0021-972X/79/4906-0917$02.00/0 Journal of Clinical Endocrinology and Metabolism Copyright © 1979 by The Endocrine Society
Vol. 49, No. 6 Printed in U.S.A.
Widespread Distribution of a Chorionic GonadotropinLike Substance in Normal Human Tissues* GLENN D. BRAUNSTEIN, VIKRAM KAMDAR, JOAN RASOR, N. SWAMINATHAN, AND MACLYN E. WADE Departments of Medicine and Obstetrics and Gynecology, Cedars-Sinai Medical Center and UCLA School of Medicine, Los Angeles, California 90048
hCG RIA. An absence of parallelism was found between extracts of nonpituitary tissues and LH in the /5>-LH RIA. Pancreatic extracts altered the [l25I]hCG used as the labeled ligand in these assays, which led to spurious results. Chromatography of the extracts on Concanavalin A-Sepharose columns revealed that the hCG-like materials from different tissues varied widely in their adsorbtion to Concanavalin A, possibly reflecting differences in their carbohydrate contents. These results indicate that an hCG-like substance is widely distributed throughout normal human tissues and further supports the concept that the fetal genome responsible for hCG production is not completely suppressed in adult tissues. (J Clin Endocrinol Metab 49: 917, 1979)
ABSTRACT. Recent studies have demonstrated that the normal human testes, colon, and liver contain a substance that resembles hCG. To extend these findings, we examined aqueous extracts of a variety of normal human tissues for the presence of this material. The /?-hCG RIA, rat Leydig cell radioreceptor assay, and a newly developed, highly specific hCG RIA were used to measure hCG activity in a serial dilutions of the extracts. Detectable concentrations of the hCG-like material were found in 146 of the 149 individual tissue samples studied. Parallelism was noted between the hCG standard and serial dilutions of extracts of testis, ovary, pituitary, lung, liver, kidney, spleen, stomach, placenta, and some small intestinal tissue samples in the /?-HCG RIA, radioreceptor assay, and the highly specific
T
used a newly developed hCG-specific RIA to demonstrate a hCG-like material in concentrated urinary extracts from postmenopausal women, the urine from a patient with Klinefelter's syndrome (4), and, more recently, the urine of normal men and women (5). Similar studies performed by other investigators have confirmed the finding of an immunoreactive hCG-like material in the testes, pituitary, and urine extracts from nonpregnant humans (6, 7). Together, these studies indicate that a hCG-like glycoprotein is produced in nonpregnant individuals. The present investigation was undertaken in order to better define the distribution and content of this substance in normal human tissues.
HE SYNCYTIOTROPHOBLASTIC cells of the human placenta secrete large quantities of hCG throughout pregnancy (1). Although gestational and nongestational trophoblastic tumors and a variety of nontrophoblastic neoplasms may produce this glycoprotein hormone, until recently the placenta has been considered to be the only physiological source of hCG. In 1975, we demonstrated the presence of an hCG-like substance in extracts of normal human testes (2). This material resembled hCG in a RIA relatively specific for the hormone bound to the plant lectin, Concanavalin A (Con A), in a manner similar to purified hCG and cochromatographed with hCG on a Sephadex G-100 column (2). Subsequent studies by Yoshimoto and associates indicated that extracts of normal human liver and colon also contained a substance that resembled hCG immunologically and which was capable of displacing 125I-labeled hCG in a radioreceptor assay (3). These investigators also noted that the hCG-like substance had a variable degree of adsorption to Con A, suggesting that the material was not completely glycosylated (3). Chen and coworkers
Materials and Methods Tissues Tissues were obtained at the time of autopsy from patients who had died of nonneoplastic disorders. Ninety-seven percent of the autopsies were performed within 36 h of death. Twelve grossly normal ovaries were obtained at surgery from women undergoing hysterectomy. The tissues were weighed and frozen at -20 C until used. Testicular tissue was decapsulated before weighing.
Received March 12, 1979. Address requests for reprints to: Dr. Glenn Braunstein, Division of Endocrinology, Cedars-Sinai Medical Center, Box 48750, Los Angeles, California 90048. * This work was supported by USPHS Research Grants CA-19392 and CA-18362. It was presented in part at the 60th Annual Meeting of The Endocrine Society, June 14, 1978, Miami, FL.
Tissue extractions The tissues were individually thawed and homogenized with 5 vol (wt/vol) 0.01 M phosphate-buffered saline (PBS), pH 7.8. 917
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 20 November 2015. at 09:27 For personal use only. No other uses without permission. . All rights reserved.
BRAUNSTEIN ET AL.
918
After extraction for 2-16 h at 4 C under constant agitation, the homogenates were centrifuged at 1,060 X g for 30 min and ultracentrifuged at 78,000 X g for 90 min at 4 C. The supernatants were lyophilized and reconstituted in distilled water just before use. Recovery of hCG by this method was determined by the addition of [125I]hCG to the organs before homogenization. After extraction and reconstitution, the remaining 125I was counted in a Searle Diagnostics model 1195 y-counter, (Searle Analytic, Chicago, IL). For the radioreceptor and Con A adsorption studies, the supernatants with high protein concentrations were ultrafiltrated through an Amicon UM-100A membrane (Amicon Corp., Lexington, MA), and the filtrates containing the hCG-like material were concentrated through a UM-10 membrane (Amicon). The protein concentrations of the partially purified extracts were measured by the method of Lowry et al. (8). Radioligand assays Immunoreactive hCG was measured in two separate RIAs. The first was the previously described /?-subunit hCG RIA which uses an antiserum generated against the purified /?-subunit of hCG (SB-6), [125I]hCG as the labeled ligand, and highly purified hCG preparations (CR 117 or CR 119) as the reference standards (9). In our laboratory, 1 mg of these preparations was equivalent to 5,000 IU immunoreactive activity and 10,000 IU biological activity relative to the Second International Standard for hCG. The amounts of purified hormone preparations required for 50% inhibition of [125I]hCG with the SB-6 antiserum are shown in Table 1. The reagents for this assay were generously supplied by the Hormone Distribution Officer, NIAMDD, NIH. The second RIA also uses an antiserum raised against a highly purified preparation of the /?-subunit of hCG. The methods of preparation of the /?-subunit and the antibody have been described (10). The antiserum, SN-4, was passed through an immunoabsorbant column containing partially purified LH (LER 1724; 1705 IU/mg) covalently coupled to Sepharose 4B by the cyanogen bromide reaction (10). The resulting SN-4TABLE 1. Cross-reaction of purified and standard hormone preparations in the £-hCG and /?-LH RIAs
Hormone hCG (CR 117) hCG-0 (CR 115/8) hCG-a (CR 117a) LH (LER 960) LH (LER 907) LH-0 (AFP-290-jS) LH-/3 (LER 1793-0) WHO IRP-HPLH (68/ 40) WHO 2nd IRP-HMG WHO 1st IRP FSH/ LH (69/104)
Reported biological potency (IU/mg) 10,600 40.3 0.8 4,620 60 7,000 4.4 15.0
Amount (ng) required for 50% inhibition SB-6
SN-4-adsorbed
Anti-LH-0
0.48 0.20 105.7 20.2 530.3 5.0 ND* 16.3
3.2 2.1 > 1,000.0 3,000.0 35,000.0 >25.0 ND ND
NP" NP >1,000.0 3.57 118.2 0.49 0.53 1.22
2,225.0 3,133.3
20,655.0 ND
110.0 350.0
The amounts of each hormone on a weight basis required to produce 50% inhibition of binding of [125I]hCG to the SB-6 and SN-4-adsorbed antisera in the /?-hCG RIAs or [12SI]LH-0 in the 0-LH RIA are shown. Each number represents the mean from 2-10 separate assays. " NP, Nonparallel to LH standard. * ND, Not determined.
JCE & M . 1979 Vol 49 , No 6
adsorbed antiserum was highly specific for hCG. This antibody required that the disulfide bonds of the /?-subunit be intact and, therefore, was responsive to the secondary structure of hCG. The sensitivity of the assay was 0.45 ng hCG/tube (11). CR 117 or CR 119 was used as standard and for iodination. The crossreactions of purified hormone preparations in this assay are shown in Table 1. The inter- and intraassay coefficients of variation were 13.9% and 8.2%, respectively. The previously described rat Leydig cell radioreceptor assay was used to demonstrate binding of the hCG-like substance to hCG receptors (12). CR 117 and the Second International Standard for hCG were used as reference preparations. Purified a- and /2-subunits of hCG did not displace [125I]hCG from the receptors (10). Immunoreactive LH was measured in a RIA which utilizes a rabbit antiserum raised against the /?-subunit of LH (NIH batch 1). The reagents for this assay were kindly supplied by the Hormone Distribution Officer NIAMDD, NIH (Bethesda, MD). The assay was performed according to the kit instructions, except that hCG was not added to the tubes. The crossreactions of purified hormone preparations in this assay are shown in Table 1. Validation of radioligand assays for examination of tissue extracts Several control studies were carried out in order to determine the presence of proteolytic enzyme or other interfering substances in the extracts that may alter the [125I]hCG, anti-/?-hCG serum, or second antibody reaction. Tracer quantities of [125I]hCG were incubated for 16 h at 4 C with extracts from each of the organs studied. After incubation, 1-ml aliquots were applied to a 1.5 x 85-cm column of Sephadex G-100 (Pharmacia Laboratories, Inc., Piscataway, NJ) equilibrated with 0.05 M Tris-hydrochloride-0.15 M sodium chloride buffer, pH 7.4. Onemilliliter samples of the effluent were collected. The void volume was determined by the peak elution volume of blue dextran measured at an optical density of 600 ran, while the volume of exclusion for the [125I]hCG incubated with tissue extracts was taken as the effluent tube with the highest counts per min of 125 I. The ratio of the volume of exclusion to the void volume was determined and compared to that obtained with [125I]hCG incubated in 2.5% normal rabbit serum before chromotography. The residual immunoprecipitability of the [125I]hCG after incubation with the organs and Sephadex G-100 chromatography was determined by the addition of excess anti-/8-hCG serum and subsequent immunoprecipitation with sheep or goat antirabbit globulin. The degree of nonspecific precipitation was assessed in the presence of excess (10 IU) unlabeled hCG. Evaluation of the residual immunoprecipitability of [125I]hCG which had been carried through the entire extraction procedure with the tissues was performed in a similar manner but without prior Sephadex G-100 chromatography. Since studies by Richert and Ryan (13, 14) and Maruo et al. (15) have shown that proteolytic enzymes may give parallel dose-response curves to the hCG standard in the hCG radioligand assays, several proteolytic enzyme inhibitors were tested for their ability to inhibit the binding of serial dilutions of the tissue extracts in the /8-hCG RIA. These inhibitors included: benzamidine, 156 /ig/tube (Aldrich Chemical Co., Inc., Milwau-
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 20 November 2015. at 09:27 For personal use only. No other uses without permission. . All rights reserved.
hCG IN NORMAL HUMAN TISSUES kee, WI); n-a-p-tosyl-L-lysine chloromethyl ketone-HCl, 1 jug/ tube (Sigma Chemical Co., St. Louis, MO); apoprotinin, 44 KlU/tube (Mobay Chemical Corp., New York, NY); ovalbumin, 100 jug/tube (Sigma); soybean trypsin inhibitor 20 jug/tube (Sigma); mercuric chloride, 272 jug/tube (J. T. Baker Chemical Co., Phillipsburg, NJ); and phenylmethylsulfonyl fluoride, 174 jug/tube (Pierce, Rockford, IL). The dose-response curves for the immunoreactive hCG in the tissue extracts with the inhibitors were compared to those generated without the inhibitors. In addition, the effect of placing the tissue extracts in a boiling water bath for 7 min before the addition of the labeled ligand and antibody in the /?-hCG RIA was examined. Two additional studies were carried out in order to detect enzymatic degradation of the [125I]hCG by the tissue extracts. After 16 h of incubation at 4 C of the labeled ligand with either boiled (7 min) or unboiled tissue extracts, 5 /il of each incubate were applied to Whatman 3 M M paper (Whatman Inc. Clifton, NJ) and subjected to high voltage electrophoresis (3000 volts for 3 h, 100-120 mA) at pH 2.5 in a Savant electrophoretic apparatus (Hicksville, NY). The paper was dried and cut into 1 X 2-cm strips and counted in a Y-counter. The patterns obtained with the boiled extracts, in which the tissue enzymes were assumed to be inactivated, were compared to those obtained with the extracts that were not boiled. In a separate experiment, 2- to 5-ju.l samples of each of the above incubates were spotted on Silica Gel G plates (Eastman Kodak, Rochester, NY) and developed for 16 h in a solvent consisting of methyl acetate, isopropanol, and ammonium hydroxide (9:7:1.5, vol/vol). The plates were dried and exposed to x-ray-sensitive film for 7 days, and the x-rays were developed. Under the conditions employed for both the high voltage electrophoresis and the thin layer chromatography, unlabeled hCG did not move from the origin. These methods have previously been shown to separate the small peptides released from controlled enzymatic digestion of hCG or its subunits (11). The tissue extracts were examined for the presence of solubilized hCG receptors that would compete with the binding protein for [125I]hCG in the various radioligand assays. The Sephadex G-100 column elution pattern of [125I]hCG chromatographed alone was compared to those obtained when [125I]hCG was chromatographed after incubation for 16 h at 4 C with tissue extracts from each of the organs. Tissue receptors were considered to be present if, after incubation of the tracer with the tissue extracts, an increased amount of radioactivity was detected in the void volume or a radioactive peak or shoulder was found between the void volume and the [I25I]hCG peak. The extracts were also examined for the presence of solubilized receptors by a modification of the method of Catt and Dufau (16). To 0.4 ml of the human tissue extracts, 0.1 ml [125I]hCG (-26,000 cpm) and 0.2 ml PBS with 1 mg bovine serum albumin (Sigma) were added. Background tubes contained the same reagents plus excessive unlabeled hCG (100IU; APL, Ayerst Laboratories, New York, NY). After incubation for 16 h at 4 C, the [I25I]hCG-receptor complex was precipitated by the addition of 0.5 ml 30% polyethylene glycol (PEG; mol wt, 6000-7500; wt/vol) in PBS, giving a final PEG concentration of 12.5% (wt/vol). The tubes were centrifuged at 2000 X g for 10 min and the supernatants were discarded. The precipitates were redissolved in 0.7 ml 0.1% Triton X-100 (Eastman Kodak)
919
in phosphate-buffered saline, incubated at 4 C for 10 min, and reprecipitated with 0.5 ml 30% PEG. After centrifugation and aspiration of the supernatants, the precipitates containing [125I]hCG bound to the receptors were counted. Positive controls for the receptor assay were prepared from rat testes by the methods described by Catt and Dufau (16). Con A column chromatography Disposible 0.7 X 4-cm columns (Bio-Rad Laboratories, Inc., Richmond, CA) containing 0.25 ml Sephadex G-25 and 1.5 ml Con A covalently coupled to Sepharose 4B (Pharmacia Laboratories) were extensively washed with PBS, pH 7.8, containing 1% normal rabbit serum. One to 4-ml aliquots of the tissue extracts or purified hormone preparations were applied to the columns, and the nonadsorbed substances were eluted with PBS. Elution of Con A-adsorbed glycoproteins was carried out with 0.2 M a,D-methylglucoside (MeG) in PBS. The eluted fractions were collected in 1-ml aliquots and analyzed in the /?-hCG RIA using the SB-6 antiserum. To control for tissue factors that may alter the degree of adsorption of hCG onto the Con A, tracer quantities of [125I]hCG were incubated with the tissue extracts for 16 h at 4 C before Con A chromatography. Recoveries of the [I25I]hCG in the PBS (nonadsorbed) and MeG (adsorbed) fractions were determined and compared to those found with [125I]hCG incubated in 2.5% normal rabbit serum. The degree of residual immunoprecipitability of the [125I]hCG after adsorption onto Con A was established by the addition of excess anti-yS-hCG serum to the peak MeG effluent tube, followed by precipitation with sheep or goat antirabbit globulin. Statistics The statistical methods of Rodbard were used for quality control, tests of parallelism, and dose interpolation in the radioligand assays (17). Immunological similarity between hCG and the material present in the tissue extracts was considered to be present when tbe inhibition curves generated by the hCG standards and serial dilutions of the tissue extract supernatants were parallel (18). Regression analysis and Pearson productmoment correlations were calculated on a Compu-Corp model 145-E statistician calculator. Results
The inhibition curves produced by serial dilutions of tissue extracts from testes, ovary, pituitary, lung, liver, kidney, spleen, stomach, and placenta were parallel to that obtained with the purified hCG reference preparation in the /?-hCG RIA which used the SB-6 antiserum (Fig. 1, top). Parallelism was also noted between these extracts and the hCG standard in the /?-hCG RIA with the SN-4-adsorbed antiserum (Fig. 1, bottom) and in the hCG radioreceptor assay (Fig. 2). Significant deviation from parallelism in these assays was noted with most pancreatic extracts, some small intestinal extracts (Fig. 1, bottom), and extracts with high protein concentrations (>70 mg/ml). With the exception of the pituitary ex-
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 20 November 2015. at 09:27 For personal use only. No other uses without permission. . All rights reserved.
BRAUNSTEIN ET AL.
920
JCE&M • 1979 Vol 49 • No 6
hCG RIA WITH SB-6 ANTISERUM -•3 -•2
PITUITARY
1
STOMACH^^
V
\
X
FIG. 1. Dose-response curves for purified hormone preparations and extracts of normal human tissues in the hCG RIAs. Top, hCG RIA with SB-6 antiserum. Bottom, hCG RIA with SN-4-adsorbed antiserum. B, Counts bound in the presence of labeled and unlabeled ligand; Bo, counts bound in the presence of labeled ligand alone. Logit (B/Bo) = logc [(B/Bo)/(1 - B/Bo)]. Coordinates are logit-normalized percentage counts precipitated (B/Bo X 100) on the ordinate, and the abscissa shows the log of the mass of hCG standards and purified hormone preparations or volume of tissue extracts added per tube.
PLACENTAt]
/JhC
• LH/FSH (LER 907)
&v hCO (LER tCO)
1 10 100 1000 10.000 STANDARDS (NANOQRAMS) AND TISSUE EXTRACTS or SERUM (MICROLITERS)
100.000
hCG RIA WITH SN-4 ADSORBED ANTISERUM -•3
9580-
-•2
8060O S02 40o
-., S
3 -
-•a
10-
-•3
1
10
100
1000
10,000
STANDARDS (NANOQRAMS) AND TISSUE EXTRACTS or SERUM (MICROLITERS)
-•3
95-
FIG. 2. Dose-response curves for the purified hCG preparations (CR 119), the Second International Standard for hCG, and serial dilutions of tissue extracts in the rat Leydig cell radioreceptor assay. CR 119 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 identical to those described in Fig. 1.
90-
-•2 PLACENTA
80-
-•1 605040-
LUNQ
- -1 202nd INT. STD. hCG
KIDNEY*
10-
.01
0.1 1 10 100 STANDARDS (IU or NANOGRAMS) AND TISSUE EXTRACTS (MICROLITERS)
tracts, the tissue concentrations of immunoreactive hCG determined by RIA with the SB-6 antiserum were highly correlated with those determined with the SN-4-adsorbed antiserum (r = 0.99; P < 0.001). The amount of immunoreactive material in pituitary extracts measured with the SN-4-adsorbed antiserum was approximately 4% of that measured with the SB-6 antiserum. In the /?-LH RIA, purified LH and pituitary extracts gave dose-response curves parallel to the LH standard (Fig. 3). Purified hCG demonstrated a two-component curve. Serial dilutions of extracts of placenta, ovary,
--2
1000
kidney, small intestine, testis, and lung were not parallel to the LH standard and had slopes that more closely resembled that seen with hCG. The pancreatic extracts had steeper slopes than the LH standards. Incubation of [125I]hCG with tissue extracts for 16 h at 4 C did not alter the Sephadex G-100 chromatographic pattern of the ligand (Table 2). The residual immunoprecipitability of the [125I]hCG after tissue incubation and Sephadex chromatography was similar to tracer incubated in normal rabbit serum (Table 2). However, prolonged incubation (>48 h) of the labeled ligand with the
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 20 November 2015. at 09:27 For personal use only. No other uses without permission. . All rights reserved.
hCG IN NORMAL HUMAN TISSUES
921
95-.
9080SMALL INTESTINE
o
FIG. 3. Dose-response curves for purified hormone preparations and serial dilutions of extracts of normal human tissues in the /?-LH RIA. See Fig. 1 for abbreviations.
2 60". • • m 40o 2010—i—
—r~ —r~ 1000 0.1 100 10 1.0 STANDARDS (NANOQRAMS) AND TISSUE EXTRACTS (MICROLITERS) TABLE 2. Results of Sephadex G-100 chromatography and immunoprecipitation studies on [l2GI]hCG after incubation with tissue extracts
[l25I]hCG in NRS
Sephadex G-100
% IP after
Ve/Vo
%IP
extraction of organs with [125I] hCG
1.23-1.40
100" (82.2-122.7)
100" (91.1-108.8)
Organ 104.4 117 1.38 Spleen 93.2 1.37 Kidney 104 111.4 Lung 114 1.35 1.34 111.7 Liver 110 1.37 43.0 Small intestine 107 Pancreas 1.1 1.37 89 ND 1.34 98 Pituitary 91.5 114 Ovary 1.32 b 1.36 109 Placenta Ve, Volume of exclusion; Vo, void volume; IP, immunoprecipitable; ND, not determined. " The immunoprecipitability of [125I]hCG was assumed to be maximum when the tracer was chromatographed or extracted with 2.5% normal rabbit serum (NRS). The average value was designated as 100%; the range of four experiments is shown in parentheses. The immunoprecipitability of the labeled ligand incubated or extracted with the organs was divided by that found with the same batch of [125I]hCG incubated in NRS. This allows for the comparison of data obtained with [125I]hCG that had been iodinated at different times. 6 Excessive quantities of hCG prevented complete immunoprecipitation.
tissues during the entire extraction procedure did result in a loss of immunoprecipitability of the tracer that had been incubated with pancreatic and small intestinal extracts (Table 2). Neither boiling nor the addition of protease inhibitors to the RIA tubes altered the dose-response curves of serial dilutions of the tissue extracts for any of the tissues studied, with the exception of the pancreatic extracts which gave inconsistent results and high nonspecific counts. High voltage electrophoresis of the [125I]hCG after incubation with boiled and unboiled tissue extracts revealed that only the tracer incubated with the unboiled pancreatic extracts was sufficiently damaged to migrate
10,000
away from the origin. Thin layer chromatography did not reveal fragments of hCG after incubation with the tissue extracts. The Sephadex G-100 elution patterns of [125I]hCG alone or after incubation with various tissue extracts were qualitatively similar. No high molecular weight [125I]hCG-receptor complex was identified in the chromatograms after tissue incubation. Similarly, the addition of PEG at a final concentration of 12.5% (wt/vol) to the tubes containing human tissue extracts that had been incubated for 16 h with [125I]hCG did not result in the precipitation of more counts per min than in the background tubes with excess unlabeled hCG (Fig. 4). A large amount of nonspecific binding was noted with the pancreatic extracts, again suggesting enzymatic damage to the labeled hCG. Specific [125I]hCG binding was present in the tubes containing the rat testicular extracts treated to concentrate hCG receptors (Fig. 4). Recovery of the immunoreactive hCG during extraction varied from 23% (small intestine) to 96.7% (ovary). The median concentration and range of the hCG-like material in the various tissues corrected for recovery are given in Table 3. A sufficient number of testicular and ovarian samples were available to allow comparison of the tissue contents of the hCG-like material at different ages. After the age of 15 yr, a significant inverse correlation between hCG content and age was noted for both the testes (Fig. 5) (r = 0.52; P < 0.01) and the ovaries (Fig. 6; r = 0.42; P< 0.05). Affinity chromatography of the tissue extracts on the Con A-Sepharose columns revealed a variable degree of adsorption of the immunoreactive substance to the lectin (Fig. 7). Extracts of the testes, ovaries, placenta, and small intestine had a greater degree of adsorption to Con A than those from the lungs, liver, and kidneys. Several pancreatic extracts were chromatographed and all of the immunoreactive material that could be recovered was eluted before the application of the 0.2 M MeG. The amount of immunoreactive material recovered after Con A-Sepharose chromatography varied with each tissue (Fig. 7). The results of the Con A column chromatography of
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 20 November 2015. at 09:27 For personal use only. No other uses without permission. . All rights reserved.
BRAUNSTEIN ET AL.
922
JCE&M • 1979 Vol 49 • No 6
• '2il-hCG • TISSUE • 100 IU hCG • '"l-hCG • TISSUE
FIG. 4. Percent counts percipitated of [l25I]hCG after incubation with tissue extracts for 16 h in the presence (•) or absence (63) of excess (100 IU) hCG. Precipitation of the labeled ligand was accomplished with 12.5% PEG. Rat testicular hCG receptor preparations were used for positive control purposes. The concentration of the rat soluble testes receptor no. 1 was equivalent to 0.1 g testes, while receptor no. 2 was equivalent to 0.46 g testes. TABLE 3. Concentration of hCG-like substance in normal human tissues
50-
hCG content (ng hCG/g tissue) n
urgan Spleen Kidney Lung Liver Small intestine Stomach Adrenal Testes Ovary Pituitary Placenta
6 8 6 7 5 1 1 77 30 3 5
Median
Range
2.63 4.51 2.69 2.23 10.75 3.45 1.31 19.93 5.55 20,847" 5,396
0.38-4.33 1.72-7.19 1.68-8.38
Vol. 49, No. 6 Printed in U.S.A.
Widespread Distribution of a Chorionic GonadotropinLike Substance in Normal Human Tissues* GLENN D. BRAUNSTEIN, VIKRAM KAMDAR, JOAN RASOR, N. SWAMINATHAN, AND MACLYN E. WADE Departments of Medicine and Obstetrics and Gynecology, Cedars-Sinai Medical Center and UCLA School of Medicine, Los Angeles, California 90048
hCG RIA. An absence of parallelism was found between extracts of nonpituitary tissues and LH in the /5>-LH RIA. Pancreatic extracts altered the [l25I]hCG used as the labeled ligand in these assays, which led to spurious results. Chromatography of the extracts on Concanavalin A-Sepharose columns revealed that the hCG-like materials from different tissues varied widely in their adsorbtion to Concanavalin A, possibly reflecting differences in their carbohydrate contents. These results indicate that an hCG-like substance is widely distributed throughout normal human tissues and further supports the concept that the fetal genome responsible for hCG production is not completely suppressed in adult tissues. (J Clin Endocrinol Metab 49: 917, 1979)
ABSTRACT. Recent studies have demonstrated that the normal human testes, colon, and liver contain a substance that resembles hCG. To extend these findings, we examined aqueous extracts of a variety of normal human tissues for the presence of this material. The /?-hCG RIA, rat Leydig cell radioreceptor assay, and a newly developed, highly specific hCG RIA were used to measure hCG activity in a serial dilutions of the extracts. Detectable concentrations of the hCG-like material were found in 146 of the 149 individual tissue samples studied. Parallelism was noted between the hCG standard and serial dilutions of extracts of testis, ovary, pituitary, lung, liver, kidney, spleen, stomach, placenta, and some small intestinal tissue samples in the /?-HCG RIA, radioreceptor assay, and the highly specific
T
used a newly developed hCG-specific RIA to demonstrate a hCG-like material in concentrated urinary extracts from postmenopausal women, the urine from a patient with Klinefelter's syndrome (4), and, more recently, the urine of normal men and women (5). Similar studies performed by other investigators have confirmed the finding of an immunoreactive hCG-like material in the testes, pituitary, and urine extracts from nonpregnant humans (6, 7). Together, these studies indicate that a hCG-like glycoprotein is produced in nonpregnant individuals. The present investigation was undertaken in order to better define the distribution and content of this substance in normal human tissues.
HE SYNCYTIOTROPHOBLASTIC cells of the human placenta secrete large quantities of hCG throughout pregnancy (1). Although gestational and nongestational trophoblastic tumors and a variety of nontrophoblastic neoplasms may produce this glycoprotein hormone, until recently the placenta has been considered to be the only physiological source of hCG. In 1975, we demonstrated the presence of an hCG-like substance in extracts of normal human testes (2). This material resembled hCG in a RIA relatively specific for the hormone bound to the plant lectin, Concanavalin A (Con A), in a manner similar to purified hCG and cochromatographed with hCG on a Sephadex G-100 column (2). Subsequent studies by Yoshimoto and associates indicated that extracts of normal human liver and colon also contained a substance that resembled hCG immunologically and which was capable of displacing 125I-labeled hCG in a radioreceptor assay (3). These investigators also noted that the hCG-like substance had a variable degree of adsorption to Con A, suggesting that the material was not completely glycosylated (3). Chen and coworkers
Materials and Methods Tissues Tissues were obtained at the time of autopsy from patients who had died of nonneoplastic disorders. Ninety-seven percent of the autopsies were performed within 36 h of death. Twelve grossly normal ovaries were obtained at surgery from women undergoing hysterectomy. The tissues were weighed and frozen at -20 C until used. Testicular tissue was decapsulated before weighing.
Received March 12, 1979. Address requests for reprints to: Dr. Glenn Braunstein, Division of Endocrinology, Cedars-Sinai Medical Center, Box 48750, Los Angeles, California 90048. * This work was supported by USPHS Research Grants CA-19392 and CA-18362. It was presented in part at the 60th Annual Meeting of The Endocrine Society, June 14, 1978, Miami, FL.
Tissue extractions The tissues were individually thawed and homogenized with 5 vol (wt/vol) 0.01 M phosphate-buffered saline (PBS), pH 7.8. 917
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 20 November 2015. at 09:27 For personal use only. No other uses without permission. . All rights reserved.
BRAUNSTEIN ET AL.
918
After extraction for 2-16 h at 4 C under constant agitation, the homogenates were centrifuged at 1,060 X g for 30 min and ultracentrifuged at 78,000 X g for 90 min at 4 C. The supernatants were lyophilized and reconstituted in distilled water just before use. Recovery of hCG by this method was determined by the addition of [125I]hCG to the organs before homogenization. After extraction and reconstitution, the remaining 125I was counted in a Searle Diagnostics model 1195 y-counter, (Searle Analytic, Chicago, IL). For the radioreceptor and Con A adsorption studies, the supernatants with high protein concentrations were ultrafiltrated through an Amicon UM-100A membrane (Amicon Corp., Lexington, MA), and the filtrates containing the hCG-like material were concentrated through a UM-10 membrane (Amicon). The protein concentrations of the partially purified extracts were measured by the method of Lowry et al. (8). Radioligand assays Immunoreactive hCG was measured in two separate RIAs. The first was the previously described /?-subunit hCG RIA which uses an antiserum generated against the purified /?-subunit of hCG (SB-6), [125I]hCG as the labeled ligand, and highly purified hCG preparations (CR 117 or CR 119) as the reference standards (9). In our laboratory, 1 mg of these preparations was equivalent to 5,000 IU immunoreactive activity and 10,000 IU biological activity relative to the Second International Standard for hCG. The amounts of purified hormone preparations required for 50% inhibition of [125I]hCG with the SB-6 antiserum are shown in Table 1. The reagents for this assay were generously supplied by the Hormone Distribution Officer, NIAMDD, NIH. The second RIA also uses an antiserum raised against a highly purified preparation of the /?-subunit of hCG. The methods of preparation of the /?-subunit and the antibody have been described (10). The antiserum, SN-4, was passed through an immunoabsorbant column containing partially purified LH (LER 1724; 1705 IU/mg) covalently coupled to Sepharose 4B by the cyanogen bromide reaction (10). The resulting SN-4TABLE 1. Cross-reaction of purified and standard hormone preparations in the £-hCG and /?-LH RIAs
Hormone hCG (CR 117) hCG-0 (CR 115/8) hCG-a (CR 117a) LH (LER 960) LH (LER 907) LH-0 (AFP-290-jS) LH-/3 (LER 1793-0) WHO IRP-HPLH (68/ 40) WHO 2nd IRP-HMG WHO 1st IRP FSH/ LH (69/104)
Reported biological potency (IU/mg) 10,600 40.3 0.8 4,620 60 7,000 4.4 15.0
Amount (ng) required for 50% inhibition SB-6
SN-4-adsorbed
Anti-LH-0
0.48 0.20 105.7 20.2 530.3 5.0 ND* 16.3
3.2 2.1 > 1,000.0 3,000.0 35,000.0 >25.0 ND ND
NP" NP >1,000.0 3.57 118.2 0.49 0.53 1.22
2,225.0 3,133.3
20,655.0 ND
110.0 350.0
The amounts of each hormone on a weight basis required to produce 50% inhibition of binding of [125I]hCG to the SB-6 and SN-4-adsorbed antisera in the /?-hCG RIAs or [12SI]LH-0 in the 0-LH RIA are shown. Each number represents the mean from 2-10 separate assays. " NP, Nonparallel to LH standard. * ND, Not determined.
JCE & M . 1979 Vol 49 , No 6
adsorbed antiserum was highly specific for hCG. This antibody required that the disulfide bonds of the /?-subunit be intact and, therefore, was responsive to the secondary structure of hCG. The sensitivity of the assay was 0.45 ng hCG/tube (11). CR 117 or CR 119 was used as standard and for iodination. The crossreactions of purified hormone preparations in this assay are shown in Table 1. The inter- and intraassay coefficients of variation were 13.9% and 8.2%, respectively. The previously described rat Leydig cell radioreceptor assay was used to demonstrate binding of the hCG-like substance to hCG receptors (12). CR 117 and the Second International Standard for hCG were used as reference preparations. Purified a- and /2-subunits of hCG did not displace [125I]hCG from the receptors (10). Immunoreactive LH was measured in a RIA which utilizes a rabbit antiserum raised against the /?-subunit of LH (NIH batch 1). The reagents for this assay were kindly supplied by the Hormone Distribution Officer NIAMDD, NIH (Bethesda, MD). The assay was performed according to the kit instructions, except that hCG was not added to the tubes. The crossreactions of purified hormone preparations in this assay are shown in Table 1. Validation of radioligand assays for examination of tissue extracts Several control studies were carried out in order to determine the presence of proteolytic enzyme or other interfering substances in the extracts that may alter the [125I]hCG, anti-/?-hCG serum, or second antibody reaction. Tracer quantities of [125I]hCG were incubated for 16 h at 4 C with extracts from each of the organs studied. After incubation, 1-ml aliquots were applied to a 1.5 x 85-cm column of Sephadex G-100 (Pharmacia Laboratories, Inc., Piscataway, NJ) equilibrated with 0.05 M Tris-hydrochloride-0.15 M sodium chloride buffer, pH 7.4. Onemilliliter samples of the effluent were collected. The void volume was determined by the peak elution volume of blue dextran measured at an optical density of 600 ran, while the volume of exclusion for the [125I]hCG incubated with tissue extracts was taken as the effluent tube with the highest counts per min of 125 I. The ratio of the volume of exclusion to the void volume was determined and compared to that obtained with [125I]hCG incubated in 2.5% normal rabbit serum before chromotography. The residual immunoprecipitability of the [125I]hCG after incubation with the organs and Sephadex G-100 chromatography was determined by the addition of excess anti-/8-hCG serum and subsequent immunoprecipitation with sheep or goat antirabbit globulin. The degree of nonspecific precipitation was assessed in the presence of excess (10 IU) unlabeled hCG. Evaluation of the residual immunoprecipitability of [125I]hCG which had been carried through the entire extraction procedure with the tissues was performed in a similar manner but without prior Sephadex G-100 chromatography. Since studies by Richert and Ryan (13, 14) and Maruo et al. (15) have shown that proteolytic enzymes may give parallel dose-response curves to the hCG standard in the hCG radioligand assays, several proteolytic enzyme inhibitors were tested for their ability to inhibit the binding of serial dilutions of the tissue extracts in the /8-hCG RIA. These inhibitors included: benzamidine, 156 /ig/tube (Aldrich Chemical Co., Inc., Milwau-
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 20 November 2015. at 09:27 For personal use only. No other uses without permission. . All rights reserved.
hCG IN NORMAL HUMAN TISSUES kee, WI); n-a-p-tosyl-L-lysine chloromethyl ketone-HCl, 1 jug/ tube (Sigma Chemical Co., St. Louis, MO); apoprotinin, 44 KlU/tube (Mobay Chemical Corp., New York, NY); ovalbumin, 100 jug/tube (Sigma); soybean trypsin inhibitor 20 jug/tube (Sigma); mercuric chloride, 272 jug/tube (J. T. Baker Chemical Co., Phillipsburg, NJ); and phenylmethylsulfonyl fluoride, 174 jug/tube (Pierce, Rockford, IL). The dose-response curves for the immunoreactive hCG in the tissue extracts with the inhibitors were compared to those generated without the inhibitors. In addition, the effect of placing the tissue extracts in a boiling water bath for 7 min before the addition of the labeled ligand and antibody in the /?-hCG RIA was examined. Two additional studies were carried out in order to detect enzymatic degradation of the [125I]hCG by the tissue extracts. After 16 h of incubation at 4 C of the labeled ligand with either boiled (7 min) or unboiled tissue extracts, 5 /il of each incubate were applied to Whatman 3 M M paper (Whatman Inc. Clifton, NJ) and subjected to high voltage electrophoresis (3000 volts for 3 h, 100-120 mA) at pH 2.5 in a Savant electrophoretic apparatus (Hicksville, NY). The paper was dried and cut into 1 X 2-cm strips and counted in a Y-counter. The patterns obtained with the boiled extracts, in which the tissue enzymes were assumed to be inactivated, were compared to those obtained with the extracts that were not boiled. In a separate experiment, 2- to 5-ju.l samples of each of the above incubates were spotted on Silica Gel G plates (Eastman Kodak, Rochester, NY) and developed for 16 h in a solvent consisting of methyl acetate, isopropanol, and ammonium hydroxide (9:7:1.5, vol/vol). The plates were dried and exposed to x-ray-sensitive film for 7 days, and the x-rays were developed. Under the conditions employed for both the high voltage electrophoresis and the thin layer chromatography, unlabeled hCG did not move from the origin. These methods have previously been shown to separate the small peptides released from controlled enzymatic digestion of hCG or its subunits (11). The tissue extracts were examined for the presence of solubilized hCG receptors that would compete with the binding protein for [125I]hCG in the various radioligand assays. The Sephadex G-100 column elution pattern of [125I]hCG chromatographed alone was compared to those obtained when [125I]hCG was chromatographed after incubation for 16 h at 4 C with tissue extracts from each of the organs. Tissue receptors were considered to be present if, after incubation of the tracer with the tissue extracts, an increased amount of radioactivity was detected in the void volume or a radioactive peak or shoulder was found between the void volume and the [I25I]hCG peak. The extracts were also examined for the presence of solubilized receptors by a modification of the method of Catt and Dufau (16). To 0.4 ml of the human tissue extracts, 0.1 ml [125I]hCG (-26,000 cpm) and 0.2 ml PBS with 1 mg bovine serum albumin (Sigma) were added. Background tubes contained the same reagents plus excessive unlabeled hCG (100IU; APL, Ayerst Laboratories, New York, NY). After incubation for 16 h at 4 C, the [I25I]hCG-receptor complex was precipitated by the addition of 0.5 ml 30% polyethylene glycol (PEG; mol wt, 6000-7500; wt/vol) in PBS, giving a final PEG concentration of 12.5% (wt/vol). The tubes were centrifuged at 2000 X g for 10 min and the supernatants were discarded. The precipitates were redissolved in 0.7 ml 0.1% Triton X-100 (Eastman Kodak)
919
in phosphate-buffered saline, incubated at 4 C for 10 min, and reprecipitated with 0.5 ml 30% PEG. After centrifugation and aspiration of the supernatants, the precipitates containing [125I]hCG bound to the receptors were counted. Positive controls for the receptor assay were prepared from rat testes by the methods described by Catt and Dufau (16). Con A column chromatography Disposible 0.7 X 4-cm columns (Bio-Rad Laboratories, Inc., Richmond, CA) containing 0.25 ml Sephadex G-25 and 1.5 ml Con A covalently coupled to Sepharose 4B (Pharmacia Laboratories) were extensively washed with PBS, pH 7.8, containing 1% normal rabbit serum. One to 4-ml aliquots of the tissue extracts or purified hormone preparations were applied to the columns, and the nonadsorbed substances were eluted with PBS. Elution of Con A-adsorbed glycoproteins was carried out with 0.2 M a,D-methylglucoside (MeG) in PBS. The eluted fractions were collected in 1-ml aliquots and analyzed in the /?-hCG RIA using the SB-6 antiserum. To control for tissue factors that may alter the degree of adsorption of hCG onto the Con A, tracer quantities of [125I]hCG were incubated with the tissue extracts for 16 h at 4 C before Con A chromatography. Recoveries of the [I25I]hCG in the PBS (nonadsorbed) and MeG (adsorbed) fractions were determined and compared to those found with [125I]hCG incubated in 2.5% normal rabbit serum. The degree of residual immunoprecipitability of the [125I]hCG after adsorption onto Con A was established by the addition of excess anti-yS-hCG serum to the peak MeG effluent tube, followed by precipitation with sheep or goat antirabbit globulin. Statistics The statistical methods of Rodbard were used for quality control, tests of parallelism, and dose interpolation in the radioligand assays (17). Immunological similarity between hCG and the material present in the tissue extracts was considered to be present when tbe inhibition curves generated by the hCG standards and serial dilutions of the tissue extract supernatants were parallel (18). Regression analysis and Pearson productmoment correlations were calculated on a Compu-Corp model 145-E statistician calculator. Results
The inhibition curves produced by serial dilutions of tissue extracts from testes, ovary, pituitary, lung, liver, kidney, spleen, stomach, and placenta were parallel to that obtained with the purified hCG reference preparation in the /?-hCG RIA which used the SB-6 antiserum (Fig. 1, top). Parallelism was also noted between these extracts and the hCG standard in the /?-hCG RIA with the SN-4-adsorbed antiserum (Fig. 1, bottom) and in the hCG radioreceptor assay (Fig. 2). Significant deviation from parallelism in these assays was noted with most pancreatic extracts, some small intestinal extracts (Fig. 1, bottom), and extracts with high protein concentrations (>70 mg/ml). With the exception of the pituitary ex-
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 20 November 2015. at 09:27 For personal use only. No other uses without permission. . All rights reserved.
BRAUNSTEIN ET AL.
920
JCE&M • 1979 Vol 49 • No 6
hCG RIA WITH SB-6 ANTISERUM -•3 -•2
PITUITARY
1
STOMACH^^
V
\
X
FIG. 1. Dose-response curves for purified hormone preparations and extracts of normal human tissues in the hCG RIAs. Top, hCG RIA with SB-6 antiserum. Bottom, hCG RIA with SN-4-adsorbed antiserum. B, Counts bound in the presence of labeled and unlabeled ligand; Bo, counts bound in the presence of labeled ligand alone. Logit (B/Bo) = logc [(B/Bo)/(1 - B/Bo)]. Coordinates are logit-normalized percentage counts precipitated (B/Bo X 100) on the ordinate, and the abscissa shows the log of the mass of hCG standards and purified hormone preparations or volume of tissue extracts added per tube.
PLACENTAt]
/JhC
• LH/FSH (LER 907)
&v hCO (LER tCO)
1 10 100 1000 10.000 STANDARDS (NANOQRAMS) AND TISSUE EXTRACTS or SERUM (MICROLITERS)
100.000
hCG RIA WITH SN-4 ADSORBED ANTISERUM -•3
9580-
-•2
8060O S02 40o
-., S
3 -
-•a
10-
-•3
1
10
100
1000
10,000
STANDARDS (NANOQRAMS) AND TISSUE EXTRACTS or SERUM (MICROLITERS)
-•3
95-
FIG. 2. Dose-response curves for the purified hCG preparations (CR 119), the Second International Standard for hCG, and serial dilutions of tissue extracts in the rat Leydig cell radioreceptor assay. CR 119 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 identical to those described in Fig. 1.
90-
-•2 PLACENTA
80-
-•1 605040-
LUNQ
- -1 202nd INT. STD. hCG
KIDNEY*
10-
.01
0.1 1 10 100 STANDARDS (IU or NANOGRAMS) AND TISSUE EXTRACTS (MICROLITERS)
tracts, the tissue concentrations of immunoreactive hCG determined by RIA with the SB-6 antiserum were highly correlated with those determined with the SN-4-adsorbed antiserum (r = 0.99; P < 0.001). The amount of immunoreactive material in pituitary extracts measured with the SN-4-adsorbed antiserum was approximately 4% of that measured with the SB-6 antiserum. In the /?-LH RIA, purified LH and pituitary extracts gave dose-response curves parallel to the LH standard (Fig. 3). Purified hCG demonstrated a two-component curve. Serial dilutions of extracts of placenta, ovary,
--2
1000
kidney, small intestine, testis, and lung were not parallel to the LH standard and had slopes that more closely resembled that seen with hCG. The pancreatic extracts had steeper slopes than the LH standards. Incubation of [125I]hCG with tissue extracts for 16 h at 4 C did not alter the Sephadex G-100 chromatographic pattern of the ligand (Table 2). The residual immunoprecipitability of the [125I]hCG after tissue incubation and Sephadex chromatography was similar to tracer incubated in normal rabbit serum (Table 2). However, prolonged incubation (>48 h) of the labeled ligand with the
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 20 November 2015. at 09:27 For personal use only. No other uses without permission. . All rights reserved.
hCG IN NORMAL HUMAN TISSUES
921
95-.
9080SMALL INTESTINE
o
FIG. 3. Dose-response curves for purified hormone preparations and serial dilutions of extracts of normal human tissues in the /?-LH RIA. See Fig. 1 for abbreviations.
2 60". • • m 40o 2010—i—
—r~ —r~ 1000 0.1 100 10 1.0 STANDARDS (NANOQRAMS) AND TISSUE EXTRACTS (MICROLITERS) TABLE 2. Results of Sephadex G-100 chromatography and immunoprecipitation studies on [l2GI]hCG after incubation with tissue extracts
[l25I]hCG in NRS
Sephadex G-100
% IP after
Ve/Vo
%IP
extraction of organs with [125I] hCG
1.23-1.40
100" (82.2-122.7)
100" (91.1-108.8)
Organ 104.4 117 1.38 Spleen 93.2 1.37 Kidney 104 111.4 Lung 114 1.35 1.34 111.7 Liver 110 1.37 43.0 Small intestine 107 Pancreas 1.1 1.37 89 ND 1.34 98 Pituitary 91.5 114 Ovary 1.32 b 1.36 109 Placenta Ve, Volume of exclusion; Vo, void volume; IP, immunoprecipitable; ND, not determined. " The immunoprecipitability of [125I]hCG was assumed to be maximum when the tracer was chromatographed or extracted with 2.5% normal rabbit serum (NRS). The average value was designated as 100%; the range of four experiments is shown in parentheses. The immunoprecipitability of the labeled ligand incubated or extracted with the organs was divided by that found with the same batch of [125I]hCG incubated in NRS. This allows for the comparison of data obtained with [125I]hCG that had been iodinated at different times. 6 Excessive quantities of hCG prevented complete immunoprecipitation.
tissues during the entire extraction procedure did result in a loss of immunoprecipitability of the tracer that had been incubated with pancreatic and small intestinal extracts (Table 2). Neither boiling nor the addition of protease inhibitors to the RIA tubes altered the dose-response curves of serial dilutions of the tissue extracts for any of the tissues studied, with the exception of the pancreatic extracts which gave inconsistent results and high nonspecific counts. High voltage electrophoresis of the [125I]hCG after incubation with boiled and unboiled tissue extracts revealed that only the tracer incubated with the unboiled pancreatic extracts was sufficiently damaged to migrate
10,000
away from the origin. Thin layer chromatography did not reveal fragments of hCG after incubation with the tissue extracts. The Sephadex G-100 elution patterns of [125I]hCG alone or after incubation with various tissue extracts were qualitatively similar. No high molecular weight [125I]hCG-receptor complex was identified in the chromatograms after tissue incubation. Similarly, the addition of PEG at a final concentration of 12.5% (wt/vol) to the tubes containing human tissue extracts that had been incubated for 16 h with [125I]hCG did not result in the precipitation of more counts per min than in the background tubes with excess unlabeled hCG (Fig. 4). A large amount of nonspecific binding was noted with the pancreatic extracts, again suggesting enzymatic damage to the labeled hCG. Specific [125I]hCG binding was present in the tubes containing the rat testicular extracts treated to concentrate hCG receptors (Fig. 4). Recovery of the immunoreactive hCG during extraction varied from 23% (small intestine) to 96.7% (ovary). The median concentration and range of the hCG-like material in the various tissues corrected for recovery are given in Table 3. A sufficient number of testicular and ovarian samples were available to allow comparison of the tissue contents of the hCG-like material at different ages. After the age of 15 yr, a significant inverse correlation between hCG content and age was noted for both the testes (Fig. 5) (r = 0.52; P < 0.01) and the ovaries (Fig. 6; r = 0.42; P< 0.05). Affinity chromatography of the tissue extracts on the Con A-Sepharose columns revealed a variable degree of adsorption of the immunoreactive substance to the lectin (Fig. 7). Extracts of the testes, ovaries, placenta, and small intestine had a greater degree of adsorption to Con A than those from the lungs, liver, and kidneys. Several pancreatic extracts were chromatographed and all of the immunoreactive material that could be recovered was eluted before the application of the 0.2 M MeG. The amount of immunoreactive material recovered after Con A-Sepharose chromatography varied with each tissue (Fig. 7). The results of the Con A column chromatography of
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 20 November 2015. at 09:27 For personal use only. No other uses without permission. . All rights reserved.
BRAUNSTEIN ET AL.
922
JCE&M • 1979 Vol 49 • No 6
• '2il-hCG • TISSUE • 100 IU hCG • '"l-hCG • TISSUE
FIG. 4. Percent counts percipitated of [l25I]hCG after incubation with tissue extracts for 16 h in the presence (•) or absence (63) of excess (100 IU) hCG. Precipitation of the labeled ligand was accomplished with 12.5% PEG. Rat testicular hCG receptor preparations were used for positive control purposes. The concentration of the rat soluble testes receptor no. 1 was equivalent to 0.1 g testes, while receptor no. 2 was equivalent to 0.46 g testes. TABLE 3. Concentration of hCG-like substance in normal human tissues
50-
hCG content (ng hCG/g tissue) n
urgan Spleen Kidney Lung Liver Small intestine Stomach Adrenal Testes Ovary Pituitary Placenta
6 8 6 7 5 1 1 77 30 3 5
Median
Range
2.63 4.51 2.69 2.23 10.75 3.45 1.31 19.93 5.55 20,847" 5,396
0.38-4.33 1.72-7.19 1.68-8.38
