3
Section ofEndocrinology
Symposium: Gonadotrophins
half-lives in circulation, and hence different potencies in the two assays. These specific methods of FSH and LH measurement soon demonstrated that the two hormones may fluctuate independently. This was well illustrated by the careful studies of Fukushima et al. (1964) on samples of urine obtained from normal subjects throughout the menstrual cycle. The patterns reported are remarkably similar to those that have subsequently been described using radioimmunoassays. A disadvantage of the biological methods is their lack of sensitivity which limited their application to extracted samples of urine and it was not until the more sensitive radioimmunoassays became available that results could be obtained on small samples of serum. The development of specific radioimmunoassays for the gonadotrophins proved difficult, however, because of the considerable cross-reactivity between antisera, probably related to the fact that the glycoprotein hormones contain a common a-subunit. Satisfactory assays were first developed for LH while those for FSH encountered greater problems. Antisera to FSH had usually to be produced to relatively crude preparations but this practice was justified because the antibodies produced might be specific to the :-subunit or nonspecific to the a-subunit even with highly purified preparations. Furthermore, for the milligram quantities needed for immunization, it was necessary to use crude preparations because of the scarcity of the more purified material. However, recent refinements in immunization techniques such as the use of intracutaneous injections at multiple sites in rabbits have enabled Ross et al. (1971) to produce antisera of high titre to the subunits of HCG using only 20-50 ,ug of im.munogen for the primary immunization. Butt et al. (1974) employed this multiple site technique with preparations of pituitary FSH with some interesting results. In the first experiments relatively crude immunogen was injected intradermally into 30-40 sites on the backs of the rabbits. The total dose was 100 ,ug and a second set of injections containing only 20 ,ug was given after about twelve weeks, by which time the antibody titre had risen and then fallen. Within two weeks high titres were obtained, but the antisera were nonspecific in all animals and showed complete cross-reactions with each of the other glycoproteins. When the experiment was repeated with highly purified FSH (CPDS/6, Butt & Lynch 1972) 3 out of 4 animals produced antisera of high titre, affinity and specificity. Cross-reactivity with LH or with thyrotrophin (TSH) could be accounted for by the known contamination of these preparations with small amounts of FSH. The same technique was successfully applied to LH and other hormones.
Dr W R Butt, Dr S S Lynch, Mr W Robinson, Mr J Williams (Birmingham and Midland Hospital for Women, Birmingham, Bl 4HL) and Dr D R London (Queen Elizabeth Hospital, Edgbaston, Birmingham, B15 2TH) Assays of Gonadotrophins and Their Applications The early biological assays for the gonadotrophins were very elaborate as their specificity depended upon histological end-points in hypophysectomized animals. Quantitative estimates were, therefore, difficult to obtain, and for routine use simpler blat nonspecific assays were employed. The best known of these used the uterine weight of immature mice as the endpoint. This was a satisfactorily precise assay and was widely applied. Unfortunately, besides being nonspecific it was affected by synergistic action between follicle-stimulating hormone (FSH) and luteinizing hormone (LH) (Butt et al. 1964). It was not until quantitative specific assays had been developed that the uncertainty as to whether there were one or two gonadotrophins in human urine could be resolved. A convenient and specific assay for FSH known as the ovarian augmentation assay, was described by Steelman & Pohley (1953). Immature rats were given injections of FSH or unknown in the presence of a large excess of human chorionic gonadotrophin (HCG): the effect of any LH in the preparation of FSH was therefore minimized. Brown (1955) described a similar assay in which mice were used which tended to be more sensitive but was less precise than the assay using rats. Greep et al. (1941) described a specific assay by which LH measurements were correlated with an increase in weight of the ventral prostate in hypophysectomized rats. The use of hyposectomized animals was avoided in a method described by Parlow (1961), in which LH acutely depletes ascorbic acid in the luteinized ovaries of pseudo-pregnant immature rats (OAAD assay). These two assays for LH have been widely used but they do not always give the same estimate of potency. This is because the ventral prostate assay measures the effect of the hormone over a period of four days, whereas the OAAD assay measures the effect over a short period of two to four hours. Thus differences in the structure of the hormone as may occur between different species or between pituitary, serum and urinary preparations may result in different biological
71
72 Proc. roy. Soc. Med. Volume 68 February 1975
4
These antisera have made available specific and sensitive assays for both FSH and LH. Differences in potency estimates which have been noted between laboratories (Taymor & Miyata 1970) may be caused by the use of different antigens for labelling, by different antisera, and by differences between standards. Now that a Supra Regional Assay Service has been set up for routine radioimmunoassays in hospitals in this country it is important that these differences are minimized as far as possible by making use of common reagents and standards. Standards which have been recommended for serum assays are the pituitary preparations MRC 69/104 for FSH and MRC 68/40 for LH, distributed by the Division of Biological Standards, Hampstead. The first is a crude but stable preparation containing both FSH and LH: each ampoule is assumed to contain 10 units FSH. MRC 68/40 is a purer preparation and is assumed to contain 77 units LH per ampoule. Normal ranges obtained in this laboratory using these standards are given in Table 1. Some of the more common clinical applications of radioimmunoassays for FSH and LH are as follows:
response in females depends on the stage of the cycle: the increase in LH at 30 and 60 minutes ranges from twofold to fivefold in the early follicular phase, and up to thirtyfold in the luteal phase. (2) (Estrogen stimulation test (Shaw 1975): 1 mg aestradiol benzoate is given by intramuscular injection. Females with normal hypothalamic-pituitary function show a release of LH within the next 72 hours. (3) Clomiphene: given orally for 5 to 7 days (maximum dose 150 mg) stimulates the release of gonadotrophins. Serum samples are taken at the beginning and at intervals until the end of treatment with clomiphene. Single determinations are particularly useful in the diagnosis of conditions in which there is excessive production of gonadotrophins, but when this is normal or subnormal dynamic test procedures are required. Patients of both sexes with gonadal failure may have grossly elevated levels of gonadotrophins. FSH tends to be raised more often than LH as it is in normal postmenopausal women (Wide et al. 1973) and so is the more useful of the two determinations in this respect. In dynamic tests, however, LH increases in response to the various stimuli more than FSH and the measurement of both hormones may give clinically useful information. Future developments in the assays of gonadotrophins may be summarized into three categories. Firstly there may be improvements in the immunoassays themselves. Specificity and possibly sensitivity appear to be no longer very great problems, but in all radioimmunoassays the labelling of the antigen is expensive and has to be done at regular intervals. Alternative procedures for labelling with radioiodine have been suggested (Marchalonis 1969, Redshaw & Lynch 1974) and labelling of the antibody instead of antigen is another approach (Woodhead et al. 1974). In addition, methods have been described in which nonradioactive end-points have been used; thus Van Weeman & Schuurs (1971) coupled peroxidase to HCG and demonstrated that in principle an enzyme-substrate reaction served as an alternative end-point to gamma counting. Such a system should provide a stable method of labelling the antigen and provide a method suitable for automation, but at present it appears that the problem of achieving adequate sensitivity has still to be overcome. Secondly radioimmunoassay is applicable to the individual subunits of the gonadotrophins and it may be that these will be shown to be of some clinical importance. Thus some interesting observations on the production of subunits by the placenta and by tumours of the trophoblast have been reported by Franchimont et al. (1972). Thirdly immunoassays do not measure the biological activity of the hormone and there are many examples in which modifications of the structure
Single estimation: Single estimations of FSH are usually of more value than single estimations of LH. A high result may indicate primary gonadal failure in conditions covering infertility and eunuchoidism in the male and primary or secondary amenorrhoea in the female. An estimation of FSH in the assessment of postmenopausal patients with suspected pituitary disease or following hypophysectomy may also be of value. A low or normal result with hypogonadism suggests hypothalamic or pituitary dysfunction and a subsequent dynamic test procedure may then be of value. Dynamic test procedures: Changes in LH are usually greater than in FSH in these procedures. Examples are: (1) Gonadotrophin-releasing hormone (LH-RH) test (London et al. 1974): 100 ,ug is given by intravenous injection and serum samples are taken at 0, 30 and 60 minutes. The Table I Gonadotrophins in the serum of normal subjects Normal basal ranges
FSH u/l (MRC 69/104)
Children: Age I year to puberty * < 2.5 Adult males
1-7
Females:
Reproductive years Midcycle peak Postmenopausal
1-5 4-15 > 15
* FSH rises towards puberty before LH
LIH
u/t (MRC 68/40) < 2.5
2.5-20
2.5-15 20-100 > 40
5
Section of Endocrinology
lead to losses of biological but not of immunological activities (Butt 1969). The complete hormonal activity of the gonadotrophin can only be indicated by bioassays in vivo, but methods which depend upon binding to a receptor site will at least measure one facet of the biological activity, and such assays using testicular or ovarian receptors have been described (Catt et al. 1972, Reichert & Bhalla 1974). Alternatively those steroids produced in vitro under the influence of the gonadotrophin may be used as the end-point (Watson 1971, Shirley & Stephenson 1973). There is also the interesting approach of Rees et al. (1973) using a system for LH similar to that originally described for ACTH (Chayen et al. 1972). The effect of LH on ascorbic acid in the luteinized ovary used in the assay of Parlow (1961) has been adapted to this in vitro technique. Reducing activity in the luteinized ovaries treated with LH is estimated by microdensitometry after staining with Prussian blue. Although this assay is tedious in comparison with radioimmunoassay it is even more sensitive and thus gives the possibility of measuring levels which are less than normal.
Dr R W Shaw (Department of Clinical Endocrinology, The Women's Hospital, Sparkhill, Birmingham, Bll 4HL)
REFERENCES Brown P S (1955) Journal ofEndocrinology 13, 59 Butt W R (1969) Acta endocrinologica (Copenhagen) Suppl. 142, p 13 Butt W R, Cunningham F J & Hartree A S (1964) Proceedings of the Royal Society ofMedicine 57, 107 Butt W R & Lynch S S (1972) In: Hormones Glycoproteiques Hypophysaires. Ed. M Jutisz. Inserm, Paris; p 83 Butt W R, Lynch S S & Shirley A (1974) Journal of Endocrinology 61, xliii Catt K J, Dufau M L & Tsuruhara T (I 972) Journal ofClinical Endocrinology and Metabolismn 34, 123 Chayen J, Loveridge N & Daly J R (1972) Clinical Endocrinology 1, 219 Franchimont P, Gaspart U, Reuter A & Heynen G (1972) Clinical Endocrinology 1, 315 Fukushima M, Stevens V C, Gantt C L & Vorys V (1964) Journal of Clinical Endocrinology and Metabolisnm 24, 205 Greep R 0, Van Dyke H B & Chow B F (1941) Proceedings of the Society of Experimental Biology and Medicine 46, 644 London D R, Butt W R, Rudd B T, Holder G, Robinson W, Duignan N & Logan-Edwards R (1975) Proceedings of the Royal Society of Medicine 68, 75-76 Marchalonis J J (1969) Biochemical Journal 113, 299 Parlow A F (1961) In: Human Pituitary Gonadotropins. Ed. A Albert. Thomas, New York; p 300 Redshaw M R & Lynch S S (1974) Journal of Endocrinology 60, 527 Rees L H, Holdaway I M, Kramer R, McNeilly A S & Chard T (1973) Nature (London) 244, 232 Reichert L E jr & Bhalla V K (1974) Endocrinology 94,483 Ross G T, Vaitukaitis J L & Robbins J B (1971) In: Structureactivity Relationships of Protein and Polypeptide Hormones. Ed. M Margoulies & F C Greenwood. Excerpta Medica Foundation, Amsterdam; Part 1, p 153 Shaw R W (1975) Proceedings of the Royal Society of Medicine 68, 73-75 Shirley A & Stephenson J (1973) Journal of Endocrinology 58, 345 Steelman S L & Pohley F M (1953) Endocrinology 53, 604 Taymor M L & Miyata J (1970) Acta endocrinologica (Copenhagen) Suppl. 142, p 324 Van Weeman B K & Schuurs A H W M (1971) FEBS Letters 15, 232 Watson J (1971) Journal of Endocrinology 50, 711 Woodhead J S, Addison G M & Hales C N (1974) British Medical Bulletin 30, 44 Wide L, Nillius S J, Gemzell C & Roos P (1973) Acta endocrinologica (Copenhagen) Suppl. 174; p 41
73
(Estrogen Modulation of Gonadotrophin Release CEstrogens play an important role in the control of the menstrual cycle in the human female. Although the understanding of cestrogen feedback circuits is far from complete, it appears that at different phases of the cycle aestrogens have different effects upon gonadotrophin release. During the early and mid-follicular phase, the increasing cestradiol levels suppress follicle stimulating hormone (FSH) and luteinizing hormone (LH) release. At midcycle, however, a positive release of gonadotrophins is seen in response to a higher but falling cestradiol level. These feedback effects are thought to be mediated via control of the secretion of hypothalamic gonadotrophin hormone-releasing hormone (LHRH), as well as by direct action on the pituitary. To study the effects of cestrogen upon gonadotrophin release in vivo in the human female we have been using two approaches; first testing the pituitary's response to a standard dose of synthetic LH-RH in differing steroid environments, and second, testing the effects of injected cestrogens upon spontaneous LH and FSH release in various clinical states. Effect of wstradiol upon pituitary responsiveness to LH-RH: Throughout the normal menstrual cycle cestrogen levels change dramatically. The effects of synthetic LH-RH given at different stages during the cycle have previously been studied (Nillius & Wide 1972, Thomas et al. 1972, Yen et al. 1972) but in most instances the tests were not performed during the same cycle. Four normal women received 100 /.kg LH-RH intravenously on Days 4, 9, 14 and 22 of the same cycle. The LH response was always greater on Day 22 than on Days 4 or 9, but in each case the largest effect was seen on Day 14. The FSH responses showed no such differences throughout the cycle and certainly no marked increase at midcycle. The cestradiol levels on the various test days did not correlate with LH or FSH responses, although the highest cestradiol values were seen on Day 14, when the greatest LH response was observed (Shaw et al. 1974). To investigate whether alterations in cestradiol levels would bring about changes in response to LH-RH, normal women were treated with cestradiol benzoate in doses of 0.5 mg to 2.5 mg intramuscularly during the follicular phase of their cycles and the effects of LH-RH before and 48 hours after the aestrogen injection were measured. Control subjects, receiving no oestrogen, showed no significant change in response, whereas those who received the cestradiol in most cases showed
Section ofEndocrinology
Symposium: Gonadotrophins
half-lives in circulation, and hence different potencies in the two assays. These specific methods of FSH and LH measurement soon demonstrated that the two hormones may fluctuate independently. This was well illustrated by the careful studies of Fukushima et al. (1964) on samples of urine obtained from normal subjects throughout the menstrual cycle. The patterns reported are remarkably similar to those that have subsequently been described using radioimmunoassays. A disadvantage of the biological methods is their lack of sensitivity which limited their application to extracted samples of urine and it was not until the more sensitive radioimmunoassays became available that results could be obtained on small samples of serum. The development of specific radioimmunoassays for the gonadotrophins proved difficult, however, because of the considerable cross-reactivity between antisera, probably related to the fact that the glycoprotein hormones contain a common a-subunit. Satisfactory assays were first developed for LH while those for FSH encountered greater problems. Antisera to FSH had usually to be produced to relatively crude preparations but this practice was justified because the antibodies produced might be specific to the :-subunit or nonspecific to the a-subunit even with highly purified preparations. Furthermore, for the milligram quantities needed for immunization, it was necessary to use crude preparations because of the scarcity of the more purified material. However, recent refinements in immunization techniques such as the use of intracutaneous injections at multiple sites in rabbits have enabled Ross et al. (1971) to produce antisera of high titre to the subunits of HCG using only 20-50 ,ug of im.munogen for the primary immunization. Butt et al. (1974) employed this multiple site technique with preparations of pituitary FSH with some interesting results. In the first experiments relatively crude immunogen was injected intradermally into 30-40 sites on the backs of the rabbits. The total dose was 100 ,ug and a second set of injections containing only 20 ,ug was given after about twelve weeks, by which time the antibody titre had risen and then fallen. Within two weeks high titres were obtained, but the antisera were nonspecific in all animals and showed complete cross-reactions with each of the other glycoproteins. When the experiment was repeated with highly purified FSH (CPDS/6, Butt & Lynch 1972) 3 out of 4 animals produced antisera of high titre, affinity and specificity. Cross-reactivity with LH or with thyrotrophin (TSH) could be accounted for by the known contamination of these preparations with small amounts of FSH. The same technique was successfully applied to LH and other hormones.
Dr W R Butt, Dr S S Lynch, Mr W Robinson, Mr J Williams (Birmingham and Midland Hospital for Women, Birmingham, Bl 4HL) and Dr D R London (Queen Elizabeth Hospital, Edgbaston, Birmingham, B15 2TH) Assays of Gonadotrophins and Their Applications The early biological assays for the gonadotrophins were very elaborate as their specificity depended upon histological end-points in hypophysectomized animals. Quantitative estimates were, therefore, difficult to obtain, and for routine use simpler blat nonspecific assays were employed. The best known of these used the uterine weight of immature mice as the endpoint. This was a satisfactorily precise assay and was widely applied. Unfortunately, besides being nonspecific it was affected by synergistic action between follicle-stimulating hormone (FSH) and luteinizing hormone (LH) (Butt et al. 1964). It was not until quantitative specific assays had been developed that the uncertainty as to whether there were one or two gonadotrophins in human urine could be resolved. A convenient and specific assay for FSH known as the ovarian augmentation assay, was described by Steelman & Pohley (1953). Immature rats were given injections of FSH or unknown in the presence of a large excess of human chorionic gonadotrophin (HCG): the effect of any LH in the preparation of FSH was therefore minimized. Brown (1955) described a similar assay in which mice were used which tended to be more sensitive but was less precise than the assay using rats. Greep et al. (1941) described a specific assay by which LH measurements were correlated with an increase in weight of the ventral prostate in hypophysectomized rats. The use of hyposectomized animals was avoided in a method described by Parlow (1961), in which LH acutely depletes ascorbic acid in the luteinized ovaries of pseudo-pregnant immature rats (OAAD assay). These two assays for LH have been widely used but they do not always give the same estimate of potency. This is because the ventral prostate assay measures the effect of the hormone over a period of four days, whereas the OAAD assay measures the effect over a short period of two to four hours. Thus differences in the structure of the hormone as may occur between different species or between pituitary, serum and urinary preparations may result in different biological
71
72 Proc. roy. Soc. Med. Volume 68 February 1975
4
These antisera have made available specific and sensitive assays for both FSH and LH. Differences in potency estimates which have been noted between laboratories (Taymor & Miyata 1970) may be caused by the use of different antigens for labelling, by different antisera, and by differences between standards. Now that a Supra Regional Assay Service has been set up for routine radioimmunoassays in hospitals in this country it is important that these differences are minimized as far as possible by making use of common reagents and standards. Standards which have been recommended for serum assays are the pituitary preparations MRC 69/104 for FSH and MRC 68/40 for LH, distributed by the Division of Biological Standards, Hampstead. The first is a crude but stable preparation containing both FSH and LH: each ampoule is assumed to contain 10 units FSH. MRC 68/40 is a purer preparation and is assumed to contain 77 units LH per ampoule. Normal ranges obtained in this laboratory using these standards are given in Table 1. Some of the more common clinical applications of radioimmunoassays for FSH and LH are as follows:
response in females depends on the stage of the cycle: the increase in LH at 30 and 60 minutes ranges from twofold to fivefold in the early follicular phase, and up to thirtyfold in the luteal phase. (2) (Estrogen stimulation test (Shaw 1975): 1 mg aestradiol benzoate is given by intramuscular injection. Females with normal hypothalamic-pituitary function show a release of LH within the next 72 hours. (3) Clomiphene: given orally for 5 to 7 days (maximum dose 150 mg) stimulates the release of gonadotrophins. Serum samples are taken at the beginning and at intervals until the end of treatment with clomiphene. Single determinations are particularly useful in the diagnosis of conditions in which there is excessive production of gonadotrophins, but when this is normal or subnormal dynamic test procedures are required. Patients of both sexes with gonadal failure may have grossly elevated levels of gonadotrophins. FSH tends to be raised more often than LH as it is in normal postmenopausal women (Wide et al. 1973) and so is the more useful of the two determinations in this respect. In dynamic tests, however, LH increases in response to the various stimuli more than FSH and the measurement of both hormones may give clinically useful information. Future developments in the assays of gonadotrophins may be summarized into three categories. Firstly there may be improvements in the immunoassays themselves. Specificity and possibly sensitivity appear to be no longer very great problems, but in all radioimmunoassays the labelling of the antigen is expensive and has to be done at regular intervals. Alternative procedures for labelling with radioiodine have been suggested (Marchalonis 1969, Redshaw & Lynch 1974) and labelling of the antibody instead of antigen is another approach (Woodhead et al. 1974). In addition, methods have been described in which nonradioactive end-points have been used; thus Van Weeman & Schuurs (1971) coupled peroxidase to HCG and demonstrated that in principle an enzyme-substrate reaction served as an alternative end-point to gamma counting. Such a system should provide a stable method of labelling the antigen and provide a method suitable for automation, but at present it appears that the problem of achieving adequate sensitivity has still to be overcome. Secondly radioimmunoassay is applicable to the individual subunits of the gonadotrophins and it may be that these will be shown to be of some clinical importance. Thus some interesting observations on the production of subunits by the placenta and by tumours of the trophoblast have been reported by Franchimont et al. (1972). Thirdly immunoassays do not measure the biological activity of the hormone and there are many examples in which modifications of the structure
Single estimation: Single estimations of FSH are usually of more value than single estimations of LH. A high result may indicate primary gonadal failure in conditions covering infertility and eunuchoidism in the male and primary or secondary amenorrhoea in the female. An estimation of FSH in the assessment of postmenopausal patients with suspected pituitary disease or following hypophysectomy may also be of value. A low or normal result with hypogonadism suggests hypothalamic or pituitary dysfunction and a subsequent dynamic test procedure may then be of value. Dynamic test procedures: Changes in LH are usually greater than in FSH in these procedures. Examples are: (1) Gonadotrophin-releasing hormone (LH-RH) test (London et al. 1974): 100 ,ug is given by intravenous injection and serum samples are taken at 0, 30 and 60 minutes. The Table I Gonadotrophins in the serum of normal subjects Normal basal ranges
FSH u/l (MRC 69/104)
Children: Age I year to puberty * < 2.5 Adult males
1-7
Females:
Reproductive years Midcycle peak Postmenopausal
1-5 4-15 > 15
* FSH rises towards puberty before LH
LIH
u/t (MRC 68/40) < 2.5
2.5-20
2.5-15 20-100 > 40
5
Section of Endocrinology
lead to losses of biological but not of immunological activities (Butt 1969). The complete hormonal activity of the gonadotrophin can only be indicated by bioassays in vivo, but methods which depend upon binding to a receptor site will at least measure one facet of the biological activity, and such assays using testicular or ovarian receptors have been described (Catt et al. 1972, Reichert & Bhalla 1974). Alternatively those steroids produced in vitro under the influence of the gonadotrophin may be used as the end-point (Watson 1971, Shirley & Stephenson 1973). There is also the interesting approach of Rees et al. (1973) using a system for LH similar to that originally described for ACTH (Chayen et al. 1972). The effect of LH on ascorbic acid in the luteinized ovary used in the assay of Parlow (1961) has been adapted to this in vitro technique. Reducing activity in the luteinized ovaries treated with LH is estimated by microdensitometry after staining with Prussian blue. Although this assay is tedious in comparison with radioimmunoassay it is even more sensitive and thus gives the possibility of measuring levels which are less than normal.
Dr R W Shaw (Department of Clinical Endocrinology, The Women's Hospital, Sparkhill, Birmingham, Bll 4HL)
REFERENCES Brown P S (1955) Journal ofEndocrinology 13, 59 Butt W R (1969) Acta endocrinologica (Copenhagen) Suppl. 142, p 13 Butt W R, Cunningham F J & Hartree A S (1964) Proceedings of the Royal Society ofMedicine 57, 107 Butt W R & Lynch S S (1972) In: Hormones Glycoproteiques Hypophysaires. Ed. M Jutisz. Inserm, Paris; p 83 Butt W R, Lynch S S & Shirley A (1974) Journal of Endocrinology 61, xliii Catt K J, Dufau M L & Tsuruhara T (I 972) Journal ofClinical Endocrinology and Metabolismn 34, 123 Chayen J, Loveridge N & Daly J R (1972) Clinical Endocrinology 1, 219 Franchimont P, Gaspart U, Reuter A & Heynen G (1972) Clinical Endocrinology 1, 315 Fukushima M, Stevens V C, Gantt C L & Vorys V (1964) Journal of Clinical Endocrinology and Metabolisnm 24, 205 Greep R 0, Van Dyke H B & Chow B F (1941) Proceedings of the Society of Experimental Biology and Medicine 46, 644 London D R, Butt W R, Rudd B T, Holder G, Robinson W, Duignan N & Logan-Edwards R (1975) Proceedings of the Royal Society of Medicine 68, 75-76 Marchalonis J J (1969) Biochemical Journal 113, 299 Parlow A F (1961) In: Human Pituitary Gonadotropins. Ed. A Albert. Thomas, New York; p 300 Redshaw M R & Lynch S S (1974) Journal of Endocrinology 60, 527 Rees L H, Holdaway I M, Kramer R, McNeilly A S & Chard T (1973) Nature (London) 244, 232 Reichert L E jr & Bhalla V K (1974) Endocrinology 94,483 Ross G T, Vaitukaitis J L & Robbins J B (1971) In: Structureactivity Relationships of Protein and Polypeptide Hormones. Ed. M Margoulies & F C Greenwood. Excerpta Medica Foundation, Amsterdam; Part 1, p 153 Shaw R W (1975) Proceedings of the Royal Society of Medicine 68, 73-75 Shirley A & Stephenson J (1973) Journal of Endocrinology 58, 345 Steelman S L & Pohley F M (1953) Endocrinology 53, 604 Taymor M L & Miyata J (1970) Acta endocrinologica (Copenhagen) Suppl. 142, p 324 Van Weeman B K & Schuurs A H W M (1971) FEBS Letters 15, 232 Watson J (1971) Journal of Endocrinology 50, 711 Woodhead J S, Addison G M & Hales C N (1974) British Medical Bulletin 30, 44 Wide L, Nillius S J, Gemzell C & Roos P (1973) Acta endocrinologica (Copenhagen) Suppl. 174; p 41
73
(Estrogen Modulation of Gonadotrophin Release CEstrogens play an important role in the control of the menstrual cycle in the human female. Although the understanding of cestrogen feedback circuits is far from complete, it appears that at different phases of the cycle aestrogens have different effects upon gonadotrophin release. During the early and mid-follicular phase, the increasing cestradiol levels suppress follicle stimulating hormone (FSH) and luteinizing hormone (LH) release. At midcycle, however, a positive release of gonadotrophins is seen in response to a higher but falling cestradiol level. These feedback effects are thought to be mediated via control of the secretion of hypothalamic gonadotrophin hormone-releasing hormone (LHRH), as well as by direct action on the pituitary. To study the effects of cestrogen upon gonadotrophin release in vivo in the human female we have been using two approaches; first testing the pituitary's response to a standard dose of synthetic LH-RH in differing steroid environments, and second, testing the effects of injected cestrogens upon spontaneous LH and FSH release in various clinical states. Effect of wstradiol upon pituitary responsiveness to LH-RH: Throughout the normal menstrual cycle cestrogen levels change dramatically. The effects of synthetic LH-RH given at different stages during the cycle have previously been studied (Nillius & Wide 1972, Thomas et al. 1972, Yen et al. 1972) but in most instances the tests were not performed during the same cycle. Four normal women received 100 /.kg LH-RH intravenously on Days 4, 9, 14 and 22 of the same cycle. The LH response was always greater on Day 22 than on Days 4 or 9, but in each case the largest effect was seen on Day 14. The FSH responses showed no such differences throughout the cycle and certainly no marked increase at midcycle. The cestradiol levels on the various test days did not correlate with LH or FSH responses, although the highest cestradiol values were seen on Day 14, when the greatest LH response was observed (Shaw et al. 1974). To investigate whether alterations in cestradiol levels would bring about changes in response to LH-RH, normal women were treated with cestradiol benzoate in doses of 0.5 mg to 2.5 mg intramuscularly during the follicular phase of their cycles and the effects of LH-RH before and 48 hours after the aestrogen injection were measured. Control subjects, receiving no oestrogen, showed no significant change in response, whereas those who received the cestradiol in most cases showed
