Control of progesterone and inhibin secretion luteal phase in the macaque
during the
K. B. Smith and H. M. Fraser MRC Reproductive Biology Unit, Centre for Reproductive Biology, 37 Chalmers Street, Edinburgh eh3 9ew received
21
May
1990
ABSTRACT
We
the temporal relationship between concentrations of progesterone and immuno\x=req-\ reactive inhibin after treatment with an LHRH
investigated
serum
antagonist ([N-Ac-D-Nal(2)1,D-pCl-Phe2,D-Trp3,D\x=req-\ hArg(Et2)6,D-Ala10]-LHRH), during the mid-luteal phase in the macaque. Further, in an attempt to obtain a model of transitory suppression of luteal function, the effect of
treatment with the LHRH
antagonist for 1, 2 or 3 days during the mid-luteal phase on serum concentrations of progesterone and immunoreactive inhibin was compared. Differences in
the pattern of decline of the two hormones were observed. Progesterone concentrations fell by 6 h after antagonist administration while inhibin was not significantly suppressed until 48 h. Treatment with three injections of LHRH antagonist caused a sustained suppression of luteal function as shown by low serum concentrations of progesterone and inhibin. Recovery of progesterone and inhibin secretion was observed in two out of six macaques treated with two injections of antagonist and in three out of six treated with a single
INTRODUCTION
Inhibin,
a
thought
to
heterodimeric glycoprotein, was originally be a product of the developing follicle with the function of suppressing follicle-stimulating hormone (FSH) secretion. However, it has become apparent that a major source of immunoreactive inhibin secretion is the corpus luteum. Using a heter¬ ologous radioimmunoassay it has been demonstrated that serum concentrations of immunoreactive inhibin are highest during the luteal phase in women (McLachlan, Robertson, Healy et al. 1987; Buckler, McLachlan, McLachlan et al. 1988), stumptailed macaques (Fraser, Robertson & de Kretser, 1989) and marmoset monkeys (Smith, Lunn & Fraser, 1990). Furthermore, the mRNA for the subunit has been identified in RNA isolated from luteal tissue of these primates (Davis, Krozowski, McLachlan & Burger,
injection. Therefore, with the regimens of LHRH antagonist which we employed this approach was not conducive to obtaining a reliable transitory suppression of luteal function. To elucidate further the gonadotrophin control of inhibin, six macaques were treated with three injections of the LHRH antagonist to induce a permanent suppression ofluteal function but received concomitantly either human chorionic gonadotrophin (hCG) or human FSH daily for 5 days (n= 3 per group). FSH failed to prevent the antagonist-induced fall in progesterone and inhibin while hCG
treatment com-
pletely reversed the inhibitory effects of the LHRH antagonist. These results give further support to the concept that the secretion of inhibin, like progester-
integrated with the LH control of the corpus luteum. The slower decline in inhibin after LHRH antagonist suggests that the gonadotrophic stimulus to the corpus luteum results in a more prolonged stimulus for inhibin than for progesterone secretion, or that inhibin has a longer metabolic clearance rate. Journal ofEndocrinology (1991) 128, 107\p=n-\113 one, is
1987; Hillier, Wickings, Saunders et al. 1989; Basseti, Winters, Keeping & Zeleznik, 1990). In addition, evi¬
dence for both and ß subunit expression using insitu hybridization has recently been demonstrated in the primate corpus luteum (Schwall, Mason, Wilcox etal. 1990). A convenient approach for investigations of the gonadotrophic control of luteal function is by treat¬
luteinizing hormone-releasing hormone (LHRH) antagonists to block pituitary secretion of luteinizing hormone (LH) and FSH. Various studies involving the use of LHRH antagonists have estab¬ lished that progesterone secretion by the corpus luteum of women and non-human primates is depen¬ dent on pituitary LH secretion (Fraser, Baird, McRae ment with
al. 1985; Collins, Sopelak, Williams & Hodgen, 1986; Fraser, Abbott, Laird et al. 1986; Mais, Kazer, Cetel et al. 1986; Fraser, Nestor & Vickery, 1987;
et
Hodges, Green, Cottingham et al. 1988). Treatment with LHRH antagonist for 3 days starting during the mid-luteal phase also caused permanent suppression
of serum inhibin concentrations for the remainder of the cycle (Fraser et al. 1989; McLachlan, Cohen, Vale et al. 1989). The ability of the corpus luteum to recover from transitory suppression of gonadotrophin has also been investigated in monkeys with induced hypothalamic lesions. In the rhesus monkey, where endogenous gonadotrophin was abolished using radiofrequency lesions and restored by chronic pul¬ satile infusion of LHRH, withdrawal of gonado¬ trophic support for 3 days resulted in luteolysis or recovery of luteal function depending on the age of the corpus luteum (Hutchison & Zeleznik, 1985). The aims ofthe present study were threefold. First, to examine the temporal relationship between the serum concentrations of inhibin and progesterone after LHRH antagonist treatment. Secondly, since treat¬ ment with LHRH antagonist for 3 days during the midluteal phase to induce continued luteal suppression can be used to test the luteal response to exogenous fac¬ tors, we also determined whether suppression of inhibin secretion induced by LHRH antagonist in the macaque could be prevented by replacement of gonadotrophin with either human chorionic gonado¬ trophin (hCG) or FSH. Thirdly, since the results of Hutchison & Zeleznik (1985) indicated that a with¬ drawal of LHRH for 3 days during the mid-luteal phase could create a period of suppression of luteal function followed by recovery to normal for the remainder of the luteal phase, we investigated whether this could be achieved by reducing the administration of LHRH antagonist to 1 or 2 days. Induction of such a response could be used to identify the action of putative luteolytic agents, the effects of which might be obscured in the presence of endogenous gonado¬ trophins, but which could have a deleterious effect on luteal function when gonadotrophins were transitorily suppressed, resulting in abolition of the recovery phase. Such an approach might help to identify strat¬ egies for overcoming the ability of hCG to 'rescue' the corpus luteum and lead to improved methods of post-ovulatory fertility control. MATERIALS AND METHODS
Animals
Eighteen adult female macaques (Macaca arctoides) weighing 8-13 kg were used in these studies. Details of their management have been published previously (Fraser et al. 1986). All animals demonstrated regular menstrual cycles with normal luteal phases as deter¬ mined by hormonal estimations three times per week, fulfilling the criteria previously described (Fraser et al.
1986). When animals
were
used in
more
than
one
aspect of the study, intervals of at least 3 months were
elapse between treatments. Blood samples collected daily without anaesthesia (4 ml) late follicular and luteal phases. The the throughout LH of the was considered as day 0 of the surge day luteal phase if followed immediately by an increase in subsequent daily serum progesterone concentrations. Treatment with an LHRH antagonist To determine the ability of the corpus luteum to recover from various periods of suppression of gonadotrophin, macaques were injected with the LHRH antagonist allowed to
were
[N-Ac-D-Nal(2)'D-pCl-Phe2,D-Trp3,D-hArg(Et2)6,
D-Ala'°]-LHRH (Detirelix; Syntex,
Palo Alto, CA, U.S.A.) dissolved in 0-9% (w/v) NaCl/propylene glycol (1:1, v/v) and administered s.c. at a dose of 300 pg/kg, once daily for 1,2 or 3 days (n 6 per group), beginning on day 6-8 after the mid-cycle LH surge. Previous =
studies have shown that 300 pg/kg causes suppression of the pituitary-gonadal axis for 24 h in both female and male monkeys while lower doses result in partial steroid production during the 24-h recovery of period (Fraser et al. 1985, 1986; Adams, Bremner, Nestor et al. 1986). Daily samples were taken through¬ out the luteal phase with further samples collected at 0, 2,4,6,8 and 12 h from the time offirst administration of antagonist in 12 of the treated animals. Six macaques receiving vehicle alone and studied over the same timeperiod acted as controls. Blood samples were centri¬ fuged at 1000# for 30 min and stored at -20 °C until assayed for progesterone, LH and inhibin. Luteal function was considered to have recovered if serum progesterone concentrations rose to > 5 nmol/1 for 2 consecutive days after starting treatment.
gonadal
Gonadotrophin replacement To elucidate the gonadotrophic control of the corpus luteum
further, six macaques were treated with 300 µg
antagonist/kg once daily for 3 consecutive days beginning on day 6-8 of the luteal phase, i.e. a regimen shown previously to produce a permanent suppression of luteal progesterone and inhibin secretion (Fraser et al. 1989). In addition, they received concomitantly either hCG (Chorulon; Intervet, Cambridge, Cambs, U.K.) in incremental doses of 30, 60, 90, 180 and 360 IU i.m. for 5 days, or FSH (Metrodin: Serono Laboratories Ltd, Welwyn Garden City, Herts, U.K.) at 36IU/day for 5 days (n 3 per group). Blood samples were collected daily for 10 days after starting =
treatment.
Assays determined by use of a for progesterone together rapid radioimmunoassay with an LH radioimmunoassay to determine the day Occurrence of ovulation
was
of the
mid-cycle LH surge as described previously (Fraser et al. 1986). The LH assay utilized an ovine LH antiserum (provided by Dr G. Niswender), ovine
LH as tracer and NICHD-rhLH-RPI as standard. Detection limit for the assay was 40 pg/1 and interand intra-assay coefficients of variation were < 10%. The sensitivity of the progesterone assay was 0-7 nmol/1 and inter- and intra-assay coefficients of variation were 15% and 4% respectively. Inhibin concentrations were measured using
heterologous radioimmunoassay (Robertson, Hayward, Irby et al. 1988; Fraser et al. 1989; Fraser, Smith, Reddi & Lunn, 1990; Smith et al. 1990). This
Statistical
analysis Data were subjected to one and two factor analysis of variance following log transformation to reduce hetero¬ geneity of variance. Where significant differences were observed, data were further analysed using NewmanKeuls test, the level of significance being set at < 0-05.
RESULTS
a
rabbit antiserum (no. 1989) raised bovine 31 kDa inhibin at an initial dilution of 1 : 3000 and iodinated bovine 31 kDa inhibin as a tracer. The standard was prepared from partially purified human follicular fluid as a macaque inhibin preparation is not available (Reddi, Wickings, Hillier & Baird, 1989). This standard has a mean activity of 26 +31 U/l (mean ± s.d.) measured by an in-vitro sheep pituitary cell bioassay (Tsonis, McNeilly & Baird, 1986) calibrated against the activity of ovine rete testis fluid (IU/mg protein). The dose range used was 3-57-230 mU/tube and the detec¬ tion limit of the assay based on the 80% effective dose level was lOmU/tube. Maximum binding of tracer at the working dilution of the antibody ranged between 25 and 30%. Serum from female macaques gave a dose-response curve parallel to that of human serum, as shown by lack of significant differences in the slopes of the logit-log dose-response curves. The intra-assay coefficient of variation was 4% and between assay variance was 21 %. Initially, it was stated that inhibin subunits obtained after reductive alkylation as well as other inhibin-related proteins showed limited cross-reaction in this assay (Robertson et al. 1988). However, it has since been found that an subunit precursor (pro-aC) cross-reacts 288% in the assay. Pro-aC consists of the pro region of the a subunit, disulphide-linked to the mature subunit (Robertson, Giacometti, Foulds et al. 1989). In addition, supernatants containing an admix¬ ture of partially purified recombinant -inhibin subunits of 21 000-57 000 Mr have also been shown to displace binding (Schneyer, Mason, Burton et al. 1990). However, the presence of pro-aC and free subunit has not been demonstrated in the circulation or in other biological fluids in primates and therefore the biological significance of these findings has yet to be determined. We were unable to determine bioactive concentrations ofinhibin in these studies as previous experience showed that use of macaque samples in the sheep pituitary cell bioassay resulted in cell death due to the toxic effect of macaque serum in this system (C. G. Tsonis, R. Leask & H. M. Fraser, unpublished observations). assay utilizes
a
against purified
The short-term (24 h) response of progesterone and inhibin to the effects of LHRH antagonist treatment are shown in Fig. 1. Serum progesterone was signifi¬ cantly (P
during the
K. B. Smith and H. M. Fraser MRC Reproductive Biology Unit, Centre for Reproductive Biology, 37 Chalmers Street, Edinburgh eh3 9ew received
21
May
1990
ABSTRACT
We
the temporal relationship between concentrations of progesterone and immuno\x=req-\ reactive inhibin after treatment with an LHRH
investigated
serum
antagonist ([N-Ac-D-Nal(2)1,D-pCl-Phe2,D-Trp3,D\x=req-\ hArg(Et2)6,D-Ala10]-LHRH), during the mid-luteal phase in the macaque. Further, in an attempt to obtain a model of transitory suppression of luteal function, the effect of
treatment with the LHRH
antagonist for 1, 2 or 3 days during the mid-luteal phase on serum concentrations of progesterone and immunoreactive inhibin was compared. Differences in
the pattern of decline of the two hormones were observed. Progesterone concentrations fell by 6 h after antagonist administration while inhibin was not significantly suppressed until 48 h. Treatment with three injections of LHRH antagonist caused a sustained suppression of luteal function as shown by low serum concentrations of progesterone and inhibin. Recovery of progesterone and inhibin secretion was observed in two out of six macaques treated with two injections of antagonist and in three out of six treated with a single
INTRODUCTION
Inhibin,
a
thought
to
heterodimeric glycoprotein, was originally be a product of the developing follicle with the function of suppressing follicle-stimulating hormone (FSH) secretion. However, it has become apparent that a major source of immunoreactive inhibin secretion is the corpus luteum. Using a heter¬ ologous radioimmunoassay it has been demonstrated that serum concentrations of immunoreactive inhibin are highest during the luteal phase in women (McLachlan, Robertson, Healy et al. 1987; Buckler, McLachlan, McLachlan et al. 1988), stumptailed macaques (Fraser, Robertson & de Kretser, 1989) and marmoset monkeys (Smith, Lunn & Fraser, 1990). Furthermore, the mRNA for the subunit has been identified in RNA isolated from luteal tissue of these primates (Davis, Krozowski, McLachlan & Burger,
injection. Therefore, with the regimens of LHRH antagonist which we employed this approach was not conducive to obtaining a reliable transitory suppression of luteal function. To elucidate further the gonadotrophin control of inhibin, six macaques were treated with three injections of the LHRH antagonist to induce a permanent suppression ofluteal function but received concomitantly either human chorionic gonadotrophin (hCG) or human FSH daily for 5 days (n= 3 per group). FSH failed to prevent the antagonist-induced fall in progesterone and inhibin while hCG
treatment com-
pletely reversed the inhibitory effects of the LHRH antagonist. These results give further support to the concept that the secretion of inhibin, like progester-
integrated with the LH control of the corpus luteum. The slower decline in inhibin after LHRH antagonist suggests that the gonadotrophic stimulus to the corpus luteum results in a more prolonged stimulus for inhibin than for progesterone secretion, or that inhibin has a longer metabolic clearance rate. Journal ofEndocrinology (1991) 128, 107\p=n-\113 one, is
1987; Hillier, Wickings, Saunders et al. 1989; Basseti, Winters, Keeping & Zeleznik, 1990). In addition, evi¬
dence for both and ß subunit expression using insitu hybridization has recently been demonstrated in the primate corpus luteum (Schwall, Mason, Wilcox etal. 1990). A convenient approach for investigations of the gonadotrophic control of luteal function is by treat¬
luteinizing hormone-releasing hormone (LHRH) antagonists to block pituitary secretion of luteinizing hormone (LH) and FSH. Various studies involving the use of LHRH antagonists have estab¬ lished that progesterone secretion by the corpus luteum of women and non-human primates is depen¬ dent on pituitary LH secretion (Fraser, Baird, McRae ment with
al. 1985; Collins, Sopelak, Williams & Hodgen, 1986; Fraser, Abbott, Laird et al. 1986; Mais, Kazer, Cetel et al. 1986; Fraser, Nestor & Vickery, 1987;
et
Hodges, Green, Cottingham et al. 1988). Treatment with LHRH antagonist for 3 days starting during the mid-luteal phase also caused permanent suppression
of serum inhibin concentrations for the remainder of the cycle (Fraser et al. 1989; McLachlan, Cohen, Vale et al. 1989). The ability of the corpus luteum to recover from transitory suppression of gonadotrophin has also been investigated in monkeys with induced hypothalamic lesions. In the rhesus monkey, where endogenous gonadotrophin was abolished using radiofrequency lesions and restored by chronic pul¬ satile infusion of LHRH, withdrawal of gonado¬ trophic support for 3 days resulted in luteolysis or recovery of luteal function depending on the age of the corpus luteum (Hutchison & Zeleznik, 1985). The aims ofthe present study were threefold. First, to examine the temporal relationship between the serum concentrations of inhibin and progesterone after LHRH antagonist treatment. Secondly, since treat¬ ment with LHRH antagonist for 3 days during the midluteal phase to induce continued luteal suppression can be used to test the luteal response to exogenous fac¬ tors, we also determined whether suppression of inhibin secretion induced by LHRH antagonist in the macaque could be prevented by replacement of gonadotrophin with either human chorionic gonado¬ trophin (hCG) or FSH. Thirdly, since the results of Hutchison & Zeleznik (1985) indicated that a with¬ drawal of LHRH for 3 days during the mid-luteal phase could create a period of suppression of luteal function followed by recovery to normal for the remainder of the luteal phase, we investigated whether this could be achieved by reducing the administration of LHRH antagonist to 1 or 2 days. Induction of such a response could be used to identify the action of putative luteolytic agents, the effects of which might be obscured in the presence of endogenous gonado¬ trophins, but which could have a deleterious effect on luteal function when gonadotrophins were transitorily suppressed, resulting in abolition of the recovery phase. Such an approach might help to identify strat¬ egies for overcoming the ability of hCG to 'rescue' the corpus luteum and lead to improved methods of post-ovulatory fertility control. MATERIALS AND METHODS
Animals
Eighteen adult female macaques (Macaca arctoides) weighing 8-13 kg were used in these studies. Details of their management have been published previously (Fraser et al. 1986). All animals demonstrated regular menstrual cycles with normal luteal phases as deter¬ mined by hormonal estimations three times per week, fulfilling the criteria previously described (Fraser et al.
1986). When animals
were
used in
more
than
one
aspect of the study, intervals of at least 3 months were
elapse between treatments. Blood samples collected daily without anaesthesia (4 ml) late follicular and luteal phases. The the throughout LH of the was considered as day 0 of the surge day luteal phase if followed immediately by an increase in subsequent daily serum progesterone concentrations. Treatment with an LHRH antagonist To determine the ability of the corpus luteum to recover from various periods of suppression of gonadotrophin, macaques were injected with the LHRH antagonist allowed to
were
[N-Ac-D-Nal(2)'D-pCl-Phe2,D-Trp3,D-hArg(Et2)6,
D-Ala'°]-LHRH (Detirelix; Syntex,
Palo Alto, CA, U.S.A.) dissolved in 0-9% (w/v) NaCl/propylene glycol (1:1, v/v) and administered s.c. at a dose of 300 pg/kg, once daily for 1,2 or 3 days (n 6 per group), beginning on day 6-8 after the mid-cycle LH surge. Previous =
studies have shown that 300 pg/kg causes suppression of the pituitary-gonadal axis for 24 h in both female and male monkeys while lower doses result in partial steroid production during the 24-h recovery of period (Fraser et al. 1985, 1986; Adams, Bremner, Nestor et al. 1986). Daily samples were taken through¬ out the luteal phase with further samples collected at 0, 2,4,6,8 and 12 h from the time offirst administration of antagonist in 12 of the treated animals. Six macaques receiving vehicle alone and studied over the same timeperiod acted as controls. Blood samples were centri¬ fuged at 1000# for 30 min and stored at -20 °C until assayed for progesterone, LH and inhibin. Luteal function was considered to have recovered if serum progesterone concentrations rose to > 5 nmol/1 for 2 consecutive days after starting treatment.
gonadal
Gonadotrophin replacement To elucidate the gonadotrophic control of the corpus luteum
further, six macaques were treated with 300 µg
antagonist/kg once daily for 3 consecutive days beginning on day 6-8 of the luteal phase, i.e. a regimen shown previously to produce a permanent suppression of luteal progesterone and inhibin secretion (Fraser et al. 1989). In addition, they received concomitantly either hCG (Chorulon; Intervet, Cambridge, Cambs, U.K.) in incremental doses of 30, 60, 90, 180 and 360 IU i.m. for 5 days, or FSH (Metrodin: Serono Laboratories Ltd, Welwyn Garden City, Herts, U.K.) at 36IU/day for 5 days (n 3 per group). Blood samples were collected daily for 10 days after starting =
treatment.
Assays determined by use of a for progesterone together rapid radioimmunoassay with an LH radioimmunoassay to determine the day Occurrence of ovulation
was
of the
mid-cycle LH surge as described previously (Fraser et al. 1986). The LH assay utilized an ovine LH antiserum (provided by Dr G. Niswender), ovine
LH as tracer and NICHD-rhLH-RPI as standard. Detection limit for the assay was 40 pg/1 and interand intra-assay coefficients of variation were < 10%. The sensitivity of the progesterone assay was 0-7 nmol/1 and inter- and intra-assay coefficients of variation were 15% and 4% respectively. Inhibin concentrations were measured using
heterologous radioimmunoassay (Robertson, Hayward, Irby et al. 1988; Fraser et al. 1989; Fraser, Smith, Reddi & Lunn, 1990; Smith et al. 1990). This
Statistical
analysis Data were subjected to one and two factor analysis of variance following log transformation to reduce hetero¬ geneity of variance. Where significant differences were observed, data were further analysed using NewmanKeuls test, the level of significance being set at < 0-05.
RESULTS
a
rabbit antiserum (no. 1989) raised bovine 31 kDa inhibin at an initial dilution of 1 : 3000 and iodinated bovine 31 kDa inhibin as a tracer. The standard was prepared from partially purified human follicular fluid as a macaque inhibin preparation is not available (Reddi, Wickings, Hillier & Baird, 1989). This standard has a mean activity of 26 +31 U/l (mean ± s.d.) measured by an in-vitro sheep pituitary cell bioassay (Tsonis, McNeilly & Baird, 1986) calibrated against the activity of ovine rete testis fluid (IU/mg protein). The dose range used was 3-57-230 mU/tube and the detec¬ tion limit of the assay based on the 80% effective dose level was lOmU/tube. Maximum binding of tracer at the working dilution of the antibody ranged between 25 and 30%. Serum from female macaques gave a dose-response curve parallel to that of human serum, as shown by lack of significant differences in the slopes of the logit-log dose-response curves. The intra-assay coefficient of variation was 4% and between assay variance was 21 %. Initially, it was stated that inhibin subunits obtained after reductive alkylation as well as other inhibin-related proteins showed limited cross-reaction in this assay (Robertson et al. 1988). However, it has since been found that an subunit precursor (pro-aC) cross-reacts 288% in the assay. Pro-aC consists of the pro region of the a subunit, disulphide-linked to the mature subunit (Robertson, Giacometti, Foulds et al. 1989). In addition, supernatants containing an admix¬ ture of partially purified recombinant -inhibin subunits of 21 000-57 000 Mr have also been shown to displace binding (Schneyer, Mason, Burton et al. 1990). However, the presence of pro-aC and free subunit has not been demonstrated in the circulation or in other biological fluids in primates and therefore the biological significance of these findings has yet to be determined. We were unable to determine bioactive concentrations ofinhibin in these studies as previous experience showed that use of macaque samples in the sheep pituitary cell bioassay resulted in cell death due to the toxic effect of macaque serum in this system (C. G. Tsonis, R. Leask & H. M. Fraser, unpublished observations). assay utilizes
a
against purified
The short-term (24 h) response of progesterone and inhibin to the effects of LHRH antagonist treatment are shown in Fig. 1. Serum progesterone was signifi¬ cantly (P
