0021-972X/78/4702-0326$02.00/0 Journal of Clinical Endocrinology and Metabolism Copyright © 1978 by The Endocrine Society
Vol. 47, No. 2 Printed in U.S.A.
Properties of Human Chorionic Gonadotropin Produced in Vitro by Ovarian Carcinoma Cells* GLENN D. BRAUNSTEIN, VIKRAM V. KAMDAR, JAN KANABUS, AND JOAN RASOR Departments of Medicine and Obstetrics and Gynecology, Cedars-Sinai Medical Center and UCLA School of Medicine; and the Department of Obstetrics and Gynecology (J.K.), University of Southern California School of Medicine, Los Angeles, California 90033 a-D-methylglucoside. The biological activity of the ectopic hCG was 9289 IU/mg, as determined by the ventral prostate weight (VPW) method in hypophysectomized immature male rats. The biological to immunological ratios by the ventral prostate weight method and RRA were 1.79 and 2.17, respectively. The in vivo disappearance rate of ectopic hCG after injection into immature female rats was significantly faster than that of placental or urinary hCG, but was considerably slower than the disappearance rate of human LH. These studies demonstrate that the immunoreactive and biologically active portions of the hCG produced by the ovarian adenocarcinoma cell line and native hCG are similar or identical. The faster disappearance rate of the ectopic hCG in the rat model may be due to incomplete sialylation of the oligosaccharide moiety of the hCG molecule. (JClin Endocrinol Metab 47: 326, 1978)
ABSTRACT. The present study was designed to compare the immunological, physical, and biological properties of native hCG with an hCG molecule secreted ectopically in vitro by an ovarian adenocarcinoma cell line maintained in long term tissue culture. The hCG produced by the cell line was concentrated by ultrafiltration of the tissue culture medium. The inhibition curves generated by serial dilutions of the culture medium concentrates were parallel to those obtained with purified urinary hCG in the /?-hCG RIA and the rat Leydig cell radioreceptor assay (RRA). The ectopic hCG also reacted with an antibody generated against the carboxyl-terminal peptide (109-145) of 0-hCG. The immunoreactive material cochromatographed with urinary hCG on a Sephadex G-100 column, as determined by the /?-hCG RIA and RRA. Neither free a nor free 0 subunits were found in the tissue culture medium. The tissue culture gonadotropin was adsorbed onto a Concanavalin A-Sepharose column and could be eluted with
S
EVERAL recent studies have demonstrated that approximately 16% of patients with a wide variety of malignant nontrophoblastic neoplasms have measurable quantities of immunoreactive hCG present in their serum or plasma (1). The quantities of the hormone in the circulation of the majority of these patients with ectopic hCG production have generally been quite small. The difficulties in obtaining sufficient amounts of the ectopically produced hormone from the body fluids or tumors of these patients probably accounts for the paucity of immunological, biological, and physical studies performed on Received June 20, 1977. Address requests for reprints to: Glenn D. Braunstein, M.D., Room 6736, South Tower, Department of Medicine, Cedars-Sinai Medical Center, 8700 Beverly Boulevard, Los Angeles, California 90048. * This work was supported in part by the Robert E. and Mary R. Wright Foundation, through the University of Southern California Medical School, and USPHS Grant CA-18362-02.
the hCG of neoplastic origin. The development of techniques for the establishment and maintenance of long term in vitro cultures of tumors provides a potential source of relatively large amounts of humoral products from these tumors. The ectopic production of hCG or its subunits by nontrophoblastic tumors in vitro has previously been described for cell lines derived from a hepatoblastoma (2), a bronchogenic carcinoma (3-5), and cervical carcinomas (6-9). The present report describes the physical and biological characteristics of the hCG produced in vitro by an ovarian papillary cystadenocarcinoma which has been maintained in long term tissue culture. Materials and Methods Cell cultures The ovarian carcinoma cell line, designated 163, was established in June, 1972, from cells obtained
326
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 14 November 2015. at 16:43 For personal use only. No other uses without permission. . All rights reserved.
PROPERTIES OF hCG FROM OVARIAN CARCINOMA from the ascitic fluid of a 45-yr-old woman with a papillary cystadenocarcinoma of the ovary. The cells have a modal chromosome number of 66. This line has a type B glucose-6-phosphate dehydrogenase isoenzyme mobility, distinguishing it from HeLa cells which have a type A isoenzyme mobility (10). For the present study, the cells were grown in multiple culture flasks containing Ham's F-10 medium (Flow Laboratories, Rockville, MD) supplemented with 20% fetal bovine serum, penicillin (100 U/ml), streptomycin (100 /ig/ml), and kanamycin (50 jug/ml). They were maintained in a humidified atmosphere of 5% CO2 and 95% air at 37 C. The medium was decanted and centrifuged at 500 X g to eliminate detached cells, and the supernatant was frozen at —20 C until further study. Concentration method The tissue culture medium was ultrafiltered through an Amicon UM-100 A membrane (Amicon Corporation, Lexington, MA) and the filtrate was lyophilized and reconstituted in distilled water to approximately 5% of the starting volume. Recovery by this method of hCG and the free a and /? subunits of hCG after addition of purified materials to tissue culture medium was 72, 72, and 63%, y respectively. Radioligand assays
i
Three separate RIAs were used to measure hCG. The double antibody hCG RIA, which does not discriminate between hCG and human LH (hLH) (hCG/hLH RIA), and the /?-hCG RIA, which is relatively specific for the /? subunit of hCG and intact hCG, have been previously described (11). The reagents used for these assays were obtained rom the Hormone Distribution Officer, NIAMDD, 4IH, Bethesda, MD. The highly purified preparaion of hCG (CR 117) was used as the labeled ligand lid reference standard for both assays. This maerial contained 5200 IU/mg immunological activity nd 10,600 IU/mg biological activity relative to the •econd International Standard for hCG. A recently eveloped, double antibody RIA for hCG which utilizes an antiserum generated against the carboxyl-terminal peptide (amino acids 109-145) of the /? subunit of hCG was used to further examine the immunological identity of the ectopic hCG. This assay was developed and performed by Dr. Albert Parlow and the methodological details of the assay will be described in a separate publication by Dr. Parlow. There is less than 0.001% crossreaction with equimolar amounts of highly purified pituitary glycoprotein hormones (Parlow, unpub-
327
lished observations). The reference standard used in the carboxyl-terminal /MiCG RIA was a purified preparation of urinary hCG, supplied by Dr. O. P. Bahl, containing 6000 IU immunological activity/mg relative to the Second International Standard for hCG. The a subunit of hCG was measured by a homologous RIA, by the method of Vaitukaitis and Ross (12), with reagents provided by the Hormone Distribution Officer, NIAMDD, NIH, Bethesda, MD. The Leydig cell radioreceptor assay (RRA) method of Catt et al. (13) was used to measure gonadotropin-binding activity. The hCG standard (CR 117) had a specific activity of 10,000 IU/mg in the RRA relative to the Second International Standard for hCG. The potency estimate for the tissue culture hCG in the RRA was based on the weight of immunoreactive hCG, as determined by the /? subunit RIA using CR 117 (5200 IU/mg) as the reference standard. Bioassay The ventral prostate weight assay was used to measure the in vitro biological activity of hCG (14). Immature hypophysectomized Sprague-Dawley rats (Hilltop Laboratories, Bilroy, CA), weighing 40-50 g each, were injected with the test materials beginning 3 days after hypophysectomy. All dilutions of standard or tissue culture extracts were made with 1% bovine serum albumin in isotonic saline and the total volume injected over 4 days was 4 ml. Four or five rats were used at each dose level. Autopsy was performed on the morning of the 5th day and the ventral lobe of the prostate was removed and weighed on a torsion balance. Three assays were performed with three or four dilutions of the tissue culture extracts and four to seven doses of the Second International Standard for hCG in each assay. The potency estimates for the tissue culture hCG were based on the weight of immunoreactive hCG, as determined by the fi subunit RIA using CR 117 (5200 IU/mg) as the reference standard. Chromatography A 1.5 X 85.0-cm column of Sephadex G-100 (Pharmacia Laboratories, Piscataway, NJ) equilibrated with 0.05 M Tris-hydrochloride-0.15 M sodium chloride buffer, pH 7.4, was used to separate hCG, hLH, and the a and fi subunits of hCG. Onemilliliter samples of test or reference materials were applied to the column and 1-ml aliquots of the effluent were collected. The volume of the column
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 14 November 2015. at 16:43 For personal use only. No other uses without permission. . All rights reserved.
328
JCE&M • 1978 Vol47 • No2
BRAUNSTEIN ET AL.
was determined by the peak elution volume of blue dextran, measured spectrophotometrically at 600 nm. The column was calibrated with [125I]hCG, purified hCG (CR 117), pregnancy serum, term placental tissue extract, [125I]hCG /? subunit, purified hCG /? subunit (CR 115-/3), purified hCG a subunit (CR 115-a), [125I]hLH, purified hLH (LER 960), and postmenopausal serum. The volume of exclusion for each hormone corresponded to the tube with the highest counts per min of 125I for the radioiodinated hormones or with the highest concentration of immunoreactive hormone, as determined by RIA. Because of the heterogeneity of the hormones used to calibrate the column, elution ranges were established for hCG, the /? subunit of hCG, and hLH. Disposable columns (BioRad Laboratories, Richmond, CA), 0.7 X 4.0 cm., containing 1.5 ml Concanavalin A (Con A) covalently coupled to Sepharose 4B (Pharmacia Laboratories, Piscataway, NJ) were extensively washed with phosphate-buffered saline, pH 7.8, containing 2.5% normal rabbit serum. One-milliliter aliquots of pregnancy serum, [125I]hCG, or purified hCG dissolved in fresh tissue culture medium and 1 ml concentrated tissue culture medium from cell line 163 were then filtered through the columns with the use of phosphatebuffered saline with normal rabbit serum for elution of unadsorbed substances. Elution of Con A-adsorbed glycoproteins was accomplished with 0.2 M a-D-methylglucoside (15). The eluted fractions were collected in 0.5-ml aliquots and analyzed in the /?hCG RIA. Disappearance studies
The logarithm of the hormone concentration in the rat serum was plotted against time and then extrapolated back to the ordinate (zero time). These results were then plotted in terms of a percentage of zero time concentrations. Each disappearance curve was best fit by a biexponential model and the two components, designated fast and slow, were each fit by the method of least squares and half-lives (ti^) in serum were determined. Statistics The statistical methods of Rodbard and Frazier (16) were used for dose interpolation in the radioligand assays. Bioassay dose interpolation and tests for parallelism between standard and test materials were performed according to the bioassay statistical program developed by Faden and Rodbard (17). The fast and slow components of the disappearance curves for hCG, hLH, and tissue culture hCG were compared to each other by analysis of covariance (18). A significant difference between disappearance curves was considered to be present if P < 0.05.
Results Serial dilution of the concentrated material from cell line 163 tissue culture medium exhibited parallelism with the purified urinary hCG reference standard in the /8-hCG RIA (Fig. 1) and in the RIA which specifically detects the 109-145 carboxyl-terminal portion 90 80 -
A comparison of the disappearance rates of the tissue culture hCG, purified urinary hCG (CR117), the hCG extracted from term placentas, purified pituitary hLH (LER 960), and the hLH present in a pituitary extract was carried out in 22-day-old female Sprague-Dawley rats (Hilltop Laboratories, Gilroy, CA), weighing 50-60 g. The test material was injected via the tail vein. At multiple points in time, three to five animals were bled and the serum was separated and frozen until assay. The /?-hCG RIA was used to measure the hCG concentration in the serum from rats injected with the tissue culture extracts, urinary hCG, and placental hCG, while the nonspecific hCG/hLH RIA was used to measure the immunoreactive hLH in the rat serum after injection of the pituitary hLH or pituitary extracts. [125I]hLH (LER 960) was used as a labeled ligand and LER 907 was used for reference purposes for the hLH measurements.
60 40 -
TISSUE CULTURE EXTRACTS
2010-
hCG STANDARD (CR 117)
4 2 -
hCG
NANOGRAMS/TUBE
TISSUE CULTURE EXTRACTS
MICROLITERS/TUBE
FIG. 1. Dose-response curves for the hCG reference preparation (CR 117) and serial dilutions of culture extracts in the /?-hCG RIA. B, Counts bound in the presence of labeled and unlabeled ligand; Bo, counts bound in the presence of labeled ligand alone; Logit (B/Bo), loge [B/Bo/(1 - B/Bo)]. Coordinates are Logit normalized percentage counts precipitated (B/Bo X 100) (on the ordinate) and log of mass of hCG standard or volume of tissue culture extract added per tube (on the abscissa).
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 14 November 2015. at 16:43 For personal use only. No other uses without permission. . All rights reserved.
PROPERTIES OF hCG FROM OVARIAN CARCINOMA
329
99-1
FIG. 2. Dose-response curves for the purified hCG preparation (CR 117), the Second International Standard for hCG, and serial dilutions of tissue culture extracts in the rat Leydig cell RRA. CR 117 is expressed in terms of nanograms per tube, while the Second International Standard for hCG is depicted as International Units per tube. The other abbreviations are described in Fig. 1.
2nd INTERNATIONAL STANDARD hCG hCG (CR117)
TISSUE CULTURE EXTRACT
1.0
10
10
100
hCG IU or NANOGRAMS/TUBE 0.1
1.0
TISSUE CULTURE EXTRACTS
of the /? subunit of hCG. These results suggest that the material secreted by the ovarian adenocarcinoma cell line and urinary hCG are immunologically similar. In the rat Leydig cell RRA, serial dilutions of the tissue culture concentrates were parallel to the purified hCG standard and the Second International Standard for hCG (Fig. 2). The ratio of radioreceptor activity to immunological activity for the hCG standard (CR 117) was 2.37, while the ratio for the tissue culture extract was 2.17. Control fresh tissue culture medium concentrated by identical procedures, as described for the cell line 163 media, demonstrated no activity in either the /?-hCG RIA or RRA. Parallelism was also demonstrated between the dose-response curves generated by the tissue culture extracts and that produced by the Second International Standard for hCG in the ventral prostate weight bioassay (Fig. 3). The mean potency estimate of the tissue culture hCG in three assays was 9,289 IU/mg (95% confidence limits, 7,909-10,669). The ratio of biological to immunological activity was 1.79 for the tissue culture hCG compared to 2.37 for the highly purified urinary hCG reference preparation. The tissue culture gonadotropin eluted from the Sephadex G-100 column within the established hCG region (Fig. 4). No qualitative differences were found when the hCG in the
£ so i
MICROLITERS/TUBE
2nd INTERNATIONAL STANDARD hCG
hCG
TISSUE CULTURE EXTRACT
IU INJECTED
TISSUE CULTURE EXTRACT
100 ul INJECTED
FIG. 3. Dose-response curves for the Second International Standard for hCG and serial dilutions of tissue culture extracts in the ventral prostate weight bioassay. The ventral prostate weight in milligrams is depicted on the ordinate, while the log of IU of the Second International Standard for hCG or volume of tissue culture extracts injected per rat is shown on the abscissa.
effluent tubes was measured by the nonspecific hCG/hLH RIA, the /?-hCG RIA, or the RRA. No free a subunit of hCG was detected in the column effluent by the homologous ahCG RIA. As the ultrafiltration method does not selectively concentrate hCG over its subunits, it is unlikely that significant amounts of free a or /? subunits were secreted into the medium. The immunoreactive material in the tissue culture extracts was adsorbed onto Con A and was eluted with a-D-methylglucoside with a
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 14 November 2015. at 16:43 For personal use only. No other uses without permission. . All rights reserved.
330
BRAUNSTEIN ET AL. Vo
hCG /3 hLH a
i
mziDczii
JCE&M • 1978 Vol47 • No 2
hCG/hLH R1A
hCG TISSUE CULTURE
60120 240 360 MINUTES AFTER INJECTION
5040-
hCG/hLH RRA
302010-
0
10
20
30
40
SO
60
70
80
90
100
ELUTION VOLUME ml
FIG. 4. Sephadex G-100 elution profile of tissue culture extracts expressed in terms of nanograms per hCG, using CR 117 as a reference standard. The void volume (Vo), peak hCG, /?-hCG, hLH elution ranges, and the region of peak a subunit elution are depicted on the top of the graph. The hCG/hLH and the /?-hCG RIAs were performed on aliquots from the effluent from a single chromatographical run, while the hCG/hLH RRA was performed on the effluent tubes from a separate chromatographical procedure in which a greater concentration of the tissue culture extract was applied to the column.
recovery of 94%. Recovery of the hCG present in pregnancy serum, [125I]hCG, or purified hCG added to fresh tissue culture medium and treated in identical fashion was 91.8% ± 2.9 (SEM).
The disappearance curves of the purified urinary hCG, the hCG present in placental extracts, purified pituitary hLH, the LH present in extracts of the pituitary, and the hCG found in the tissue culture medium are illustrated in Fig. 5. There were no significant differences between either the fast or slow components of the purified hCG (tested on two occasions) and the hCG in the placental extract and, therefore, the data from three experiments were combined. The ti /2 of the fast component of hCG was 65.4 min, while
FIG. 5. Disappearance of hCG, tissue culture hCG (TC), and hLH from the plasma of immature female rats after iv injection. Each symbol represents the mean of the percentage of the zero time concentration vs. time derived from: 1) two separate experiments with unlabeled hCG (CR 117) and the hCG in a placental extract (•); 2) two separate experiments with the tissue culture extracts (O); and 3) purified hLH (LER 960) and the hLH present in a pituitary extract (A).
that for the slow component was 143.3 min. The individual disappearance curves for the purified pituitary hLH and the hLH from the crude pituitary extract were identical, with a ti/2 of 21.1 and 177.1 min for the respective components. The results from two separate studies of the disappearance of tissue culture hCG from the serum of the rats were similar and, therefore, the data was combined. The ti/2 of the fast and slow components for the tissue culture hCG was 37.6 and 250.8 min, respectively. The analysis of covariance of the disappearance curves of each substance indicated that hLH left the plasma compartment significantly faster than hCG or the tissue culture hCG and that the tissue culture hCG disappeared from the serum more rapidly than did urinary or placental hCG but at a significantly slower rate than hLH. Discussion Although a considerable amount of information has accumulated concerning the biological, immunological, and structural properties of hCG from the urine of pregnant
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 14 November 2015. at 16:43 For personal use only. No other uses without permission. . All rights reserved.
PROPERTIES OF hCG FROM OVARIAN CARCINOMA women and patients with hCG-producing trophoblastic tumors (eutopic secretion) (19, 20), » only a few studies have been carried out on the hCG secreted by nontrophoblastic neoplasms (ectopic secretion) (21, 22). Vaitukaitis examined the plasma, urine, and tumor extracts of patients with hCG-secreting trophoblastic and nontrophoblastic neoplasms » and found intact hCG as well as free subunits and altered molecular forms of these substances (21). Similarly, Weintraub and coworkers found differences between the a subx units of the naturally occurring glycoprotein hormones and the purified a subunits derived ^from two nontrophoblastic neoplasms (22). Both groups of investigators have raised the possibility that altered protein synthesis may be a concomitant of neoplasia, although the described alterations in hormone synthesis *" may reflect a quantitative difference in the ^-secretion of naturally occurring isohormones of hCG or its subunits (23-28). The hCG secreted by the ovarian adenocarcinoma cell line 163 clearly resembles urinary hCG by several immunological, biological, and r physical criteria. In particular, the results of > the Con A chromatography indicate the glycoprotein nature of the tissue culture gonadotropin (15) and the results from the /?-hCG RIA, RRA, and ventral prostate weight assays demonstrate the close immunological and biological similarity of these species. Because of • the small degree of cross-reactivity between hLH and hCG in the /?-hCG RIA and the * similar reactivity of these hormones in the RRA and ventral prostate weight bioassay, it was conceivable that the activity in the above assays noted with the tissue culture media t extracts could be due to the secretion of hLH. The prime evidence that the material pro* duced by the cell line is hCG and not hLH is y provided by the column chromatographical elution pattern of the tissue culture material which resembles that of hCG and not hLH, ^ and the parallelism between the tissue culture hCG and native hCG in the 109-145 carboxyl" ^terminal /?-hCG RIA. The sequence of the first 115 amino acids of the ft subunits of hCG and hLH demonstrates significant homology (29). However, the ft subunit of hCG contains
331
30 additional carboxyl-terminal amino acids in excess of hLH, which confer immunological specificity to hCG (29). As the tissue culture hCG is recognized by an antiserum to this unique peptide, the immunological identity of the tissue culture gonadotropin as hCG would seem to be well established. However, there are some differences in biological behavior between the tissue culture hCG and the predominant molecular species of hCG present in the highly purified hCG (CR 117) used in these studies. The tissue culture hCG contains approximately 88% of the biological activity of CR 117 and disappears from the serum after injection into intact immature female rats significantly faster than does CR 117. Both of these observations may be due to a decrease in the content of the sialic acid moiety of the tissue culture hCG relative to the purified urinary hCG. Van Hall et al. (30) have previously demonstrated that stepwise desialylation of hCG leads to a progressive reduction of hCG biological activity in the ventral prostate weight assay with maintenance of immunological activity, resulting in a progressive decrease in the biological to immunological ratio. These workers also showed that the half-life of the initial component of the plasma disappearance curve after injection in rats decreased concurrently with the reduction of the sialic acid content of the molecule (31). As much of the well known heterogeneity of hCG found in the urine of pregnant women seems to reside in differences in the sialic acid content of the hCG molecule (27, 28), it is reasonable to expect that similar differences may be found in hCG molecules secreted by tumors. That tumors may secrete hCG molecules with varying amounts of sialic acid and, hence, varying biological to immunological ratios has previously been shown for clonal cell lines of gestational choriocarcinoma (32, 33). Although the hCG secreted by the ovarian adenocarcinoma cell line resembles an incompletely sialylated form of hCG, further purification and amino acid and carbohydrate analysis of the hCG secreted by cell line 163 will be needed to establish with certainty the structural identity of the ectopic hCG to a naturally occurring isohormone of hCG.
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 14 November 2015. at 16:43 For personal use only. No other uses without permission. . All rights reserved.
BRAUNSTEIN ET AL.
332
Acknowledgments We gratefully acknowledge Dr. Albert Parlow for the performance of the carboxyl-terminal hCG RIA. The glucose-6-phosphate dehydrogenase isoenzyme mobility of the cell line was kindly performed by Dr. Robert Sparks. The technical assistance of Mr. G. V. Scarboro and G. Firestein, the secretarial assistance of Ms. M. C. Brill, and the statistical analyses performed by Mr. P. Schroth are greatly appreciated.
References 1. BRAUNSTEIN, G. D., Use of human chorionic gonadotropin as a tumor marker in cancer, In Herberman, R. B., and K. R. Mclntire, (eds.), Immunodiagnosis of Cancer, New York, Marcel Dekker, 1978. 2. BRAUNSTEIN, G. D., W. E. BRIDSON, A. GLASS, E. W. HULL,
AND K. R. MCINTIRE, In vivo and in vitro production of human chorionic gonadotropin and a-fetoprotein by a virilizing hepatoblastoma, J Clin Endocrinol Metab 35: 857, 1972. 3. RABSON, A. S., S. W. ROSE, A. H. TASHJIAN, JR., AND B. D.
WEINTRAUB, Production of human chorionic gonadotropin in vitro by a cell line derived from a carcinoma of the lung, J Natl Cancer Inst 50: 669, 1973. 4. TASHJIAN, A. H., JR., B. D. WEINTRAUB, N. J. BAROWSKY, A.
S. RABSON, AND S. W. ROSEN, Subunits of human chorionic gonadotropin: unbalanced synthesis and secretion by clonal cell strains derived from a bronchogenic carcinoma, Proc Natl Acad Sci 70: 1419, 1973. 5. LIEBLICH, J. M., B. D. WEINTRAUB, G. H. KRAUTH, P. 0. KOHLER, A. S. RABSON, AND S. W. ROSEN, Ectopic and
eutopic secretion of chorionic gonadotropin and its subunits in vitro: comparison of clonal strains from carcinomas of lung and placenta, J Natl Cancer Inst 56: 911,1976. 6. GHOSH, N. K., AND R. P. Cox, Production of human chorionic gonadotropin in HeLa cell cultures, Nature 259: 416, 1976. 7. LIEBLICH, J. M., B. D. WEINTRAUB, S. W. ROSEN, J. Y. CHOU,
AND J. C. ROBINSON, HeLa cells secrete a-subunit of glycoprotein trophic hormones, Nature 260: 530, 1976. 8. LIEBLICH, J. M., B. D. WEINTRAUB, S. W. ROSEN, N. K.
GHOSH, AND R. P. Cox, Secretion of hCG-asubunit and hCG by HeLa strains, Nature 265: 746, 1977. 9. PATILLO, R. A., R. 0. HUSSA, M. T. STORY, A. C. F. RUCKERT,
M. R. SHALABY, AND R. F. MATTINGLY, Tumor antigen and human chorionic gonadotropin in CaSKi cells: a new epidermoid cervical cancer cell line, Science 196: 1456, 1977. 10. NELSON-REES, W. A., AND R. R. FLANDERMEYER, HeLa cul-
tures defined, Science 161: 96, 1976. 11. VAITUKAITIS, J. L., G. D. BRAUNSTEIN, AND G. T. ROSS,
Radioimmunoassay which specifically measures human chorionic gonadotropin in the presence of human luteinizing hormone, Am J Obstet Gynecol 113: 731, 1972. 12. VAITUKAITIS, J. L., AND G. T. ROSS, Immunologic cross-reactivity among the human glycoprotein hormones and their subunits, In Saxena, B. B., C. G. Beling, and H. M. Gandy (eds.), Gonadotropins, New York, John Wiley and Sons, 1972, p. 435. 13. CATT, K. J., M. L. DUFAU, AND T. TSURUHARA, Studies on a
radioligand, receptor assay system for luteinizing hormone and chorionic gonadotropin, J Clin Endocrinol Metab 32: 860, 1971. 14. MCARTHUR, J. W., Identification of pituitary interstitial cell stimulating hormone in human urine, Endocrinology 50: 304, 1952.
JCE & M Vol47
1978 No 2
15. DUFAU, M. L., T. TSURUHARA, AND K. J. CATT, Interaction of
„,
glycoprotein hormones with agarose-Concanavalin A, ' Biochim Biophys Ada 278: 281, 1972. 16. RODBARD, D., AND G. R. FRAZIER, Radioimmunoassay Data *• Processing, ed. 2, Springfield, VA, National Technical Information Service, 1973. 17. FADEN, V. B., AND D. RODBARD, Bioprog. Bioassay Data Processing, Bethesda, NICHHD, NIH, 1975. 18. DIXON, W. J., BMD-PIV, Biomedical Computer Program, P Series, Los Angeles, University of California Los Angeles : Press, 1975, p. 683. 19. CANFIELD, R. E., F. J. MORGAN, S. KAMMERMAN, J. J. BELL,
AND G. M. AGOSTO, Studies of human chorionic gonadotropin, « Recent Prog Horm Res 27: 121, 1971. 20. VAITUKAITIS, J. L., G. T. Ross, G. D. BRAUNSTEIN, AND P. L.
RAYFORD, Gonadotropins and their subunits: basic and clinical studies, Recent Prog Horm Res 32: 289, 1976. 21. VAITUKAITIS, J. L., Immunologic and physical characterization of human chorionic gonadotropin (hCG) secreted by tumors, J Clin Endocrinol Metab 37: 505, 1973. 22. WEINTRAUB, B. D., G. KRAUTH, S. W. ROSEN, AND A. S.
RABSON, Differences between purified ectopic and normalir alpha subunits of human glycoprotein hormones, J Clin Invest 56: 1043, 1975. 23. WEINTRAUB, B. D., B. S. STANNARD, AND S. W. ROSEN,
Combination of ectopic and standard human glycoprotein hormone alpha with beta subunits: discordance of immunologic and receptor-binding activity, Endocrinology 101: 225, 1977. ** 24. LANDEFELD, T., S. BOGUSLAWSKI, L. CORASH, AND I. BOIME,
The cell-free synthesis of the alpha subunit of human cho--^ rionic gonadotropin, Endocrinology 98: 1220, 1976. 25. VAITUKAITIS, J. L., Changing placental concentrations of human chorionic gonadotropin and its subunits during gestation, J Clin Endocrinol Metab 38: 755, 1974. 26. MARUO, T., Y. ASHITAKA, M. MOCHIZUKI, AND S. TOJO,
Chorionic gonadotropin synthesized in cultivated trophoblast, Endocrinol Jap 21: 499, 1974.
t
27. VAN HELL, H., R. MATTHIJSEN, AND J. D. H. HOMAN, Studies
on human chorionic gonadotrophin. I. Purification and some < physicochemical properties, Ada Endocrinol (Kbh) 59: 89, 1968. 28. SCHUURS, A. H., W. M. E. DE JAGER, AND J. D. H. HOMAN,
Studies on human chorionic gonadotrophin. III. Immunochemical characterisation, Ada Endocrinol (Kbh) 59: 120, 1968. 29. MORGAN, F. J., S. BIRKEN, AND R. E. CANFIELD, Comparison
of chorionic gonadotropin and luteinizing hormone: a note on a proposed significant structural difference in the beta sub- -4 unit, FEBS Lett 31: 101, 1973. 30. VAN HALL, E. V., J. L. VAITUKAITIS, G. T. ROSS, J. W.
HICKMAN, AND G. ASHWELL, Immunologic and biologic activity of HCG following progressive desialylation, Endocrinology 88: 456, 1971. 31. VAN HALL, E. V., J. L. VAITUKAITIS, G. T. ROSS, J. W.
HICKMAN, AND G. ASHWELL, Effects of progressive desialylation on the rate of disappearance of immunoreactive HCG ^ from plasma in rats, Endocrinology 89: 11, 1971. 32. BRIDSON, W. E., G. T. ROSS, AND P. 0. KOHLER, Immunologic
and biologic activity of chorionic gonadotropin synthesized by cloned choriocarcinoma cells in culture, J Clin Endocrinol Metab 33: 145, 1971. 33. HAMMOND, J. M., W. E. BRIDSON, P. O. KOHLER, AND A.
CHRAMBACH, Physical characteristics of immunoreactive chorionic gonadotropin produced in culture, Endocrinology 89: 801, 1971.
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 14 November 2015. at 16:43 For personal use only. No other uses without permission. . All rights reserved.
i
-
Vol. 47, No. 2 Printed in U.S.A.
Properties of Human Chorionic Gonadotropin Produced in Vitro by Ovarian Carcinoma Cells* GLENN D. BRAUNSTEIN, VIKRAM V. KAMDAR, JAN KANABUS, AND JOAN RASOR Departments of Medicine and Obstetrics and Gynecology, Cedars-Sinai Medical Center and UCLA School of Medicine; and the Department of Obstetrics and Gynecology (J.K.), University of Southern California School of Medicine, Los Angeles, California 90033 a-D-methylglucoside. The biological activity of the ectopic hCG was 9289 IU/mg, as determined by the ventral prostate weight (VPW) method in hypophysectomized immature male rats. The biological to immunological ratios by the ventral prostate weight method and RRA were 1.79 and 2.17, respectively. The in vivo disappearance rate of ectopic hCG after injection into immature female rats was significantly faster than that of placental or urinary hCG, but was considerably slower than the disappearance rate of human LH. These studies demonstrate that the immunoreactive and biologically active portions of the hCG produced by the ovarian adenocarcinoma cell line and native hCG are similar or identical. The faster disappearance rate of the ectopic hCG in the rat model may be due to incomplete sialylation of the oligosaccharide moiety of the hCG molecule. (JClin Endocrinol Metab 47: 326, 1978)
ABSTRACT. The present study was designed to compare the immunological, physical, and biological properties of native hCG with an hCG molecule secreted ectopically in vitro by an ovarian adenocarcinoma cell line maintained in long term tissue culture. The hCG produced by the cell line was concentrated by ultrafiltration of the tissue culture medium. The inhibition curves generated by serial dilutions of the culture medium concentrates were parallel to those obtained with purified urinary hCG in the /?-hCG RIA and the rat Leydig cell radioreceptor assay (RRA). The ectopic hCG also reacted with an antibody generated against the carboxyl-terminal peptide (109-145) of 0-hCG. The immunoreactive material cochromatographed with urinary hCG on a Sephadex G-100 column, as determined by the /?-hCG RIA and RRA. Neither free a nor free 0 subunits were found in the tissue culture medium. The tissue culture gonadotropin was adsorbed onto a Concanavalin A-Sepharose column and could be eluted with
S
EVERAL recent studies have demonstrated that approximately 16% of patients with a wide variety of malignant nontrophoblastic neoplasms have measurable quantities of immunoreactive hCG present in their serum or plasma (1). The quantities of the hormone in the circulation of the majority of these patients with ectopic hCG production have generally been quite small. The difficulties in obtaining sufficient amounts of the ectopically produced hormone from the body fluids or tumors of these patients probably accounts for the paucity of immunological, biological, and physical studies performed on Received June 20, 1977. Address requests for reprints to: Glenn D. Braunstein, M.D., Room 6736, South Tower, Department of Medicine, Cedars-Sinai Medical Center, 8700 Beverly Boulevard, Los Angeles, California 90048. * This work was supported in part by the Robert E. and Mary R. Wright Foundation, through the University of Southern California Medical School, and USPHS Grant CA-18362-02.
the hCG of neoplastic origin. The development of techniques for the establishment and maintenance of long term in vitro cultures of tumors provides a potential source of relatively large amounts of humoral products from these tumors. The ectopic production of hCG or its subunits by nontrophoblastic tumors in vitro has previously been described for cell lines derived from a hepatoblastoma (2), a bronchogenic carcinoma (3-5), and cervical carcinomas (6-9). The present report describes the physical and biological characteristics of the hCG produced in vitro by an ovarian papillary cystadenocarcinoma which has been maintained in long term tissue culture. Materials and Methods Cell cultures The ovarian carcinoma cell line, designated 163, was established in June, 1972, from cells obtained
326
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 14 November 2015. at 16:43 For personal use only. No other uses without permission. . All rights reserved.
PROPERTIES OF hCG FROM OVARIAN CARCINOMA from the ascitic fluid of a 45-yr-old woman with a papillary cystadenocarcinoma of the ovary. The cells have a modal chromosome number of 66. This line has a type B glucose-6-phosphate dehydrogenase isoenzyme mobility, distinguishing it from HeLa cells which have a type A isoenzyme mobility (10). For the present study, the cells were grown in multiple culture flasks containing Ham's F-10 medium (Flow Laboratories, Rockville, MD) supplemented with 20% fetal bovine serum, penicillin (100 U/ml), streptomycin (100 /ig/ml), and kanamycin (50 jug/ml). They were maintained in a humidified atmosphere of 5% CO2 and 95% air at 37 C. The medium was decanted and centrifuged at 500 X g to eliminate detached cells, and the supernatant was frozen at —20 C until further study. Concentration method The tissue culture medium was ultrafiltered through an Amicon UM-100 A membrane (Amicon Corporation, Lexington, MA) and the filtrate was lyophilized and reconstituted in distilled water to approximately 5% of the starting volume. Recovery by this method of hCG and the free a and /? subunits of hCG after addition of purified materials to tissue culture medium was 72, 72, and 63%, y respectively. Radioligand assays
i
Three separate RIAs were used to measure hCG. The double antibody hCG RIA, which does not discriminate between hCG and human LH (hLH) (hCG/hLH RIA), and the /?-hCG RIA, which is relatively specific for the /? subunit of hCG and intact hCG, have been previously described (11). The reagents used for these assays were obtained rom the Hormone Distribution Officer, NIAMDD, 4IH, Bethesda, MD. The highly purified preparaion of hCG (CR 117) was used as the labeled ligand lid reference standard for both assays. This maerial contained 5200 IU/mg immunological activity nd 10,600 IU/mg biological activity relative to the •econd International Standard for hCG. A recently eveloped, double antibody RIA for hCG which utilizes an antiserum generated against the carboxyl-terminal peptide (amino acids 109-145) of the /? subunit of hCG was used to further examine the immunological identity of the ectopic hCG. This assay was developed and performed by Dr. Albert Parlow and the methodological details of the assay will be described in a separate publication by Dr. Parlow. There is less than 0.001% crossreaction with equimolar amounts of highly purified pituitary glycoprotein hormones (Parlow, unpub-
327
lished observations). The reference standard used in the carboxyl-terminal /MiCG RIA was a purified preparation of urinary hCG, supplied by Dr. O. P. Bahl, containing 6000 IU immunological activity/mg relative to the Second International Standard for hCG. The a subunit of hCG was measured by a homologous RIA, by the method of Vaitukaitis and Ross (12), with reagents provided by the Hormone Distribution Officer, NIAMDD, NIH, Bethesda, MD. The Leydig cell radioreceptor assay (RRA) method of Catt et al. (13) was used to measure gonadotropin-binding activity. The hCG standard (CR 117) had a specific activity of 10,000 IU/mg in the RRA relative to the Second International Standard for hCG. The potency estimate for the tissue culture hCG in the RRA was based on the weight of immunoreactive hCG, as determined by the /? subunit RIA using CR 117 (5200 IU/mg) as the reference standard. Bioassay The ventral prostate weight assay was used to measure the in vitro biological activity of hCG (14). Immature hypophysectomized Sprague-Dawley rats (Hilltop Laboratories, Bilroy, CA), weighing 40-50 g each, were injected with the test materials beginning 3 days after hypophysectomy. All dilutions of standard or tissue culture extracts were made with 1% bovine serum albumin in isotonic saline and the total volume injected over 4 days was 4 ml. Four or five rats were used at each dose level. Autopsy was performed on the morning of the 5th day and the ventral lobe of the prostate was removed and weighed on a torsion balance. Three assays were performed with three or four dilutions of the tissue culture extracts and four to seven doses of the Second International Standard for hCG in each assay. The potency estimates for the tissue culture hCG were based on the weight of immunoreactive hCG, as determined by the fi subunit RIA using CR 117 (5200 IU/mg) as the reference standard. Chromatography A 1.5 X 85.0-cm column of Sephadex G-100 (Pharmacia Laboratories, Piscataway, NJ) equilibrated with 0.05 M Tris-hydrochloride-0.15 M sodium chloride buffer, pH 7.4, was used to separate hCG, hLH, and the a and fi subunits of hCG. Onemilliliter samples of test or reference materials were applied to the column and 1-ml aliquots of the effluent were collected. The volume of the column
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 14 November 2015. at 16:43 For personal use only. No other uses without permission. . All rights reserved.
328
JCE&M • 1978 Vol47 • No2
BRAUNSTEIN ET AL.
was determined by the peak elution volume of blue dextran, measured spectrophotometrically at 600 nm. The column was calibrated with [125I]hCG, purified hCG (CR 117), pregnancy serum, term placental tissue extract, [125I]hCG /? subunit, purified hCG /? subunit (CR 115-/3), purified hCG a subunit (CR 115-a), [125I]hLH, purified hLH (LER 960), and postmenopausal serum. The volume of exclusion for each hormone corresponded to the tube with the highest counts per min of 125I for the radioiodinated hormones or with the highest concentration of immunoreactive hormone, as determined by RIA. Because of the heterogeneity of the hormones used to calibrate the column, elution ranges were established for hCG, the /? subunit of hCG, and hLH. Disposable columns (BioRad Laboratories, Richmond, CA), 0.7 X 4.0 cm., containing 1.5 ml Concanavalin A (Con A) covalently coupled to Sepharose 4B (Pharmacia Laboratories, Piscataway, NJ) were extensively washed with phosphate-buffered saline, pH 7.8, containing 2.5% normal rabbit serum. One-milliliter aliquots of pregnancy serum, [125I]hCG, or purified hCG dissolved in fresh tissue culture medium and 1 ml concentrated tissue culture medium from cell line 163 were then filtered through the columns with the use of phosphatebuffered saline with normal rabbit serum for elution of unadsorbed substances. Elution of Con A-adsorbed glycoproteins was accomplished with 0.2 M a-D-methylglucoside (15). The eluted fractions were collected in 0.5-ml aliquots and analyzed in the /?hCG RIA. Disappearance studies
The logarithm of the hormone concentration in the rat serum was plotted against time and then extrapolated back to the ordinate (zero time). These results were then plotted in terms of a percentage of zero time concentrations. Each disappearance curve was best fit by a biexponential model and the two components, designated fast and slow, were each fit by the method of least squares and half-lives (ti^) in serum were determined. Statistics The statistical methods of Rodbard and Frazier (16) were used for dose interpolation in the radioligand assays. Bioassay dose interpolation and tests for parallelism between standard and test materials were performed according to the bioassay statistical program developed by Faden and Rodbard (17). The fast and slow components of the disappearance curves for hCG, hLH, and tissue culture hCG were compared to each other by analysis of covariance (18). A significant difference between disappearance curves was considered to be present if P < 0.05.
Results Serial dilution of the concentrated material from cell line 163 tissue culture medium exhibited parallelism with the purified urinary hCG reference standard in the /8-hCG RIA (Fig. 1) and in the RIA which specifically detects the 109-145 carboxyl-terminal portion 90 80 -
A comparison of the disappearance rates of the tissue culture hCG, purified urinary hCG (CR117), the hCG extracted from term placentas, purified pituitary hLH (LER 960), and the hLH present in a pituitary extract was carried out in 22-day-old female Sprague-Dawley rats (Hilltop Laboratories, Gilroy, CA), weighing 50-60 g. The test material was injected via the tail vein. At multiple points in time, three to five animals were bled and the serum was separated and frozen until assay. The /?-hCG RIA was used to measure the hCG concentration in the serum from rats injected with the tissue culture extracts, urinary hCG, and placental hCG, while the nonspecific hCG/hLH RIA was used to measure the immunoreactive hLH in the rat serum after injection of the pituitary hLH or pituitary extracts. [125I]hLH (LER 960) was used as a labeled ligand and LER 907 was used for reference purposes for the hLH measurements.
60 40 -
TISSUE CULTURE EXTRACTS
2010-
hCG STANDARD (CR 117)
4 2 -
hCG
NANOGRAMS/TUBE
TISSUE CULTURE EXTRACTS
MICROLITERS/TUBE
FIG. 1. Dose-response curves for the hCG reference preparation (CR 117) and serial dilutions of culture extracts in the /?-hCG RIA. B, Counts bound in the presence of labeled and unlabeled ligand; Bo, counts bound in the presence of labeled ligand alone; Logit (B/Bo), loge [B/Bo/(1 - B/Bo)]. Coordinates are Logit normalized percentage counts precipitated (B/Bo X 100) (on the ordinate) and log of mass of hCG standard or volume of tissue culture extract added per tube (on the abscissa).
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 14 November 2015. at 16:43 For personal use only. No other uses without permission. . All rights reserved.
PROPERTIES OF hCG FROM OVARIAN CARCINOMA
329
99-1
FIG. 2. Dose-response curves for the purified hCG preparation (CR 117), the Second International Standard for hCG, and serial dilutions of tissue culture extracts in the rat Leydig cell RRA. CR 117 is expressed in terms of nanograms per tube, while the Second International Standard for hCG is depicted as International Units per tube. The other abbreviations are described in Fig. 1.
2nd INTERNATIONAL STANDARD hCG hCG (CR117)
TISSUE CULTURE EXTRACT
1.0
10
10
100
hCG IU or NANOGRAMS/TUBE 0.1
1.0
TISSUE CULTURE EXTRACTS
of the /? subunit of hCG. These results suggest that the material secreted by the ovarian adenocarcinoma cell line and urinary hCG are immunologically similar. In the rat Leydig cell RRA, serial dilutions of the tissue culture concentrates were parallel to the purified hCG standard and the Second International Standard for hCG (Fig. 2). The ratio of radioreceptor activity to immunological activity for the hCG standard (CR 117) was 2.37, while the ratio for the tissue culture extract was 2.17. Control fresh tissue culture medium concentrated by identical procedures, as described for the cell line 163 media, demonstrated no activity in either the /?-hCG RIA or RRA. Parallelism was also demonstrated between the dose-response curves generated by the tissue culture extracts and that produced by the Second International Standard for hCG in the ventral prostate weight bioassay (Fig. 3). The mean potency estimate of the tissue culture hCG in three assays was 9,289 IU/mg (95% confidence limits, 7,909-10,669). The ratio of biological to immunological activity was 1.79 for the tissue culture hCG compared to 2.37 for the highly purified urinary hCG reference preparation. The tissue culture gonadotropin eluted from the Sephadex G-100 column within the established hCG region (Fig. 4). No qualitative differences were found when the hCG in the
£ so i
MICROLITERS/TUBE
2nd INTERNATIONAL STANDARD hCG
hCG
TISSUE CULTURE EXTRACT
IU INJECTED
TISSUE CULTURE EXTRACT
100 ul INJECTED
FIG. 3. Dose-response curves for the Second International Standard for hCG and serial dilutions of tissue culture extracts in the ventral prostate weight bioassay. The ventral prostate weight in milligrams is depicted on the ordinate, while the log of IU of the Second International Standard for hCG or volume of tissue culture extracts injected per rat is shown on the abscissa.
effluent tubes was measured by the nonspecific hCG/hLH RIA, the /?-hCG RIA, or the RRA. No free a subunit of hCG was detected in the column effluent by the homologous ahCG RIA. As the ultrafiltration method does not selectively concentrate hCG over its subunits, it is unlikely that significant amounts of free a or /? subunits were secreted into the medium. The immunoreactive material in the tissue culture extracts was adsorbed onto Con A and was eluted with a-D-methylglucoside with a
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 14 November 2015. at 16:43 For personal use only. No other uses without permission. . All rights reserved.
330
BRAUNSTEIN ET AL. Vo
hCG /3 hLH a
i
mziDczii
JCE&M • 1978 Vol47 • No 2
hCG/hLH R1A
hCG TISSUE CULTURE
60120 240 360 MINUTES AFTER INJECTION
5040-
hCG/hLH RRA
302010-
0
10
20
30
40
SO
60
70
80
90
100
ELUTION VOLUME ml
FIG. 4. Sephadex G-100 elution profile of tissue culture extracts expressed in terms of nanograms per hCG, using CR 117 as a reference standard. The void volume (Vo), peak hCG, /?-hCG, hLH elution ranges, and the region of peak a subunit elution are depicted on the top of the graph. The hCG/hLH and the /?-hCG RIAs were performed on aliquots from the effluent from a single chromatographical run, while the hCG/hLH RRA was performed on the effluent tubes from a separate chromatographical procedure in which a greater concentration of the tissue culture extract was applied to the column.
recovery of 94%. Recovery of the hCG present in pregnancy serum, [125I]hCG, or purified hCG added to fresh tissue culture medium and treated in identical fashion was 91.8% ± 2.9 (SEM).
The disappearance curves of the purified urinary hCG, the hCG present in placental extracts, purified pituitary hLH, the LH present in extracts of the pituitary, and the hCG found in the tissue culture medium are illustrated in Fig. 5. There were no significant differences between either the fast or slow components of the purified hCG (tested on two occasions) and the hCG in the placental extract and, therefore, the data from three experiments were combined. The ti /2 of the fast component of hCG was 65.4 min, while
FIG. 5. Disappearance of hCG, tissue culture hCG (TC), and hLH from the plasma of immature female rats after iv injection. Each symbol represents the mean of the percentage of the zero time concentration vs. time derived from: 1) two separate experiments with unlabeled hCG (CR 117) and the hCG in a placental extract (•); 2) two separate experiments with the tissue culture extracts (O); and 3) purified hLH (LER 960) and the hLH present in a pituitary extract (A).
that for the slow component was 143.3 min. The individual disappearance curves for the purified pituitary hLH and the hLH from the crude pituitary extract were identical, with a ti/2 of 21.1 and 177.1 min for the respective components. The results from two separate studies of the disappearance of tissue culture hCG from the serum of the rats were similar and, therefore, the data was combined. The ti/2 of the fast and slow components for the tissue culture hCG was 37.6 and 250.8 min, respectively. The analysis of covariance of the disappearance curves of each substance indicated that hLH left the plasma compartment significantly faster than hCG or the tissue culture hCG and that the tissue culture hCG disappeared from the serum more rapidly than did urinary or placental hCG but at a significantly slower rate than hLH. Discussion Although a considerable amount of information has accumulated concerning the biological, immunological, and structural properties of hCG from the urine of pregnant
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 14 November 2015. at 16:43 For personal use only. No other uses without permission. . All rights reserved.
PROPERTIES OF hCG FROM OVARIAN CARCINOMA women and patients with hCG-producing trophoblastic tumors (eutopic secretion) (19, 20), » only a few studies have been carried out on the hCG secreted by nontrophoblastic neoplasms (ectopic secretion) (21, 22). Vaitukaitis examined the plasma, urine, and tumor extracts of patients with hCG-secreting trophoblastic and nontrophoblastic neoplasms » and found intact hCG as well as free subunits and altered molecular forms of these substances (21). Similarly, Weintraub and coworkers found differences between the a subx units of the naturally occurring glycoprotein hormones and the purified a subunits derived ^from two nontrophoblastic neoplasms (22). Both groups of investigators have raised the possibility that altered protein synthesis may be a concomitant of neoplasia, although the described alterations in hormone synthesis *" may reflect a quantitative difference in the ^-secretion of naturally occurring isohormones of hCG or its subunits (23-28). The hCG secreted by the ovarian adenocarcinoma cell line 163 clearly resembles urinary hCG by several immunological, biological, and r physical criteria. In particular, the results of > the Con A chromatography indicate the glycoprotein nature of the tissue culture gonadotropin (15) and the results from the /?-hCG RIA, RRA, and ventral prostate weight assays demonstrate the close immunological and biological similarity of these species. Because of • the small degree of cross-reactivity between hLH and hCG in the /?-hCG RIA and the * similar reactivity of these hormones in the RRA and ventral prostate weight bioassay, it was conceivable that the activity in the above assays noted with the tissue culture media t extracts could be due to the secretion of hLH. The prime evidence that the material pro* duced by the cell line is hCG and not hLH is y provided by the column chromatographical elution pattern of the tissue culture material which resembles that of hCG and not hLH, ^ and the parallelism between the tissue culture hCG and native hCG in the 109-145 carboxyl" ^terminal /?-hCG RIA. The sequence of the first 115 amino acids of the ft subunits of hCG and hLH demonstrates significant homology (29). However, the ft subunit of hCG contains
331
30 additional carboxyl-terminal amino acids in excess of hLH, which confer immunological specificity to hCG (29). As the tissue culture hCG is recognized by an antiserum to this unique peptide, the immunological identity of the tissue culture gonadotropin as hCG would seem to be well established. However, there are some differences in biological behavior between the tissue culture hCG and the predominant molecular species of hCG present in the highly purified hCG (CR 117) used in these studies. The tissue culture hCG contains approximately 88% of the biological activity of CR 117 and disappears from the serum after injection into intact immature female rats significantly faster than does CR 117. Both of these observations may be due to a decrease in the content of the sialic acid moiety of the tissue culture hCG relative to the purified urinary hCG. Van Hall et al. (30) have previously demonstrated that stepwise desialylation of hCG leads to a progressive reduction of hCG biological activity in the ventral prostate weight assay with maintenance of immunological activity, resulting in a progressive decrease in the biological to immunological ratio. These workers also showed that the half-life of the initial component of the plasma disappearance curve after injection in rats decreased concurrently with the reduction of the sialic acid content of the molecule (31). As much of the well known heterogeneity of hCG found in the urine of pregnant women seems to reside in differences in the sialic acid content of the hCG molecule (27, 28), it is reasonable to expect that similar differences may be found in hCG molecules secreted by tumors. That tumors may secrete hCG molecules with varying amounts of sialic acid and, hence, varying biological to immunological ratios has previously been shown for clonal cell lines of gestational choriocarcinoma (32, 33). Although the hCG secreted by the ovarian adenocarcinoma cell line resembles an incompletely sialylated form of hCG, further purification and amino acid and carbohydrate analysis of the hCG secreted by cell line 163 will be needed to establish with certainty the structural identity of the ectopic hCG to a naturally occurring isohormone of hCG.
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 14 November 2015. at 16:43 For personal use only. No other uses without permission. . All rights reserved.
BRAUNSTEIN ET AL.
332
Acknowledgments We gratefully acknowledge Dr. Albert Parlow for the performance of the carboxyl-terminal hCG RIA. The glucose-6-phosphate dehydrogenase isoenzyme mobility of the cell line was kindly performed by Dr. Robert Sparks. The technical assistance of Mr. G. V. Scarboro and G. Firestein, the secretarial assistance of Ms. M. C. Brill, and the statistical analyses performed by Mr. P. Schroth are greatly appreciated.
References 1. BRAUNSTEIN, G. D., Use of human chorionic gonadotropin as a tumor marker in cancer, In Herberman, R. B., and K. R. Mclntire, (eds.), Immunodiagnosis of Cancer, New York, Marcel Dekker, 1978. 2. BRAUNSTEIN, G. D., W. E. BRIDSON, A. GLASS, E. W. HULL,
AND K. R. MCINTIRE, In vivo and in vitro production of human chorionic gonadotropin and a-fetoprotein by a virilizing hepatoblastoma, J Clin Endocrinol Metab 35: 857, 1972. 3. RABSON, A. S., S. W. ROSE, A. H. TASHJIAN, JR., AND B. D.
WEINTRAUB, Production of human chorionic gonadotropin in vitro by a cell line derived from a carcinoma of the lung, J Natl Cancer Inst 50: 669, 1973. 4. TASHJIAN, A. H., JR., B. D. WEINTRAUB, N. J. BAROWSKY, A.
S. RABSON, AND S. W. ROSEN, Subunits of human chorionic gonadotropin: unbalanced synthesis and secretion by clonal cell strains derived from a bronchogenic carcinoma, Proc Natl Acad Sci 70: 1419, 1973. 5. LIEBLICH, J. M., B. D. WEINTRAUB, G. H. KRAUTH, P. 0. KOHLER, A. S. RABSON, AND S. W. ROSEN, Ectopic and
eutopic secretion of chorionic gonadotropin and its subunits in vitro: comparison of clonal strains from carcinomas of lung and placenta, J Natl Cancer Inst 56: 911,1976. 6. GHOSH, N. K., AND R. P. Cox, Production of human chorionic gonadotropin in HeLa cell cultures, Nature 259: 416, 1976. 7. LIEBLICH, J. M., B. D. WEINTRAUB, S. W. ROSEN, J. Y. CHOU,
AND J. C. ROBINSON, HeLa cells secrete a-subunit of glycoprotein trophic hormones, Nature 260: 530, 1976. 8. LIEBLICH, J. M., B. D. WEINTRAUB, S. W. ROSEN, N. K.
GHOSH, AND R. P. Cox, Secretion of hCG-asubunit and hCG by HeLa strains, Nature 265: 746, 1977. 9. PATILLO, R. A., R. 0. HUSSA, M. T. STORY, A. C. F. RUCKERT,
M. R. SHALABY, AND R. F. MATTINGLY, Tumor antigen and human chorionic gonadotropin in CaSKi cells: a new epidermoid cervical cancer cell line, Science 196: 1456, 1977. 10. NELSON-REES, W. A., AND R. R. FLANDERMEYER, HeLa cul-
tures defined, Science 161: 96, 1976. 11. VAITUKAITIS, J. L., G. D. BRAUNSTEIN, AND G. T. ROSS,
Radioimmunoassay which specifically measures human chorionic gonadotropin in the presence of human luteinizing hormone, Am J Obstet Gynecol 113: 731, 1972. 12. VAITUKAITIS, J. L., AND G. T. ROSS, Immunologic cross-reactivity among the human glycoprotein hormones and their subunits, In Saxena, B. B., C. G. Beling, and H. M. Gandy (eds.), Gonadotropins, New York, John Wiley and Sons, 1972, p. 435. 13. CATT, K. J., M. L. DUFAU, AND T. TSURUHARA, Studies on a
radioligand, receptor assay system for luteinizing hormone and chorionic gonadotropin, J Clin Endocrinol Metab 32: 860, 1971. 14. MCARTHUR, J. W., Identification of pituitary interstitial cell stimulating hormone in human urine, Endocrinology 50: 304, 1952.
JCE & M Vol47
1978 No 2
15. DUFAU, M. L., T. TSURUHARA, AND K. J. CATT, Interaction of
„,
glycoprotein hormones with agarose-Concanavalin A, ' Biochim Biophys Ada 278: 281, 1972. 16. RODBARD, D., AND G. R. FRAZIER, Radioimmunoassay Data *• Processing, ed. 2, Springfield, VA, National Technical Information Service, 1973. 17. FADEN, V. B., AND D. RODBARD, Bioprog. Bioassay Data Processing, Bethesda, NICHHD, NIH, 1975. 18. DIXON, W. J., BMD-PIV, Biomedical Computer Program, P Series, Los Angeles, University of California Los Angeles : Press, 1975, p. 683. 19. CANFIELD, R. E., F. J. MORGAN, S. KAMMERMAN, J. J. BELL,
AND G. M. AGOSTO, Studies of human chorionic gonadotropin, « Recent Prog Horm Res 27: 121, 1971. 20. VAITUKAITIS, J. L., G. T. Ross, G. D. BRAUNSTEIN, AND P. L.
RAYFORD, Gonadotropins and their subunits: basic and clinical studies, Recent Prog Horm Res 32: 289, 1976. 21. VAITUKAITIS, J. L., Immunologic and physical characterization of human chorionic gonadotropin (hCG) secreted by tumors, J Clin Endocrinol Metab 37: 505, 1973. 22. WEINTRAUB, B. D., G. KRAUTH, S. W. ROSEN, AND A. S.
RABSON, Differences between purified ectopic and normalir alpha subunits of human glycoprotein hormones, J Clin Invest 56: 1043, 1975. 23. WEINTRAUB, B. D., B. S. STANNARD, AND S. W. ROSEN,
Combination of ectopic and standard human glycoprotein hormone alpha with beta subunits: discordance of immunologic and receptor-binding activity, Endocrinology 101: 225, 1977. ** 24. LANDEFELD, T., S. BOGUSLAWSKI, L. CORASH, AND I. BOIME,
The cell-free synthesis of the alpha subunit of human cho--^ rionic gonadotropin, Endocrinology 98: 1220, 1976. 25. VAITUKAITIS, J. L., Changing placental concentrations of human chorionic gonadotropin and its subunits during gestation, J Clin Endocrinol Metab 38: 755, 1974. 26. MARUO, T., Y. ASHITAKA, M. MOCHIZUKI, AND S. TOJO,
Chorionic gonadotropin synthesized in cultivated trophoblast, Endocrinol Jap 21: 499, 1974.
t
27. VAN HELL, H., R. MATTHIJSEN, AND J. D. H. HOMAN, Studies
on human chorionic gonadotrophin. I. Purification and some < physicochemical properties, Ada Endocrinol (Kbh) 59: 89, 1968. 28. SCHUURS, A. H., W. M. E. DE JAGER, AND J. D. H. HOMAN,
Studies on human chorionic gonadotrophin. III. Immunochemical characterisation, Ada Endocrinol (Kbh) 59: 120, 1968. 29. MORGAN, F. J., S. BIRKEN, AND R. E. CANFIELD, Comparison
of chorionic gonadotropin and luteinizing hormone: a note on a proposed significant structural difference in the beta sub- -4 unit, FEBS Lett 31: 101, 1973. 30. VAN HALL, E. V., J. L. VAITUKAITIS, G. T. ROSS, J. W.
HICKMAN, AND G. ASHWELL, Immunologic and biologic activity of HCG following progressive desialylation, Endocrinology 88: 456, 1971. 31. VAN HALL, E. V., J. L. VAITUKAITIS, G. T. ROSS, J. W.
HICKMAN, AND G. ASHWELL, Effects of progressive desialylation on the rate of disappearance of immunoreactive HCG ^ from plasma in rats, Endocrinology 89: 11, 1971. 32. BRIDSON, W. E., G. T. ROSS, AND P. 0. KOHLER, Immunologic
and biologic activity of chorionic gonadotropin synthesized by cloned choriocarcinoma cells in culture, J Clin Endocrinol Metab 33: 145, 1971. 33. HAMMOND, J. M., W. E. BRIDSON, P. O. KOHLER, AND A.
CHRAMBACH, Physical characteristics of immunoreactive chorionic gonadotropin produced in culture, Endocrinology 89: 801, 1971.
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 14 November 2015. at 16:43 For personal use only. No other uses without permission. . All rights reserved.
i
-
