Steroid Formation and Differentiation of Cortical Cells in Tissue Culture of Human Fetal Adrenals in the Presence and Absence of ACTH1 A. I. KAHRI, I. HUHTANIEMI, AND M. SALMENPERA Second Department of Pathology, and Department of Medical Chemistry, University of Helsinki, Helsinki, Finland and their pattern was transformed into the adult type with a 30-200 times higher secretion rate of cortisol. Cortical cells capable of proliferation in the culture had the ultrastructure of the permanent zone cells of the fetal adrenal or adult zona glomerulosa type. ACTH stimulation induced a differentiation of these cells into zona fasciculata type. The results suggest that ACTH is the main hormonal regulator in the genesis of the adult human adrenal cortex and that there is a factor during fetal life which inhibits the synthesis of the 3/3-hydroxysteroid dehydrogenase system. (Endocrinology 98: 33, 1976)
ABSTRACT. Steroid secretion and ultrastructural differentiation of human fetal adrenal cortical cells were analyzed in tissue culture with and without ACTH. The unconjugated and sulfated endogenous neutral steroids were analyzed by gas-liquid chromatography and gas chromatography-mass spectrometry. A fetal pattern of neutral steroids, including high concentrations of sulfate conjugates, was found during the first five days of the cultivation. At 6 to 11 days of cultivation, a decrease was seen in concentrations of these steroids. However, when stimulated with ACTH, an increasing amount of steroids was secreted during days 6 to 11
T
HE steroidogenic capacity of human fetal adrenals in vitro is well understood (1-4). The lack of enzymatic activity of the 3/3-hydroxysteroid dehydrogenase system in the human fetal adrenal cortex induces the production of steroids of the 3j3-hydroxy-A5 series and an insufficient ability to form other steroids of the 3-ketoA4 series. The regulation of the development of the fetal adrenal cortex, especially of the fetal zone, is still an unsolved problem. Tissue culture studies have shown that the basic cell type growing in vitro is a typical undifferentiated zona glomerulosa cell (5). This corresponds to cells of the permanent zone of human fetal adrenals and adult zona glomerulosa cells. Cortisol is the main steroid produced in culture of human fetal adrenals if analyzed by the fluorescent method (6-9). The object of this study was to examine the steroidogenic properties of human fetal adrenals under in vitro conditions in a tissue culture system suitable for long-term
cultivation of the adrenal cortex (10). The basal steroid secretion of the undifferentiated cultivated cortical cells during different growth periods was analyzed and the effects of ACTH on the ultrastructural and functional differentiation of cortical cells of human fetal adrenals in tissue culture were compared.
Received March 3, 1975. 1 Supported by a grant from the Sigrid Juselius Foundation.
Reference steroids were purchased from Ikapharm (Ramat-Gan, Israel) and Steraloids Inc. (Pawling, N.Y., U.S.A.).
Materials and Methods Tissue. Human early mid-term fetuses were obtained from interruptions of pregnancy for sociomedical and therapeutic reasons. The crownrump lengths of the three fetuses used in this study were 13, 14 and 20 cm, corresponding to gestational ages of 15, 16 and 20 weeks (11). The fetus was delivered by abdominal hysterectomy and immediately bled from the umbilical cord. When the blood flow had ceased, the adrenals were removed aseptically and transported to the laboratory in ice-cold Hanks' BSS solution. Solvents were as described previously (12).
33
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 15 November 2015. at 21:35 For personal use only. No other uses without permission. . All rights reserved.
34
KAHRI, HUHTANIEMI AND SALMENPERA
Trivial and systematic names. 11/3-hydroxyandrostenedione, 11/3-hydroxyandrost - 4 - ene - 3, 17-dione; 16a-hydroxydehydroepiandrosterone, 3)8,16a - dihydroxyandrost - 5 - en - 17 - one; 17a - hydroxyprogesterone, 17a - hydroxypregn 4 - ene - 3,20 - dione; 17a-hydroxypregnenolone, 3/3,17a - dihydroxypregn - 5 - en - 20 - one; prednisolone, ll/3,17a,21 - trihydroxypregn - 1, 4 - diene - 3,20 - dione; stigmasterol, (24S) - 24ethylcholesta-5,22-diene-3/3-ol. Electron microscopy. The tissue cultures were carefully studied by phase contrast microscopy during the experiment and before fixation. The cultures were fixed in situ in 2.5% glutaraldehyde in Hanks' BSS adjusted to pH 7.3 at the beginning of fixation and postfixed in 1% osmium tetroxide with phosphate buffer (13). After dehydration in ethyl alcohol the cultures were embedded in situ in the culture vessels in a mixture of Epon 812 and Araldite 6005. Thin sections were stained with 0.2% lead citrate for 20 seconds (14). Electron micrographs were made at original magnifications of 1,500 to 30,000 with a Hitachi HS-7S electron microscope. Gas-liquid chromatography (GLC) of steroid trimethylsilyl (TMS) (15) or 0-methyloxime trimethylsilyl (MO-TMS) (16) ethers was performed on 3% QF-1 and 2.2% SE-30 columns using flame ionization detection as described previously (17). Gas chromatography-mass spectrometry (GCMS) was carried out using a computerized (Varian Spectro System 100 MS) gas chromatograph-mass spectrometer (Varian, Model CH-7). The same steroid derivatives and GLC columns were used as in conventional GLC. The energy of the bombarding electrons was 70 eV and the ionizing current 300 fiA.
Encln < 1976 Vol 98 . No 1
culture was approximately 10 mg. Adrenocorticotrophic hormone (Cortrophine, Organon, The Netherlands) was added to the culture medium at the rate of 100 mU/ml daily for six days from the 6th cultivation day up to and including the 11th day. A part of the fetal adrenals was cultivated as a suspension culture. It was found that approximately half of the original tissue was released as a suspension of single cells or cell groups without enzymatic digestion when the adrenals were chopped up with scissors until the fragments were fine enough. Six to 10 cultures of the suspended cells were made from the adrenals of fetus together with the 15 to 30 explant cultures. The culture medium was changed every fifth day during cultivation. Steroid analysis. The media (4-5 ml) from 2-10 similarly cultured dishes were pooled and analyzed for neutral steroids. The analysis was performed essentially as described earlier (12, 18). In short, the procedure was as follows: The culture medium was diluted with four volumes of acetone/ethanol (1:1, v/v) and incubated overnight at 39 C. After filtration, the sample was chromatographed on a 4 g Sephadex LH-20 column and fractions of unconjugated, mono- and disulfated steroids were obtained (19). The unconjugated steroid fraction was purified by solvent partition between 50 ml of ethylacetate and 10 ml of 0.1 M sodium hydroxide and the sulfate conjugates were solvolysed. Final purification of the steroid fractions was carried out by silicic acid chromatography (200 mg) (4). The compounds in the fraction of unconjugated steroids were converted to their MO-TMS ethers and those in the sulfate conjugate fractions to the TMS ethers and analyzed by GLC and GCMS (see above). Quantitative measurements of the compounds were carried out using stigmasterol as an internal standard (17).
Results Tissue culture. The culture method employed was found in earlier studies (5) to be suitable for long-term cultivation of rat and human adrenals. The medium consisted of 50% Melnick's solution A (Hanks' BSS + 0.5% lactalbumin hydrolyzate), 20% calf serum (ultrafiltered and heat inactivated for 30 min at 56C) (State Serum Institute and 25% Eagle's Minimum Essential Medium, (Pharmaceutical Manufacture Orion Oy, Finland). The average amount of tissue per
General growth characteristics In both culture types (explant cultures and suspension cultures) growth started typically from small colonies of cortical cells (5). Outgrowth from explants was very slow or nil. In both culture types a small number of fetal zone cells was still present at the beginning of cultivation. However,
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 15 November 2015. at 21:35 For personal use only. No other uses without permission. . All rights reserved.
DIFFERENTIATION OF HUMAN FETAL ADRENAL
35
most of them had damaged plasma membranes or the cytoplasmic organelles had dispersed into the surroundings. In both types of culture, cells from the permanent zone seemed to tolerate best the preparation of cultures. Examined from the very beginning of cultivation, the permanent zone cells with an ultrastructure of the adult zona glomerulosa cells seemed to be the only cortical cells growing in colonies touching the bottom of the petri dishes (Fig. 1). The ultrastructure of cortical cells in the second and third week of culture The ultrastructure of cultured cortical cells of human fetal adrenals has been described in detail earlier (5). Generally, the cortical cells in vitro had the ultrastructure of zona glomerulosa cells of the permanent zone of the adrenal cortex.
FlG. 2. Cortical cells in tissue culture of human fetal adrenals stimulated with ACTH (0.1 U/ml/6 days) and cultivated 11 days. Note the increased size of the cells and very prominent nucleoli. (xl570)
The majority of the cortical cells in suspension had the ultrastructure of the cells of the permanent zone of the fetal adrenals. Effects ofACTH on the ultrastructure of the cortical cells of human fetal adrenals in tissue culture
FlG. 1. Cortical cells in tissue culture of human fetal adrenals cultivated 11 days. (x!570)
ACTH treatment induced an increase in the size of the cortical cells (Fig. 2). This appeared mainly as an enlargement of the cytoplasmic volume. The nuclei became spherical; the nucleoli were greatly increased in size and were usually localized in the middle of the nucleus as spherical bodies (Fig. 3). The nuclear heterochromatin disappeared or was seen only around the nucleus. The cortical cells were separated from each other and a large amount of well developed microvilli was seen between the
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 15 November 2015. at 21:35 For personal use only. No other uses without permission. . All rights reserved.
36
KAHRI, HUHTANIEMI AND SALMENPERA
En do • 1976 Vol 98 • No 1
FlG. 3. Cortical cells in tissue culture of human fetal adrenals stimulated with ACTH (0.1 U/ml/6 days) and cultivated 11 days. Note the development of microvilli and the lack of heterochroniatin in the nucleoli. (x4200)
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 15 November 2015. at 21:35 For personal use only. No other uses without permission. . All rights reserved.
DIFFERENTIATION OF HUMAN FETAL ADRENAL
37
FlC. 4. ACTH induced development of smooth-suifaced endoplasmic reticulum in cortical cells of human fetal adrenals cultivated 11 days. Note the lack of dense granules in the mitochondrial matrix, (x 15000)
cortical cells as protrusions of the cytoplasmic membrane. Some development of junctional complexes between the cortical cells was found. One of the most striking effects of ACTH on the ultrastructure of the cortical cells was the appearance and development of smooth-surfaced endoplasmic reticulum membranes (Fig. 4). The cytoplasm was packed with tubular profiles of smooth-surfaced endoplasmic reticulum membranes. A typical feature was the disappearance of groups of lamellar, roughsurfaced endoplasmic reticulum membranes. Rough membranes became a part of the tubular system of the smooth-surfaced endoplasmic reticulum. They formed a continuous system of tubular membranes which
were partly smooth-, partly rough-surfaced. Varying amounts of glycogen particles were still found in the cytoplasm during ACTH stimulation. Ribosomes appeared as polysomes which formed beautifully developed spirals in large groups. Some helical polysomes containing seven helices were also found. Constant findings included lipid droplets throughout the cytoplasm. There was an increase in the number of mitochondrial profiles in the ACTH-treated cortical cells. The mitochondria had tubular or lamellar inner membranes. No morphological differences in the shape of the inner membranes were observed between ACTH-stimulated cortical cells and cortical cells cultivated without ACTH. ACTH in-
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 15 November 2015. at 21:35 For personal use only. No other uses without permission. . All rights reserved.
E ndo • 1976 Vo 1 9 8 • No 1
KAHR1, HUHTANIEMI AND SALMENPERA
38
TABLE 1. Neutral steroids secreted into the culture medium by human fetal adrenal1 cortical cells grownin tissue culture. The amounts are expressed as ng/ml of medium Experiment II
Experiment i*
Steroids identified Unconjuituted steroids Androsteiiedione 1 1/3-hydroxyandrostenedione Dehyclroepiandrosterone 17
)
1
0-5 days culture
100 37 180
6- 11 days culture
29 — 38
•
6- 11 clays culture +ACTHI
0-5 days culture
31
840
1,700 490
_ 100
6-11 clays culture
21 — 61
1 600
17a-hydro.\ypregnenolone Cortisol
29 110
Sulfatc conjugates DehydroepiaiKlrosterone 5-androstene-3/3,l7ft-diol PreKnenolone 17a-liyclroxypregneiiolone
1,500 25 380
Total free steroids Total sulfate conjugates Total steroids
456 2,065 2,521
160
32 200
7,400
15
2,400
350
1,900
83 10 71
1,300 25 920
1,100
4.2
500
69
299 30.2 329.5
12,380 4,100 16,480
230 400 120 180 540
37 49
11
100
6-11 days culture + ACT II
28 48
1,000 6.0 200
_
Experiment III
207
1,275 1,482
9.0
168 173 351
0 - 5 clays culture
120
83 190 97
72 240
500
12 59 7.2 13
3,370 2,745 6,115
732 91.2 823.2
6 - 1 1 clay.s culture
190 21 150 21 51 38
13 89 4.0 11 471 1 14 585
6- 1 1 clays culture + ACTH
1,700 3,600 780 250 830
6,500 5.800 82
2,100 270
13,660 8,252 21,912
: Experiment 1 — adrenals lioin a 13 cm fetus (cr-lengtli), abortion lor socio-medical reasons. Experiment II = adrenals from a 14 cm fetus (cr-length), abortion for socio-medical reasons. Experiment III = adrenals from a 20 cm (cr-length), therapeutic abortion for the mother's rheumatoid arthritis, mother treated with prednisolone ( 3 x 5 nig/day) before the operation,
f ACTH added 0.1 lU/ml/clav.
duced the disappearance of electron-dense granules from the mitochondria] matrix.
when, on both liquid phases used, the relative retention times (calculated against 5a-cholestane) and mass spectra of its Steroid analysis TMS or MO-TMS ether were the same Ten neutral steroids were found in the as those of the same derivatives of the culture media analyzed (Table 1). Six of authentic compound. Figure 5 shows the these were in the fraction of unconjugated mass spectrum of the MO-TMS ether of a steroids and four as monosulfates. No compound identified as 11/3-hydroxyandrodisulfate conjugates were present within stenedione. (For mass spectra of the other the range of detection (about 100 ng). A compounds detected, see references 20-22). GLC analysis of the calf serum used in compound was considered to be identified 100
M 432
M-(90*3D 311
125 M-90 342 M-(9O15) 327 M-(9O31+15) I, 296
200
300
250
M-31 401 M-15 417
M-(30*31) 371
350
400
450
m/e FIG. 5. Mass spectrum of the MO-TMS ether of the compound identified as 11/3-hydioxyanclrostenedione.
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 15 November 2015. at 21:35 For personal use only. No other uses without permission. . All rights reserved.
DIFFERENTIATION OF HUMAN FETAL ADRENAL
the incubation medium revealed that this material contained neutral steroids in such low concentrations that their contribution, both qualitative and quantitative, to the results can be disregarded. C19 and C2i steroids of both the 3/3hydroxy-A5 and the 3-keto-A4 series were present in the free steroid fraction. In addition to cortisol, traces of several other compounds with a mass spectrum typical of C2i steroids of the 3-keto-A4 series were observed, but their concentrations were too low for accurate identification. No progesterone could be detected. Both C19 and C2i steroids were present in the monosulfate fraction, but all of them had a 3/3-hydroxyA5 structure. Distinct changes were seen with advancing culture time in the concentrations of the steroids identified. In experiments I and II, a dominant part of the steroids was monosulfated during the first five culture days. During the subsequent five days there was a decrease in the total concentrations of both free and sulfated steroids. This decrease was especially clear in the concentrations of the sulfate conjugates. In contrast to this decrease, when ACTH was added to the culture dishes during culture days 6—11, a dramatic increase was seen in the quantities of each steroid identified. This increase was especially distinct for the unconjugated steroids. Compared with the values for days 0-5, the concentration of cortisol increased 30-200 fold. The results of experiment III are not fully comparable with those of the two other experiments since in III the mother had received corticosteroid therapy for rheumatoid arthritis prior to and during the operation (Table 1). The initial high concentration of steroid monosulfates was not seen here in the medium. However, during culture days 6-11, after ACTH stimulation, an increase in the steroid concentrations was seen similar to that in the other two experiments. In experiment I, the cortisol concentration of six individual dishes was determined. The mean concentration in the medium during the initial five days was 148 ng/ml (SEM
39
= 22 ng/ml). After ACTH stimulation, the cortisol concentration was 7,100 ng/ml (SEM = 510 ng/ml). The low variability between individual culture dishes justified making subsequent determinations from pooled samples. Discussion The formation of all the steroids detected in this study has been previously demonstrated in human fetal adrenal in vitro incubations and whole fetal perfusions (1-3). The steroid pattern present endogenously in the fetal adrenal tissue has been determined in this laboratory (12,23). The nature of the conjugation of the endogenous steroids was the same as in the present study; both unconjugated and monosulfated steroids were found. The clearest difference was that all the endogenous steroids detected had a 3/3-hydroxy-A5 structure, while here several steroids of the 3-keto-A4 series were also found. An explanation for this could be the difference in 3/3-hydroxysteroid dehydrogenase activity in the fetal and adult zones of the fetal adrenal cortex. The former part lacks 3j3-hydroxysteroid dehydrogenase (24,25) and forms in vivo the main part of this organ during early and mid-pregnancy (26). In tissue culture the new generation of cultured cells has characteristics of the adult cell type which has an active 3/3-hydroxysteroid dehydrogenase system (24). Of the individual steroids detected, 5androstene-3/3,17a-diol monosulfate is of special interest. This steroid is a known constituent of fetal body fluids (cord blood, amniotic fluid, bile and meconium) (18,27, 29). Its concentration in all these sources exceeds that of the respective 17/3-isomer. The formation of the 17/3-isomer has been demonstrated in several fetal tissues (30), but the site of synthesis of 5-androstene3j8,17a-diol is still unknown. This study suggests that the fetal adrenal cortex has a role in the formation of this steroid. After ACTH administration there is a clear increase in the concentration of the steroids
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 15 November 2015. at 21:35 For personal use only. No other uses without permission. . All rights reserved.
40
KAHRI, HUHTANIEMI AND SALiMENPERA
accumulated in the culture media (Table 1). This increase is especially clear in the amount of cortisol. The pathway of its formation, however, remains unclear since one of its possible precursors, progesterone, was not detected. Whether this suggests that in the experimental conditions used progesterone has a very rapid turnover time or that the other 3-keto-A4 steroids are formed via pathways excluding progesterone (i.e., via 17a-hydroxypregnenolone and 17a-hydroxyprogesterone) remains to be solved. In experiment III, the mother was treated for rheumatoid arthritis with prednisolone prior to the operation. In this case, in contrast to the other adrenals cultured, low steroid concentrations were observed in the culture medium of the initial five days. There is evidence that corticosteroids can cross the placenta (31) and that the steroidogenic capacity of the fetal adrenals decreases when corticosteroids are given to the mother (32-33). Thus, it seems likely that the initial low steroid production in experiment III was caused by the corticosteroid therapy given to the mother. After ACTH stimulation, however, the new generation of adrenal cortical cells seemed to retain their steroidogenic capacity. The origin of cortical cells in tissue culture of human fetal adrenals The human adrenal cortex consists, during fetal life, mainly of two different cell types, fetal zone and permanent zone (26). The permanent zone cells have an ultrastructure similar to that of adult zona glomerulosa cells (34). According to the present study, at the beginning of cultivation there are definitely two different types in culture, even in the suspension culture: fetal zone cells and permanent zone cells. But the present observations revealed no significant growth in the fetal zone cells and no mitoses. Instead, there occurred continuous depression of the fetal zone cells and disappearance of their ultrastructural characteristics. Following this develop-
Endo • 1976 \'ol 98 • No 1
ment, it was difficult after one week's cultivation to find original fetal zone cells in the cultures (5). It seems to be clear that fetal zone cells are responsible for the steroids secreted during this period and whose typical characteristic is lack of 3/3hydroxysteroid dehydrogenase and whose pattern is similar that of the in vivo adrenals (4). This inability of the fetal adrenal cortex to secrete cortisol in the absence of ACTH at the beginning of cultivation has been reported by others (6,9). All these results indicate that the basal cell type capable of proliferation in the human adrenal cortex is a cortical cell of the zona glomerulosa type, as has been found in the rat adrenal cortex under similar conditions (10). The role of ACTH in the development of human adrenal cortex ACTH induced a very dramatic increase in the secretion of steroids of adult type, namely cortisol. The quantity of cortisol increased from 0.1 /ug/ml to 10 /xg/ml. This rise in functional capacity was reflected in the increased stability of the cortical cell colonies and changes in the ultrastructure of individual cortical cells. The time needed for ultrastructural differentiation and for increased capacity is similar to that in cultures of fetal rat adrenal cortex (10,35). During six days' stimulation the cortical cells gained the capacity of fasciculata cells of adult human adrenal cortex. Comments on the regulation of fetal zone development and steroid secretion As presented in our study, the proliferation of a cell population which has the ultrastructural characteristics of the permanent zone cells of human fetal adrenals, also has the capacity for 3/3-hydroxysteroid dehydrogenation if cultivated in vitro. This can be interpreted to mean that these cells have no genetic inability to form 3-keto-A4 steroids in the fetal phase. Stimulation with ACTH produced normal differentiation and
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 15 November 2015. at 21:35 For personal use only. No other uses without permission. . All rights reserved.
DIFFERENTIATION OF HUMAN FETAL ADRENAL an increased capacity for secretion of 3-ketoA4 steroids of adult zona fasciculata type. However, it seems warranted to suggest that the assumed similar response to ACTH of the permanent zone cells in vivo and the formation of highly differentiated fetal zone cells are the result of ACTH secreted by the fetal hypophysis. During fetal development at least the 3/3-hydroxysteroid dehydrogenase system is inhibited. This inhibition possibly represents a depression of a genetic locus during differentiation, because there is no activation of this enzyme system when fetal zone cells are taken from the fetus and cultivated in vitro. The cortical cells do not regain their enzymatic capacity for the formation of 3-keto-A4 steroids. Hence, this inhibition of the synthesis of the 3/3hydroxysteroid dehydrogenase system is irreversible and occurs in highly differentiated fetal zone cells in vivo and in vitro. The results of steroid secretion during the different cultivation periods provide some evidence that the 3/3-hydroxysteroid dehydrogenase system may not be the only enzyme which is depressed in the formation of corticosteroids. References 1. Diczfalusy, E., Excerpta Medico, Int Congr Ser 183: 65, 1969. 2. Eberlein, W. R., In Christy, N. P. (ed.), The Human Adrenal Cortex, Harper & Row, New York, 1971, p. 317. 3. Schindler, A. E., Dtsch Med Wochenschr 97: 1712, 1972. 4. Huhtaniemi, I., Steroids 23: 145, 1974. 5. Kahri, A., and H. Halinen, Acta Anat 88: 541, 1974. 6. Stark, E., A. Gyevai, K. Szalay, and Zs. Acs, Can J Phijsiol Pharmacol 43: 1, 1965. 7. Jost, A., Recent Frog Horm Res 22: 541, 1966. 8. Gyevai, A., B. Bukylya, K. Mihaly, K. Szalay, and E. Stark, Symp Biol Hung 14: 73, 1972. 9. Roos, B. R., Endocrinology 94: 685, 1974. 10. Kahri, A., Ada Endocrinol (Kbh) 52: Suppl 108, 1966.
41
11. Tanimura, T., T. Nelson, R. R. Hollingsworth, and T. H. Shepard, Anat Rec 171: 227, 1971. 12. Huhtaniemi, I., T. Luukkainen, and R. Vihko, Acta Endocrinol (Kbh) 64: 273, 1970. 13. Millonig, C , In Breese, S. S. (ed.), Fifth International Congress for Electron Microscopy, vol. 2, Academic Press, New York, 1962, p. 8. 14. Venable, J., and R. Coggeshall, J Cell Biol 25: 407, 1965. 15. Luukkainen, T., W. J. A. van den Heuvel, E. O. A. Haahti, and E. C. Homing, Biochim Biophys Ada 52: 599, 1961. 16. Fales, H. M., and T. Luukkainen, Anal Chem 37: 955, 1965. 17. Vihko, R., Acta Endocrinol (Kbh) 52: Suppl 109, 1966. 18. Huhtaniemi, I., and R. Vihko, J Endocrinol 57: 143, 1973. 19. Jiinne, O., Clin Chim Acta 29: 529, 1970. 20. Aringer, L., P. Eneroth, and J.-A. Gustafsson, Steroids 17: 377, 1971. 21. Baillie, T. A., C. J. W. Brooks, and B. S. Middleditsh, Anal Chem 44: 30, 1972. 22. Thompson, R. H., Jr., N. D. Young, J. E. Harten, T. A. Springer, R. Vihko, and C. C. Sweeley, Gas Chromatography and Mass Spectrometry of Selected C1B and C2, Steroids, East Lansing, Michigan, 1973. 23. Huhtaniemi, I., Steroids 21: 511, 1973. 24. Goldman, A. S., W. C. Yakovac, and A. M. Bongiovanni,/ Clin Endocrinol Metab 26: 14, 1966. 25. Cavallero, C , and U. Magvini, Excerfita Medica Int Congr Ser 132: 667, 1967. 26. Johannisson, E., Acta Endocrinol (Kbh) 58: Suppl 130, 1968. 27. Huhtaniemi, I., and R. Vihko, Steroids 16: 197, 1970. 28. , and , Ann Med Exp Biol Fenn 48: 188, 1970. 29. J Endocrinol 59: 503, 1973. 30. Lisboa, B. P., Excerpta Medica Int Congr Ser 219: 511, 1971. 31. Migeon, C. J., J. Bertrand, and C. A. Gunzell, Recent Prog Horm Res 17: 207, 1961. 32. Wray, P. M., and C. S. Russell, J Obstet Gynaecol Br Commonio 71: 97, 1964. 33. Simmer, H. H., W. J. Dignam, W. E. Easterling, M. V. Frankland, and F. Naftolin, Steroids 8: 1966. 34. Long, J. A., and A. L. Jones, Lab Invest 17: 355, 1967. 35. Kahri, A., S. Pesonen, and A. Saure, Steroidologia 1: 25, 1970.
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 15 November 2015. at 21:35 For personal use only. No other uses without permission. . All rights reserved.
ABSTRACT. Steroid secretion and ultrastructural differentiation of human fetal adrenal cortical cells were analyzed in tissue culture with and without ACTH. The unconjugated and sulfated endogenous neutral steroids were analyzed by gas-liquid chromatography and gas chromatography-mass spectrometry. A fetal pattern of neutral steroids, including high concentrations of sulfate conjugates, was found during the first five days of the cultivation. At 6 to 11 days of cultivation, a decrease was seen in concentrations of these steroids. However, when stimulated with ACTH, an increasing amount of steroids was secreted during days 6 to 11
T
HE steroidogenic capacity of human fetal adrenals in vitro is well understood (1-4). The lack of enzymatic activity of the 3/3-hydroxysteroid dehydrogenase system in the human fetal adrenal cortex induces the production of steroids of the 3j3-hydroxy-A5 series and an insufficient ability to form other steroids of the 3-ketoA4 series. The regulation of the development of the fetal adrenal cortex, especially of the fetal zone, is still an unsolved problem. Tissue culture studies have shown that the basic cell type growing in vitro is a typical undifferentiated zona glomerulosa cell (5). This corresponds to cells of the permanent zone of human fetal adrenals and adult zona glomerulosa cells. Cortisol is the main steroid produced in culture of human fetal adrenals if analyzed by the fluorescent method (6-9). The object of this study was to examine the steroidogenic properties of human fetal adrenals under in vitro conditions in a tissue culture system suitable for long-term
cultivation of the adrenal cortex (10). The basal steroid secretion of the undifferentiated cultivated cortical cells during different growth periods was analyzed and the effects of ACTH on the ultrastructural and functional differentiation of cortical cells of human fetal adrenals in tissue culture were compared.
Received March 3, 1975. 1 Supported by a grant from the Sigrid Juselius Foundation.
Reference steroids were purchased from Ikapharm (Ramat-Gan, Israel) and Steraloids Inc. (Pawling, N.Y., U.S.A.).
Materials and Methods Tissue. Human early mid-term fetuses were obtained from interruptions of pregnancy for sociomedical and therapeutic reasons. The crownrump lengths of the three fetuses used in this study were 13, 14 and 20 cm, corresponding to gestational ages of 15, 16 and 20 weeks (11). The fetus was delivered by abdominal hysterectomy and immediately bled from the umbilical cord. When the blood flow had ceased, the adrenals were removed aseptically and transported to the laboratory in ice-cold Hanks' BSS solution. Solvents were as described previously (12).
33
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 15 November 2015. at 21:35 For personal use only. No other uses without permission. . All rights reserved.
34
KAHRI, HUHTANIEMI AND SALMENPERA
Trivial and systematic names. 11/3-hydroxyandrostenedione, 11/3-hydroxyandrost - 4 - ene - 3, 17-dione; 16a-hydroxydehydroepiandrosterone, 3)8,16a - dihydroxyandrost - 5 - en - 17 - one; 17a - hydroxyprogesterone, 17a - hydroxypregn 4 - ene - 3,20 - dione; 17a-hydroxypregnenolone, 3/3,17a - dihydroxypregn - 5 - en - 20 - one; prednisolone, ll/3,17a,21 - trihydroxypregn - 1, 4 - diene - 3,20 - dione; stigmasterol, (24S) - 24ethylcholesta-5,22-diene-3/3-ol. Electron microscopy. The tissue cultures were carefully studied by phase contrast microscopy during the experiment and before fixation. The cultures were fixed in situ in 2.5% glutaraldehyde in Hanks' BSS adjusted to pH 7.3 at the beginning of fixation and postfixed in 1% osmium tetroxide with phosphate buffer (13). After dehydration in ethyl alcohol the cultures were embedded in situ in the culture vessels in a mixture of Epon 812 and Araldite 6005. Thin sections were stained with 0.2% lead citrate for 20 seconds (14). Electron micrographs were made at original magnifications of 1,500 to 30,000 with a Hitachi HS-7S electron microscope. Gas-liquid chromatography (GLC) of steroid trimethylsilyl (TMS) (15) or 0-methyloxime trimethylsilyl (MO-TMS) (16) ethers was performed on 3% QF-1 and 2.2% SE-30 columns using flame ionization detection as described previously (17). Gas chromatography-mass spectrometry (GCMS) was carried out using a computerized (Varian Spectro System 100 MS) gas chromatograph-mass spectrometer (Varian, Model CH-7). The same steroid derivatives and GLC columns were used as in conventional GLC. The energy of the bombarding electrons was 70 eV and the ionizing current 300 fiA.
Encln < 1976 Vol 98 . No 1
culture was approximately 10 mg. Adrenocorticotrophic hormone (Cortrophine, Organon, The Netherlands) was added to the culture medium at the rate of 100 mU/ml daily for six days from the 6th cultivation day up to and including the 11th day. A part of the fetal adrenals was cultivated as a suspension culture. It was found that approximately half of the original tissue was released as a suspension of single cells or cell groups without enzymatic digestion when the adrenals were chopped up with scissors until the fragments were fine enough. Six to 10 cultures of the suspended cells were made from the adrenals of fetus together with the 15 to 30 explant cultures. The culture medium was changed every fifth day during cultivation. Steroid analysis. The media (4-5 ml) from 2-10 similarly cultured dishes were pooled and analyzed for neutral steroids. The analysis was performed essentially as described earlier (12, 18). In short, the procedure was as follows: The culture medium was diluted with four volumes of acetone/ethanol (1:1, v/v) and incubated overnight at 39 C. After filtration, the sample was chromatographed on a 4 g Sephadex LH-20 column and fractions of unconjugated, mono- and disulfated steroids were obtained (19). The unconjugated steroid fraction was purified by solvent partition between 50 ml of ethylacetate and 10 ml of 0.1 M sodium hydroxide and the sulfate conjugates were solvolysed. Final purification of the steroid fractions was carried out by silicic acid chromatography (200 mg) (4). The compounds in the fraction of unconjugated steroids were converted to their MO-TMS ethers and those in the sulfate conjugate fractions to the TMS ethers and analyzed by GLC and GCMS (see above). Quantitative measurements of the compounds were carried out using stigmasterol as an internal standard (17).
Results Tissue culture. The culture method employed was found in earlier studies (5) to be suitable for long-term cultivation of rat and human adrenals. The medium consisted of 50% Melnick's solution A (Hanks' BSS + 0.5% lactalbumin hydrolyzate), 20% calf serum (ultrafiltered and heat inactivated for 30 min at 56C) (State Serum Institute and 25% Eagle's Minimum Essential Medium, (Pharmaceutical Manufacture Orion Oy, Finland). The average amount of tissue per
General growth characteristics In both culture types (explant cultures and suspension cultures) growth started typically from small colonies of cortical cells (5). Outgrowth from explants was very slow or nil. In both culture types a small number of fetal zone cells was still present at the beginning of cultivation. However,
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 15 November 2015. at 21:35 For personal use only. No other uses without permission. . All rights reserved.
DIFFERENTIATION OF HUMAN FETAL ADRENAL
35
most of them had damaged plasma membranes or the cytoplasmic organelles had dispersed into the surroundings. In both types of culture, cells from the permanent zone seemed to tolerate best the preparation of cultures. Examined from the very beginning of cultivation, the permanent zone cells with an ultrastructure of the adult zona glomerulosa cells seemed to be the only cortical cells growing in colonies touching the bottom of the petri dishes (Fig. 1). The ultrastructure of cortical cells in the second and third week of culture The ultrastructure of cultured cortical cells of human fetal adrenals has been described in detail earlier (5). Generally, the cortical cells in vitro had the ultrastructure of zona glomerulosa cells of the permanent zone of the adrenal cortex.
FlG. 2. Cortical cells in tissue culture of human fetal adrenals stimulated with ACTH (0.1 U/ml/6 days) and cultivated 11 days. Note the increased size of the cells and very prominent nucleoli. (xl570)
The majority of the cortical cells in suspension had the ultrastructure of the cells of the permanent zone of the fetal adrenals. Effects ofACTH on the ultrastructure of the cortical cells of human fetal adrenals in tissue culture
FlG. 1. Cortical cells in tissue culture of human fetal adrenals cultivated 11 days. (x!570)
ACTH treatment induced an increase in the size of the cortical cells (Fig. 2). This appeared mainly as an enlargement of the cytoplasmic volume. The nuclei became spherical; the nucleoli were greatly increased in size and were usually localized in the middle of the nucleus as spherical bodies (Fig. 3). The nuclear heterochromatin disappeared or was seen only around the nucleus. The cortical cells were separated from each other and a large amount of well developed microvilli was seen between the
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 15 November 2015. at 21:35 For personal use only. No other uses without permission. . All rights reserved.
36
KAHRI, HUHTANIEMI AND SALMENPERA
En do • 1976 Vol 98 • No 1
FlG. 3. Cortical cells in tissue culture of human fetal adrenals stimulated with ACTH (0.1 U/ml/6 days) and cultivated 11 days. Note the development of microvilli and the lack of heterochroniatin in the nucleoli. (x4200)
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 15 November 2015. at 21:35 For personal use only. No other uses without permission. . All rights reserved.
DIFFERENTIATION OF HUMAN FETAL ADRENAL
37
FlC. 4. ACTH induced development of smooth-suifaced endoplasmic reticulum in cortical cells of human fetal adrenals cultivated 11 days. Note the lack of dense granules in the mitochondrial matrix, (x 15000)
cortical cells as protrusions of the cytoplasmic membrane. Some development of junctional complexes between the cortical cells was found. One of the most striking effects of ACTH on the ultrastructure of the cortical cells was the appearance and development of smooth-surfaced endoplasmic reticulum membranes (Fig. 4). The cytoplasm was packed with tubular profiles of smooth-surfaced endoplasmic reticulum membranes. A typical feature was the disappearance of groups of lamellar, roughsurfaced endoplasmic reticulum membranes. Rough membranes became a part of the tubular system of the smooth-surfaced endoplasmic reticulum. They formed a continuous system of tubular membranes which
were partly smooth-, partly rough-surfaced. Varying amounts of glycogen particles were still found in the cytoplasm during ACTH stimulation. Ribosomes appeared as polysomes which formed beautifully developed spirals in large groups. Some helical polysomes containing seven helices were also found. Constant findings included lipid droplets throughout the cytoplasm. There was an increase in the number of mitochondrial profiles in the ACTH-treated cortical cells. The mitochondria had tubular or lamellar inner membranes. No morphological differences in the shape of the inner membranes were observed between ACTH-stimulated cortical cells and cortical cells cultivated without ACTH. ACTH in-
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 15 November 2015. at 21:35 For personal use only. No other uses without permission. . All rights reserved.
E ndo • 1976 Vo 1 9 8 • No 1
KAHR1, HUHTANIEMI AND SALMENPERA
38
TABLE 1. Neutral steroids secreted into the culture medium by human fetal adrenal1 cortical cells grownin tissue culture. The amounts are expressed as ng/ml of medium Experiment II
Experiment i*
Steroids identified Unconjuituted steroids Androsteiiedione 1 1/3-hydroxyandrostenedione Dehyclroepiandrosterone 17
)
1
0-5 days culture
100 37 180
6- 11 days culture
29 — 38
•
6- 11 clays culture +ACTHI
0-5 days culture
31
840
1,700 490
_ 100
6-11 clays culture
21 — 61
1 600
17a-hydro.\ypregnenolone Cortisol
29 110
Sulfatc conjugates DehydroepiaiKlrosterone 5-androstene-3/3,l7ft-diol PreKnenolone 17a-liyclroxypregneiiolone
1,500 25 380
Total free steroids Total sulfate conjugates Total steroids
456 2,065 2,521
160
32 200
7,400
15
2,400
350
1,900
83 10 71
1,300 25 920
1,100
4.2
500
69
299 30.2 329.5
12,380 4,100 16,480
230 400 120 180 540
37 49
11
100
6-11 days culture + ACT II
28 48
1,000 6.0 200
_
Experiment III
207
1,275 1,482
9.0
168 173 351
0 - 5 clays culture
120
83 190 97
72 240
500
12 59 7.2 13
3,370 2,745 6,115
732 91.2 823.2
6 - 1 1 clay.s culture
190 21 150 21 51 38
13 89 4.0 11 471 1 14 585
6- 1 1 clays culture + ACTH
1,700 3,600 780 250 830
6,500 5.800 82
2,100 270
13,660 8,252 21,912
: Experiment 1 — adrenals lioin a 13 cm fetus (cr-lengtli), abortion lor socio-medical reasons. Experiment II = adrenals from a 14 cm fetus (cr-length), abortion for socio-medical reasons. Experiment III = adrenals from a 20 cm (cr-length), therapeutic abortion for the mother's rheumatoid arthritis, mother treated with prednisolone ( 3 x 5 nig/day) before the operation,
f ACTH added 0.1 lU/ml/clav.
duced the disappearance of electron-dense granules from the mitochondria] matrix.
when, on both liquid phases used, the relative retention times (calculated against 5a-cholestane) and mass spectra of its Steroid analysis TMS or MO-TMS ether were the same Ten neutral steroids were found in the as those of the same derivatives of the culture media analyzed (Table 1). Six of authentic compound. Figure 5 shows the these were in the fraction of unconjugated mass spectrum of the MO-TMS ether of a steroids and four as monosulfates. No compound identified as 11/3-hydroxyandrodisulfate conjugates were present within stenedione. (For mass spectra of the other the range of detection (about 100 ng). A compounds detected, see references 20-22). GLC analysis of the calf serum used in compound was considered to be identified 100
M 432
M-(90*3D 311
125 M-90 342 M-(9O15) 327 M-(9O31+15) I, 296
200
300
250
M-31 401 M-15 417
M-(30*31) 371
350
400
450
m/e FIG. 5. Mass spectrum of the MO-TMS ether of the compound identified as 11/3-hydioxyanclrostenedione.
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 15 November 2015. at 21:35 For personal use only. No other uses without permission. . All rights reserved.
DIFFERENTIATION OF HUMAN FETAL ADRENAL
the incubation medium revealed that this material contained neutral steroids in such low concentrations that their contribution, both qualitative and quantitative, to the results can be disregarded. C19 and C2i steroids of both the 3/3hydroxy-A5 and the 3-keto-A4 series were present in the free steroid fraction. In addition to cortisol, traces of several other compounds with a mass spectrum typical of C2i steroids of the 3-keto-A4 series were observed, but their concentrations were too low for accurate identification. No progesterone could be detected. Both C19 and C2i steroids were present in the monosulfate fraction, but all of them had a 3/3-hydroxyA5 structure. Distinct changes were seen with advancing culture time in the concentrations of the steroids identified. In experiments I and II, a dominant part of the steroids was monosulfated during the first five culture days. During the subsequent five days there was a decrease in the total concentrations of both free and sulfated steroids. This decrease was especially clear in the concentrations of the sulfate conjugates. In contrast to this decrease, when ACTH was added to the culture dishes during culture days 6—11, a dramatic increase was seen in the quantities of each steroid identified. This increase was especially distinct for the unconjugated steroids. Compared with the values for days 0-5, the concentration of cortisol increased 30-200 fold. The results of experiment III are not fully comparable with those of the two other experiments since in III the mother had received corticosteroid therapy for rheumatoid arthritis prior to and during the operation (Table 1). The initial high concentration of steroid monosulfates was not seen here in the medium. However, during culture days 6-11, after ACTH stimulation, an increase in the steroid concentrations was seen similar to that in the other two experiments. In experiment I, the cortisol concentration of six individual dishes was determined. The mean concentration in the medium during the initial five days was 148 ng/ml (SEM
39
= 22 ng/ml). After ACTH stimulation, the cortisol concentration was 7,100 ng/ml (SEM = 510 ng/ml). The low variability between individual culture dishes justified making subsequent determinations from pooled samples. Discussion The formation of all the steroids detected in this study has been previously demonstrated in human fetal adrenal in vitro incubations and whole fetal perfusions (1-3). The steroid pattern present endogenously in the fetal adrenal tissue has been determined in this laboratory (12,23). The nature of the conjugation of the endogenous steroids was the same as in the present study; both unconjugated and monosulfated steroids were found. The clearest difference was that all the endogenous steroids detected had a 3/3-hydroxy-A5 structure, while here several steroids of the 3-keto-A4 series were also found. An explanation for this could be the difference in 3/3-hydroxysteroid dehydrogenase activity in the fetal and adult zones of the fetal adrenal cortex. The former part lacks 3j3-hydroxysteroid dehydrogenase (24,25) and forms in vivo the main part of this organ during early and mid-pregnancy (26). In tissue culture the new generation of cultured cells has characteristics of the adult cell type which has an active 3/3-hydroxysteroid dehydrogenase system (24). Of the individual steroids detected, 5androstene-3/3,17a-diol monosulfate is of special interest. This steroid is a known constituent of fetal body fluids (cord blood, amniotic fluid, bile and meconium) (18,27, 29). Its concentration in all these sources exceeds that of the respective 17/3-isomer. The formation of the 17/3-isomer has been demonstrated in several fetal tissues (30), but the site of synthesis of 5-androstene3j8,17a-diol is still unknown. This study suggests that the fetal adrenal cortex has a role in the formation of this steroid. After ACTH administration there is a clear increase in the concentration of the steroids
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 15 November 2015. at 21:35 For personal use only. No other uses without permission. . All rights reserved.
40
KAHRI, HUHTANIEMI AND SALiMENPERA
accumulated in the culture media (Table 1). This increase is especially clear in the amount of cortisol. The pathway of its formation, however, remains unclear since one of its possible precursors, progesterone, was not detected. Whether this suggests that in the experimental conditions used progesterone has a very rapid turnover time or that the other 3-keto-A4 steroids are formed via pathways excluding progesterone (i.e., via 17a-hydroxypregnenolone and 17a-hydroxyprogesterone) remains to be solved. In experiment III, the mother was treated for rheumatoid arthritis with prednisolone prior to the operation. In this case, in contrast to the other adrenals cultured, low steroid concentrations were observed in the culture medium of the initial five days. There is evidence that corticosteroids can cross the placenta (31) and that the steroidogenic capacity of the fetal adrenals decreases when corticosteroids are given to the mother (32-33). Thus, it seems likely that the initial low steroid production in experiment III was caused by the corticosteroid therapy given to the mother. After ACTH stimulation, however, the new generation of adrenal cortical cells seemed to retain their steroidogenic capacity. The origin of cortical cells in tissue culture of human fetal adrenals The human adrenal cortex consists, during fetal life, mainly of two different cell types, fetal zone and permanent zone (26). The permanent zone cells have an ultrastructure similar to that of adult zona glomerulosa cells (34). According to the present study, at the beginning of cultivation there are definitely two different types in culture, even in the suspension culture: fetal zone cells and permanent zone cells. But the present observations revealed no significant growth in the fetal zone cells and no mitoses. Instead, there occurred continuous depression of the fetal zone cells and disappearance of their ultrastructural characteristics. Following this develop-
Endo • 1976 \'ol 98 • No 1
ment, it was difficult after one week's cultivation to find original fetal zone cells in the cultures (5). It seems to be clear that fetal zone cells are responsible for the steroids secreted during this period and whose typical characteristic is lack of 3/3hydroxysteroid dehydrogenase and whose pattern is similar that of the in vivo adrenals (4). This inability of the fetal adrenal cortex to secrete cortisol in the absence of ACTH at the beginning of cultivation has been reported by others (6,9). All these results indicate that the basal cell type capable of proliferation in the human adrenal cortex is a cortical cell of the zona glomerulosa type, as has been found in the rat adrenal cortex under similar conditions (10). The role of ACTH in the development of human adrenal cortex ACTH induced a very dramatic increase in the secretion of steroids of adult type, namely cortisol. The quantity of cortisol increased from 0.1 /ug/ml to 10 /xg/ml. This rise in functional capacity was reflected in the increased stability of the cortical cell colonies and changes in the ultrastructure of individual cortical cells. The time needed for ultrastructural differentiation and for increased capacity is similar to that in cultures of fetal rat adrenal cortex (10,35). During six days' stimulation the cortical cells gained the capacity of fasciculata cells of adult human adrenal cortex. Comments on the regulation of fetal zone development and steroid secretion As presented in our study, the proliferation of a cell population which has the ultrastructural characteristics of the permanent zone cells of human fetal adrenals, also has the capacity for 3/3-hydroxysteroid dehydrogenation if cultivated in vitro. This can be interpreted to mean that these cells have no genetic inability to form 3-keto-A4 steroids in the fetal phase. Stimulation with ACTH produced normal differentiation and
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 15 November 2015. at 21:35 For personal use only. No other uses without permission. . All rights reserved.
DIFFERENTIATION OF HUMAN FETAL ADRENAL an increased capacity for secretion of 3-ketoA4 steroids of adult zona fasciculata type. However, it seems warranted to suggest that the assumed similar response to ACTH of the permanent zone cells in vivo and the formation of highly differentiated fetal zone cells are the result of ACTH secreted by the fetal hypophysis. During fetal development at least the 3/3-hydroxysteroid dehydrogenase system is inhibited. This inhibition possibly represents a depression of a genetic locus during differentiation, because there is no activation of this enzyme system when fetal zone cells are taken from the fetus and cultivated in vitro. The cortical cells do not regain their enzymatic capacity for the formation of 3-keto-A4 steroids. Hence, this inhibition of the synthesis of the 3/3hydroxysteroid dehydrogenase system is irreversible and occurs in highly differentiated fetal zone cells in vivo and in vitro. The results of steroid secretion during the different cultivation periods provide some evidence that the 3/3-hydroxysteroid dehydrogenase system may not be the only enzyme which is depressed in the formation of corticosteroids. References 1. Diczfalusy, E., Excerpta Medico, Int Congr Ser 183: 65, 1969. 2. Eberlein, W. R., In Christy, N. P. (ed.), The Human Adrenal Cortex, Harper & Row, New York, 1971, p. 317. 3. Schindler, A. E., Dtsch Med Wochenschr 97: 1712, 1972. 4. Huhtaniemi, I., Steroids 23: 145, 1974. 5. Kahri, A., and H. Halinen, Acta Anat 88: 541, 1974. 6. Stark, E., A. Gyevai, K. Szalay, and Zs. Acs, Can J Phijsiol Pharmacol 43: 1, 1965. 7. Jost, A., Recent Frog Horm Res 22: 541, 1966. 8. Gyevai, A., B. Bukylya, K. Mihaly, K. Szalay, and E. Stark, Symp Biol Hung 14: 73, 1972. 9. Roos, B. R., Endocrinology 94: 685, 1974. 10. Kahri, A., Ada Endocrinol (Kbh) 52: Suppl 108, 1966.
41
11. Tanimura, T., T. Nelson, R. R. Hollingsworth, and T. H. Shepard, Anat Rec 171: 227, 1971. 12. Huhtaniemi, I., T. Luukkainen, and R. Vihko, Acta Endocrinol (Kbh) 64: 273, 1970. 13. Millonig, C , In Breese, S. S. (ed.), Fifth International Congress for Electron Microscopy, vol. 2, Academic Press, New York, 1962, p. 8. 14. Venable, J., and R. Coggeshall, J Cell Biol 25: 407, 1965. 15. Luukkainen, T., W. J. A. van den Heuvel, E. O. A. Haahti, and E. C. Homing, Biochim Biophys Ada 52: 599, 1961. 16. Fales, H. M., and T. Luukkainen, Anal Chem 37: 955, 1965. 17. Vihko, R., Acta Endocrinol (Kbh) 52: Suppl 109, 1966. 18. Huhtaniemi, I., and R. Vihko, J Endocrinol 57: 143, 1973. 19. Jiinne, O., Clin Chim Acta 29: 529, 1970. 20. Aringer, L., P. Eneroth, and J.-A. Gustafsson, Steroids 17: 377, 1971. 21. Baillie, T. A., C. J. W. Brooks, and B. S. Middleditsh, Anal Chem 44: 30, 1972. 22. Thompson, R. H., Jr., N. D. Young, J. E. Harten, T. A. Springer, R. Vihko, and C. C. Sweeley, Gas Chromatography and Mass Spectrometry of Selected C1B and C2, Steroids, East Lansing, Michigan, 1973. 23. Huhtaniemi, I., Steroids 21: 511, 1973. 24. Goldman, A. S., W. C. Yakovac, and A. M. Bongiovanni,/ Clin Endocrinol Metab 26: 14, 1966. 25. Cavallero, C , and U. Magvini, Excerfita Medica Int Congr Ser 132: 667, 1967. 26. Johannisson, E., Acta Endocrinol (Kbh) 58: Suppl 130, 1968. 27. Huhtaniemi, I., and R. Vihko, Steroids 16: 197, 1970. 28. , and , Ann Med Exp Biol Fenn 48: 188, 1970. 29. J Endocrinol 59: 503, 1973. 30. Lisboa, B. P., Excerpta Medica Int Congr Ser 219: 511, 1971. 31. Migeon, C. J., J. Bertrand, and C. A. Gunzell, Recent Prog Horm Res 17: 207, 1961. 32. Wray, P. M., and C. S. Russell, J Obstet Gynaecol Br Commonio 71: 97, 1964. 33. Simmer, H. H., W. J. Dignam, W. E. Easterling, M. V. Frankland, and F. Naftolin, Steroids 8: 1966. 34. Long, J. A., and A. L. Jones, Lab Invest 17: 355, 1967. 35. Kahri, A., S. Pesonen, and A. Saure, Steroidologia 1: 25, 1970.
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 15 November 2015. at 21:35 For personal use only. No other uses without permission. . All rights reserved.
