0021-972X/91/7204-0841$03.00/0 Journal of Clinical Endocrinology and Metabolism Copyright © 1991 by The Endocrine Society
Vol. 72, No. 4 Printed in U.S.A.
Purified Preparations of Human Luteinizing Hormone Are Contaminated with Small Amounts of a Chorionic Gonadotropin-Like Material ARLEEN L. SAWITZKE, JEANINE GRIFFIN, AND WILLIAM D. ODELL Departments of Physiology and Medicine, University of Utah School of Medicine, Salt Lake City, Utah 84132
ABSTRACT. We have identified a CG-like protein contaminating a purified human LH preparation of immunochemical grade. This CG-like material is estimated to comprise 0.17%, by weight, of the LH and reacts in specific, sequential-type, twomonoclonal antibody, immunoradiometric assays for CG as well as in the carboxyl-tail CG RIA. The CG-like material is not separable from LH by size exclusion or ion exchange chromatography. The LH can be freed of this small contamination of CG-like material by immunopurification employing specific
monoclonal antibodies. Sephadex G-100 chromatography shows this material to have a mol wt of 40.0K. Western blot analysis of the LH run under nonreducing conditions, using an anti-CG carboxyl-tail primary antibody, reveals two bands of this CGlike material, one at 60.8K and one at 50.7K. When electrophoresed under reducing conditions, the material reacts with the anti-CG carboxyl-tail antibody at several mol wt, ranging from 10.5-64K. (J Clin Endocrinol Metab 72: 841-846, 1991)
H
While developing the sensitive IRMAs used in some of these studies on CG, a consistent small cross-reaction of purified human LH preparations in the CG IRMAs was noted. The studies reported herein were performed to evaluate whether this reaction was a true cross-reaction in the assay or whether the LH preparation was contaminated with small amounts of a pituitary CG-like material.
UMAN LH and hCG are very similar glycoproteins. Each is composed of an a- and a /3-chain. The enchains are identical; the /3-chains share 82% amino acid homology. LH is produced by the anterior pituitary gland. CG was traditionally thought to be produced only by the syncytiotrophoblast of the placenta or from neoplasms derived from these cells (gestational trophoblast neoplasms) (1). However, many, if not all, nontrophoblast carcinomas have been shown to secrete CG (2-4), and CG has now been extracted from virtually all tissues from humans dying of traumatic causes (5-8) and from serum and urine of nonpregnant normal humans (9-11). Matsuura et al. (12, 13) extracted a CG-like material from normal pituitary glands. With the advent of very sensitive and specific immunoradiometric assays (IRMAs), CG was detected in unextracted serum of normal nonpregnant men and women (14, 15). This CG was reported to be secreted in a pulsatile fashion in postmenopausal women and during the normal menstrual cycle (16). Furthermore, this CG was stimulated by GnRH and suppressed by continuous administration of GnRH agonists and estrogens (15, 17). Odell et al. have demonstrated a unique CG-producing cell in the human pituitary (18) and have shown that CG is secreted by human fetal pituitary cells in tissue culture (19).
Materials and Methods LH (AFP-0642B) was kindly supplied by the Hormone Distribution Program of the NIDDK. This LH has an immunopotency of 11,559 IU/mg (Second International Reference Preparation for human menopausal gonadotropin). Samples of LH from the NIDDK used without immunopurification will be referred to as LH. After immunopurification has been performed on the LH, the sample will be referred to as immunopurified LH (i-LH). CG (CR-121), with an immunopotency of 13,450 IU/mg (Second International Standard for hCG), was kindly supplied by the NICHHD and Dr. Robert Canfield, and was used as the reference preparation in the CG IRMAs. All experiments were performed at least twice to insure reproducibility. Immunopurification A monoclonal antibody (mab CG4) produced in this laboratory and shown to be specific for CG (14) was conjugated to CNBr-activated Sepharose beads (Pharmacia, Uppsala, Sweden). One milligram of antibody was conjugated to 1 mL Sepharose beads. Fifty-eight international units (5 ng) of LH were dissolved in 5 mL horse serum (Hyclone Laboratories,
Received February 26,1990. Address all correspondence and requests for reprints to: Arleen L. Sawitzke, Department of Medicine, University of Utah School of Medicine, 50 North Medical Drive, Salt Lake City, Utah 84132.
841
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 16 November 2015. at 14:46 For personal use only. No other uses without permission. . All rights reserved.
SAWITZKE, GRIFFIN, AND ODELL
842
Ogden, UT) to help prevent nonspecific binding of LH to the Sepharose beads and mixing tube. The antibodies used in this study have been shown not to cross-react with horse serum (14), and horse serum is used as a blank in the IRMAs. An aliquot of this solution was saved for later assay, and the remaining 4.5 mL of solution were mixed with 0.5 mL antibodyconjugated beads in a rotating shaker overnight at 4 C. The beads were then poured into a column, and the effluent was collected. The column was extensively washed (100 mL) with phosphate-buffered saline (0.01 M sodium phosphate-0.15 M sodium chloride, pH 7.4; PBS) until the CG-like material, measurable by the CG sequential IRMA, in the wash had reached a plateau at 0.070-0.135 mlU/mL (5-10 pg/mL). The material remaining on the column was eluted with 3.5 M MgCl2. The MgCl2 was then removed from the eluted protein by passing the solution through a Beckman buffer exchange column (Beckman, Fullerton, CA) that had been equilibrated with PBS. Samples of the effluent, wash, and eluate were saved for assay in the CG and LH IRMAs (14, 20). The samples were measured in the LH IRMA to assess the exact concentration of LH present after immunopurification. This concentration was used to determine the percent reaction in the CG IRMA. The purification of the LH preparation was performed independently four times. IRMAs The IRMAs for CG and LH that were developed in this laboratory (14,20) were used with some modifications. In assays designated simultaneous assays, the biotinylated and radiolabeled monoclonal antibodies were both added at the onset of the assay. In assays designated sequential assays, the biotinylated antibody was incubated with the assay components (CG or LH, horse serum, polystyrene bead) for 16-20 h. The beads were then washed twice with PBS. A 1:1 mixture of radiolabeled antibody and horse serum (total volume, 200 /xL) was added, and the beads were incubated for 2 h (LH assay) or 5 h (CG assay). The beads were then washed with washing buffer (0.9% NaCl-0.1% Triton X-100) and counted on a Packard 7-counter (Packard, Downers Grove, IL). The sensitivities were less than 0.02 mlU/mL for both the LH and CG IRMAs. The intraassay variations averaged 10.5% for the CG IRMA and 3.7% for the LH IRMA, while the interassay variations averaged 4% for the CG IRMA and 6% for the LH IRMA. Polyacrylamide gel electrophoresis (PAGE) LH and CG were run on 15% polyacrylamide gels by the method of Laemmli (21) and were stained with Coomassie blue. Western blotting (22) was carried out onto charged HybondECL nitrocellulose paper (Amersham, United Kingdom), which was then incubated with an anti-CG carboxyl-tail antibody (12) (1:250 final dilution). This primary antibody was diluted in a 1% solution of dried milk. The second antibody, goat antirabbit immunoglobulin G heavy and light chains (61-610, ICN, Costa Mesa, CA; 350,000 cpm/mL), was radiolabeled by the chloramine-T method and incubated overnight with the nitrocellulose paper with shaking at room temperature. The paper was then washed and exposed to Hyperfilm-ECL (Amersham) film for
JCE & M • 1991 Vol 72 • No 4
16 h at —80 C. An intensifier screen was also present in the film cassette. Three hundred international units (25 ^g) of LH were run in each lane. Gels were run under reducing (mercaptoethanol in the loading buffer) and nonreducing conditions. A control Western blot was performed using anti-CG carboxyltail antibody that had been presaturated with 50 /tig/mL carboxyl-tail peptide (Beckman). Carboxyl-tail RIA The LH preparation was assayed in a double antibody equilibrium-type RIA (12). The primary antibody (used at a 1:2500 final dilution) was a rabbit antibody produced against the unique 37-amino acid carboxyl-tail segment present in CG but absent in LH. The second antibody was a goat antirabbit serum used at a 1:50 final dilution. The labeled tracer was CG (CR121), which had been labeled with the chloramine-T method and purified on a Concanavilin-A (Pharmacia) column (23). The intra- and interassay variations of the carboxyl-tail assay were 7% and 11%, respectively. Sephadex G-100 and ion exchange purification Three hundred international units (25 ng) of the LH preparation were diluted in 1 mL PBS and applied to a 1.5 X 100cm Sephadex G-100 column. Fractions (0.5 mL) were collected and measured in the sequential CG and LH IRMAs. The column was standardized with bovine albumin, ovalbumin, chymotrypsin, and ribonuclease. The mol wt of the LH and CG-like material was determined from the Rf calculated for the peak fraction of each protein. One hundred and eighty international units (15 ng) of the LH preparation were diluted in 25 HIM Tris-HCl buffer, pH 7.5 (TB), and applied to a Whatman DE-52 anion exchange column that had been equilibrated with the same buffer. The column was washed with this buffer before beginning elution with a 100-mM sodium chloride gradient in TB. The 0.5-mL fractions that were collected were measured in the sequential CG and LH IRMAs. Chromatofocusing Chromatofocusing was performed on the LH preparation with a polybuffer exchanger (Pharmacia). Three hundred international units (25 ng) LH were dissolved in 0.5 mL starting buffer (0.025 M imidazole-HCl, pH 7.0) and applied to a 1 X 30 cm column of polybuffer exchanger 94 that had been equilibrated with starting buffer. The eluting buffer was polybuffer 74-HC1 at a pH of 3. Three-milliliter fractions were collected and assayed in the sequential LH and CG IRMAs
Results IRMAs Figure 1, A and B, demonstrates the dose-response curves in the CG IRMA produced by urinary CG obtained from pregnant females (CR-121), LH, and i-LH. Figure 1A represents the results obtained when the assay is run as a simultaneous assay, while Fig. IB shows the results
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 16 November 2015. at 14:46 For personal use only. No other uses without permission. . All rights reserved.
CG-LIKE MATERIAL IN A LH PREPARATION
843
100i
150001
80
10000-
605000
400 10 2 10-'10° 1 0 ' 1 0 2 1 0 3 10"
10000;
hCG LH Purified LH
hCX3 LH
20-
0
10'
10 3
10 4
105
10e
mlU HORMONE
8000-
FIG. 2. Dose-response curves produced by CG (CR-121) and LH in the carboxyl-tail RIA. The LH used in this assay had not been immunopurified. %B/BO, percent specific binding.
6000ffl
10 2
400020000 0
10 2 1 O ' 1 O ° 1 0 ' 1 0 2 1 0 3 1 0" mlU HORMONE
FIG. 1. Dose-response curves produced in the CG IRMA by CG obtained from the urine of pregnant humans (CR-121), LH, and purified LH (i-LH). A, Simultaneous assay. B, Sequential assay. TABLE 1. Percent reaction and percent contamination of LH (AFP0642B) with a CG-like material, as assessed by simultaneous and sequential CG IRMAs Simultaneous Sequential assay assay (n = 8) (n = 4) Cross-reaction 0.03 ± 0.02° 0.13 ± 0.02 Contamination 0.15 ± 0.04 0.17 ± 0.03 n=4 n=8 n, Number of times assay was performed. Results are expressed as the mean ± SEM from independent assays. 0 The cross-reaction in the sequential assay is not statistically different from zero.
obtained in a sequential assay. In both types of assays, the CG-like material measured in the LH preparations is reduced after immunopurification. The percent reaction of LH before immunopurification was accepted as a combination of cross-reaction and contamination. Since greater than 85% of the CG can be removed from a sample with the purification employed here (our unpublished data), the percent reaction remaining of the i-LH in the assays was accepted as cross-reaction. The percent contamination of the LH with a CG-like material was determined as the difference in reaction between LH and i-LH. These results are summarized in Table 1. Notice that the contamination of LH with CG-like material is calculated to be approximately 0.16% whether calculated from simultaneous or sequential assays. Carboxyl-tail RIA Figure 2 shows the dose-response curves obtained with urinary CG obtained from pregnant females (CR-121)
1
2
3
FIG. 3. Fifteen percent SDS-polyacrylamide gel electrophoresis of: lane 1, LH with no mercaptoethanol present (i.e. nonreduced); lane 2, LH with mercaptoethanol present (i.e. reduced); and lane 3, mol wt markers (42K, 30K, 21.5K, and 14K). The LH characterized on this gel had not been immunopurified.
and LH in the carboxyl-tail RIA. The percent reaction of LH in this assay was 0.013% by weight. Sodium dodecyl sulfate-PAGE (SDS-PAGE) and Western blots Figure 3 shows that nonreduced LH produced a stainable band at 21K (lane 1). Under reducing conditions, LH separated into stainable bands of 17K, 15K, and 10.5K (lane 2). A Western blot of the LH preparation is shown in Fig. 4. In the reduced condition (lane 1), the anti-CG carboxyl-tail antibody highlighted six bands,
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 16 November 2015. at 14:46 For personal use only. No other uses without permission. . All rights reserved.
SAWITZKE, GRIFFIN, AND ODELL
844
JCE & M • 1991 Vol 72 • No 4
3500
75K —
3000 2500 UJ
50K —
2000 1500" 1000-
39K —
500-
oo
27 K —
50
17K —
70
90
110
130
I 5000i u ^ 4000" 30002000-
1
1OOO-
2
FIG. 4. Western blot of: lane 1, LH with mercaptoethanol present (i.e. reduced); and lane 2, LH without mercaptoethanol present (i.e. nonreduced). The primary antibody was a polyclonal antibody against the carboxyl-tail of CG. The LH characterized by this blot had not been immunopurified.
64.4K, 56.8K, 34.8K, 29.0K, 20.7K, and 10.5K. The major bands were at 20.7K and 10.5K. Only the 10.5K band corresponds to a band that was present in sufficient quantity to be stained with Coomassie blue. In the nonreduced condition (lane 2), only two bands, 60.8K and 50.7K, were highlighted. Neither of these bands was shown with Coomassie blue staining. The negative control blot, using anti-CG carboxyl-tail antibody that had been presaturated with carboxyl-tail peptide, showed no highlighted bands (data not shown).
50
Chromatofocusing Figure 6 demonstrates the LH and CG activities measured in fractions collected from a polybuffer exchange
130
FIG. 5. LH and CG activity measured by the specific LH and CG IRMAs in fractions collected from a Sephadex G-100 column. A, LH was applied to the Sephadex column. B, After immunopurification on an anti-CG affinity column, i-LH was applied to the Sephadex column.
Sephadex G-100 and ion exchange purification Figure 5, A and B, demonstrates the LH and CG activities measured in fractions from a Sephadex G-100 column. In Fig. 5A, CG-like material was measured in fractions 85-112. LH was also measured over this range. The mol wt marker ovalbumin (43K) was detected by absorbance at 280 nm in fractions 85-110. The mol wt of the LH component was 41.5K, while that of the CGlike component was 40.0K. Figure 5B shows that there was no longer any detectable CG-like activity in the fractions when i-LH was applied to the Sephadex G-100 column. Both the LH and the CG activities were eluted from the ion exchange column at the same salt concentration. Thus, the two components were not separable by this method.
70 90 110 FRACTION NUMBER
^o
0
20
40
20
40
60
60
80
100
80
FRACTION NUMBER
°^
FIG. 6. LH and CG activities measured by the specific LH and CG IRMAs in fractions collected from a polybuffer chromatofocusing column. The upper panel represents the pH of the fractions collected from the column. The lower panel represents the activity of LH or CG in the fractions collected from the column. The LH applied to this column had not been immunopurified.
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 16 November 2015. at 14:46 For personal use only. No other uses without permission. . All rights reserved.
CG-LIKE MATERIAL IN A LH PREPARATION column. The top panel depicts the pH values for the fractions collected. The LH activity was eluted over a pH range of 7.3-6.3, while the majority of the CG-like material was eluted from pH 7.3-6.8, indicating that it has a slightly higher pi value than LH.
Discussion The CG-like material identified in LH is similar enough to urinary pregnancy CG to react with a very specific monoclonal antibody and with the carboxyl-tail antibody; however, it differs from urinary pregnancy CG in that it cannot be separated from the LH by size or ion exchange chromatography. The CG-like material can be partially separated from LH by chromatofocusing. As demonstrated by the Western blots in Fig. 4, more than one form of this CG-like material exists. One possible explanation for this size heterogeneity may be differences in glycosylation. Wang et al. (24) have purified and characterized a small form (43K by gel filtration) of CG from acetone powder extracts of first trimester placentas. The pi of this small CG is 10.0, and upon reduction, two /3-subunits at 23K and 20K and an a-subunit at 21K appear on SDS-PAGE. Carbohydrate studies demonstrate no 0linked and shortened JV-linked oligosaccarides on this small CG (24). The CG-like material we have identified in the pituitary LH preparation, with a mol wt by gel chromatography of 40.0K and a pi slightly higher than that of LH, may be similar or identical to this small placental CG. The CG measured in the serum of nonpregnant subjects by Stenman et al. (15) also showed a 43K mol wt upon gel filtration. The Western blots shown in this study indicate mol wt of 60.8K and 50.7K for the CG; however, these values are not incompatible with a mol wt of 40-43K determined by gel filtration. The true mol wt of glycosylated proteins cannot be determined by SDS-PAGE due to the effect of the carbohydrates. When urinary pregnancy CG (CR-121) is analyzed with SDSPAGE, the a- and /3-chains stain at 18K and 39K for a combined mol wt of 57K. Analysis of this preparation on Sephadex G-100 results in a calculated mol wt of 73K. The very small size (10.5K) of the SDS-PAGE band formed by LH under reducing conditions, which reacts with the anti-CG carboxyl-tail antibody, is intriguing. This band also reacts with a monoclonal antibody that is specific for the /3-chain of LH/CG (data not shown). The observation of low mol wt fragments from LH after disulfide reduction has been reported previously (25-27). Interestingly, Ward et al (28) reported that LH (AFP0642B) contained more of this small fragment than other LH preparations. This small fragment presumably originates from protein nicking during processing of the pituitary glands. Because the 10.5K band reacts with the
845
anti-CG carboxyl-tail antibody, we feel the CG-like material must also be nicked during processing. The stainability of the band indicates that a relatively large quantity of this fragment is present in the LH preparation; thus, both the LH and the CG-like material may be nicked to give 10.5K fragments. Nishimura et al. (29) found that when urinary CG was subjected to PAGE under reducing conditions, a small mol wt peptide (25K, or 18K after neuraminidase treatment) was separated from the /3-subunit. This small peptide, which bound the carboxyl-tail antibody in Western blots, was more apparent in CG from a patient with choriocarcinoma than in normal pregnancy CG (29). Puisieux et al. (30) also demonstrated the existence of several fragments of the CG /3-subunit on SDS-PAGE and Western blots. Both of these research groups worked with urinary CG, but the small fragments they demonstrated may correspond to the 29K and 20.7K bands that we have described. The above studies indicate that both LH and CG are subject to specific protein nicking. The site of production and/or the amount of glycosylation of the hormone may alter its susceptibility to this nicking. Stenman et al. (15) showed an increase in LH cross-reaction in their CG assay after storage or purification of the LH, perhaps reflecting the unveiling of a previously masked epitope upon breakdown of the hormone. There is only one known gene that codes for the (3chain of LH, while as many as six to eight genes or pseudogenes have been identified for the /3-chain of CG (31, 32). It is not known how many of these CG genes are translated. Possibly, some of the genes are expressed in the placenta, which lead to the traditional, highly glycosylated CG, and different genes are expressed in the pituitary gland, which lead to a slightly different and perhaps less glycosylated form of CG. This would explain some of the differences we have encountered between urinary pregnancy CG and the CG-like material that exists in the pituitary. Further studies, including carbohydrate and amino acid analysis will help determine the true nature of this CG-like material.
References 1. Odell WD, Hertz R, Lipsett MB, Ross GT, Hammond CB. Endocrine aspects of trophoblastic neoplasms. Clin Obstet Gynecol. 1967;10(Suppl 2):290-302. 2. Braunstein GD, Vaitukaitis JL, Carbone PP, Ross GT. Ectopic production of human chorionic gonadotropin by neoplasms. Ann Intern Med. 1973;78:39-45. 3. Odell WD, Wolfsen A, Yoshimoto Y, Weitzman D, Fisher D, Hirose F. Ectopic peptide synthesis: a universal concomitant of neoplasia. Trans Assoc Am Physicians. 1977;90:204-27. 4. Yoshimoto Y, Wolfsen AR, Odell WD. Glycosylation: a variable in the production of hCG by cancers. Am J Med. 1979;67:414-20. 5. Braunstein GD, Rasor J, Wade ME. Presence in normal human testes of a chorionic-gonadotropin-like substance distinct from human luteinizing hormone. N Engl J Med. 1975;293(Suppl
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 16 November 2015. at 14:46 For personal use only. No other uses without permission. . All rights reserved.
846
SAWITZKE, GRIFFIN, AND ODELL
26):1339-43. 6. Yoshimoto Y, Wolfsen AR, Odell WD. Human chorionic gonadotropin-like material in nonendocrine tissues of normal subjects. Science. 1977;197:575-7. 7. Braunstein GD, Kamdar V, Rasor J, Swaminathan N, Wade ME. Widespread distribution of a chorionic gonadotropin-like substance in normal human tissues. J Clin Endocrinol Metab. 1979;49(Suppl 6):917-25. 8. Yoshimoto Y, Wolfsen AR, Hirose F, Odell WD. Human chorionic gonadotropin-like material: presence in normal human tissues. Am J Obstet Gynecol. 1979b;134(Suppl 7):729-33. 9. Borkowski A, Muquardt C. Human chorionic gonadotropin in the plasma of normal, nonpregnant subjects. N Engl J Med. 1979;301(Suppl 6):298-302. 10. Borkowski A, Puttaert V, Gyling M, Muquardt C, Body JJ. Human chorionic gonadotropin-like substance in plasma of normal nonpregnant subjects and women with breast cancer. J Clin Endocrinol Metab. 1984;58(Suppl 6):1171-8. 11. Chen H, Hodgen GD, Matsuura S, et al. Evidence for a gonadotropin from nonpregnant subjects that has physical, immunological and biological similarities to human chorionic gonadotropin. Proc Natl Acad Sci USA. 1976;73(Suppl 8):2885-9. 12. Matsuura S, Chen H-C, Hodgen GD. Antibodies to the carboxylterminal fragment of human chorionic gonadotropin B-subunit: characterization of antibody recognition sites using synthetic peptide analogues. Biochemistry. 1978;17(Suppl 4):575-80. 13. Matsuura S, Ohashi M, Chen H-C, et al. Physicochemical and immunological characterization of an hCG-like substance from human pituitary glands. Nature. 1980;286:740-l. 14. Griffin J, Odell WD. Ultrasensitive immunoradiometric assay for chorionic gonadotropin which does not cross-react with luteinizing hormone nor free beta chain of hCG and which detects hCG in blood of non pregnant humans. J Immunol Methods. 1987;103:27583. 15. Stenman U, Alfthan H, Ranta T, Vartiainen E, Jalkanen J, Seppala M. Serum levels of human chorionic gonadotropin in nonpregnant women and men are modulated by gonadotropin-releasing hormone and sex steroids. J Clin Endocrinol Metab. 1987;64(Suppl 4):7306. 16. Odell WD, Griffin J. Pulsatile secretion of chorionic gonadotropin during the normal menstrual cycle. J Clin Endocrinol Metab. 1989;69:528-32. 17. Kyle CV, Griffin J, Odell WD. GnRH stimulation of human chorionic gonadotropin secretion in normal men and women [Abstract]. Clin Res. 1988;36(Suppl 1):181A. 18. Hammond E, Griffin J, Odell WD. A chorionic gonadotropin
Vol72«No4
secreting human pituitary cell. J Clin Endocrinol Metab. In Press. 19. Odell WD, Griffin J, Bashey HM, Snyder PJ. Secretion of chorionic gonadotropin by cultured human pituitary cells. J Clin Endocrinol Metab. 1990;71:1318-21. 20. Odell WD, Griffin J. Two-monoclonal-antibody "sandwich"-type assay of human lutropin, with no cross reaction with choriogonadotropin. Clin Chem. 1987;33(Suppl 9):1603-7. 21. Laemmli UK. Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature (Lond). 1970;227:680-5. 22. Towbin H, Staehelin T, Gordon J. Electrophoretic transfer of proteins from polyacrylamide gels to nitrocellulose sheets: procedure and some applications. Proc Natl Acad Sci USA. 1979;76(Suppl 9):4350-4. 23. Patritti-Laborde N, Yoshimoto Y, Wolfsen AR, Odell WD. Improved method of purifying some radiolabeled glycopeptide hormones. Clin Chem. 1979;25(Suppl l):163-5. 24. Wang HY, Segal SJ, Koide SS. Classification and characterization of an incompletely glycosylated form of human chorionic gonadotropin in human placenta. Endocrinology. 1988;123(Suppl 2):795803. 25. Courte C, Willimot J. Heterogeneity of porcine pituitary luteinizing hormone subunits (LH alpha and LH beta) in 6 M guanidine hydrochloride agarose gel chromatography. J Biol Chem. 1972;247(Suppl 13):4429-31. 26. Ward DN, Reichert Jr LE, Liu W-K, et al. Chemical studies of luteinizing hormone from human and ovine pituitaries. Recent Prog Horm Res. 1973;29:533-61. 27. Reichert Jr LE, Lawson Jr GM. Molecular weight relationships among the subunits of human glycoprotein hormones. Endocrinology. 1973;92:1034-42. 28. Ward DN, Glenn SD, Nahm HS, Wen T. Characterization of cleavage products in selected human lutropin preparations. Int J Peptide Protein Res. 1986;27:70-8. 29. Nishimura R, Ide K, Utsunomiya T, et al. Fragmentation of the /?subunit of human chorionic gonadotropin produced by choriocarcinoma. Endocrinology. 1988;123:420-5. 30. Puisieux A, Bellet D, Troalen F, et al. Occurrence of fragmentation of free and combined forms of the /3-subunit of human chorionic gonadotropin. Endocrinology. 1990;126(Suppl 2):687-94. 31. Policastro P, Ovitt CD, Hoshina M, Fukuoka H, Boothby M, Boime I. The subunit of human chorionic gonadotropin is encoded by multiple genes. J Biol Chem. 1983;258(Suppl 19):11492-9. 32. Talmadge K, Boorstein WR, Fiddes JC. The human genome contains seven genes for the B-subunit of chorionic gonadotropin but only one gene for the B-subunit of luteinizing hormone. DNA. 1983;2:279-87.
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 16 November 2015. at 14:46 For personal use only. No other uses without permission. . All rights reserved.
Vol. 72, No. 4 Printed in U.S.A.
Purified Preparations of Human Luteinizing Hormone Are Contaminated with Small Amounts of a Chorionic Gonadotropin-Like Material ARLEEN L. SAWITZKE, JEANINE GRIFFIN, AND WILLIAM D. ODELL Departments of Physiology and Medicine, University of Utah School of Medicine, Salt Lake City, Utah 84132
ABSTRACT. We have identified a CG-like protein contaminating a purified human LH preparation of immunochemical grade. This CG-like material is estimated to comprise 0.17%, by weight, of the LH and reacts in specific, sequential-type, twomonoclonal antibody, immunoradiometric assays for CG as well as in the carboxyl-tail CG RIA. The CG-like material is not separable from LH by size exclusion or ion exchange chromatography. The LH can be freed of this small contamination of CG-like material by immunopurification employing specific
monoclonal antibodies. Sephadex G-100 chromatography shows this material to have a mol wt of 40.0K. Western blot analysis of the LH run under nonreducing conditions, using an anti-CG carboxyl-tail primary antibody, reveals two bands of this CGlike material, one at 60.8K and one at 50.7K. When electrophoresed under reducing conditions, the material reacts with the anti-CG carboxyl-tail antibody at several mol wt, ranging from 10.5-64K. (J Clin Endocrinol Metab 72: 841-846, 1991)
H
While developing the sensitive IRMAs used in some of these studies on CG, a consistent small cross-reaction of purified human LH preparations in the CG IRMAs was noted. The studies reported herein were performed to evaluate whether this reaction was a true cross-reaction in the assay or whether the LH preparation was contaminated with small amounts of a pituitary CG-like material.
UMAN LH and hCG are very similar glycoproteins. Each is composed of an a- and a /3-chain. The enchains are identical; the /3-chains share 82% amino acid homology. LH is produced by the anterior pituitary gland. CG was traditionally thought to be produced only by the syncytiotrophoblast of the placenta or from neoplasms derived from these cells (gestational trophoblast neoplasms) (1). However, many, if not all, nontrophoblast carcinomas have been shown to secrete CG (2-4), and CG has now been extracted from virtually all tissues from humans dying of traumatic causes (5-8) and from serum and urine of nonpregnant normal humans (9-11). Matsuura et al. (12, 13) extracted a CG-like material from normal pituitary glands. With the advent of very sensitive and specific immunoradiometric assays (IRMAs), CG was detected in unextracted serum of normal nonpregnant men and women (14, 15). This CG was reported to be secreted in a pulsatile fashion in postmenopausal women and during the normal menstrual cycle (16). Furthermore, this CG was stimulated by GnRH and suppressed by continuous administration of GnRH agonists and estrogens (15, 17). Odell et al. have demonstrated a unique CG-producing cell in the human pituitary (18) and have shown that CG is secreted by human fetal pituitary cells in tissue culture (19).
Materials and Methods LH (AFP-0642B) was kindly supplied by the Hormone Distribution Program of the NIDDK. This LH has an immunopotency of 11,559 IU/mg (Second International Reference Preparation for human menopausal gonadotropin). Samples of LH from the NIDDK used without immunopurification will be referred to as LH. After immunopurification has been performed on the LH, the sample will be referred to as immunopurified LH (i-LH). CG (CR-121), with an immunopotency of 13,450 IU/mg (Second International Standard for hCG), was kindly supplied by the NICHHD and Dr. Robert Canfield, and was used as the reference preparation in the CG IRMAs. All experiments were performed at least twice to insure reproducibility. Immunopurification A monoclonal antibody (mab CG4) produced in this laboratory and shown to be specific for CG (14) was conjugated to CNBr-activated Sepharose beads (Pharmacia, Uppsala, Sweden). One milligram of antibody was conjugated to 1 mL Sepharose beads. Fifty-eight international units (5 ng) of LH were dissolved in 5 mL horse serum (Hyclone Laboratories,
Received February 26,1990. Address all correspondence and requests for reprints to: Arleen L. Sawitzke, Department of Medicine, University of Utah School of Medicine, 50 North Medical Drive, Salt Lake City, Utah 84132.
841
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 16 November 2015. at 14:46 For personal use only. No other uses without permission. . All rights reserved.
SAWITZKE, GRIFFIN, AND ODELL
842
Ogden, UT) to help prevent nonspecific binding of LH to the Sepharose beads and mixing tube. The antibodies used in this study have been shown not to cross-react with horse serum (14), and horse serum is used as a blank in the IRMAs. An aliquot of this solution was saved for later assay, and the remaining 4.5 mL of solution were mixed with 0.5 mL antibodyconjugated beads in a rotating shaker overnight at 4 C. The beads were then poured into a column, and the effluent was collected. The column was extensively washed (100 mL) with phosphate-buffered saline (0.01 M sodium phosphate-0.15 M sodium chloride, pH 7.4; PBS) until the CG-like material, measurable by the CG sequential IRMA, in the wash had reached a plateau at 0.070-0.135 mlU/mL (5-10 pg/mL). The material remaining on the column was eluted with 3.5 M MgCl2. The MgCl2 was then removed from the eluted protein by passing the solution through a Beckman buffer exchange column (Beckman, Fullerton, CA) that had been equilibrated with PBS. Samples of the effluent, wash, and eluate were saved for assay in the CG and LH IRMAs (14, 20). The samples were measured in the LH IRMA to assess the exact concentration of LH present after immunopurification. This concentration was used to determine the percent reaction in the CG IRMA. The purification of the LH preparation was performed independently four times. IRMAs The IRMAs for CG and LH that were developed in this laboratory (14,20) were used with some modifications. In assays designated simultaneous assays, the biotinylated and radiolabeled monoclonal antibodies were both added at the onset of the assay. In assays designated sequential assays, the biotinylated antibody was incubated with the assay components (CG or LH, horse serum, polystyrene bead) for 16-20 h. The beads were then washed twice with PBS. A 1:1 mixture of radiolabeled antibody and horse serum (total volume, 200 /xL) was added, and the beads were incubated for 2 h (LH assay) or 5 h (CG assay). The beads were then washed with washing buffer (0.9% NaCl-0.1% Triton X-100) and counted on a Packard 7-counter (Packard, Downers Grove, IL). The sensitivities were less than 0.02 mlU/mL for both the LH and CG IRMAs. The intraassay variations averaged 10.5% for the CG IRMA and 3.7% for the LH IRMA, while the interassay variations averaged 4% for the CG IRMA and 6% for the LH IRMA. Polyacrylamide gel electrophoresis (PAGE) LH and CG were run on 15% polyacrylamide gels by the method of Laemmli (21) and were stained with Coomassie blue. Western blotting (22) was carried out onto charged HybondECL nitrocellulose paper (Amersham, United Kingdom), which was then incubated with an anti-CG carboxyl-tail antibody (12) (1:250 final dilution). This primary antibody was diluted in a 1% solution of dried milk. The second antibody, goat antirabbit immunoglobulin G heavy and light chains (61-610, ICN, Costa Mesa, CA; 350,000 cpm/mL), was radiolabeled by the chloramine-T method and incubated overnight with the nitrocellulose paper with shaking at room temperature. The paper was then washed and exposed to Hyperfilm-ECL (Amersham) film for
JCE & M • 1991 Vol 72 • No 4
16 h at —80 C. An intensifier screen was also present in the film cassette. Three hundred international units (25 ^g) of LH were run in each lane. Gels were run under reducing (mercaptoethanol in the loading buffer) and nonreducing conditions. A control Western blot was performed using anti-CG carboxyltail antibody that had been presaturated with 50 /tig/mL carboxyl-tail peptide (Beckman). Carboxyl-tail RIA The LH preparation was assayed in a double antibody equilibrium-type RIA (12). The primary antibody (used at a 1:2500 final dilution) was a rabbit antibody produced against the unique 37-amino acid carboxyl-tail segment present in CG but absent in LH. The second antibody was a goat antirabbit serum used at a 1:50 final dilution. The labeled tracer was CG (CR121), which had been labeled with the chloramine-T method and purified on a Concanavilin-A (Pharmacia) column (23). The intra- and interassay variations of the carboxyl-tail assay were 7% and 11%, respectively. Sephadex G-100 and ion exchange purification Three hundred international units (25 ng) of the LH preparation were diluted in 1 mL PBS and applied to a 1.5 X 100cm Sephadex G-100 column. Fractions (0.5 mL) were collected and measured in the sequential CG and LH IRMAs. The column was standardized with bovine albumin, ovalbumin, chymotrypsin, and ribonuclease. The mol wt of the LH and CG-like material was determined from the Rf calculated for the peak fraction of each protein. One hundred and eighty international units (15 ng) of the LH preparation were diluted in 25 HIM Tris-HCl buffer, pH 7.5 (TB), and applied to a Whatman DE-52 anion exchange column that had been equilibrated with the same buffer. The column was washed with this buffer before beginning elution with a 100-mM sodium chloride gradient in TB. The 0.5-mL fractions that were collected were measured in the sequential CG and LH IRMAs. Chromatofocusing Chromatofocusing was performed on the LH preparation with a polybuffer exchanger (Pharmacia). Three hundred international units (25 ng) LH were dissolved in 0.5 mL starting buffer (0.025 M imidazole-HCl, pH 7.0) and applied to a 1 X 30 cm column of polybuffer exchanger 94 that had been equilibrated with starting buffer. The eluting buffer was polybuffer 74-HC1 at a pH of 3. Three-milliliter fractions were collected and assayed in the sequential LH and CG IRMAs
Results IRMAs Figure 1, A and B, demonstrates the dose-response curves in the CG IRMA produced by urinary CG obtained from pregnant females (CR-121), LH, and i-LH. Figure 1A represents the results obtained when the assay is run as a simultaneous assay, while Fig. IB shows the results
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 16 November 2015. at 14:46 For personal use only. No other uses without permission. . All rights reserved.
CG-LIKE MATERIAL IN A LH PREPARATION
843
100i
150001
80
10000-
605000
400 10 2 10-'10° 1 0 ' 1 0 2 1 0 3 10"
10000;
hCG LH Purified LH
hCX3 LH
20-
0
10'
10 3
10 4
105
10e
mlU HORMONE
8000-
FIG. 2. Dose-response curves produced by CG (CR-121) and LH in the carboxyl-tail RIA. The LH used in this assay had not been immunopurified. %B/BO, percent specific binding.
6000ffl
10 2
400020000 0
10 2 1 O ' 1 O ° 1 0 ' 1 0 2 1 0 3 1 0" mlU HORMONE
FIG. 1. Dose-response curves produced in the CG IRMA by CG obtained from the urine of pregnant humans (CR-121), LH, and purified LH (i-LH). A, Simultaneous assay. B, Sequential assay. TABLE 1. Percent reaction and percent contamination of LH (AFP0642B) with a CG-like material, as assessed by simultaneous and sequential CG IRMAs Simultaneous Sequential assay assay (n = 8) (n = 4) Cross-reaction 0.03 ± 0.02° 0.13 ± 0.02 Contamination 0.15 ± 0.04 0.17 ± 0.03 n=4 n=8 n, Number of times assay was performed. Results are expressed as the mean ± SEM from independent assays. 0 The cross-reaction in the sequential assay is not statistically different from zero.
obtained in a sequential assay. In both types of assays, the CG-like material measured in the LH preparations is reduced after immunopurification. The percent reaction of LH before immunopurification was accepted as a combination of cross-reaction and contamination. Since greater than 85% of the CG can be removed from a sample with the purification employed here (our unpublished data), the percent reaction remaining of the i-LH in the assays was accepted as cross-reaction. The percent contamination of the LH with a CG-like material was determined as the difference in reaction between LH and i-LH. These results are summarized in Table 1. Notice that the contamination of LH with CG-like material is calculated to be approximately 0.16% whether calculated from simultaneous or sequential assays. Carboxyl-tail RIA Figure 2 shows the dose-response curves obtained with urinary CG obtained from pregnant females (CR-121)
1
2
3
FIG. 3. Fifteen percent SDS-polyacrylamide gel electrophoresis of: lane 1, LH with no mercaptoethanol present (i.e. nonreduced); lane 2, LH with mercaptoethanol present (i.e. reduced); and lane 3, mol wt markers (42K, 30K, 21.5K, and 14K). The LH characterized on this gel had not been immunopurified.
and LH in the carboxyl-tail RIA. The percent reaction of LH in this assay was 0.013% by weight. Sodium dodecyl sulfate-PAGE (SDS-PAGE) and Western blots Figure 3 shows that nonreduced LH produced a stainable band at 21K (lane 1). Under reducing conditions, LH separated into stainable bands of 17K, 15K, and 10.5K (lane 2). A Western blot of the LH preparation is shown in Fig. 4. In the reduced condition (lane 1), the anti-CG carboxyl-tail antibody highlighted six bands,
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 16 November 2015. at 14:46 For personal use only. No other uses without permission. . All rights reserved.
SAWITZKE, GRIFFIN, AND ODELL
844
JCE & M • 1991 Vol 72 • No 4
3500
75K —
3000 2500 UJ
50K —
2000 1500" 1000-
39K —
500-
oo
27 K —
50
17K —
70
90
110
130
I 5000i u ^ 4000" 30002000-
1
1OOO-
2
FIG. 4. Western blot of: lane 1, LH with mercaptoethanol present (i.e. reduced); and lane 2, LH without mercaptoethanol present (i.e. nonreduced). The primary antibody was a polyclonal antibody against the carboxyl-tail of CG. The LH characterized by this blot had not been immunopurified.
64.4K, 56.8K, 34.8K, 29.0K, 20.7K, and 10.5K. The major bands were at 20.7K and 10.5K. Only the 10.5K band corresponds to a band that was present in sufficient quantity to be stained with Coomassie blue. In the nonreduced condition (lane 2), only two bands, 60.8K and 50.7K, were highlighted. Neither of these bands was shown with Coomassie blue staining. The negative control blot, using anti-CG carboxyl-tail antibody that had been presaturated with carboxyl-tail peptide, showed no highlighted bands (data not shown).
50
Chromatofocusing Figure 6 demonstrates the LH and CG activities measured in fractions collected from a polybuffer exchange
130
FIG. 5. LH and CG activity measured by the specific LH and CG IRMAs in fractions collected from a Sephadex G-100 column. A, LH was applied to the Sephadex column. B, After immunopurification on an anti-CG affinity column, i-LH was applied to the Sephadex column.
Sephadex G-100 and ion exchange purification Figure 5, A and B, demonstrates the LH and CG activities measured in fractions from a Sephadex G-100 column. In Fig. 5A, CG-like material was measured in fractions 85-112. LH was also measured over this range. The mol wt marker ovalbumin (43K) was detected by absorbance at 280 nm in fractions 85-110. The mol wt of the LH component was 41.5K, while that of the CGlike component was 40.0K. Figure 5B shows that there was no longer any detectable CG-like activity in the fractions when i-LH was applied to the Sephadex G-100 column. Both the LH and the CG activities were eluted from the ion exchange column at the same salt concentration. Thus, the two components were not separable by this method.
70 90 110 FRACTION NUMBER
^o
0
20
40
20
40
60
60
80
100
80
FRACTION NUMBER
°^
FIG. 6. LH and CG activities measured by the specific LH and CG IRMAs in fractions collected from a polybuffer chromatofocusing column. The upper panel represents the pH of the fractions collected from the column. The lower panel represents the activity of LH or CG in the fractions collected from the column. The LH applied to this column had not been immunopurified.
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 16 November 2015. at 14:46 For personal use only. No other uses without permission. . All rights reserved.
CG-LIKE MATERIAL IN A LH PREPARATION column. The top panel depicts the pH values for the fractions collected. The LH activity was eluted over a pH range of 7.3-6.3, while the majority of the CG-like material was eluted from pH 7.3-6.8, indicating that it has a slightly higher pi value than LH.
Discussion The CG-like material identified in LH is similar enough to urinary pregnancy CG to react with a very specific monoclonal antibody and with the carboxyl-tail antibody; however, it differs from urinary pregnancy CG in that it cannot be separated from the LH by size or ion exchange chromatography. The CG-like material can be partially separated from LH by chromatofocusing. As demonstrated by the Western blots in Fig. 4, more than one form of this CG-like material exists. One possible explanation for this size heterogeneity may be differences in glycosylation. Wang et al. (24) have purified and characterized a small form (43K by gel filtration) of CG from acetone powder extracts of first trimester placentas. The pi of this small CG is 10.0, and upon reduction, two /3-subunits at 23K and 20K and an a-subunit at 21K appear on SDS-PAGE. Carbohydrate studies demonstrate no 0linked and shortened JV-linked oligosaccarides on this small CG (24). The CG-like material we have identified in the pituitary LH preparation, with a mol wt by gel chromatography of 40.0K and a pi slightly higher than that of LH, may be similar or identical to this small placental CG. The CG measured in the serum of nonpregnant subjects by Stenman et al. (15) also showed a 43K mol wt upon gel filtration. The Western blots shown in this study indicate mol wt of 60.8K and 50.7K for the CG; however, these values are not incompatible with a mol wt of 40-43K determined by gel filtration. The true mol wt of glycosylated proteins cannot be determined by SDS-PAGE due to the effect of the carbohydrates. When urinary pregnancy CG (CR-121) is analyzed with SDSPAGE, the a- and /3-chains stain at 18K and 39K for a combined mol wt of 57K. Analysis of this preparation on Sephadex G-100 results in a calculated mol wt of 73K. The very small size (10.5K) of the SDS-PAGE band formed by LH under reducing conditions, which reacts with the anti-CG carboxyl-tail antibody, is intriguing. This band also reacts with a monoclonal antibody that is specific for the /3-chain of LH/CG (data not shown). The observation of low mol wt fragments from LH after disulfide reduction has been reported previously (25-27). Interestingly, Ward et al (28) reported that LH (AFP0642B) contained more of this small fragment than other LH preparations. This small fragment presumably originates from protein nicking during processing of the pituitary glands. Because the 10.5K band reacts with the
845
anti-CG carboxyl-tail antibody, we feel the CG-like material must also be nicked during processing. The stainability of the band indicates that a relatively large quantity of this fragment is present in the LH preparation; thus, both the LH and the CG-like material may be nicked to give 10.5K fragments. Nishimura et al. (29) found that when urinary CG was subjected to PAGE under reducing conditions, a small mol wt peptide (25K, or 18K after neuraminidase treatment) was separated from the /3-subunit. This small peptide, which bound the carboxyl-tail antibody in Western blots, was more apparent in CG from a patient with choriocarcinoma than in normal pregnancy CG (29). Puisieux et al. (30) also demonstrated the existence of several fragments of the CG /3-subunit on SDS-PAGE and Western blots. Both of these research groups worked with urinary CG, but the small fragments they demonstrated may correspond to the 29K and 20.7K bands that we have described. The above studies indicate that both LH and CG are subject to specific protein nicking. The site of production and/or the amount of glycosylation of the hormone may alter its susceptibility to this nicking. Stenman et al. (15) showed an increase in LH cross-reaction in their CG assay after storage or purification of the LH, perhaps reflecting the unveiling of a previously masked epitope upon breakdown of the hormone. There is only one known gene that codes for the (3chain of LH, while as many as six to eight genes or pseudogenes have been identified for the /3-chain of CG (31, 32). It is not known how many of these CG genes are translated. Possibly, some of the genes are expressed in the placenta, which lead to the traditional, highly glycosylated CG, and different genes are expressed in the pituitary gland, which lead to a slightly different and perhaps less glycosylated form of CG. This would explain some of the differences we have encountered between urinary pregnancy CG and the CG-like material that exists in the pituitary. Further studies, including carbohydrate and amino acid analysis will help determine the true nature of this CG-like material.
References 1. Odell WD, Hertz R, Lipsett MB, Ross GT, Hammond CB. Endocrine aspects of trophoblastic neoplasms. Clin Obstet Gynecol. 1967;10(Suppl 2):290-302. 2. Braunstein GD, Vaitukaitis JL, Carbone PP, Ross GT. Ectopic production of human chorionic gonadotropin by neoplasms. Ann Intern Med. 1973;78:39-45. 3. Odell WD, Wolfsen A, Yoshimoto Y, Weitzman D, Fisher D, Hirose F. Ectopic peptide synthesis: a universal concomitant of neoplasia. Trans Assoc Am Physicians. 1977;90:204-27. 4. Yoshimoto Y, Wolfsen AR, Odell WD. Glycosylation: a variable in the production of hCG by cancers. Am J Med. 1979;67:414-20. 5. Braunstein GD, Rasor J, Wade ME. Presence in normal human testes of a chorionic-gonadotropin-like substance distinct from human luteinizing hormone. N Engl J Med. 1975;293(Suppl
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 16 November 2015. at 14:46 For personal use only. No other uses without permission. . All rights reserved.
846
SAWITZKE, GRIFFIN, AND ODELL
26):1339-43. 6. Yoshimoto Y, Wolfsen AR, Odell WD. Human chorionic gonadotropin-like material in nonendocrine tissues of normal subjects. Science. 1977;197:575-7. 7. Braunstein GD, Kamdar V, Rasor J, Swaminathan N, Wade ME. Widespread distribution of a chorionic gonadotropin-like substance in normal human tissues. J Clin Endocrinol Metab. 1979;49(Suppl 6):917-25. 8. Yoshimoto Y, Wolfsen AR, Hirose F, Odell WD. Human chorionic gonadotropin-like material: presence in normal human tissues. Am J Obstet Gynecol. 1979b;134(Suppl 7):729-33. 9. Borkowski A, Muquardt C. Human chorionic gonadotropin in the plasma of normal, nonpregnant subjects. N Engl J Med. 1979;301(Suppl 6):298-302. 10. Borkowski A, Puttaert V, Gyling M, Muquardt C, Body JJ. Human chorionic gonadotropin-like substance in plasma of normal nonpregnant subjects and women with breast cancer. J Clin Endocrinol Metab. 1984;58(Suppl 6):1171-8. 11. Chen H, Hodgen GD, Matsuura S, et al. Evidence for a gonadotropin from nonpregnant subjects that has physical, immunological and biological similarities to human chorionic gonadotropin. Proc Natl Acad Sci USA. 1976;73(Suppl 8):2885-9. 12. Matsuura S, Chen H-C, Hodgen GD. Antibodies to the carboxylterminal fragment of human chorionic gonadotropin B-subunit: characterization of antibody recognition sites using synthetic peptide analogues. Biochemistry. 1978;17(Suppl 4):575-80. 13. Matsuura S, Ohashi M, Chen H-C, et al. Physicochemical and immunological characterization of an hCG-like substance from human pituitary glands. Nature. 1980;286:740-l. 14. Griffin J, Odell WD. Ultrasensitive immunoradiometric assay for chorionic gonadotropin which does not cross-react with luteinizing hormone nor free beta chain of hCG and which detects hCG in blood of non pregnant humans. J Immunol Methods. 1987;103:27583. 15. Stenman U, Alfthan H, Ranta T, Vartiainen E, Jalkanen J, Seppala M. Serum levels of human chorionic gonadotropin in nonpregnant women and men are modulated by gonadotropin-releasing hormone and sex steroids. J Clin Endocrinol Metab. 1987;64(Suppl 4):7306. 16. Odell WD, Griffin J. Pulsatile secretion of chorionic gonadotropin during the normal menstrual cycle. J Clin Endocrinol Metab. 1989;69:528-32. 17. Kyle CV, Griffin J, Odell WD. GnRH stimulation of human chorionic gonadotropin secretion in normal men and women [Abstract]. Clin Res. 1988;36(Suppl 1):181A. 18. Hammond E, Griffin J, Odell WD. A chorionic gonadotropin
Vol72«No4
secreting human pituitary cell. J Clin Endocrinol Metab. In Press. 19. Odell WD, Griffin J, Bashey HM, Snyder PJ. Secretion of chorionic gonadotropin by cultured human pituitary cells. J Clin Endocrinol Metab. 1990;71:1318-21. 20. Odell WD, Griffin J. Two-monoclonal-antibody "sandwich"-type assay of human lutropin, with no cross reaction with choriogonadotropin. Clin Chem. 1987;33(Suppl 9):1603-7. 21. Laemmli UK. Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature (Lond). 1970;227:680-5. 22. Towbin H, Staehelin T, Gordon J. Electrophoretic transfer of proteins from polyacrylamide gels to nitrocellulose sheets: procedure and some applications. Proc Natl Acad Sci USA. 1979;76(Suppl 9):4350-4. 23. Patritti-Laborde N, Yoshimoto Y, Wolfsen AR, Odell WD. Improved method of purifying some radiolabeled glycopeptide hormones. Clin Chem. 1979;25(Suppl l):163-5. 24. Wang HY, Segal SJ, Koide SS. Classification and characterization of an incompletely glycosylated form of human chorionic gonadotropin in human placenta. Endocrinology. 1988;123(Suppl 2):795803. 25. Courte C, Willimot J. Heterogeneity of porcine pituitary luteinizing hormone subunits (LH alpha and LH beta) in 6 M guanidine hydrochloride agarose gel chromatography. J Biol Chem. 1972;247(Suppl 13):4429-31. 26. Ward DN, Reichert Jr LE, Liu W-K, et al. Chemical studies of luteinizing hormone from human and ovine pituitaries. Recent Prog Horm Res. 1973;29:533-61. 27. Reichert Jr LE, Lawson Jr GM. Molecular weight relationships among the subunits of human glycoprotein hormones. Endocrinology. 1973;92:1034-42. 28. Ward DN, Glenn SD, Nahm HS, Wen T. Characterization of cleavage products in selected human lutropin preparations. Int J Peptide Protein Res. 1986;27:70-8. 29. Nishimura R, Ide K, Utsunomiya T, et al. Fragmentation of the /?subunit of human chorionic gonadotropin produced by choriocarcinoma. Endocrinology. 1988;123:420-5. 30. Puisieux A, Bellet D, Troalen F, et al. Occurrence of fragmentation of free and combined forms of the /3-subunit of human chorionic gonadotropin. Endocrinology. 1990;126(Suppl 2):687-94. 31. Policastro P, Ovitt CD, Hoshina M, Fukuoka H, Boothby M, Boime I. The subunit of human chorionic gonadotropin is encoded by multiple genes. J Biol Chem. 1983;258(Suppl 19):11492-9. 32. Talmadge K, Boorstein WR, Fiddes JC. The human genome contains seven genes for the B-subunit of chorionic gonadotropin but only one gene for the B-subunit of luteinizing hormone. DNA. 1983;2:279-87.
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 16 November 2015. at 14:46 For personal use only. No other uses without permission. . All rights reserved.
