Metabolism of Testosterone in Previable Human Fetuses : M. D. STERN,1 W. LING,2 J. R. T. COUTTS,3 M. C. MACNAUGHTON,3 AND S. SOLOMON Departments of Biochemistry and Experimental Medicine, McGill University and the University Clinic, Royal Victoria Hospital, Montreal, Canada and the Department of Obstetrics and Gynaecology, University of Dundee, Dundee, Scotland drosterone, 5a-androstanediol, 5/3-androstanediol and testosterone were isolated from the livers. In addition, etiocholanolone and 5/3-androstanediol were isolated from the sulfate and glucuronide fractions of the livers. A sex difference in fetal testosterone metabolism was observed in the liver where the amounts of unconjugated 5/3-androstanediol in the female livers were some 50% higher than those in the male livers with no overlap in values. This conclusion has to be viewed with caution as only 4 male and 3 female fetuses were examined in these studies. (J Clin Endocrinol Metab 40: 1057, 1975)
ABSTRACT. Labeled testosterone was injected into the umbilical vein of each of 4 male and 3 female previable fetuses at the time of laparotomy. After leaving the circulation intact for 3 min, the fetal tissues of each fetus were removed and the labeled metabolites present in each tissue were extracted, separated and isolated. Aliquots of tissue extracts were used for the qualitative analysis of metabolites following which metabolites were reisolated quantitatively from other tissue aliquots with the aid of the corresponding 3H-labeled recovery markers. Testosterone and 11/3-hydroxyandrostenedione4 were isolated from the adrenals and 5/3-androstanedione, androstenedione, etiocholanolone, an-
A
NUMBER of reports in the literature describe differences in steroid metabolism in male and female rats. Rubin and Strecker (1) have demonstrated that higher levels of 3/3-hydroxysteroid dehydrogenase activity are present in the livers of male rats than in those of female rats, while Kraulis and Clayton (2) have reported that male rats castrated 2 weeks postnatally convert testosterone to 5c*-androstane-3/3, Received December 16, 1974. * Supported by grants from the Medical Research Council of Canada (MT-1658), the U.S. Public Health Service (HDO-4365) and from the Lalor Foundation to JRTC. 1 Present address: Ayerst Laboratories, P.O. Box 6115, St. Laurent, P.Q., Canada. 2 Present address: Endocrine Laboratory, School of Medicine, University of Miami, Florida 33152. 3 Present address: Department of Obstetrics and Gynaecology, Royal Maternity Hospital, Rottenrow, Glasgow C4, Scotland. 4 Trivial names: 11/3-hydroxyandrostenedione = 11/3hydroxy-4-androstene-3,20-dione; etiocholanolone = 5/3-androstan-3a-ol-17-one; androsterone = 5a-androstan-3a-ol-17-one; 5/3-androstanedione = 5/3-androstane-3,17-dione; 5a-androstanedione = 5a-androstane-3,17-dione; androstenedione = 4-androstene3,17-dione; 5a-androstanediol = 5a-androstane-3a,17j8diol; 5/3-androstanediol = 5/3-androstane-3a,17/3-diol.
17a-diol-3- sulfate to a higher degree than do rats castrated at birth. These latter authors concluded that the metabolism of testosterone in the rat is influenced by its own testicular secretions. In human adults, the metabolism of testosterone is also sex-dependent as evidenced by the higher proportion of 5areduced metabolites in males (3). In view of the fact that certain metabolites of testosterone have biological activity distinct from the parent hormone, Baulieu and Robel (4) have suggested that the peripheral metabolism of testosterone is closely related to hormonal activity in biological systems. The study reported in this paper was conducted to compare the metabolism of testosterone by tissues from human previable male and female fetuses. Materials and Methods Methods used for counting radioactivity, solvent purification, chromatography, formation of derivatives and hydrolysis of steroid conjugates have been previously described (5,6). The solvent systems used in chromatography are given in Table 1. The labeled substrate used in these studies, 4-14C-testosterone (SA 50.6 mCi/mM) was pur-
1057
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 14 November 2015. at 01:22 For personal use only. No other uses without permission. . All rights reserved.
TABLE 1. Solvent systems used in chromatography
System
Type of chromatography
A
Ppca
Toluene-propylene glycol
B
Ppc
Benzene-cyclohexane (1:1)propylene glycol
C
Ppc
Skellysolve C-propylene glycol
D
Ppc
2,2,4-Trimethylpentane-tbutanol-methanol-water (10:4:7:1)
E
Ppc
2,2,4-Trimethylpentanetoluene-methanol-water (5:5:7:3)
F
Ppc
Skellysolve C-benzene-
Solvent composition
methanol-water (10:5:8:2) b
2,2,4-Trimethylpentanemethanol-water (20:9:1) Gradient: 2,2,4-Trimethylpentaneethylene dichloride (1:1)
G
Cpc
H
Tlcc
Benzene-ethanol (9:1)
I
Tic
Cyclohexane-ethyl acetate (1:1)
a b c
JCE & M • 1975 Vol 40 • No 6
STERN ETAL.
1058
Paper partition chromatography. Celite partition chromatography. Thin layer chromatography.
chased from New England Nuclear Corp. An aliquot of this material was mixed with carrier testosterone and the mixture was crystallized three times. The data indicate a radiochemical purity of at least 97%. Tritium labeled 5/3-androstanediol was prepared by the NaBH 4 reduction of etiocholanolone-l,2-3H (NEN). The product formed was purified by thin layer and paper in systems I and B respectively. Similarly tritiated 5aandrostanediol was prepared from androsterone-l,2-3H (NEN) and purified in systems I and B. Tritiated 11/3-hydroxyandrostenedione was prepared by the sodium bismuthate oxidation of cortisol-l,2-3H (NEN). The product formed was purified in systems G, F and D. The other tritium labeled steroids used in this study were obtained from commercial sources. All tritium labeled steroids had a radiochemical purity of not less than 98% as determined by reverse isotope dilution. The injection of labeled substrate into the umbilical vein of the fetus was carried out using
a slight modification of the procedure of Mikhail et al. (7). At the time of hysterotomy for therapeutic abortion and tubal ligation, a small incision was made in the amniotic sac and a short segment of the umbilical cord was carefully lifted out. The labeled steroid, dissolved in 0.5 ml ethanol:physiological saline (1:1) was slowly injected into the umbilical vein over a period of 30-40 s. The cord was then replaced in the sac, which was clipped and the fetoplacental circulation was allowed to continue for 3 min before the cord was clamped at both the fetal and the placental ends and the fetus and placenta were delivered. The placenta, cord and fetal tissues were then dissected and the labeled metabolites analyzed. The subsequent steps for extracting labeled metabolites from the tissues and defatting the tissue extracts have been previously described (8). Tissue extracts were divided into two equal parts. One half was used for the isolation of metabolites by chromatography and reverse isotope dilution. The results obtained by this procedure provided information as to the major metabolites present in these extracts. However, these results were not quantitative since no correction was made for procedural losses. Therefore, for the quantitative determination of metabolites, the following procedure was adopted using each remaining 50% of the tissue extracts. Appropriate tritium labeled recovery markers were added to the seven tissue extracts. Each extract was subjected to an ether-water partition and isolated metabolites were purified by chromatography using Celite partition columns, papers and silica gel thin layer plates. Metabolites eluted from the final chromatograms were mixed with the appropriate carrier steroid and the radiochemical purity determined as previously described (8). The percent conversion of 14C-testosterone to a metabolite was then calculated using the formula A/B percent conversion =
x 100 C where A is the amount of 3H-labeled recovery marker added to the initial extract, B is the 3 H/14C ratio of the radiochemically pure metabolites and C is the amount of 14C present in the extract. Each result is therefore the quantitative conversion of testosterone to a given metabolite expressed as a percentage of the total present in the tissue extract.
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 14 November 2015. at 01:22 For personal use only. No other uses without permission. . All rights reserved.
TESTOSTERONE METABOLISM IN FETAL TISSUES In one set of tissues (the livers) metabolites were also isolated from the aqueous soluble fraction, both in the sulfate and glucuronide forms. Since it was not feasible to use tritium labeled steroid conjugates as recovery markers, the metabolites in each aqueous extract were hydrolyzed separately, as previously described (8), yielding sulfate and glucuronide fractions. Unconjugated tritium labeled recovery markers were added after hydrolysis in this instance.
1059
TABLE 2. Description of fetuses injected with 14 C-testosterone Fetus
Sex
Crown-rump length (cm)
Gestational age (weeks)
Du
M M M F F M F
14.5 11.0 12.5 14.0 11.0 14.0 15.0
16 11 17 17 12 17
Do Me Ag Gi Bu
Ro
18
Results
The 7 fetuses used in this investigation are described in Table 2. A total of 10 /u.Ci of HC-testosterone was injected into the umbilical vein of each fetus at the time of hysterotomy in patients undergoing therapeutic abortion and tubal ligation. Permission for these procedures was obtained from the patients and from the Isotopic Advisory Panel of the Medical Research Council of Great Britain. In Table 3 is listed the distribution of radioactivity in the defatted extracts of the various tissues of the seven fetuses. The procedure for isolating metabolites from tissue extracts was identical for each of the fetuses, and consequently a detailed description will be given for only one set of tissues.
The adrenal extract of Do contained 106,000 dpm 14C. To this extract was added 57,000 dpm 3H-llj3-hydroxyandrostenedione, 10,470 dpm 3H-androstenedione and 25,300 dpm 3H-testosterone. This extract was then partitioned between ether and water following which 85,500 dpm 3H and 84,000 dpm 14C were found in the ether phase. The residue from the ether extract was chromatographed on paper to the solvent front in system B. The material eluted from the area corresponding in mobility to androstenedione contained 8,240 dpm 3H and 380 dpm 14C and was not further processed. The residues obtained after eluting the zone corresponding to testosterone (18,200 dpm 3H and 3,990 dpm 14C) and the area
TABLE 3. Radioactivity present in defatted tissue extracts from fetuses injected with
14
C-testosterone*
(dpm x 103)
Fetus
Du
Do
Me
3.5 6,095
106 4 3,496
155 16 2,920
316
113
181
56 763 325 40
83 1,038 72 28 7 6 80 153 748 1,227
33 551 253
Gi
Bu
173
140
111
3
141
4,210 130
4,840 52
3 7,326 312
Ag
Ro
Tissues Adrenals Gonads Liver Intestine Heart Brain Lungs Kidney Thymus Thyroid Blood Cord Placenta Residue
100
*•
9
-f
422 364 2,597
14 14 ••
308 108 1,068 2,053
60
223
55
763 311 40 15 3 355 173 928 3,397
2,273 107 37 ** 15 108 95 722 1,915
478 329 52 18 ** 362 124 1,533 2,700
129 4 9,183 235 69 876 374 26 12 ** 214 265 759 2,353
* In each fetus a total of 10 jtCi 14C-testosterone was injected into the umbilical vein. ** In some experiments it was not possible to locate the thymus or thyroid. t In this experiment the blood was combined with the placenta.
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 14 November 2015. at 01:22 For personal use only. No other uses without permission. . All rights reserved.
STERN ETAL.
1060
TABLE 4. Radiochemical purity of testosterone and 11/3-hydroxyandrostenedione isolated from the adrenal extract of Do Isolated steroid Crystallization
Xll*
Derivative Xll
ML** 3
1 2 3 Calculated
ML
14
H/ C
11/3-Hydroxyandrostenedione
11-Ketoandrostenedione
6.80 6.30 6.35 7.80
6.40 6.28
4.10 6.29 6.25
6.40 6.30
6.35 Andro-
1 2 3 Calculated
Testosterone
stenedione
6.50 6.40 6.38 5.62
6.44 6.38
4.50 4.70 6.45
6.51 6.52
6.38
*X11 = crystals, **ML = mother liquors.
corresponding to 11/3-hydroxyandrostenedione (42,160 dpm 3H and 7,010 dpm 14 C) were each rechromatographed in system D for 5 hours where symmetrical peaks were observed for both testosterone and 11/3-hydroxyandrostenedione. The residues obtained after elution of these two areas were mixed with carrier and the mixtures crystallized both before and after formation of an appropriate derivative (see Table 4). The essential data relative to the quantitative isolation of 11/3-hydroxyandrostenedione and testosterone from the six other adrenal extracts is given in Table 6. To the organic extract of the liver (3,565,000 dpm 14C) the following tritiated reference standards were added (dpm 3 H x 103): androsterone, 152; androstenedione, 1,885; etiocholanolone, 3,192; 5a-androstanediol, 502; 5j3-androstanediol, 1,225; 5a-androstenedione, 264; 5/3-androstanedione, 651; and testosterone, 241. The extract containing the tritiated steroids was then partitioned between ether and water following which 8,652,000 dpm 3H and 1,831,000 dpm 14C were recov-
JCE & M • 1975 Vol 40 • No 6
ered in the ether phase and 1,794,000 dpm 14C in the aqueous phase. The residue from the ether phase was chromatographed on a 35 g Celite partition column. A plot of dpm per fraction versus fraction number is shown in Fig. 1. Fractions eluted from this column were combined into pools as indicated in Fig. 1. The residue from pool I was chromatographed in system A for 5 hours where a symmetrical peak of radioactivity was observed with a mobility corresponding to both 5a-androstanedione and 5/3-androstanedione (14 cm). The material eluted from this zone (357,000 dpm 3H and 9,500 dpm 14C) was mixed with carrier 5/3-androstanedione and the mixture crystallized both before and after the formation of a derivative (see Table 5). The labeled material present in pool II was chromatographed in system C for 24 h. The scan of radioactivity revealed two major peaks, the less polar one corresponding in mobility to androstenedione (24 cm) and the more polar one to androsterone (12 cm). The material eluted from the less polar zone (141,900 dpm 3H, 14,200 dpm 14 C) was mixed with the carrier androstenedione and the mixture crystallized three times (Table 5). In view of the virtual absence of 14C in the third crystals, a derivative was not formed. The material from the more polar zone (141,900 dpm 3H and 14,200 dpm 14C) was identified as androsterone by reverse isotope dilution (Table 5). The residue from pool III of the Celite column (Fig. 1) was chromatographed for 24 h in system C where a large symmetrical peak of radioactivity was observed 12 cm from the origin with the mobility of etiocholanolone. The labeled material eluted from this zone (1,882,000 dpm 3H and 322,000 dpm 14C) was mixed with carrier etiocholanolone and the radiochemical purity established as described in Table 5. The labeled material present in pool V (Fig. 1) was chromatographed for 7 h in system A where a single peak of radio-
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 14 November 2015. at 01:22 For personal use only. No other uses without permission. . All rights reserved.
1061
TESTOSTERONE METABOLISM IN FETAL TISSUES 3
II androsterone
14, System: IsooctaneMtthanol Water ( 2 0 : 9 . 1 ) Gradient: Isooctane.Ethylene Oichloride ( 1 : 1 ) Fraction Volume 10ml Holdback Volume 50ml
androstenedione
2IE
etiocholanolone
300,000 200000 90,000 80,000
I
r
50 - androstanediol ' 5p-androstanedione
60,000 40,000
Q 20,000
10
20
30
40 50 60 70 Fraction Number
80
90
100
110
FIG. 1. Celite partition chromatography of the ether extract of liver from Bu.
activity was observed 9 cm from the origin. The material eluted from this zone, which had the mobility of testosterone contained 593,200 dpm 3 H and 8,900 dpm 14C. This material was mixed with carrier testosterone and the mixture crystallized both before and after the formation of a derivative (Table 5). The material in pool VI was chromatographed for 24 h in system B where a peak of radioactivity with the mobility of 5aandrostane-3a, 17/8-diol was observed 20 cm from the origin. The material eluted from this zone (348,000 dpm 3H and 16,000 dpm 14C) was identified as 5a-androstane3a, 17/8-diol by reverse isotope dilution (see Table 5). The material present in pool VII of the Celite column was run in system B for 24 h and the labeled material ran as a peak with the mobility of 5/3-androstane-3a,17/3-diol. The labeled material eluted from this zone was mixed with carrier 5j3-androstane-3a,
17/3-diol and its radiochemical purity established as described in Table 5. The isolation of the water soluble metabolites from the liver of Ro was achieved with an aliquot of the liver extract containing 4,549,500 dpm 14C. The extract was partitioned between ether and water from which 1,746,800 dpm 14C were recovered in the aqueous phase. To the residue of the aqueous extract was added 1,263,500 dpm 3H-etiocholanolone, 79,500 dpm 3 Htestosterone and 1,669,600 dpm 3H-5/3androstanediol. This mixture was then solvolyzed and a total of 3,180,000 dpm 3 H and 1,303,000 dpm 14C was found in the organic soluble material ("sulfate fraction"). To the aqueous soluble material remaining was added 1,669,900 dpm 3Hr 5/3-androstanediol and 1,261,500 dpm 3 Hetiocholanolone. This mixture was then hydrolyzed with /3-glucuronidase resulting in an organic extract containing 3,195,250 dpm 3 H and 270,500 dpm 14C.
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 14 November 2015. at 01:22 For personal use only. No other uses without permission. . All rights reserved.
1062
JCE & M • 1975 Vol 40 • No 6
STERN ET AL.
TABLE 5. Radiochemical purity of liver metabolites isolated from the ether soluble extract of Bu 3
Crystallization
Crystals
Mother liquors
44.7 41.9 41.9 37.9
37.1 41.3 40.1
Crystals
Mother liquors
5/3- Androstane-3a, 17/3-diol' 42.1 41.4
40.9 41.9
41.9 5a-Androstane-3a,17/3-diol'
Androsterone 1 2 3 Calculated
Derivative
Isolated Steroids
5/3- Androstenedione 1 2 3 Calculated*
H
3.4 15.4 24.9
21.9 24.9 25.7 10.0
26.2 25.7
27.1 26.0
25.7 Androstenedione
1 2 3 Calculated
790 1,700 3,000 465
287 603 2,250 Etiocholanolone
1 2 Calculated
5.0 5.9
5.9 5.9 5.8
5/3-Androstane-3a, 17/3-diol' 5.9 5.9 5.9
Androstenedione2
Testosterone 1 2 Calculated
228 262 66.6
49.2 72.6
5a-Androstanediol 1 2 3 4 Calculated
52.6 68.5 73.5 74.9 21.7
14.5 40.9 40.8 73.8
1.7 1.7 1.6
265 270 272
258 275
5a-Androstanediol Diacetate3 75.9 74.4
76.2 74.4
74.9 5/3-Androstanediol
1 2 Calculated
5.9 5.9
1.5 1-7
5/3-Androstanediol Diacetate3 1.7 1.7
1.7 1.7
* 5/8-Androstanedione = 357,000 dpm 3 H, 9,500 dpm 14C mixed with 18.0 mg carrier. Androsterone = 141,900 dpm 3 H, 14,200 dpm 14C mixed with 20.5 mg carrier. Androstenedione = 1,442,900 dpm 3 H, 3,100 dpm U C mixed with 30.5 mg carrier. Etiocholanolone = 1,882,000 dpm 3H, 322,000 dpm 14C mixed with 28.8 mg carrier. Testosterone = 593,200 dpm 3 H, 8,900 dpm I4C mixed with 25.4 mg carrier. 5a-Androstanediol = 348,000 dpm 3 H, 16,000 dpm 14C mixed with 22.1 mg carrier. 5/3-Androstanediol = 900,000 dpm 3 H, 547,000 dpm I4C mixed with 26.2 mg carrier. 1 The final crystals and mother liquors were combined and the mixture reduced with NaBH4 prior to crystallization. 2 The final crystals and mother liquors were combined and the mixture oxidized with CrO3 prior to crystallization. 3 The final crystals and mother liquors were combined and the mixture acetylated. The calculated specific activity is adjusted for the change in molecular weight.
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 14 November 2015. at 01:22 For personal use only. No other uses without permission. . All rights reserved.
TESTOSTERONE METABOLISM IN FETAL TISSUES
The residue from the sulfate fraction (8.1 mg) was chromatographed on paper in system A for 5.5 h and two major peaks of radioactivity were detected. The more polar of the two (9 cm from the origin) had the mobility of 5/3-androstanediol (1,081,250 dpm 3H and 363,500 dpm 14C) while the less polar area (22 cm from the origin) had the mobility of etiocholanolone (867, 200 dpm 3H and 229,900 dpm 14C). The residue obtained after eluting the area corresponding in mobility to testosterone contained 52,600 dpm 3H and 1,800 dpm 14 C and was not further processed. The labeled material which had the mobility of 5/3-androstanediol in system A was rechromatographed in system B for 24 hours when a single peak of radioactivity 9 cm from the origin was observed. The material eluted from this area was identified as 5/3-androstanediol by reverse isotope dilution analysis. The labeled material which migrated with the mobility of etiocholanolone in system A was rechromatographed in system C for 24 h. The scan of radioactivity revealed a single peak migrating with the mobility of etiocholanolone (12 cm from the origin). The material eluted from this zone was mixed with carrier etiocholanolone and the mixture recrystallized to constant specific activity. The residue from the glucuronide fraction (3,195,250 dpm 3H and 270,500 dpm 14 C) was chromatographed on paper in system A for 5.5 h where two bands of radioactive material were detected. The more polar one (10 cm from the origin) had the mobility of 5/3-androstanediol and the less polar peak (22 cm from the origin) the mobility of etiocholanolone. The residue from the more polar peak (1,406,000 dpm 3 H and 83,750 dpm 14C) was chromatographed for 24 h in system B to yield a single peak of radioactive material with the mobility of 5/3-androstanediol. The labeled material eluted from this zone (1,019,750 dpm 3H and 56,750 dpm 14C) was identified as 5/3-androstanediol by reverse isotope dilution analysis. The material eluted from
1063
TABLE 6. Percent conversion* of testosterone to 11/3-hydroxyandrostenedione and recovery of testosterone in adrenals 11/3-Hydroxyandrostenedione Male fetuses: Bu 17 weeks Du 16 weeks Do 11 weeks Me 17 weeks Average
10.9 11.9 8.6 11.3 9.9
Female fetuses: Ro 18 weeks Ag 17 weeks Gi 12 weeks Average
10.3 12.6 1.8 8.3
Testosterone 0.7 3.3
3.7 5.3 3.2 1.7 6.5 3.2 3.8
* Expressed as a percentage of the total radioactivity present in the tissue extract.
the less polar peak observed in system A was rechromatographed in system C for 24 h. The only peak of radioactive material found (12 cm from the origin) had the mobility of etiocholanolone. The residue from the elution of this area (822,000 dpm 3 H and 19,750 dpm 14C) was identified as etiocholanolone. The results obtained for unconjugated metabolites from the livers and adrenals are presented in Tables 6 and 7. Each value quoted represents the conversion of testosterone to the metabolites as a percentage of the total radioactivity present in the tissue extract. Discussion In Table 6 are shown the amounts of testosterone and 11/3-hydroxyandrostenedione isolated from the adrenals. As can be seen from this data, there is no trend indicative of a sex difference in the formation of 11/3-hydroxyandrostenedione or in the amount of unrnetabolized testosterone isolated. However, the two lowest values for this metabolite (8.6 and 1.8%) were obtained from two fetuses of 11 and 12 weeks gestational age, while the remaining five fetuses in the 16-18-week age group all had higher values (10.3-12.6%) for 11/3-hydroxyandrostenedione formed. These results are suggestive of a correla-
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 14 November 2015. at 01:22 For personal use only. No other uses without permission. . All rights reserved.
1064
JCE&M • 1975 VoUO • No 6
STERN ET AL.
TABLE 7. Percent conversion* of testosterone to ether-soluble metabolites in livers of fetuses
Female
Male
Testosterone Androstenedione 5/3-Androstanedione Androsterone Etiocholanolone 5a-Androstanediol 5/3-Androstanediol
Me
Bu
Du
Do
Avg
Ag
Gi
Ro
Avg
4.0 0.1 0.4 0.3 16.0 0.3 24.8
0.1 0.0 0.4 0.1 15.0 0.8 20.1
2.7 3.5 1.5 0.4 28.5 1.1 15.9
0.6 0.4 1.2 0.4 27.0 0.3 24.8
1.8 0.1 0.9 0.3 21.6 0.6 21.4
0.3 0.1 0.4 0.3 21.3 0.9 34.3
1.2 1.1 2.2 0.6 29.8 0.2 26.5
0.2 0.0 0.3 0.1 19.3 0.4 36.8
0.6 0.4 1.0 0.3 23.5 0.5 32.5
* Expressed as a percentage of the total present in the tissue extract.
tion of gestational age with the ability to form 11/3-hydroxyandrostenedione. It is interesting that Milner and Mills (9) showed a similar age correlation when measuring the conversion of progesterone to 11/3hydroxyandrostenedione in human fetal adrenal homogenates. The results obtained from the isolation of estrone and estradiol from the placentas of male and female fetuses indicated that there was no sex difference in the conversion of testosterone to the estrogens. Estrone and estradiol account for the bulk of the labeled material present in the placentas (> 60%). The results obtained from the ether solTABLE 8. Percent conversion* of testosterone to etiocholanolone and 5/J-androstanediol in sulfate and glucuronide fractions of liver Male
Bu
Du
Do
Me
Average
Etiocholanolone: sulfate glucuronide
5.9 0.2
3.7 0.3
20.9 0.3
3.2 1.1
8.4 0.5
5/3-Androstanediol: sulfate glucuronide
9.8 1.4
3.5 0.6
9.3 0.5
5.7 2.1
7.1 1.2
Female
Ro
Ag
Gi
Average
Etiocholanolone: sulfate glucuronide
6.2 0.6
4.2 1.5
6.5 0.3
5.8 0.8
12.5 2.0
8.6 6.0
5.4 1.0
8.8 3.0
5/3-Androstanediol: sulfate glucuronide
* Expressed as a percentage of the total present in the tissue extract.
uble fraction of the livers is shown in Table 7. Most of the labeled material was in the form of etiocholanolone and 5/3-androstanediol. Moreover, there appears to be a sex difference in the amounts of 5/3-androstanediol formed. In the three female fetal livers, the amounts of this metabolite are some 50% higher than those found in the male fetal livers (32.5 vs 21.4%). This is the only instance in these studies of a sex difference in testosterone metabolism. With the limited number of determinations it is difficult to make firm conclusions but it is noteworthy that the range of values for 5/3-androstanediol in the livers of females (26.5-36.8%) does not overlap in any way with that observed in the livers of male fetuses (15.9-24.8%). The above findings were pursued by analyzing the metabolites present in the aqueous fraction of each liver extract. These results are recorded in Table 8. One must bear in mind that these results are not fully quantitative since no correction was made for losses due to incomplete hydrolysis of conjugates and their extraction. The sex difference observed in the levels of unconjugated 5/3-androstanediol is not apparent when examining the amounts of conjugated 5/3-androstanediol. Thus it is unlikely that the sex difference observed for unconjugated 5/3-androstanediol was due to differences in the levels of conjugating enzymes in male and female fetal livers. It is possible that there is a sex difference in the levels of 17/3-ol-
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 14 November 2015. at 01:22 For personal use only. No other uses without permission. . All rights reserved.
TESTOSTERONE METABOLISM IN FETAL TISSUES dehydrogenase or in the levels of A4-5/3reductase. It is interesting to note that Ghraf et al. (10) demonstrated a higher conversion of testosterone to 5/3-reduced metabolites in the livers of newborn male rats compared to newborn female rats and that the neonatal administration of estradiol benzoate to male rats resulted in a loss of the 5/3-reduced metabolites. Recently Denef (11) showed that the sexual differentiation of testosterone metabolism in the rat is related to a sex difference in pituitary hormone secretion, the control of which is probably located in the hypothalamus. There have been previous attempts to look for a sex difference in the metabolism of testosterone in perfused previable human fetuses (12). In those studies the data were not corrected for procedural losses. Thus the losses encountered in isolating a given metabolite could not always have been exactly the same. Furthermore, the sex differences reported by these authors were apparently due to the anomalous results obtained with one of the eight fetuses. In the present communication the results obtained are quantitative since corrections were made for procedural losses. The preponderance of 5/3-H steroids in the fetal liver and the sex difference in the formation of 5/3-androstanediol focused our attention on the work of Kappas and Granick (13,14) who had described the stimulatory effect of steroids on prophyrin synthesis in chick liver cells in culture. These investigators found that 5/3-reduced metabolites of testosterone, progesterone and 17a-hydroxyprogesterone produced a marked increase in prophyrin synthesis while 5a-H metabolites of a wide class of steroids and sterols were devoid of this action. Furthermore, 5/3-androstanediol was the most active of the 5/3-H steroids
1065
studied in the chick liver preparation employed (14). It is well known that erythropoiesis in the midterm human fetus occurs mainly in the fetal liver (15) and this fact along with the findings of Kappas and Granick led us to consider that 5/3-H steroids may play an important role in hemoglobin synthesis in the human fetus. The results obtained on the stimulation of heme and hemoglobin synthesis by testosterone and 5/3- and 5a-H metabolites in human fetal liver in culture have recently been published (16,17). References 1. Rubin, B. L., and H. J. Strecker, Endocrinology 69: 257, 1961. 2. Kraulis, I., and R. B. Clayton, J Biol Chem 243: 3546, 1968. 3. Baulieu, E. E., and P. Mauvais-Jarvis,/ Biol Chem 239: 1578, 1964. 4. , and P. Robel In Eik-Nes, K. B. (ed.), The Androgens of the Testis, Marcel-Dekker, New York, 1970, p. 68. 5. Ruse, J. L., and S. Solomon, Biochemistry 5: 1065, 1966. 6. YoungLai, E., and S. Solomon, Biochemistry 6: 2040, 1967. 7. Mikhail, G., N. Wiqvist, and E. Diczfalusy, Ada Endocrinol (Kbh) 43: 213, 1963. 8. Bird, C. E., N. Wiqvist, E. Diczfalusy, and S. Solomon,7 Clin Endocrinol Metab 26: 1144, 1966. 9. Milner, A. J., and I. H. Mills, J Endocrinol 47: 379, 1970. 10. Ghraf, R., H-G. Hoff, E. R. Lax, and H. Schriefers, Ada Endocrinol 73: 577, 1973. 11. Denef, C., Endocrinology 94: 1577, 1974. 12. Benagiano, G., F. A. Kind, F. Zielske, N. Wiqvist, and E. Diczfalusy, Ada Endocrinol 56: 203, 1967. 13. Granick, S., and A. Kappas, J Biol Chem 242: 4587, 1967. 14. Kappas, A., and S. Granick, 7 Biol Chem 243: 346, 1968. 15. Wintrobe, M., In Clinical Hematology, Lea and Febiger, Philadelphia, 1961, p. 2. 16. Congote, L. F., M. D. Stern, and S. Solomon, Biochemistry 13: 4255, 1974. 17. , and S. Solomon, Proc Natl Acad Sci USA 72: 523, 1975.
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 14 November 2015. at 01:22 For personal use only. No other uses without permission. . All rights reserved.
ABSTRACT. Labeled testosterone was injected into the umbilical vein of each of 4 male and 3 female previable fetuses at the time of laparotomy. After leaving the circulation intact for 3 min, the fetal tissues of each fetus were removed and the labeled metabolites present in each tissue were extracted, separated and isolated. Aliquots of tissue extracts were used for the qualitative analysis of metabolites following which metabolites were reisolated quantitatively from other tissue aliquots with the aid of the corresponding 3H-labeled recovery markers. Testosterone and 11/3-hydroxyandrostenedione4 were isolated from the adrenals and 5/3-androstanedione, androstenedione, etiocholanolone, an-
A
NUMBER of reports in the literature describe differences in steroid metabolism in male and female rats. Rubin and Strecker (1) have demonstrated that higher levels of 3/3-hydroxysteroid dehydrogenase activity are present in the livers of male rats than in those of female rats, while Kraulis and Clayton (2) have reported that male rats castrated 2 weeks postnatally convert testosterone to 5c*-androstane-3/3, Received December 16, 1974. * Supported by grants from the Medical Research Council of Canada (MT-1658), the U.S. Public Health Service (HDO-4365) and from the Lalor Foundation to JRTC. 1 Present address: Ayerst Laboratories, P.O. Box 6115, St. Laurent, P.Q., Canada. 2 Present address: Endocrine Laboratory, School of Medicine, University of Miami, Florida 33152. 3 Present address: Department of Obstetrics and Gynaecology, Royal Maternity Hospital, Rottenrow, Glasgow C4, Scotland. 4 Trivial names: 11/3-hydroxyandrostenedione = 11/3hydroxy-4-androstene-3,20-dione; etiocholanolone = 5/3-androstan-3a-ol-17-one; androsterone = 5a-androstan-3a-ol-17-one; 5/3-androstanedione = 5/3-androstane-3,17-dione; 5a-androstanedione = 5a-androstane-3,17-dione; androstenedione = 4-androstene3,17-dione; 5a-androstanediol = 5a-androstane-3a,17j8diol; 5/3-androstanediol = 5/3-androstane-3a,17/3-diol.
17a-diol-3- sulfate to a higher degree than do rats castrated at birth. These latter authors concluded that the metabolism of testosterone in the rat is influenced by its own testicular secretions. In human adults, the metabolism of testosterone is also sex-dependent as evidenced by the higher proportion of 5areduced metabolites in males (3). In view of the fact that certain metabolites of testosterone have biological activity distinct from the parent hormone, Baulieu and Robel (4) have suggested that the peripheral metabolism of testosterone is closely related to hormonal activity in biological systems. The study reported in this paper was conducted to compare the metabolism of testosterone by tissues from human previable male and female fetuses. Materials and Methods Methods used for counting radioactivity, solvent purification, chromatography, formation of derivatives and hydrolysis of steroid conjugates have been previously described (5,6). The solvent systems used in chromatography are given in Table 1. The labeled substrate used in these studies, 4-14C-testosterone (SA 50.6 mCi/mM) was pur-
1057
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 14 November 2015. at 01:22 For personal use only. No other uses without permission. . All rights reserved.
TABLE 1. Solvent systems used in chromatography
System
Type of chromatography
A
Ppca
Toluene-propylene glycol
B
Ppc
Benzene-cyclohexane (1:1)propylene glycol
C
Ppc
Skellysolve C-propylene glycol
D
Ppc
2,2,4-Trimethylpentane-tbutanol-methanol-water (10:4:7:1)
E
Ppc
2,2,4-Trimethylpentanetoluene-methanol-water (5:5:7:3)
F
Ppc
Skellysolve C-benzene-
Solvent composition
methanol-water (10:5:8:2) b
2,2,4-Trimethylpentanemethanol-water (20:9:1) Gradient: 2,2,4-Trimethylpentaneethylene dichloride (1:1)
G
Cpc
H
Tlcc
Benzene-ethanol (9:1)
I
Tic
Cyclohexane-ethyl acetate (1:1)
a b c
JCE & M • 1975 Vol 40 • No 6
STERN ETAL.
1058
Paper partition chromatography. Celite partition chromatography. Thin layer chromatography.
chased from New England Nuclear Corp. An aliquot of this material was mixed with carrier testosterone and the mixture was crystallized three times. The data indicate a radiochemical purity of at least 97%. Tritium labeled 5/3-androstanediol was prepared by the NaBH 4 reduction of etiocholanolone-l,2-3H (NEN). The product formed was purified by thin layer and paper in systems I and B respectively. Similarly tritiated 5aandrostanediol was prepared from androsterone-l,2-3H (NEN) and purified in systems I and B. Tritiated 11/3-hydroxyandrostenedione was prepared by the sodium bismuthate oxidation of cortisol-l,2-3H (NEN). The product formed was purified in systems G, F and D. The other tritium labeled steroids used in this study were obtained from commercial sources. All tritium labeled steroids had a radiochemical purity of not less than 98% as determined by reverse isotope dilution. The injection of labeled substrate into the umbilical vein of the fetus was carried out using
a slight modification of the procedure of Mikhail et al. (7). At the time of hysterotomy for therapeutic abortion and tubal ligation, a small incision was made in the amniotic sac and a short segment of the umbilical cord was carefully lifted out. The labeled steroid, dissolved in 0.5 ml ethanol:physiological saline (1:1) was slowly injected into the umbilical vein over a period of 30-40 s. The cord was then replaced in the sac, which was clipped and the fetoplacental circulation was allowed to continue for 3 min before the cord was clamped at both the fetal and the placental ends and the fetus and placenta were delivered. The placenta, cord and fetal tissues were then dissected and the labeled metabolites analyzed. The subsequent steps for extracting labeled metabolites from the tissues and defatting the tissue extracts have been previously described (8). Tissue extracts were divided into two equal parts. One half was used for the isolation of metabolites by chromatography and reverse isotope dilution. The results obtained by this procedure provided information as to the major metabolites present in these extracts. However, these results were not quantitative since no correction was made for procedural losses. Therefore, for the quantitative determination of metabolites, the following procedure was adopted using each remaining 50% of the tissue extracts. Appropriate tritium labeled recovery markers were added to the seven tissue extracts. Each extract was subjected to an ether-water partition and isolated metabolites were purified by chromatography using Celite partition columns, papers and silica gel thin layer plates. Metabolites eluted from the final chromatograms were mixed with the appropriate carrier steroid and the radiochemical purity determined as previously described (8). The percent conversion of 14C-testosterone to a metabolite was then calculated using the formula A/B percent conversion =
x 100 C where A is the amount of 3H-labeled recovery marker added to the initial extract, B is the 3 H/14C ratio of the radiochemically pure metabolites and C is the amount of 14C present in the extract. Each result is therefore the quantitative conversion of testosterone to a given metabolite expressed as a percentage of the total present in the tissue extract.
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 14 November 2015. at 01:22 For personal use only. No other uses without permission. . All rights reserved.
TESTOSTERONE METABOLISM IN FETAL TISSUES In one set of tissues (the livers) metabolites were also isolated from the aqueous soluble fraction, both in the sulfate and glucuronide forms. Since it was not feasible to use tritium labeled steroid conjugates as recovery markers, the metabolites in each aqueous extract were hydrolyzed separately, as previously described (8), yielding sulfate and glucuronide fractions. Unconjugated tritium labeled recovery markers were added after hydrolysis in this instance.
1059
TABLE 2. Description of fetuses injected with 14 C-testosterone Fetus
Sex
Crown-rump length (cm)
Gestational age (weeks)
Du
M M M F F M F
14.5 11.0 12.5 14.0 11.0 14.0 15.0
16 11 17 17 12 17
Do Me Ag Gi Bu
Ro
18
Results
The 7 fetuses used in this investigation are described in Table 2. A total of 10 /u.Ci of HC-testosterone was injected into the umbilical vein of each fetus at the time of hysterotomy in patients undergoing therapeutic abortion and tubal ligation. Permission for these procedures was obtained from the patients and from the Isotopic Advisory Panel of the Medical Research Council of Great Britain. In Table 3 is listed the distribution of radioactivity in the defatted extracts of the various tissues of the seven fetuses. The procedure for isolating metabolites from tissue extracts was identical for each of the fetuses, and consequently a detailed description will be given for only one set of tissues.
The adrenal extract of Do contained 106,000 dpm 14C. To this extract was added 57,000 dpm 3H-llj3-hydroxyandrostenedione, 10,470 dpm 3H-androstenedione and 25,300 dpm 3H-testosterone. This extract was then partitioned between ether and water following which 85,500 dpm 3H and 84,000 dpm 14C were found in the ether phase. The residue from the ether extract was chromatographed on paper to the solvent front in system B. The material eluted from the area corresponding in mobility to androstenedione contained 8,240 dpm 3H and 380 dpm 14C and was not further processed. The residues obtained after eluting the zone corresponding to testosterone (18,200 dpm 3H and 3,990 dpm 14C) and the area
TABLE 3. Radioactivity present in defatted tissue extracts from fetuses injected with
14
C-testosterone*
(dpm x 103)
Fetus
Du
Do
Me
3.5 6,095
106 4 3,496
155 16 2,920
316
113
181
56 763 325 40
83 1,038 72 28 7 6 80 153 748 1,227
33 551 253
Gi
Bu
173
140
111
3
141
4,210 130
4,840 52
3 7,326 312
Ag
Ro
Tissues Adrenals Gonads Liver Intestine Heart Brain Lungs Kidney Thymus Thyroid Blood Cord Placenta Residue
100
*•
9
-f
422 364 2,597
14 14 ••
308 108 1,068 2,053
60
223
55
763 311 40 15 3 355 173 928 3,397
2,273 107 37 ** 15 108 95 722 1,915
478 329 52 18 ** 362 124 1,533 2,700
129 4 9,183 235 69 876 374 26 12 ** 214 265 759 2,353
* In each fetus a total of 10 jtCi 14C-testosterone was injected into the umbilical vein. ** In some experiments it was not possible to locate the thymus or thyroid. t In this experiment the blood was combined with the placenta.
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 14 November 2015. at 01:22 For personal use only. No other uses without permission. . All rights reserved.
STERN ETAL.
1060
TABLE 4. Radiochemical purity of testosterone and 11/3-hydroxyandrostenedione isolated from the adrenal extract of Do Isolated steroid Crystallization
Xll*
Derivative Xll
ML** 3
1 2 3 Calculated
ML
14
H/ C
11/3-Hydroxyandrostenedione
11-Ketoandrostenedione
6.80 6.30 6.35 7.80
6.40 6.28
4.10 6.29 6.25
6.40 6.30
6.35 Andro-
1 2 3 Calculated
Testosterone
stenedione
6.50 6.40 6.38 5.62
6.44 6.38
4.50 4.70 6.45
6.51 6.52
6.38
*X11 = crystals, **ML = mother liquors.
corresponding to 11/3-hydroxyandrostenedione (42,160 dpm 3H and 7,010 dpm 14 C) were each rechromatographed in system D for 5 hours where symmetrical peaks were observed for both testosterone and 11/3-hydroxyandrostenedione. The residues obtained after elution of these two areas were mixed with carrier and the mixtures crystallized both before and after formation of an appropriate derivative (see Table 4). The essential data relative to the quantitative isolation of 11/3-hydroxyandrostenedione and testosterone from the six other adrenal extracts is given in Table 6. To the organic extract of the liver (3,565,000 dpm 14C) the following tritiated reference standards were added (dpm 3 H x 103): androsterone, 152; androstenedione, 1,885; etiocholanolone, 3,192; 5a-androstanediol, 502; 5j3-androstanediol, 1,225; 5a-androstenedione, 264; 5/3-androstanedione, 651; and testosterone, 241. The extract containing the tritiated steroids was then partitioned between ether and water following which 8,652,000 dpm 3H and 1,831,000 dpm 14C were recov-
JCE & M • 1975 Vol 40 • No 6
ered in the ether phase and 1,794,000 dpm 14C in the aqueous phase. The residue from the ether phase was chromatographed on a 35 g Celite partition column. A plot of dpm per fraction versus fraction number is shown in Fig. 1. Fractions eluted from this column were combined into pools as indicated in Fig. 1. The residue from pool I was chromatographed in system A for 5 hours where a symmetrical peak of radioactivity was observed with a mobility corresponding to both 5a-androstanedione and 5/3-androstanedione (14 cm). The material eluted from this zone (357,000 dpm 3H and 9,500 dpm 14C) was mixed with carrier 5/3-androstanedione and the mixture crystallized both before and after the formation of a derivative (see Table 5). The labeled material present in pool II was chromatographed in system C for 24 h. The scan of radioactivity revealed two major peaks, the less polar one corresponding in mobility to androstenedione (24 cm) and the more polar one to androsterone (12 cm). The material eluted from the less polar zone (141,900 dpm 3H, 14,200 dpm 14 C) was mixed with the carrier androstenedione and the mixture crystallized three times (Table 5). In view of the virtual absence of 14C in the third crystals, a derivative was not formed. The material from the more polar zone (141,900 dpm 3H and 14,200 dpm 14C) was identified as androsterone by reverse isotope dilution (Table 5). The residue from pool III of the Celite column (Fig. 1) was chromatographed for 24 h in system C where a large symmetrical peak of radioactivity was observed 12 cm from the origin with the mobility of etiocholanolone. The labeled material eluted from this zone (1,882,000 dpm 3H and 322,000 dpm 14C) was mixed with carrier etiocholanolone and the radiochemical purity established as described in Table 5. The labeled material present in pool V (Fig. 1) was chromatographed for 7 h in system A where a single peak of radio-
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 14 November 2015. at 01:22 For personal use only. No other uses without permission. . All rights reserved.
1061
TESTOSTERONE METABOLISM IN FETAL TISSUES 3
II androsterone
14, System: IsooctaneMtthanol Water ( 2 0 : 9 . 1 ) Gradient: Isooctane.Ethylene Oichloride ( 1 : 1 ) Fraction Volume 10ml Holdback Volume 50ml
androstenedione
2IE
etiocholanolone
300,000 200000 90,000 80,000
I
r
50 - androstanediol ' 5p-androstanedione
60,000 40,000
Q 20,000
10
20
30
40 50 60 70 Fraction Number
80
90
100
110
FIG. 1. Celite partition chromatography of the ether extract of liver from Bu.
activity was observed 9 cm from the origin. The material eluted from this zone, which had the mobility of testosterone contained 593,200 dpm 3 H and 8,900 dpm 14C. This material was mixed with carrier testosterone and the mixture crystallized both before and after the formation of a derivative (Table 5). The material in pool VI was chromatographed for 24 h in system B where a peak of radioactivity with the mobility of 5aandrostane-3a, 17/8-diol was observed 20 cm from the origin. The material eluted from this zone (348,000 dpm 3H and 16,000 dpm 14C) was identified as 5a-androstane3a, 17/8-diol by reverse isotope dilution (see Table 5). The material present in pool VII of the Celite column was run in system B for 24 h and the labeled material ran as a peak with the mobility of 5/3-androstane-3a,17/3-diol. The labeled material eluted from this zone was mixed with carrier 5j3-androstane-3a,
17/3-diol and its radiochemical purity established as described in Table 5. The isolation of the water soluble metabolites from the liver of Ro was achieved with an aliquot of the liver extract containing 4,549,500 dpm 14C. The extract was partitioned between ether and water from which 1,746,800 dpm 14C were recovered in the aqueous phase. To the residue of the aqueous extract was added 1,263,500 dpm 3H-etiocholanolone, 79,500 dpm 3 Htestosterone and 1,669,600 dpm 3H-5/3androstanediol. This mixture was then solvolyzed and a total of 3,180,000 dpm 3 H and 1,303,000 dpm 14C was found in the organic soluble material ("sulfate fraction"). To the aqueous soluble material remaining was added 1,669,900 dpm 3Hr 5/3-androstanediol and 1,261,500 dpm 3 Hetiocholanolone. This mixture was then hydrolyzed with /3-glucuronidase resulting in an organic extract containing 3,195,250 dpm 3 H and 270,500 dpm 14C.
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 14 November 2015. at 01:22 For personal use only. No other uses without permission. . All rights reserved.
1062
JCE & M • 1975 Vol 40 • No 6
STERN ET AL.
TABLE 5. Radiochemical purity of liver metabolites isolated from the ether soluble extract of Bu 3
Crystallization
Crystals
Mother liquors
44.7 41.9 41.9 37.9
37.1 41.3 40.1
Crystals
Mother liquors
5/3- Androstane-3a, 17/3-diol' 42.1 41.4
40.9 41.9
41.9 5a-Androstane-3a,17/3-diol'
Androsterone 1 2 3 Calculated
Derivative
Isolated Steroids
5/3- Androstenedione 1 2 3 Calculated*
H
3.4 15.4 24.9
21.9 24.9 25.7 10.0
26.2 25.7
27.1 26.0
25.7 Androstenedione
1 2 3 Calculated
790 1,700 3,000 465
287 603 2,250 Etiocholanolone
1 2 Calculated
5.0 5.9
5.9 5.9 5.8
5/3-Androstane-3a, 17/3-diol' 5.9 5.9 5.9
Androstenedione2
Testosterone 1 2 Calculated
228 262 66.6
49.2 72.6
5a-Androstanediol 1 2 3 4 Calculated
52.6 68.5 73.5 74.9 21.7
14.5 40.9 40.8 73.8
1.7 1.7 1.6
265 270 272
258 275
5a-Androstanediol Diacetate3 75.9 74.4
76.2 74.4
74.9 5/3-Androstanediol
1 2 Calculated
5.9 5.9
1.5 1-7
5/3-Androstanediol Diacetate3 1.7 1.7
1.7 1.7
* 5/8-Androstanedione = 357,000 dpm 3 H, 9,500 dpm 14C mixed with 18.0 mg carrier. Androsterone = 141,900 dpm 3 H, 14,200 dpm 14C mixed with 20.5 mg carrier. Androstenedione = 1,442,900 dpm 3 H, 3,100 dpm U C mixed with 30.5 mg carrier. Etiocholanolone = 1,882,000 dpm 3H, 322,000 dpm 14C mixed with 28.8 mg carrier. Testosterone = 593,200 dpm 3 H, 8,900 dpm I4C mixed with 25.4 mg carrier. 5a-Androstanediol = 348,000 dpm 3 H, 16,000 dpm 14C mixed with 22.1 mg carrier. 5/3-Androstanediol = 900,000 dpm 3 H, 547,000 dpm I4C mixed with 26.2 mg carrier. 1 The final crystals and mother liquors were combined and the mixture reduced with NaBH4 prior to crystallization. 2 The final crystals and mother liquors were combined and the mixture oxidized with CrO3 prior to crystallization. 3 The final crystals and mother liquors were combined and the mixture acetylated. The calculated specific activity is adjusted for the change in molecular weight.
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 14 November 2015. at 01:22 For personal use only. No other uses without permission. . All rights reserved.
TESTOSTERONE METABOLISM IN FETAL TISSUES
The residue from the sulfate fraction (8.1 mg) was chromatographed on paper in system A for 5.5 h and two major peaks of radioactivity were detected. The more polar of the two (9 cm from the origin) had the mobility of 5/3-androstanediol (1,081,250 dpm 3H and 363,500 dpm 14C) while the less polar area (22 cm from the origin) had the mobility of etiocholanolone (867, 200 dpm 3H and 229,900 dpm 14C). The residue obtained after eluting the area corresponding in mobility to testosterone contained 52,600 dpm 3H and 1,800 dpm 14 C and was not further processed. The labeled material which had the mobility of 5/3-androstanediol in system A was rechromatographed in system B for 24 hours when a single peak of radioactivity 9 cm from the origin was observed. The material eluted from this area was identified as 5/3-androstanediol by reverse isotope dilution analysis. The labeled material which migrated with the mobility of etiocholanolone in system A was rechromatographed in system C for 24 h. The scan of radioactivity revealed a single peak migrating with the mobility of etiocholanolone (12 cm from the origin). The material eluted from this zone was mixed with carrier etiocholanolone and the mixture recrystallized to constant specific activity. The residue from the glucuronide fraction (3,195,250 dpm 3H and 270,500 dpm 14 C) was chromatographed on paper in system A for 5.5 h where two bands of radioactive material were detected. The more polar one (10 cm from the origin) had the mobility of 5/3-androstanediol and the less polar peak (22 cm from the origin) the mobility of etiocholanolone. The residue from the more polar peak (1,406,000 dpm 3 H and 83,750 dpm 14C) was chromatographed for 24 h in system B to yield a single peak of radioactive material with the mobility of 5/3-androstanediol. The labeled material eluted from this zone (1,019,750 dpm 3H and 56,750 dpm 14C) was identified as 5/3-androstanediol by reverse isotope dilution analysis. The material eluted from
1063
TABLE 6. Percent conversion* of testosterone to 11/3-hydroxyandrostenedione and recovery of testosterone in adrenals 11/3-Hydroxyandrostenedione Male fetuses: Bu 17 weeks Du 16 weeks Do 11 weeks Me 17 weeks Average
10.9 11.9 8.6 11.3 9.9
Female fetuses: Ro 18 weeks Ag 17 weeks Gi 12 weeks Average
10.3 12.6 1.8 8.3
Testosterone 0.7 3.3
3.7 5.3 3.2 1.7 6.5 3.2 3.8
* Expressed as a percentage of the total radioactivity present in the tissue extract.
the less polar peak observed in system A was rechromatographed in system C for 24 h. The only peak of radioactive material found (12 cm from the origin) had the mobility of etiocholanolone. The residue from the elution of this area (822,000 dpm 3 H and 19,750 dpm 14C) was identified as etiocholanolone. The results obtained for unconjugated metabolites from the livers and adrenals are presented in Tables 6 and 7. Each value quoted represents the conversion of testosterone to the metabolites as a percentage of the total radioactivity present in the tissue extract. Discussion In Table 6 are shown the amounts of testosterone and 11/3-hydroxyandrostenedione isolated from the adrenals. As can be seen from this data, there is no trend indicative of a sex difference in the formation of 11/3-hydroxyandrostenedione or in the amount of unrnetabolized testosterone isolated. However, the two lowest values for this metabolite (8.6 and 1.8%) were obtained from two fetuses of 11 and 12 weeks gestational age, while the remaining five fetuses in the 16-18-week age group all had higher values (10.3-12.6%) for 11/3-hydroxyandrostenedione formed. These results are suggestive of a correla-
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 14 November 2015. at 01:22 For personal use only. No other uses without permission. . All rights reserved.
1064
JCE&M • 1975 VoUO • No 6
STERN ET AL.
TABLE 7. Percent conversion* of testosterone to ether-soluble metabolites in livers of fetuses
Female
Male
Testosterone Androstenedione 5/3-Androstanedione Androsterone Etiocholanolone 5a-Androstanediol 5/3-Androstanediol
Me
Bu
Du
Do
Avg
Ag
Gi
Ro
Avg
4.0 0.1 0.4 0.3 16.0 0.3 24.8
0.1 0.0 0.4 0.1 15.0 0.8 20.1
2.7 3.5 1.5 0.4 28.5 1.1 15.9
0.6 0.4 1.2 0.4 27.0 0.3 24.8
1.8 0.1 0.9 0.3 21.6 0.6 21.4
0.3 0.1 0.4 0.3 21.3 0.9 34.3
1.2 1.1 2.2 0.6 29.8 0.2 26.5
0.2 0.0 0.3 0.1 19.3 0.4 36.8
0.6 0.4 1.0 0.3 23.5 0.5 32.5
* Expressed as a percentage of the total present in the tissue extract.
tion of gestational age with the ability to form 11/3-hydroxyandrostenedione. It is interesting that Milner and Mills (9) showed a similar age correlation when measuring the conversion of progesterone to 11/3hydroxyandrostenedione in human fetal adrenal homogenates. The results obtained from the isolation of estrone and estradiol from the placentas of male and female fetuses indicated that there was no sex difference in the conversion of testosterone to the estrogens. Estrone and estradiol account for the bulk of the labeled material present in the placentas (> 60%). The results obtained from the ether solTABLE 8. Percent conversion* of testosterone to etiocholanolone and 5/J-androstanediol in sulfate and glucuronide fractions of liver Male
Bu
Du
Do
Me
Average
Etiocholanolone: sulfate glucuronide
5.9 0.2
3.7 0.3
20.9 0.3
3.2 1.1
8.4 0.5
5/3-Androstanediol: sulfate glucuronide
9.8 1.4
3.5 0.6
9.3 0.5
5.7 2.1
7.1 1.2
Female
Ro
Ag
Gi
Average
Etiocholanolone: sulfate glucuronide
6.2 0.6
4.2 1.5
6.5 0.3
5.8 0.8
12.5 2.0
8.6 6.0
5.4 1.0
8.8 3.0
5/3-Androstanediol: sulfate glucuronide
* Expressed as a percentage of the total present in the tissue extract.
uble fraction of the livers is shown in Table 7. Most of the labeled material was in the form of etiocholanolone and 5/3-androstanediol. Moreover, there appears to be a sex difference in the amounts of 5/3-androstanediol formed. In the three female fetal livers, the amounts of this metabolite are some 50% higher than those found in the male fetal livers (32.5 vs 21.4%). This is the only instance in these studies of a sex difference in testosterone metabolism. With the limited number of determinations it is difficult to make firm conclusions but it is noteworthy that the range of values for 5/3-androstanediol in the livers of females (26.5-36.8%) does not overlap in any way with that observed in the livers of male fetuses (15.9-24.8%). The above findings were pursued by analyzing the metabolites present in the aqueous fraction of each liver extract. These results are recorded in Table 8. One must bear in mind that these results are not fully quantitative since no correction was made for losses due to incomplete hydrolysis of conjugates and their extraction. The sex difference observed in the levels of unconjugated 5/3-androstanediol is not apparent when examining the amounts of conjugated 5/3-androstanediol. Thus it is unlikely that the sex difference observed for unconjugated 5/3-androstanediol was due to differences in the levels of conjugating enzymes in male and female fetal livers. It is possible that there is a sex difference in the levels of 17/3-ol-
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 14 November 2015. at 01:22 For personal use only. No other uses without permission. . All rights reserved.
TESTOSTERONE METABOLISM IN FETAL TISSUES dehydrogenase or in the levels of A4-5/3reductase. It is interesting to note that Ghraf et al. (10) demonstrated a higher conversion of testosterone to 5/3-reduced metabolites in the livers of newborn male rats compared to newborn female rats and that the neonatal administration of estradiol benzoate to male rats resulted in a loss of the 5/3-reduced metabolites. Recently Denef (11) showed that the sexual differentiation of testosterone metabolism in the rat is related to a sex difference in pituitary hormone secretion, the control of which is probably located in the hypothalamus. There have been previous attempts to look for a sex difference in the metabolism of testosterone in perfused previable human fetuses (12). In those studies the data were not corrected for procedural losses. Thus the losses encountered in isolating a given metabolite could not always have been exactly the same. Furthermore, the sex differences reported by these authors were apparently due to the anomalous results obtained with one of the eight fetuses. In the present communication the results obtained are quantitative since corrections were made for procedural losses. The preponderance of 5/3-H steroids in the fetal liver and the sex difference in the formation of 5/3-androstanediol focused our attention on the work of Kappas and Granick (13,14) who had described the stimulatory effect of steroids on prophyrin synthesis in chick liver cells in culture. These investigators found that 5/3-reduced metabolites of testosterone, progesterone and 17a-hydroxyprogesterone produced a marked increase in prophyrin synthesis while 5a-H metabolites of a wide class of steroids and sterols were devoid of this action. Furthermore, 5/3-androstanediol was the most active of the 5/3-H steroids
1065
studied in the chick liver preparation employed (14). It is well known that erythropoiesis in the midterm human fetus occurs mainly in the fetal liver (15) and this fact along with the findings of Kappas and Granick led us to consider that 5/3-H steroids may play an important role in hemoglobin synthesis in the human fetus. The results obtained on the stimulation of heme and hemoglobin synthesis by testosterone and 5/3- and 5a-H metabolites in human fetal liver in culture have recently been published (16,17). References 1. Rubin, B. L., and H. J. Strecker, Endocrinology 69: 257, 1961. 2. Kraulis, I., and R. B. Clayton, J Biol Chem 243: 3546, 1968. 3. Baulieu, E. E., and P. Mauvais-Jarvis,/ Biol Chem 239: 1578, 1964. 4. , and P. Robel In Eik-Nes, K. B. (ed.), The Androgens of the Testis, Marcel-Dekker, New York, 1970, p. 68. 5. Ruse, J. L., and S. Solomon, Biochemistry 5: 1065, 1966. 6. YoungLai, E., and S. Solomon, Biochemistry 6: 2040, 1967. 7. Mikhail, G., N. Wiqvist, and E. Diczfalusy, Ada Endocrinol (Kbh) 43: 213, 1963. 8. Bird, C. E., N. Wiqvist, E. Diczfalusy, and S. Solomon,7 Clin Endocrinol Metab 26: 1144, 1966. 9. Milner, A. J., and I. H. Mills, J Endocrinol 47: 379, 1970. 10. Ghraf, R., H-G. Hoff, E. R. Lax, and H. Schriefers, Ada Endocrinol 73: 577, 1973. 11. Denef, C., Endocrinology 94: 1577, 1974. 12. Benagiano, G., F. A. Kind, F. Zielske, N. Wiqvist, and E. Diczfalusy, Ada Endocrinol 56: 203, 1967. 13. Granick, S., and A. Kappas, J Biol Chem 242: 4587, 1967. 14. Kappas, A., and S. Granick, 7 Biol Chem 243: 346, 1968. 15. Wintrobe, M., In Clinical Hematology, Lea and Febiger, Philadelphia, 1961, p. 2. 16. Congote, L. F., M. D. Stern, and S. Solomon, Biochemistry 13: 4255, 1974. 17. , and S. Solomon, Proc Natl Acad Sci USA 72: 523, 1975.
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 14 November 2015. at 01:22 For personal use only. No other uses without permission. . All rights reserved.
