D accumulation in epithelial and stromal cells, we compared cathepsin-D concentrations in dissociated glandular epithelial cells and in fibroblasts from two luteal phase patients. In both cases cathepsin-D concentration was higher in glandular epithelial cells (mean, 72 pmol/mg protein) than in fibroblasts (mean, 30 pmol/mg protein). The higher values for dissociated endometrial cells than for the total endometrial cytosol (Fig. 1) may simply result from the removal of plasma protein during the digestion step.
was not significantly different in follicular (mean ± SD, 5.11 ± 1.9) or luteal phases (4.75 ± 2.2; Fig. 3). After ovarian stimulation by human menopausal gonadotropin-hCG, progesterone treatment (600 mg/day) did not alter the mean level of plasma cathepsin-D (4.55 ± 1.9). Plasma levels of cathepsin D in the 17th to 18th weeks of pregnancy (6.60 ± 2.1) were slightly but statistically higher than those in patients treated with 600-mg daily doses of progesterone (P < 0.01) and those in the normal follicular phase (P < 0.04), but not those in the normal luteal phase (P < 0.06).
Cathepsin-D concentration in plasma
Hormonal regulation of cathepsin-D in primary culture
The cathepsin-D plasma concentration was stable (from 2.5 to 10 pmol/mL or 0.03 to 0.14 pmol/mg protein) and lower than the endometrial cytosol concentration. It
Cathepsin-D was immunoprecipitated from the cellular lysate of cultured epithelial endometrial cells and analyzed by sodium dodecyl sulfate-polyacrylamide gel
FIG. 2. Immunoperoxidase staining of cathepsin-D in follicular (A) and luteal (B) endometrium. Using the M1G8 monoclonal antibody, frozen sections from endometrial biopsies were fixed, permeabilized, and analyzed using the indirect peroxidase-antiperoxidase technique described in Materials and Methods. In this patient the cytosolic cathepsin-D concentrations were 8 and 17 pmol/mg protein in the follicular and luteal phases, respectively.
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 12 November 2015. at 22:53 For personal use only. No other uses without permission. . All rights reserved.
CATHEPSIN-D IN HUMAN ENDOMETRIUM (9)
(10)
(101
119
MW
(11)
10
« o
7.5
E
'*
*
5.0-
J_
t
•
52 K^
•
V •
+
•
48K 34 K—
•
FOL.
LUT.
Prog.
PREG
FIG. 3. Plasma levels of cathepsin-D in women according to hormonal status. Assay of plasmatic cathepsin-D was performed by a double determinant immunoenzymatic assay as described in Materials and Methods. The follicular (FOL.) and luteal (LUT.) phases of the menstrual cycle were defined according to clinical and histological data. PROG., The group of patients treated for sterility with progesterone (600 mg/day) after the 14th or 15th day of an artificial cycle following successive treatment with the LHRH analogs human menopausal gonadotropin and hCG. Only the group of normal pregnancies (17-18 weeks gestation; PREG) had statistically higher values than the other groups (*, P = 0.01 us. PROG., P = 0.04 us. FOL., P = 0.06 vs. LUT.). The numbers of subjects are in parentheses.
electrophoresis. Two bands were only specifically precipitated by anticathepsin-D antibody. They correspond to the precursor (52K) and the large chain (34K) of mature cathepsin-D. Generally, other bands were nonspecific, as shown by the migration of the nonspecific control. Cell treatment by progestin R5020 increased the 34K and 52K bands (3 times that of the control), and estradiol had no effect. The stimulation induced by R5020 was not inhibited by estradiol (Fig. 4). The 48K nonspecific band was also increased by R5020, suggesting that this band contains at least two proteins, one that is nonspecific and precipitated by the nonrelevant antibody and the other a 48K specific protein corresponding to the one-chain mature form of cathepsin-D (5) which is also increased by progestin. The amount of cathepsin-D secreted into the medium by endometrial cells was not detectable, in contrast with amounts secreted by the MCF7 breast cancer cells analyzed in parallel (not shown).
Discussion There are three major findings in this study. First, the hormonal regulation of cathepsin-D in human normal
NS
7o(52K + 34K) -
C
E2 R5020 R5020
100 102
320
300
FlG. 4. Fluorography of immunoprecipitated cellular cathepsin-D. Epithelial endometrial cells obtained in the luteal phase were cultured for 11 days and treated for 4 days without [control (C)] or with estradiol (E2), R5020, or E2 plus R5020, each at 10 nM. Cathepsin-D was immunoprecipitated using monoclonal antibody D7E3. The same amount of trichloroacetic acid-precipitable 36S-labeled protein was used for controls and the different treatments. Mol wt (MW) were determined from the migration of the following standard proteins: lactoglobin (18,000), carbonic anhydrase (30,000), ovalbumin (46,000), BSA (69,000), and phosphorylase-/? (92,000). 52K, 48K, and 34K correspond to three forms of the immunoprecipitated cathepsin-D. They are also expressed as the percentage of the untreated control sample after quantification as described in Materials and Methods. NS corresponds to the immunoprecipitates obtained with the nonrelevant antibody MOPC-21.
endometrium is mostly due to progesterone, not to estradiol. Second, the cathepsin-D level appears to be higher in adenocarcinoma cells than in normal cells. Third, the low level of cathepsin-D in endometrial cancer appears to be correlated with low myometrial invasion. In normal endometrium, the level of cathepsin-D was higher in the luteal phase. It would, therefore, appear that in normal human endometrium, as reported in the rat uterus (3), cathepsin-D is induced by progesterone, but not by estrogens. This suggests tissue specificity in the regulation of cathepsin-D, since it was previously shown that this lysosomal enzyme is estrogen regulated
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 12 November 2015. at 22:53 For personal use only. No other uses without permission. . All rights reserved.
120
MAUDELONDE ET AL.
in human breast cancer cells (4, 5). Similar cathepsin-D unresponsiveness to estrogens has been demonstrated in the estrogen-responsive Ishikawa endometrial cell line (22). However, it is not yet clear if this hormonal regulation difference is due to tissue specificity (uterine us. mammary cells), the transformation process (normal vs. cancer cells), or both. This question could be resolved by examining the regulation in normal mammary cells containing estrogen receptors and uterine adenocarcinoma cells responding to both progestins and estrogens (22). Examples of tissue specificity include the synthesis of PRL, which is regulated by estrogen in the pituitary but by progesterone in endometrium (17); Moulton and Koenig (18) have also shown that progesterone regulation of cathepsin-D is specific for luminal epithelial cells in
the rat uterus. This cellular specificity of hormonal regulation suggests that cathepsin-D might have a role in the biological and morphological modifications that occur during the luteal phase. Its plasma level is very stable during the menstrual cycle, and in nonpregnant women the contribution of cathepsin-D secreted by steroid-regulated tissue such as endometrium is negligible. The cathepsin-D function in endometrium is unknown. Lysosomal activity increases during the luteal phase, and its role in cyclic menstrual bleeding has been suggested (2). Cathepsin-D is a lysosomal protease and may play a role in the cyclic destructive autophagy and remodeling of endometrium in the absence of nidation. The second finding, based on a limited number of patients, suggests that this marker could be an indicator of endometrial transformation. Since progesterone is absent in all postmenopausal cancer patients, the comparison with normal endometrium in the follicular phase indicates a 3-fold increase in the level of cathepsin-D in adenocarcinoma. Further studies are required, however, since we have not excluded the possibility that the higher levels in adenocarcinoma, compared with levels in normal endometrium in the follicular phase, may result from a higher density of epithelial cells. Moreover, the cathepsin-D level is generally lower in endometrial adenocarcinoma than in breast cancer, and the proportion of procathepsin-D secretion is much lower in Ishikawa adenocarcinoma cell lines (Touitou et al., unpublished) than in breast cancer cell lines. The third finding suggests that the cathepsin-D concentration in endometrial adenocarcinoma may be an additional marker of invasiveness. Cathepsin-D values in the group with low myometrial invasion were more dispersed, and five of the eight samples had low levels, corresponding to follicular phase values. At the 15 pmol/ mg protein cut-off level, which refers to the highest value obtained in the follicular phase, there was a significant correlation between myometrial invasion and cathepsin-
JCE & M • 1990 Vol 70 • No 1
D concentrations. This cut-off level may be adjusted after more patients are studied. Macrophages produce cathepsin-D (19), and the high concentrations in adenocarcinoma without significant myometrial invasiveness may have been due to the high proportion of histiocytes. This was highly probable in the sample with abundant inflammatory cells. Myometrial invasiveness is considered to be the major prognostic parameter of endometrial carcinoma (20), and the cellular concentration of cathepsin-D may be of interest as an associated biochemical marker of invasion. This result is consistent with prospective (6) and retrospective studies (21) of breast cancer patients, indicating that cytosolic cathepsin-D concentrations have a prognostic value independent of other prognostic parameters. We could not draw conclusions on the value of cyto-
solic cathepsin-D as a tumor and prognostic marker because of the low number of patients studied. However, the results are in agreement with the hypothesis that an increased level of cathepsin-D (5) is associated with cell transformation and invasion of surrounding tissues by cancer cells, at least in hormone-dependent tissues.
Acknowledgments We are grateful to Prof. H. Pujol and Dr. D. J. Domergue (Centre Paul Lamarque, Montpellier) for providing endometrial adenocarcinoma samples and Prof. B. Hedon for providing normal endometrial samples. We thank C. Esnault and G. Salazar-Retana for technical help, and E. Barrie and M. Egea for typing the manuscript.
References 1. Schlafke S, Enders AC. Cellular basis of interaction between trophoblast and uterus at implantation. Biol Reprod. 1975;12:41-65. 2. Henzl MR, Smith RE, Boost G, Tyler ET. Lysosomal concept of menstrual bleeding in humans. J Clin Endocrinol Metab. 1972; 34:860-75. 3. Elangovan S, Moulton BC. Progesterone and estrogen control of rates of synthesis of uterine cathepsin D. J Biol Chem. 1980; 255:7474-9. 4. Westley B, Rochefort H. A secreted glycoprotein induced by estrogen in human breast cancer cell lines. Cell. 1980;20:352-62. 5. Rochefort H, Capony F, Garcia M, et al. Estrogen-induced lysosomal proteases secreted by breast cancer cells. A role in carcinogenesis? J Cell Biochem. 1987;35:17-29. 6. Maudelonde T, Khalaf S, Garcia M, et al. Immunoenzymatic assay of Mr 52,000 cathepsin D in 182 breast cancer cytosols. Low correlation with other prognostic parameters. Cancer Res. 1988;48:462-6. 7. Hill GA, Herbert III CM, Parker RA, Wentz AC. Comparison of late luteal phase endometrial biopsies using the novak curette or pipelle endometrial suction curette. Obstet Gynecol. 1989;73:4435. 8. Noyes RV, Herng AT, Rock J. Dating the endometrial biopsy. Fertil Steril. 1950;l:3-25. 9. International Federation of Gynecologists and Obstetricians. Classification and staging of malignant tumors in the female pelvis. Acta Obstet Gynecol Scand. 1971;50:l-12. 10. Morrow CP. Endometrial carcinoma stages I and II: is surgery adequate? Int J Radiation Oncol Biol Physiol. 1980;6:365-6. 11. Rogier H, Freiss G, Besse MG, et al. Two-site immunoenzymometric assay for the 52-kDa cathepsin D in cytosols of breast cancer
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 12 November 2015. at 22:53 For personal use only. No other uses without permission. . All rights reserved.
CATHEPSIN-D IN HUMAN ENDOMETRIUM tissues. Clin Chem. 1989;35:81-5. 12. Nolan C, Przywara LW, Miller LS, Suduikis VBS, Tomita JT. A sensitive solid-phase enzyme immunoassay for human estrogen receptor. In: Ames FC, Blumenschein GR, Montague ED, eds. Current controversies in breast cancer. Austin: University of Texas Press; 1984;433-41. 13. Greene GL, Harris K, Bovo R, Nolan C. Preparation and characterization of monoclonal antibodies to human progesterone receptor. Mol Endocrinol. 1988;8:714-26. 14. Maudelonde T, Rochefort H. A 51K progestin-regulated secreted by human endometrial cells in primary culture. J Clin Endocrinol Metab 1987;64:1294-1302. 15. Morisset M, Capony F, Rochefort H. Processing and estrogen regulation of the 52-kDa protein inside MCF7 breast cancer cells. Endocrinology. 1986;119:2773-83. 16. Garcia M, Salazar-Retana G, Richer G, et al. Immunohistochemical detection of the estrogen-regulated Mr 52,000 protein in primary breast cancers but not in normal breast and uterus. J Clin Endocrinol Metab. 1984;59:564-6.
121
17. Huang JR, Tseng L, Bischof P, Janne OA. Regulation of prolactin by progestin, estrogen and relaxin in human endometrial stromal cells. Endocrinology. 1987;121:2011-7. 18. Moulton BC, Koenig BB. Progestin increases cathepsin D synthesis in uterus luminal epithelial cells. Am J Physiol. 1983;244:E442-6. 19. Imort M, Ziihlsdorf M, Feige U, Hasilik H, Von Figura K. Biosynthesis and transport of lysosomal enzymes in human monocytes and macrophages. Biochem J. 1983;214:671-8. 20. Hendrickson M, Ross J, Eifel RJ, et al. Adenocarcinoma of the endometrium. Analysis of 256 cases with carcinoma limited to the uterine corpus. Gynecol Oncol. 1982;13:373-7. 21. Thorpe SM, Rochefort H, Garcia M, Freiss G, Christensen J, Khala FS, Paolucci F, Pau B, Rasmussen BB, Rose C. Association between high concentrations of 52K cathepsin-D and poor prognosis in primary breast cancer. Cancer Res. Nov. 1989, in press 22. Touitou I, Cavailles V, Garcia M, DeFrenne A, Rochefort H. Differential regulation of 52K cathepsin D by estradiol and progesterone in mammary cancer and uterine cells Mol Cell Endocrinol. 1989, in press.
International Symposium on Thyroperoxidase Fundamental and Clinical Aspects Marseille, France" June 28-30, 1990 Scientific committee: G.F. Bottazzo (London), L. Braverman (Worcester), P. Carayon (Marseille), B. Czarnocka (Warsaw), J. De Vijlder (Amsterdam), R. Mornex (Lyon), A. Pinchera (Pisa), B. Rousset (Lyon), J. Ruf (Marseille), P. Scriba (Liibeck), G. Vassart (Brussels) An international symposium on recent advances in the implication of thyroperoxydase (TPO) in autoimmune disease of the thyroid gland. Topics include: structure and expression of the TPO gene, TPO structure, catalytic mechanisms, structure-activity relationship, hormonal regulation of TPO, traffic of TPO, congenital and acquired deficiencies of iodide organification, cellular and humoral mechanisms of autoimmune disease, anti-TPO antibodies in various thyroid and autoimmune diseases, TPO and antiTPO assays, immunohistochemistry, in-situ hybridization. The program will include lectures by invited speakers, selected oral communications and poster presentations. For further information please contact: Dr. Pierre Carayon, Laboratoire des Hormones Proteiques, Faculte de Medecine 27 bd Jean Moulin 13385 Marseille Cedex 5, France Tel: 91.78.68.55, Fax: 91.25.64.07
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 12 November 2015. at 22:53 For personal use only. No other uses without permission. . All rights reserved.
