- Email: [email protected]
Circulating levels of pregnanolone isomers during the third trimester of human pregnancy
Journal of Steroid Biochemistry & Molecular Biology 105 (2007) 166–175
Circulating levels of pregnanolone isomers during the third trimester of human pregnancy Martin Hill a,∗ , David Cibula b , Helena Havl´ıkov´a a , Lyudmila Kancheva a , Tom´asˇ Fait b , Radmila Kancheva a , Anton´ın Paˇr´ızek b , Luboslav St´arka a b
a Institute of Endocrinology, N´ arodn´ı tˇr´ıda 8, CZ 116 94 Prague 1, Czech Republic Department of Gynecology and Obstetrics, 1st Faculty of Medicine, Charles University, Apolin´arˇsk´a 18, CZ 128 51 Prague 2, Czech Republic
Received 27 September 2006; accepted 26 October 2006
Abstract This study addresses the question of whether changes in the biosynthesis and metabolism of neuroactive pregnanolone isomers (PIs) might participate in the timing of parturition in humans. The time profiles of unconjugated allopregnanolone (3␣-hydroxy-5␣-pregnan-20-one, P3␣5␣), pregnanolone (3␣-hydroxy-5-pregnan-20-one, P3␣5), isopregnanolone (3-hydroxy-5␣-pregnan-20-one, P35␣) and epipregnanolone (3-hydroxy-5-pregnan-20-one, P35), pregnenolone, their polar conjugates, progesterone, 5␣-dihydroprogesterone (P5␣), and 5-dihydroprogesterone (P5) were monitored in the plasma of 30 healthy women during the third trimester of pregnancy, at 1-week intervals from the 30th week of gestation using GC–MS. Changes in the steroid levels were evaluated by two-way ANOVA with gestational age and subject as independent factors. The mean concentrations of free PIs ranged from 2 to 50 nmol/L, while the mean levels of their polar conjugates were 40–100× higher. The ratio of 5␣-PIs to progesterone significantly but inconspicuously culminated in the 35th week. The decelerating biosynthesis of free 5-PIs from the 31st week and their escalating sulfation was found from the 30th week. The changes were particularly evident in the second most abundant PI pregnanolone that may, like the allopregnanolone, sustain the pregnancy via attenuation of hypothalamic GABAA -receptors and prevent uterine contractility via binding to nuclear pregnane X receptor. © 2007 Elsevier Ltd. All rights reserved. Keywords: Pregnanolone isomers; Neurosteroids; Steroid sulfates; Human pregnancy; GC–MS
1. Introduction The initiation of parturition represents a complex system involving various factors and multiple interconnected positive and negative feedback loops [1]. The failure of one link usually does not result in a failure of the entire system. Several more or less species-specific mechanisms have been suggested to explain the primary impulse for parturition in various mammals. Some of the mechanisms involve significantly decreasing placental synthesis of progesterone, with at ∗
Corresponding author. E-mail addresses: [email protected] (M. Hill), [email protected] (D. Cibula), [email protected] (H. Havl´ıkov´a), [email protected] (L. Kancheva), [email protected] (T. Fait), [email protected] (R. Kancheva), [email protected] (A. Paˇr´ızek), [email protected] (L. St´arka). 0960-0760/$ – see front matter © 2007 Elsevier Ltd. All rights reserved. doi:10.1016/j.jsbmb.2006.10.010
the same time increasing estradiol production before the onset of parturition [1,2]. In contrast to rodents and ruminants, progesterone levels in human maternal blood do not markedly change around parturition while estradiol levels escalate up to labor, as in other mammals [3]. Nevertheless, the initiation of delivery in humans is connected rather to a changed expression of specific isoforms of estradiol- and progesterone receptors [4]. It was also reported that cortisol, the circulating levels of which increase in the late third trimester [5], may act as an endogenous inhibitor of progesterone action in the regulation of 15-hydroxyprostaglandin dehydrogenase expression at term. This process may be mediated by progesterone and glucocorticoid receptors [6]. Another mechanism suggested for humans is related to the excessive production of placental CRH near term [7]. An alternative mechanism for initiation of parturition that was recently demonstrated on mice is based on increasing
M. Hill et al. / Journal of Steroid Biochemistry & Molecular Biology 105 (2007) 166–175
secretion of fetal major lung surfactant protein into amniotic fluid that activates amniotic fluid macrophages. The macrophages migrate into uterus where they induce an inflammatory response resulting in elevated uterine contractility [8]. Despite a number of mechanism suggested, our knowledge regarding the primary impulse for the human parturition, however, is still deficient. The role of neuroactive reduced progesterone metabolites—and particularly the most abundant of them, pregnanolone isomers (PIs) in the timing of human parturition is still unclear, despite studies reporting their major effect in other mammals [9,10]. PIs are mostly effective as neuromodulators influencing the permeability of ion channels in the central nervous system (CNS) and the periphery [11–14] and some also bind on nuclear progesterone receptors [15] or pregnane X receptors [16].
167
PIs, and particularly their polar conjugates, are synthesized in large quantities in the human fetoplacental unit from progesterone (Fig. 1) [17,18]. As has been documented for 5␣-PIs, the steroids are metabolized to respective 3-oxo-derivatives (5␣/-dihydroprogesterones), which are then transported to the maternal compartment via the placenta—where they are again converted into PIs [18]. The levels of all PIs in pregnancy are strikingly higher when compared to the situation in non-pregnant women. Although several studies have brought information to light regarding the profiles of unconjugated PIs in maternal and umbilical blood [19–25], only a small number have included data regarding the respective polar conjugates [19,20,26]. To the best of the authors’ knowledge, this is the first study to evaluate the time profiles of PIs including their polar conjugates in plasma in the third trimester of pregnancy. Particular attention has been paid to the balance between parturition-
Fig. 1. Biosynthesis of pregnanolone isomers.
168
M. Hill et al. / Journal of Steroid Biochemistry & Molecular Biology 105 (2007) 166–175
sustaining unconjugated 3␣-PIs and conjugated PIs exerting the opposite effect.
The derivatization agent Sylon BFT (bis(trimethylsilyl)trifluoroacetamide 99% and trimethylchlorosilane) was purchased from Supelco (Bellefonte, PA, USA).
2. Materials and methods
2.4. Instruments
Changes in all free and conjugated isomers of pregnanolone, i.e. allopregnanolone (3␣-hydroxy-5␣-pregnan20-one, P3␣5␣), pregnanolone (3␣-hydroxy-5-pregnan-20one, P3␣5), isopregnanolone (3-hydroxy-5␣-pregnan-20one, P35␣) and epipregnanolone (3-hydroxy-5-pregnan20-one, P35) were evaluated. In addition, the levels of pregnenolone, progesterone, 5␣-dihydroprogesterone (P5␣), and 5-dihydroprogesterone (P5) in the maternal plasma during the third trimester of human pregnancy were also assessed.
The GC–MS system was supplied by Shimadzu (Kyoto, Japan). The system consisted of a GC 17A gas chromatograph equipped with automatic flow control, AOC-20 autosampler and for the MS a QP 5050A quadrupole electron-impact detector with a fixed electron voltage of 70 eV. The capillary column with a medium polarity Zebron ZB-50 (50% phenyl–50% methylpolysiloxane) from Phenomenex (St. Torrance, CA, USA) was used for the analysis. The length of the column was 15 m, the internal diameter was 0.25 mm and the film thickness was 15 m.
2.1. Subjects A total of 30 pregnant volunteers entered the study protocol according to the inclusion and exclusion criteria. The volunteers were enrolled into the study at a single institution between January 2004 and September 2005, according to the following inclusion criteria: age between 18 and 30 years, physiological pregnancy, cephalic presentation of the fetus and agreement with the study protocol. The exclusion criteria were as follows: manifestation of preeclampsia, chronic medication, anemia (glycated hemoglobin ≤ 10 g/dL), any chronic disease or medication that might influence steroid levels or delivery, premature rupture of the membranes, labor induction or operative delivery in a previous pregnancy. The blood count was checked twice, in the 34th and 38th weeks. The local Ethical Committees of the Institute of Endocrinology and the General Faculty Hospital (both in Prague, Czech Republic) approved the protocol for the study. 2.2. Sample collection After signing informed consent the patients underwent blood sampling from the cubital vein. Blood samples were collected at 1-week intervals from the 30th week of gestation up to labor or to the 42nd completed week. All blood samples were taken at the same time of day (9.00 a.m. to 12.00 noon, at least 2 h after waking), under standard conditions (elimination of stress factors, after 15 min of resting). Cooled plastic tubes containing 100 L of 5% EDTA and 50 pL of aprotinin (Antilysin from Spofa, Prague, Czech Republic) were used for blood sampling. Plasma was obtained after centrifugation for 5 min at 2000 × g at 0 ◦ C. The plasma samples were stored at −20 ◦ C until analyzed. 2.3. Steroids and chemicals The steroids were purchased from Steraloids (Wilton, NH, USA). The solvents for the extraction and HPLC were of analytical grade, sourced from Merck (Darmstadt, Germany).
2.5. Preparation of the plasma samples for GC–MS free steroids analysis Frozen samples were thawed and 1 mL of the sample was spiked with 17␣-estradiol as an internal standard to attain a concentration of 1 g/mL. The spiked sample was extracted with 3 mL of diethyl ether. The water phase was kept frozen in a mixture of solid carbon dioxide and ethanol, and the organic phase was decanted into glass tubes and evaporated to dryness. The dry organic phase residue was used for the determination of free steroids. The dry residue was partitioned between 1 mL of 80% methanol with water and 1 mL of n-pentane to eliminate the lipids and sterols. The n-pentane phase was discarded, while the methanol/water phase containing steroids for analysis was evaporated in a vacuum centrifuge. The samples were prepared twice for further processing by two different derivatization techniques. The first was used for preparation of steroids with hydroxygroups modified to trimethylsilyl (TMS) derivatives and with untouched oxo-groups. The second technique produced derivatives with hydroxy-groups modified as in the former case but, in addition, with oxo-groups modified by methoxylamine (MOX-TMS derivatives). The first derivatization technique was used for the quantification of pregnenolone and PIs. The second was applied for the measurement of progesterone, P5␣ and P5. 2.6. Sample preparation for the GC–MS analysis of steroid polar conjugates The frozen water phase in glass tubes was thawed and mixed with 1 mL of methanol. The tubes were centrifuged and the 1 mL aliquot of the supernatant was transferred into a glass tube and evaporated in a vacuum centrifuge. The steroid sulfates were hydrolyzed using a method described elsewhere [27]. The hydrolyzed sample was evaporated in a vacuum centrifuge; the dry residue was spiked with 17␣-estradiol as an internal standard to attain a concentration of 1 g/mL and further processed in the same way as the free steroids.
M. Hill et al. / Journal of Steroid Biochemistry & Molecular Biology 105 (2007) 166–175
169
and 9.0 kPa/min up to 310 ◦ C and 70 kPa, followed by a 3 min delay. The overall time taken for the analysis was 15.2 min.
2.7. Derivatization The trimethylsilyl derivatives of the steroids were prepared using the modified method of Hill et al. [20], with some of the modifications reported recently [28]. Briefly, Sylon B (99% BTSFA + 1% TMCS) (50 L) was added to the dry residues from plasma, mixed briefly and heated at 90 ◦ C for 45 min. The derivatization agent was evaporated under a stream of nitrogen. The dry residue was rinsed down with isooctane (50 L) and the mixture was evaporated again. Finally, steroid derivatives were dissolved into 20 L isooctane, and 4 L portions were injected into the GC–MS system. The MOX-TMS derivatives were prepared as follows: 50 L of 2% solution of methoxylamine-hydrochloride in pyridine was added to the dry residues from plasma, mixed briefly and heated at 60 ◦ C for 2 h. The mixture was then evaporated under a stream of nitrogen, before further trimethylsilyl derivatization proceeded as described above.
2.9. Retention times and effective masses used for the determination of steroids To exploit the sample, the individual samples were applied three times in independent courses, in each case employing a part of the steroids under investigation. The choices of the steroids measured within the individual courses, as well as the effective masses used for the measurement, were optimized to attain maximum sensitivity at sufficient selectivity. The types of gradients, effective masses for determination and quantification, the order numbers of injection, and retention times for individual steroids and are shown in Table 1. 2.10. Statistical analysis of the data To evaluate changes in the steroid levels and steroid ratios, a two-way ANOVA model was used with the week of gestation as the first factor and the subject as the second factor. Multiple testing was handled by Bonferroni multiple comparisons to evaluate the differences between individual groups. Given the non-Gaussian distribution and non-constant variance in most of the steroids, the original data underwent a power transformation to attain symmetry and homoscedasticity in the data as well as in the residuals. The group mean values and their 95% confidence intervals calculated in the transformed data were re-transformed to the original scale for graphical demonstration. Statistical computations were performed using Statgraphics Plus v5.1 statistical software (Manugistics, Rockville, MA, USA).
2.8. Temperature and pressure gradients for the GC–MS analysis of trimethylsilyl-derivatives For GC–MS analysis of TMS-derivatives, the temperature and pressure gradients used were as follows (GS): 1 min high pressure injection at 120 ◦ C and 100 kPa followed by a pressure release to 30 kPa and a rapid linear gradient of 40 ◦ C/min and 8.5 kPa/min up to 200 ◦ C and 49.3 kPa, then a slow linear gradient of 2.9 ◦ C/min and 0.5 kPa/min up to 220 ◦ C and 52.7 kPa, a medium linear gradient of 20 ◦ C/min and 8 kPa/min up to 265 ◦ C and 70 kPa, and a rapid linear gradient of 40 ◦ C/min and 10.0 kPa/min up to 310 ◦ C and 80.7 kPa, followed by a 2 min delay. The overall time taken for the analysis was 15.3 min. For the analysis of TMS-MOX-derivatives, the following temperature and pressure gradient was used (GMS): 1 min high pressure injection at 120 ◦ C and 100 kPa followed by a pressure release to 30 kPa and a rapid linear gradient of 40 ◦ C/min and 8.5 kPa/min up to 220 ◦ C and 51.0 kPa, then a slow linear gradient of 2.9 ◦ C/min and 0.5 kPa/min up to 240 ◦ C and 54.5 kPa, and a rapid linear gradient of 40 ◦ C/min
3. Results 3.1. Identification and quantification of the steroids The steroids separated well from each other and from the background. As demonstrated in Table 1, the selectivity and sensitivity were sufficient for quantification of all of the inves-
Table 1 Analytical criteria of the method for the multi-component quantification of neuroactive pregnanolone isomers and related steroids No.
Steroid
Form
Gradient/ derivatization
1 2 3 4 5 6 7 8 9
Pregnenolone Progesterone 3␣-Dihydroprogesterone (P3␣) 3-Dihydroprogesterone (P3) Allopregnanolone (P3␣5␣) Isopregnanolone (P35␣) Pregnanolone (P3␣5) Epipregnanolone (P3␣5␣) 17␣-Estradiol (internal standard)
F, Ca F F F F, C F, C F, C F, C –
Sb MS MS MS S S S S S, (MS)
a b
Injection
Retention time (min)
Effective mass (m/z)
Sensitivity (pg) (mean ± S.E.M., n = 5)
3 1 1 1 1 1 1 1 1–4
11.592 11.250, 11.392 10.975, 11.008 10.396, 10.475 10.758 11.563 10.950 10.550 9.863
298, 388 100, 341, 372 288, 343 288, 343 285, 300, 375 285, 300, 375 285, 300, 375 285, 300, 375 285, 416
3.67 2.09 1.26 3.43 1.79 1.93 2.76 2.03 1.42
F: free steroid; C: conjugated steroid. S: trimethylsilyl derivatives; MS: methoxyamine-trimethylsilyl derivatives.
± ± ± ± ± ± ± ± ±
0.35 0.27 0.11 0.61 0.26 0.24 0.39 0.29 0.17
170
M. Hill et al. / Journal of Steroid Biochemistry & Molecular Biology 105 (2007) 166–175
tigated steroids, as well as of reproducibility (the inter-assay coefficient of variance did not exceed 10% for any steroid).
3.3. Changes in the ratios of 3α- and 3β-pregnanolone isomers
3.2. Changes in unconjugated pregnane steroids
The ratios of 3␣- to 3-pregnanolone isomers did not change significantly during the third trimester (data not shown).
Pronounced differences were observed between the 5␣and 5-pregnane steroids in their profiles during the third trimester. The P5␣ and 5␣-PIs (Fig. 2A–C) showed a significant increase between the 30th and 36th weeks while the respective 5-pregnane steroids displayed either no significant change (P5) (Fig. 2D) or a decreasing trend in 5-PIs from the 36th week of gestation (Fig. 2E and F).
3.4. Changes in the ratios of 5α- and 5β-dihydroprogesterone to progesterone The ratio of P5␣ to progesterone showed a trend of borderline significance with a maximum in the 35th week (Fig. 3A),
Fig. 2. Profiles of 5␣-dihydroprogesterone (P5␣, panel A), allopregnanolone (P3␣5␣, panel B), isopregnanolone (P35␣, panel C), 5-dihydroprogesterone (P5, panel D), pregnanolone (P3␣5, panel E) and epipregnanolone (P35, panel F) in maternal plasma during the third trimester of pregnancy. F represent the Fisher’s fractiles for the factors week and subjects in ANOVA model and p denote their statistical significances. The asterisks symbolize significant differences between the 30th week of gestation and another stages, as found by Bonferroni multiple comparisons (p < 0.05).
M. Hill et al. / Journal of Steroid Biochemistry & Molecular Biology 105 (2007) 166–175
171
Fig. 3. Profiles of the ratios of 5␣-dihydroprogesterone to progesterone (P5␣/Prog, panel A), and 5-dihydroprogesterone to progesterone (P5/Prog, panel B). The drawings and symbols are as in Fig. 2.
while the respective ratio for the P5 significantly decreased from the 31st week (Fig. 3B). 3.5. Changes in conjugated pregnanolone isomers Three of the conjugated PIs showed significant increases from the 30th week to the 37th week, followed by a plateau (Fig. 4A, B and D). The only exception was P3␣5, which increased markedly up to the 39th week (Fig. 4C). 3.6. Changes in the ratios of conjugated pregnanolone isomers to free steroids Both ratios of conjugated 5␣-PIs to the respective free steroids gradually increased from the 30th to the 37th weeks, and then showed a plateau up to the 40th week (Fig. 5A and B). The ratios for 5-PIs, i.e. P35␣ (Fig. 5C) and P35 (Fig. 5D), showed a manifest increase from the 30th or 31st week.
4. Discussion The aim of the study was to evaluate the changes of neuroactive steroids in the third trimester of pregnancy, and to hypothesize their role in the timing of parturition. The supposed changes that may involve steroid oxido-reductive metabolism and the balance between steroid sulfation and hydrolysis of the sulfates were evaluated using a multicomponent analysis of steroid profiles by GC–MS. The levels of free PIs obtained in this study are comparable with the previously published data [20,26]. The concentrations of conjugates obtained here are however higher. This
discrepancy appears to be due to the less efficient method used for the hydrolysis of steroid polar conjugates in the earlier studies. The method used in this study [27] was highly efficient, with negligible losses. In general, PIs with a hydroxy group in the 3␣-position (3␣-PIs) attenuate neuronal activity via stimulation of the receptors of type A ␥-aminobutyric acid (GABAA -r) [11]. PIs hydroxylated in the 3-position (3-PIs) compete with the 3␣-PIs for the active sites on the receptors [29]. Sulfation counteracts the effect of 3␣-PIs and further amplifies the antagonistic effect of 3-PIs, forming products that negatively modulate GABAA -r on the binding sites, which are different from the sites specific to the unconjugated 3␣-PIs. For example, the neuromodulating efficiency of isopregnanolone sulfate is comparable to the neuromodulating effectiveness of allopregnanolone [13] but the steroid conjugate exerts the opposite effect. Therefore, it may be of importance that the concentrations of neuroinhibiting free 3␣-PIs are markedly lower in the maternal plasma before labor when compared to polar conjugates of PIs [19,20,26]. Besides the effect on GABAA -r, PIs and particularly the unconjugated 5-isomers also block calcium channels of type T in the rat peripheral neurons, mediating the perception of pain [14]. The polar conjugates of 5␣-PIs are also known to be positive modulators of N-methyl-d-aspartate receptors (NMDA-r), which are present both in the CNS and the periphery [30]. Conjugated 5-PIs exert the opposite effect [12,30–32]. The role of the most abundant PI allopregnanolone in the onset of parturition has been reported in rats where a positive feedback loop in oxytocin production resulting in a rapid delivery forms just before labor. A decrease in the levels of allopregnanolone triggers the production of oxytocin [9,10].
172
M. Hill et al. / Journal of Steroid Biochemistry & Molecular Biology 105 (2007) 166–175
Fig. 4. Profiles of the polar conjugates of allopregnanolone (P3␣5␣, panel A), isopregnanolone (P35␣, panel B), pregnanolone (P3␣5, panel C) and epipregnanolone (P35, panel D) in maternal plasma during the third trimester of pregnancy. The drawings and symbols are as in Fig. 2.
This substance brings the GABAA -r from a neurosteroidsensitive mode towards a condition in which the receptors are not sensitive [33], via a shift in the balance between the activities of endogenous Ser/Thr phosphatase and protein kinase C [33]. Considering the aforementioned findings, it is likely that changing biosynthesis of PIs could influence the onset of labor. Hence, we analyzed changes of product to
precursors ratios associated with the respective steroid reductases, oxidoreductases, sulfatases and sulfotransferases. The ratios of 3␣-PI and 3-PIs to their 3-oxo-precursors (5␣or 5-dihydroprogesterone) did not change during the third trimester (data not shown), negating the supposed oxidoreductive shift between 3␣-PIs and 3-PIs, which could influence the function of GABAA -r.
M. Hill et al. / Journal of Steroid Biochemistry & Molecular Biology 105 (2007) 166–175
173
Fig. 5. Profiles of the ratios of the polar conjugates of allopregnanolone (P3␣5␣, panel A), isopregnanolone (P35␣, panel B), pregnanolone (P3␣5, panel C) and epipregnanolone (P35, panel D) to their respective free steroids in maternal plasma during the third trimester of pregnancy. The drawings and symbols are as in Fig. 2.
174
M. Hill et al. / Journal of Steroid Biochemistry & Molecular Biology 105 (2007) 166–175
On the other hand, pronounced differences were found between the profiles of 5␣- (Fig. 2A–C) and 5-PIs (Fig. 2D–F). The ratios of P5␣ and P5 to progesterone indicated that the metabolism of progesterone to P5␣ inconspicuously culminated in the 35th week (Fig. 3A). Alternatively, the conversion of progesterone to P5 significantly declined from the 31st week (Fig. 3B). The latter effect accords with the recent study of Gilbert Evans et al. [34]. Excessive concentrations of neuroactivating conjugated PIs exerting an opposite effect on GABAA -r compared to the free 3␣-PIs [13] were found in maternal circulation (Fig. 4). The results clearly demonstrate increasing sulfotransferase activity in all PIs, as documented by significantly increasing levels of conjugated PIs (Fig. 4) and ratios of conjugates to the respective free steroids (Fig. 5). The ratios of conjugates to free PIs showed a gradual increase in the 5␣-isomers up to the 37th week (Fig. 5A and B) but a pronounced, accelerating increase in both 5-analogs (Fig. 5C and D). The most prominent change was apparent in the second most abundant PIs—the neuroinhibiting P3␣5 (Fig. 5C). The aforementioned findings showed that unconjugated PIs including the neuroinhibiting 3␣-PIs were more rapidly metabolized with oncoming parturition, particularly in the case of P3␣5. This steroid is chemically identical with the short-term intravenous anesthetic eltanolone that takes an effect via positive modulation of GABAA -r. Moreover, biosynthesis of the 5-pregnane steroids decelerated from the 31st week (Fig. 3B). Considering these findings, as well as the previously demonstrated pronounced decrease of the ratio of conjugated PIs to free PIs after delivery [26], one can speculate as to whether the decreasing levels of P3␣5 might influence the sustaining of parturition. In terms of the effects of circulating PIs on neuronal activity in the CNS, the authors are aware of the limitations in transport across the blood–brain barrier (BBB), particularly in the case of steroid conjugates. On the other hand, the lowpolar free PIs may pass across the BBB relatively easily. The chances of overcoming the blood–brain barrier generally increase with the decreasing polarity of the substance [35]. The peripheral effects of PIs in pregnancy should be considered, respecting the presence of receptors sensitive to pregnane steroids in tissues connected with parturition. For instance, 5-PIs may take effect through a pregnane X receptor-mediated mechanism regulating uterine contractility [16]. It is likely that the decrease in activity of 5-reductase [34,36] and the increasing sulfation of PIs during the third trimester are associated with a reduced ability to sustain the pregnancy. In conclusion, the identification of accelerating sulfation in pregnancy-sustaining steroids during the third trimester is the most fundamental outcome of this study. This metabolic step at the least eliminates the neuroinhibiting effect of 3␣PIs. Another important result is an endorsement for the previously suggested declining expression of 5-reductase from the 31st week of gestation. The enzyme participates in
the formation of almost 40% of pregnancy- sustaining pregnanolone isomers. It is likely that the contribution of the PIs to the initiation of human parturition is based in part on the descending expression of 5-reductase from the 31st week. 5-Reductase shows similar neuroinhibiting efficiency on GABAA -r to the slightly more abundant 3␣-isomer allopregnanolone, but even higher activity on the nuclear pregnane X receptors preventing uterine contractility [16]. In terms of explicit changes in the circulating levels of PIs and a plausible explanation of the respective mechanisms, a significant role for PIs may be expected in the timing of human parturition. Acknowledgements This study was supported by Grant Agency of the Czech Republic grant no. 303/04/1086. The excellent technical assistance of Mrs. Ivona Kr´alov´a and Mrs. Marta Vel´ıkov´a is gratefully acknowledged. References [1] P.W. Nathanielsz, Comparative studies on the initiation of labor, Eur. J. Obstet. Gynecol. Reprod. Biol. 78 (2) (1998) 127–132. [2] W.X. Wu, X.H. Ma, T. Coksaygan, K. Chakrabarty, V. Collins, J. Rose, P.W. Nathanielsz, Prostaglandin mediates premature delivery in pregnant sheep induced by estradiol at 121 days of gestational age, Endocrinology 145 (3) (2004) 1444–1452. [3] R.S. Mathur, S. Landgrebe, H.O. Williamson, Progesterone, 17hydroxyprogesterone, estradiol, and estriol in late pregnancy and labor, Am. J. Obstet. Gynecol. 136 (1) (1980) 25–27. [4] S. Mesiano, Myometrial progesterone responsiveness and the control of human parturition, J. Soc. Gynecol. Investig. 11 (4) (2004) 193– 202. [5] W. Jeske, P. Soszynski, W. Rogozinski, E. Lukaszewicz, W. Latoszewska, H. Snochowska, Plasma GHRH, CRH, ACTH, betaendorphin, human placental lactogen, GH and cortisol concentrations at the third trimester of pregnancy, Acta Endocrinol. (Copenh.) 120 (6) (1989) 785–789. [6] F.A. Patel, J.W. Funder, J.R. Challis, Mechanism of cortisol/ progesterone antagonism in the regulation of 15-hydroxyprostaglandin dehydrogenase activity and messenger ribonucleic acid levels in human chorion and placental trophoblast cells at term, J. Clin. Endocrinol. Metab. 88 (6) (2003) 2922–2933. [7] M. McLean, R. Smith, Corticotrophin-releasing hormone and human parturition, Reproduction 121 (4) (2001) 493–501. [8] J.C. Condon, P. Jeyasuria, J.M. Faust, C.R. Mendelson, Surfactant protein secreted by the maturing mouse fetal lung acts as a hormone that signals the initiation of parturition, Proc. Natl. Acad. Sci. U.S.A. 101 (14) (2004) 4978–4983. [9] A.B. Brussaard, J. Wossink, J.C. Lodder, K.S. Kits, Progesteronemetabolite prevents protein kinase C-dependent modulation of gamma-aminobutyric acid type A receptors in oxytocin neurons, Proc. Natl. Acad. Sci. U.S.A. 97 (7) (2000) 3625–3630. [10] A.B. Brussaard, K.S. Kits, R.E. Baker, W.P. Willems, J.W. LeytingVermeulen, P. Voorn, A.B. Smit, R.J. Bicknell, A.E. Herbison, Plasticity in fast synaptic inhibition of adult oxytocin neurons caused by switch in GABA(A) receptor subunit expression, Neuron 19 (5) (1997) 1103–1114. [11] M.D. Majewska, Steroid regulation of the GABAA receptor: ligand binding, chloride transport and behaviour, Ciba Found Symp. 153 (1990) 83–97 (discussion 97–106).
M. Hill et al. / Journal of Steroid Biochemistry & Molecular Biology 105 (2007) 166–175 [12] M. Park-Chung, F.S. Wu, R.H. Purdy, A.A. Malayev, T.T. Gibbs, D.H. Farb, Distinct sites for inverse modulation of N-methyl-daspartate receptors by sulfated steroids, Mol. Pharmacol. 52 (6) (1997) 1113–1123. [13] M. Park-Chung, A. Malayev, R.H. Purdy, T.T. Gibbs, D.H. Farb, Sulfated and unsulfated steroids modulate gamma-aminobutyric acid A receptor function through distinct sites, Brain Res. 830 (1) (1999) 72–87. [14] S.M. Todorovic, S. Pathirathna, B.C. Brimelow, M.M. Jagodic, S.H. Ko, X. Jiang, K.R. Nilsson, C.F. Zorumski, D.F. Covey, V. Jevtovic-Todorovic, 5Beta-reduced neuroactive steroids are novel voltage-dependent blockers of T-type Ca2+ channels in rat sensory neurons in vitro and potent peripheral analgesics in vivo, Mol. Pharmacol. 66 (5) (2004) 1223–1235. [15] C.D. Putnam, D.W. Brann, R.C. Kolbeck, V.B. Mahesh, Inhibition of uterine contractility by progesterone and progesterone metabolites: mediation by progesterone and gamma amino butyric acid A receptor systems, Biol. Reprod. 45 (2) (1991) 266–272. [16] B.F. Mitchell, J.M. Mitchell, J. Chowdhury, M. Tougas, S.M. Engelen, N. Senff, I. Heijnen, J.T. Moore, B. Goodwin, S. Wong, S.T. Davidge, Metabolites of progesterone and the pregnane X receptor: a novel pathway regulating uterine contractility in pregnancy? Am. J. Obstet. Gynecol. 192 (4) (2005) 1304–1313 (discussion 1303–1305). [17] L. Milewich, N.F. Gant, B.E. Schwarz, G.T. Chen, P.C. MacDonald, 5 alpha-reductase activity in human placenta, Am. J. Obstet. Gynecol. 133 (6) (1979) 611–617. [18] R.A. Dombroski, M.L. Casey, P.C. MacDonald, 5-Alpha-dihydroprogesterone formation in human placenta from 5alpha-pregnan3beta/alpha-ol-20-ones and 5-pregnan-3beta-yl-20-one sulfate, J. Steroid Biochem. Mol. Biol. 63 (1–3) (1997) 155–163. [19] H. Mickan, J. Zander, Pregnanolones, pregnenolone and progesterone in the human feto-placental circulation at term of pregnancy, J. Steroid Biochem. 11 (4) (1979) 1461–1466. [20] M. Hill, A. Parizek, M. Bicikova, H. Havlikova, J. Klak, T. Fait, D. Cibula, R. Hampl, A. Cegan, J. Sulcova, L. Starka, Neuroactive steroids, their precursors, and polar conjugates during parturition and postpartum in maternal and umbilical blood. 1. Identification and simultaneous determination of pregnanolone isomers, J. Steroid Biochem. Mol. Biol. 75 (4–5) (2000) 237–244. [21] M. Hill, P. Popov, H. Havlikova, L. Kancheva, J. Vrbikova, R. Kancheva, V. Pouzar, I. Cerny, L. Starka, Altered profiles of serum neuroactive steroids in premenopausal women treated for alcohol addiction, Steroids 70 (8) (2005) 515–524. [22] A.R. Genazzani, F. Petraglia, F. Bernardi, E. Casarosa, C. Salvestroni, A. Tonetti, R.E. Nappi, S. Luisi, M. Palumbo, R.H. Purdy, M. Luisi, Circulating levels of allopregnanolone in humans: gender, age, and endocrine influences, J. Clin. Endocrinol. Metab. 83 (6) (1998) 2099–2103. [23] S. Luisi, F. Petraglia, C. Benedetto, R.E. Nappi, F. Bernardi, M. Fadalti, F.M. Reis, M. Luisi, A.R. Genazzani, Serum allopregnanolone levels in pregnant women: changes during pregnancy, at delivery, and in hypertensive patients, J. Clin. Endocrinol. Metab. 85 (7) (2000) 2429– 2433.
175
[24] B.E. Pearson Murphy, C.M. Allison, Determination of progesterone and some of its neuroactive ring A-reduced metabolites in human serum, J. Steroid Biochem. Mol. Biol. 74 (3) (2000) 137–142. [25] B.E. Pearson Murphy, S.I. Steinberg, F.Y. Hu, C.M. Allison, Neuroactive ring A-reduced metabolites of progesterone in human plasma during pregnancy: elevated levels of 5 alpha-dihydroprogesterone in depressed patients during the latter half of pregnancy, J. Clin. Endocrinol. Metab. 86 (12) (2001) 5981–5987. [26] M. Hill, M. Bicikova, A. Parizek, H. Havlikova, J. Klak, T. Fajt, M. Meloun, D. Cibula, A. Cegan, J. Sulcova, R. Hampl, L. Starka, Neuroactive steroids, their precursors and polar conjugates during parturition and postpartum in maternal blood. 2. Time profiles of pregnanolone isomers, J. Steroid Biochem. Mol. Biol. 78 (1) (2001) 51–57. [27] L. Dehennin, P. Lafarge, P. Dailly, D. Bailloux, J.P. Lafarge, Combined profile of androgen glucuro- and sulfoconjugates in postcompetition urine of sportsmen: a simple screening procedure using gas chromatography-mass spectrometry, J. Chromatogr. B: Biomed. Appl. 687 (1) (1996) 85–91. [28] H. Havlikova, M. Hill, L. Kancheva, J. Vrbikova, V. Pouzar, I. Cerny, R. Kancheva, L. Starka, Serum profiles of free and conjugated neuroactive pregnanolone isomers in non-pregnant women of fertile age, J. Clin. Endocrinol. Metab. (2006). [29] P. Lundgren, J. Stromberg, T. Backstrom, M. Wang, Allopregnanolonestimulated GABA-mediated chloride ion flux is inhibited by 3beta-hydroxy-5alpha-pregnan-20-one (isoallopregnanolone), Brain Res. 982 (1) (2003) 45–53. [30] C.E. Weaver, M.B. Land, R.H. Purdy, K.G. Richards, T.T. Gibbs, D.H. Farb, Geometry and charge determine pharmacological effects of steroids on N-methyl-d-aspartate receptor-induced Ca(2+) accumulation and cell death, J. Pharmacol. Exp. Ther. 293 (3) (2000) 747–754. [31] A. Malayev, T.T. Gibbs, D.H. Farb, Inhibition of the NMDA response by pregnenolone sulphate reveals subtype selective modulation of NMDA receptors by sulphated steroids, Br. J. Pharmacol. 135 (4) (2002) 901–909. [32] M.D. Wang, T. Backstrom, S. Landgren, The inhibitory effects of allopregnanolone and pregnanolone on the population spike, evoked in the rat hippocampal CA1 stratum pyramidale in vitro, can be blocked selectively by epiallopregnanolone, Acta Physiol. Scand. 169 (4) (2000) 333–341. [33] J.J. Koksma, R.E. van Kesteren, T.W. Rosahl, R. Zwart, A.B. Smit, H. Luddens, A.B. Brussaard, Oxytocin regulates neurosteroid modulation of GABA(A) receptors in supraoptic nucleus around parturition, J. Neurosci. 23 (3) (2003) 788–797. [34] S.E. Gilbert Evans, L.E. Ross, E.M. Sellers, R.H. Purdy, M.K. Romach, 3Alpha-reduced neuroactive steroids and their precursors during pregnancy and the postpartum period, Gynecol. Endocrinol. 21 (5) (2005) 268–279. [35] I. Oren, S.J. Fleishman, A. Kessel, N. Ben-Tal, Free diffusion of steroid hormones across biomembranes: a simplex search with implicit solvent model calculations, Biophys. J. 87 (2) (2004) 768–779. [36] P.M. Sheehan, G.E. Rice, E.K. Moses, S.P. Brennecke, 5Beta-dihydroprogesterone and steroid 5 beta-reductase decrease in association with human parturition at term, Mol. Hum. Reprod. 11 (7) (2005) 495–501.
Circulating levels of pregnanolone isomers during the third trimester of human pregnancy Martin Hill a,∗ , David Cibula b , Helena Havl´ıkov´a a , Lyudmila Kancheva a , Tom´asˇ Fait b , Radmila Kancheva a , Anton´ın Paˇr´ızek b , Luboslav St´arka a b
a Institute of Endocrinology, N´ arodn´ı tˇr´ıda 8, CZ 116 94 Prague 1, Czech Republic Department of Gynecology and Obstetrics, 1st Faculty of Medicine, Charles University, Apolin´arˇsk´a 18, CZ 128 51 Prague 2, Czech Republic
Received 27 September 2006; accepted 26 October 2006
Abstract This study addresses the question of whether changes in the biosynthesis and metabolism of neuroactive pregnanolone isomers (PIs) might participate in the timing of parturition in humans. The time profiles of unconjugated allopregnanolone (3␣-hydroxy-5␣-pregnan-20-one, P3␣5␣), pregnanolone (3␣-hydroxy-5-pregnan-20-one, P3␣5), isopregnanolone (3-hydroxy-5␣-pregnan-20-one, P35␣) and epipregnanolone (3-hydroxy-5-pregnan-20-one, P35), pregnenolone, their polar conjugates, progesterone, 5␣-dihydroprogesterone (P5␣), and 5-dihydroprogesterone (P5) were monitored in the plasma of 30 healthy women during the third trimester of pregnancy, at 1-week intervals from the 30th week of gestation using GC–MS. Changes in the steroid levels were evaluated by two-way ANOVA with gestational age and subject as independent factors. The mean concentrations of free PIs ranged from 2 to 50 nmol/L, while the mean levels of their polar conjugates were 40–100× higher. The ratio of 5␣-PIs to progesterone significantly but inconspicuously culminated in the 35th week. The decelerating biosynthesis of free 5-PIs from the 31st week and their escalating sulfation was found from the 30th week. The changes were particularly evident in the second most abundant PI pregnanolone that may, like the allopregnanolone, sustain the pregnancy via attenuation of hypothalamic GABAA -receptors and prevent uterine contractility via binding to nuclear pregnane X receptor. © 2007 Elsevier Ltd. All rights reserved. Keywords: Pregnanolone isomers; Neurosteroids; Steroid sulfates; Human pregnancy; GC–MS
1. Introduction The initiation of parturition represents a complex system involving various factors and multiple interconnected positive and negative feedback loops [1]. The failure of one link usually does not result in a failure of the entire system. Several more or less species-specific mechanisms have been suggested to explain the primary impulse for parturition in various mammals. Some of the mechanisms involve significantly decreasing placental synthesis of progesterone, with at ∗
Corresponding author. E-mail addresses: [email protected] (M. Hill), [email protected] (D. Cibula), [email protected] (H. Havl´ıkov´a), [email protected] (L. Kancheva), [email protected] (T. Fait), [email protected] (R. Kancheva), [email protected] (A. Paˇr´ızek), [email protected] (L. St´arka). 0960-0760/$ – see front matter © 2007 Elsevier Ltd. All rights reserved. doi:10.1016/j.jsbmb.2006.10.010
the same time increasing estradiol production before the onset of parturition [1,2]. In contrast to rodents and ruminants, progesterone levels in human maternal blood do not markedly change around parturition while estradiol levels escalate up to labor, as in other mammals [3]. Nevertheless, the initiation of delivery in humans is connected rather to a changed expression of specific isoforms of estradiol- and progesterone receptors [4]. It was also reported that cortisol, the circulating levels of which increase in the late third trimester [5], may act as an endogenous inhibitor of progesterone action in the regulation of 15-hydroxyprostaglandin dehydrogenase expression at term. This process may be mediated by progesterone and glucocorticoid receptors [6]. Another mechanism suggested for humans is related to the excessive production of placental CRH near term [7]. An alternative mechanism for initiation of parturition that was recently demonstrated on mice is based on increasing
M. Hill et al. / Journal of Steroid Biochemistry & Molecular Biology 105 (2007) 166–175
secretion of fetal major lung surfactant protein into amniotic fluid that activates amniotic fluid macrophages. The macrophages migrate into uterus where they induce an inflammatory response resulting in elevated uterine contractility [8]. Despite a number of mechanism suggested, our knowledge regarding the primary impulse for the human parturition, however, is still deficient. The role of neuroactive reduced progesterone metabolites—and particularly the most abundant of them, pregnanolone isomers (PIs) in the timing of human parturition is still unclear, despite studies reporting their major effect in other mammals [9,10]. PIs are mostly effective as neuromodulators influencing the permeability of ion channels in the central nervous system (CNS) and the periphery [11–14] and some also bind on nuclear progesterone receptors [15] or pregnane X receptors [16].
167
PIs, and particularly their polar conjugates, are synthesized in large quantities in the human fetoplacental unit from progesterone (Fig. 1) [17,18]. As has been documented for 5␣-PIs, the steroids are metabolized to respective 3-oxo-derivatives (5␣/-dihydroprogesterones), which are then transported to the maternal compartment via the placenta—where they are again converted into PIs [18]. The levels of all PIs in pregnancy are strikingly higher when compared to the situation in non-pregnant women. Although several studies have brought information to light regarding the profiles of unconjugated PIs in maternal and umbilical blood [19–25], only a small number have included data regarding the respective polar conjugates [19,20,26]. To the best of the authors’ knowledge, this is the first study to evaluate the time profiles of PIs including their polar conjugates in plasma in the third trimester of pregnancy. Particular attention has been paid to the balance between parturition-
Fig. 1. Biosynthesis of pregnanolone isomers.
168
M. Hill et al. / Journal of Steroid Biochemistry & Molecular Biology 105 (2007) 166–175
sustaining unconjugated 3␣-PIs and conjugated PIs exerting the opposite effect.
The derivatization agent Sylon BFT (bis(trimethylsilyl)trifluoroacetamide 99% and trimethylchlorosilane) was purchased from Supelco (Bellefonte, PA, USA).
2. Materials and methods
2.4. Instruments
Changes in all free and conjugated isomers of pregnanolone, i.e. allopregnanolone (3␣-hydroxy-5␣-pregnan20-one, P3␣5␣), pregnanolone (3␣-hydroxy-5-pregnan-20one, P3␣5), isopregnanolone (3-hydroxy-5␣-pregnan-20one, P35␣) and epipregnanolone (3-hydroxy-5-pregnan20-one, P35) were evaluated. In addition, the levels of pregnenolone, progesterone, 5␣-dihydroprogesterone (P5␣), and 5-dihydroprogesterone (P5) in the maternal plasma during the third trimester of human pregnancy were also assessed.
The GC–MS system was supplied by Shimadzu (Kyoto, Japan). The system consisted of a GC 17A gas chromatograph equipped with automatic flow control, AOC-20 autosampler and for the MS a QP 5050A quadrupole electron-impact detector with a fixed electron voltage of 70 eV. The capillary column with a medium polarity Zebron ZB-50 (50% phenyl–50% methylpolysiloxane) from Phenomenex (St. Torrance, CA, USA) was used for the analysis. The length of the column was 15 m, the internal diameter was 0.25 mm and the film thickness was 15 m.
2.1. Subjects A total of 30 pregnant volunteers entered the study protocol according to the inclusion and exclusion criteria. The volunteers were enrolled into the study at a single institution between January 2004 and September 2005, according to the following inclusion criteria: age between 18 and 30 years, physiological pregnancy, cephalic presentation of the fetus and agreement with the study protocol. The exclusion criteria were as follows: manifestation of preeclampsia, chronic medication, anemia (glycated hemoglobin ≤ 10 g/dL), any chronic disease or medication that might influence steroid levels or delivery, premature rupture of the membranes, labor induction or operative delivery in a previous pregnancy. The blood count was checked twice, in the 34th and 38th weeks. The local Ethical Committees of the Institute of Endocrinology and the General Faculty Hospital (both in Prague, Czech Republic) approved the protocol for the study. 2.2. Sample collection After signing informed consent the patients underwent blood sampling from the cubital vein. Blood samples were collected at 1-week intervals from the 30th week of gestation up to labor or to the 42nd completed week. All blood samples were taken at the same time of day (9.00 a.m. to 12.00 noon, at least 2 h after waking), under standard conditions (elimination of stress factors, after 15 min of resting). Cooled plastic tubes containing 100 L of 5% EDTA and 50 pL of aprotinin (Antilysin from Spofa, Prague, Czech Republic) were used for blood sampling. Plasma was obtained after centrifugation for 5 min at 2000 × g at 0 ◦ C. The plasma samples were stored at −20 ◦ C until analyzed. 2.3. Steroids and chemicals The steroids were purchased from Steraloids (Wilton, NH, USA). The solvents for the extraction and HPLC were of analytical grade, sourced from Merck (Darmstadt, Germany).
2.5. Preparation of the plasma samples for GC–MS free steroids analysis Frozen samples were thawed and 1 mL of the sample was spiked with 17␣-estradiol as an internal standard to attain a concentration of 1 g/mL. The spiked sample was extracted with 3 mL of diethyl ether. The water phase was kept frozen in a mixture of solid carbon dioxide and ethanol, and the organic phase was decanted into glass tubes and evaporated to dryness. The dry organic phase residue was used for the determination of free steroids. The dry residue was partitioned between 1 mL of 80% methanol with water and 1 mL of n-pentane to eliminate the lipids and sterols. The n-pentane phase was discarded, while the methanol/water phase containing steroids for analysis was evaporated in a vacuum centrifuge. The samples were prepared twice for further processing by two different derivatization techniques. The first was used for preparation of steroids with hydroxygroups modified to trimethylsilyl (TMS) derivatives and with untouched oxo-groups. The second technique produced derivatives with hydroxy-groups modified as in the former case but, in addition, with oxo-groups modified by methoxylamine (MOX-TMS derivatives). The first derivatization technique was used for the quantification of pregnenolone and PIs. The second was applied for the measurement of progesterone, P5␣ and P5. 2.6. Sample preparation for the GC–MS analysis of steroid polar conjugates The frozen water phase in glass tubes was thawed and mixed with 1 mL of methanol. The tubes were centrifuged and the 1 mL aliquot of the supernatant was transferred into a glass tube and evaporated in a vacuum centrifuge. The steroid sulfates were hydrolyzed using a method described elsewhere [27]. The hydrolyzed sample was evaporated in a vacuum centrifuge; the dry residue was spiked with 17␣-estradiol as an internal standard to attain a concentration of 1 g/mL and further processed in the same way as the free steroids.
M. Hill et al. / Journal of Steroid Biochemistry & Molecular Biology 105 (2007) 166–175
169
and 9.0 kPa/min up to 310 ◦ C and 70 kPa, followed by a 3 min delay. The overall time taken for the analysis was 15.2 min.
2.7. Derivatization The trimethylsilyl derivatives of the steroids were prepared using the modified method of Hill et al. [20], with some of the modifications reported recently [28]. Briefly, Sylon B (99% BTSFA + 1% TMCS) (50 L) was added to the dry residues from plasma, mixed briefly and heated at 90 ◦ C for 45 min. The derivatization agent was evaporated under a stream of nitrogen. The dry residue was rinsed down with isooctane (50 L) and the mixture was evaporated again. Finally, steroid derivatives were dissolved into 20 L isooctane, and 4 L portions were injected into the GC–MS system. The MOX-TMS derivatives were prepared as follows: 50 L of 2% solution of methoxylamine-hydrochloride in pyridine was added to the dry residues from plasma, mixed briefly and heated at 60 ◦ C for 2 h. The mixture was then evaporated under a stream of nitrogen, before further trimethylsilyl derivatization proceeded as described above.
2.9. Retention times and effective masses used for the determination of steroids To exploit the sample, the individual samples were applied three times in independent courses, in each case employing a part of the steroids under investigation. The choices of the steroids measured within the individual courses, as well as the effective masses used for the measurement, were optimized to attain maximum sensitivity at sufficient selectivity. The types of gradients, effective masses for determination and quantification, the order numbers of injection, and retention times for individual steroids and are shown in Table 1. 2.10. Statistical analysis of the data To evaluate changes in the steroid levels and steroid ratios, a two-way ANOVA model was used with the week of gestation as the first factor and the subject as the second factor. Multiple testing was handled by Bonferroni multiple comparisons to evaluate the differences between individual groups. Given the non-Gaussian distribution and non-constant variance in most of the steroids, the original data underwent a power transformation to attain symmetry and homoscedasticity in the data as well as in the residuals. The group mean values and their 95% confidence intervals calculated in the transformed data were re-transformed to the original scale for graphical demonstration. Statistical computations were performed using Statgraphics Plus v5.1 statistical software (Manugistics, Rockville, MA, USA).
2.8. Temperature and pressure gradients for the GC–MS analysis of trimethylsilyl-derivatives For GC–MS analysis of TMS-derivatives, the temperature and pressure gradients used were as follows (GS): 1 min high pressure injection at 120 ◦ C and 100 kPa followed by a pressure release to 30 kPa and a rapid linear gradient of 40 ◦ C/min and 8.5 kPa/min up to 200 ◦ C and 49.3 kPa, then a slow linear gradient of 2.9 ◦ C/min and 0.5 kPa/min up to 220 ◦ C and 52.7 kPa, a medium linear gradient of 20 ◦ C/min and 8 kPa/min up to 265 ◦ C and 70 kPa, and a rapid linear gradient of 40 ◦ C/min and 10.0 kPa/min up to 310 ◦ C and 80.7 kPa, followed by a 2 min delay. The overall time taken for the analysis was 15.3 min. For the analysis of TMS-MOX-derivatives, the following temperature and pressure gradient was used (GMS): 1 min high pressure injection at 120 ◦ C and 100 kPa followed by a pressure release to 30 kPa and a rapid linear gradient of 40 ◦ C/min and 8.5 kPa/min up to 220 ◦ C and 51.0 kPa, then a slow linear gradient of 2.9 ◦ C/min and 0.5 kPa/min up to 240 ◦ C and 54.5 kPa, and a rapid linear gradient of 40 ◦ C/min
3. Results 3.1. Identification and quantification of the steroids The steroids separated well from each other and from the background. As demonstrated in Table 1, the selectivity and sensitivity were sufficient for quantification of all of the inves-
Table 1 Analytical criteria of the method for the multi-component quantification of neuroactive pregnanolone isomers and related steroids No.
Steroid
Form
Gradient/ derivatization
1 2 3 4 5 6 7 8 9
Pregnenolone Progesterone 3␣-Dihydroprogesterone (P3␣) 3-Dihydroprogesterone (P3) Allopregnanolone (P3␣5␣) Isopregnanolone (P35␣) Pregnanolone (P3␣5) Epipregnanolone (P3␣5␣) 17␣-Estradiol (internal standard)
F, Ca F F F F, C F, C F, C F, C –
Sb MS MS MS S S S S S, (MS)
a b
Injection
Retention time (min)
Effective mass (m/z)
Sensitivity (pg) (mean ± S.E.M., n = 5)
3 1 1 1 1 1 1 1 1–4
11.592 11.250, 11.392 10.975, 11.008 10.396, 10.475 10.758 11.563 10.950 10.550 9.863
298, 388 100, 341, 372 288, 343 288, 343 285, 300, 375 285, 300, 375 285, 300, 375 285, 300, 375 285, 416
3.67 2.09 1.26 3.43 1.79 1.93 2.76 2.03 1.42
F: free steroid; C: conjugated steroid. S: trimethylsilyl derivatives; MS: methoxyamine-trimethylsilyl derivatives.
± ± ± ± ± ± ± ± ±
0.35 0.27 0.11 0.61 0.26 0.24 0.39 0.29 0.17
170
M. Hill et al. / Journal of Steroid Biochemistry & Molecular Biology 105 (2007) 166–175
tigated steroids, as well as of reproducibility (the inter-assay coefficient of variance did not exceed 10% for any steroid).
3.3. Changes in the ratios of 3α- and 3β-pregnanolone isomers
3.2. Changes in unconjugated pregnane steroids
The ratios of 3␣- to 3-pregnanolone isomers did not change significantly during the third trimester (data not shown).
Pronounced differences were observed between the 5␣and 5-pregnane steroids in their profiles during the third trimester. The P5␣ and 5␣-PIs (Fig. 2A–C) showed a significant increase between the 30th and 36th weeks while the respective 5-pregnane steroids displayed either no significant change (P5) (Fig. 2D) or a decreasing trend in 5-PIs from the 36th week of gestation (Fig. 2E and F).
3.4. Changes in the ratios of 5α- and 5β-dihydroprogesterone to progesterone The ratio of P5␣ to progesterone showed a trend of borderline significance with a maximum in the 35th week (Fig. 3A),
Fig. 2. Profiles of 5␣-dihydroprogesterone (P5␣, panel A), allopregnanolone (P3␣5␣, panel B), isopregnanolone (P35␣, panel C), 5-dihydroprogesterone (P5, panel D), pregnanolone (P3␣5, panel E) and epipregnanolone (P35, panel F) in maternal plasma during the third trimester of pregnancy. F represent the Fisher’s fractiles for the factors week and subjects in ANOVA model and p denote their statistical significances. The asterisks symbolize significant differences between the 30th week of gestation and another stages, as found by Bonferroni multiple comparisons (p < 0.05).
M. Hill et al. / Journal of Steroid Biochemistry & Molecular Biology 105 (2007) 166–175
171
Fig. 3. Profiles of the ratios of 5␣-dihydroprogesterone to progesterone (P5␣/Prog, panel A), and 5-dihydroprogesterone to progesterone (P5/Prog, panel B). The drawings and symbols are as in Fig. 2.
while the respective ratio for the P5 significantly decreased from the 31st week (Fig. 3B). 3.5. Changes in conjugated pregnanolone isomers Three of the conjugated PIs showed significant increases from the 30th week to the 37th week, followed by a plateau (Fig. 4A, B and D). The only exception was P3␣5, which increased markedly up to the 39th week (Fig. 4C). 3.6. Changes in the ratios of conjugated pregnanolone isomers to free steroids Both ratios of conjugated 5␣-PIs to the respective free steroids gradually increased from the 30th to the 37th weeks, and then showed a plateau up to the 40th week (Fig. 5A and B). The ratios for 5-PIs, i.e. P35␣ (Fig. 5C) and P35 (Fig. 5D), showed a manifest increase from the 30th or 31st week.
4. Discussion The aim of the study was to evaluate the changes of neuroactive steroids in the third trimester of pregnancy, and to hypothesize their role in the timing of parturition. The supposed changes that may involve steroid oxido-reductive metabolism and the balance between steroid sulfation and hydrolysis of the sulfates were evaluated using a multicomponent analysis of steroid profiles by GC–MS. The levels of free PIs obtained in this study are comparable with the previously published data [20,26]. The concentrations of conjugates obtained here are however higher. This
discrepancy appears to be due to the less efficient method used for the hydrolysis of steroid polar conjugates in the earlier studies. The method used in this study [27] was highly efficient, with negligible losses. In general, PIs with a hydroxy group in the 3␣-position (3␣-PIs) attenuate neuronal activity via stimulation of the receptors of type A ␥-aminobutyric acid (GABAA -r) [11]. PIs hydroxylated in the 3-position (3-PIs) compete with the 3␣-PIs for the active sites on the receptors [29]. Sulfation counteracts the effect of 3␣-PIs and further amplifies the antagonistic effect of 3-PIs, forming products that negatively modulate GABAA -r on the binding sites, which are different from the sites specific to the unconjugated 3␣-PIs. For example, the neuromodulating efficiency of isopregnanolone sulfate is comparable to the neuromodulating effectiveness of allopregnanolone [13] but the steroid conjugate exerts the opposite effect. Therefore, it may be of importance that the concentrations of neuroinhibiting free 3␣-PIs are markedly lower in the maternal plasma before labor when compared to polar conjugates of PIs [19,20,26]. Besides the effect on GABAA -r, PIs and particularly the unconjugated 5-isomers also block calcium channels of type T in the rat peripheral neurons, mediating the perception of pain [14]. The polar conjugates of 5␣-PIs are also known to be positive modulators of N-methyl-d-aspartate receptors (NMDA-r), which are present both in the CNS and the periphery [30]. Conjugated 5-PIs exert the opposite effect [12,30–32]. The role of the most abundant PI allopregnanolone in the onset of parturition has been reported in rats where a positive feedback loop in oxytocin production resulting in a rapid delivery forms just before labor. A decrease in the levels of allopregnanolone triggers the production of oxytocin [9,10].
172
M. Hill et al. / Journal of Steroid Biochemistry & Molecular Biology 105 (2007) 166–175
Fig. 4. Profiles of the polar conjugates of allopregnanolone (P3␣5␣, panel A), isopregnanolone (P35␣, panel B), pregnanolone (P3␣5, panel C) and epipregnanolone (P35, panel D) in maternal plasma during the third trimester of pregnancy. The drawings and symbols are as in Fig. 2.
This substance brings the GABAA -r from a neurosteroidsensitive mode towards a condition in which the receptors are not sensitive [33], via a shift in the balance between the activities of endogenous Ser/Thr phosphatase and protein kinase C [33]. Considering the aforementioned findings, it is likely that changing biosynthesis of PIs could influence the onset of labor. Hence, we analyzed changes of product to
precursors ratios associated with the respective steroid reductases, oxidoreductases, sulfatases and sulfotransferases. The ratios of 3␣-PI and 3-PIs to their 3-oxo-precursors (5␣or 5-dihydroprogesterone) did not change during the third trimester (data not shown), negating the supposed oxidoreductive shift between 3␣-PIs and 3-PIs, which could influence the function of GABAA -r.
M. Hill et al. / Journal of Steroid Biochemistry & Molecular Biology 105 (2007) 166–175
173
Fig. 5. Profiles of the ratios of the polar conjugates of allopregnanolone (P3␣5␣, panel A), isopregnanolone (P35␣, panel B), pregnanolone (P3␣5, panel C) and epipregnanolone (P35, panel D) to their respective free steroids in maternal plasma during the third trimester of pregnancy. The drawings and symbols are as in Fig. 2.
174
M. Hill et al. / Journal of Steroid Biochemistry & Molecular Biology 105 (2007) 166–175
On the other hand, pronounced differences were found between the profiles of 5␣- (Fig. 2A–C) and 5-PIs (Fig. 2D–F). The ratios of P5␣ and P5 to progesterone indicated that the metabolism of progesterone to P5␣ inconspicuously culminated in the 35th week (Fig. 3A). Alternatively, the conversion of progesterone to P5 significantly declined from the 31st week (Fig. 3B). The latter effect accords with the recent study of Gilbert Evans et al. [34]. Excessive concentrations of neuroactivating conjugated PIs exerting an opposite effect on GABAA -r compared to the free 3␣-PIs [13] were found in maternal circulation (Fig. 4). The results clearly demonstrate increasing sulfotransferase activity in all PIs, as documented by significantly increasing levels of conjugated PIs (Fig. 4) and ratios of conjugates to the respective free steroids (Fig. 5). The ratios of conjugates to free PIs showed a gradual increase in the 5␣-isomers up to the 37th week (Fig. 5A and B) but a pronounced, accelerating increase in both 5-analogs (Fig. 5C and D). The most prominent change was apparent in the second most abundant PIs—the neuroinhibiting P3␣5 (Fig. 5C). The aforementioned findings showed that unconjugated PIs including the neuroinhibiting 3␣-PIs were more rapidly metabolized with oncoming parturition, particularly in the case of P3␣5. This steroid is chemically identical with the short-term intravenous anesthetic eltanolone that takes an effect via positive modulation of GABAA -r. Moreover, biosynthesis of the 5-pregnane steroids decelerated from the 31st week (Fig. 3B). Considering these findings, as well as the previously demonstrated pronounced decrease of the ratio of conjugated PIs to free PIs after delivery [26], one can speculate as to whether the decreasing levels of P3␣5 might influence the sustaining of parturition. In terms of the effects of circulating PIs on neuronal activity in the CNS, the authors are aware of the limitations in transport across the blood–brain barrier (BBB), particularly in the case of steroid conjugates. On the other hand, the lowpolar free PIs may pass across the BBB relatively easily. The chances of overcoming the blood–brain barrier generally increase with the decreasing polarity of the substance [35]. The peripheral effects of PIs in pregnancy should be considered, respecting the presence of receptors sensitive to pregnane steroids in tissues connected with parturition. For instance, 5-PIs may take effect through a pregnane X receptor-mediated mechanism regulating uterine contractility [16]. It is likely that the decrease in activity of 5-reductase [34,36] and the increasing sulfation of PIs during the third trimester are associated with a reduced ability to sustain the pregnancy. In conclusion, the identification of accelerating sulfation in pregnancy-sustaining steroids during the third trimester is the most fundamental outcome of this study. This metabolic step at the least eliminates the neuroinhibiting effect of 3␣PIs. Another important result is an endorsement for the previously suggested declining expression of 5-reductase from the 31st week of gestation. The enzyme participates in
the formation of almost 40% of pregnancy- sustaining pregnanolone isomers. It is likely that the contribution of the PIs to the initiation of human parturition is based in part on the descending expression of 5-reductase from the 31st week. 5-Reductase shows similar neuroinhibiting efficiency on GABAA -r to the slightly more abundant 3␣-isomer allopregnanolone, but even higher activity on the nuclear pregnane X receptors preventing uterine contractility [16]. In terms of explicit changes in the circulating levels of PIs and a plausible explanation of the respective mechanisms, a significant role for PIs may be expected in the timing of human parturition. Acknowledgements This study was supported by Grant Agency of the Czech Republic grant no. 303/04/1086. The excellent technical assistance of Mrs. Ivona Kr´alov´a and Mrs. Marta Vel´ıkov´a is gratefully acknowledged. References [1] P.W. Nathanielsz, Comparative studies on the initiation of labor, Eur. J. Obstet. Gynecol. Reprod. Biol. 78 (2) (1998) 127–132. [2] W.X. Wu, X.H. Ma, T. Coksaygan, K. Chakrabarty, V. Collins, J. Rose, P.W. Nathanielsz, Prostaglandin mediates premature delivery in pregnant sheep induced by estradiol at 121 days of gestational age, Endocrinology 145 (3) (2004) 1444–1452. [3] R.S. Mathur, S. Landgrebe, H.O. Williamson, Progesterone, 17hydroxyprogesterone, estradiol, and estriol in late pregnancy and labor, Am. J. Obstet. Gynecol. 136 (1) (1980) 25–27. [4] S. Mesiano, Myometrial progesterone responsiveness and the control of human parturition, J. Soc. Gynecol. Investig. 11 (4) (2004) 193– 202. [5] W. Jeske, P. Soszynski, W. Rogozinski, E. Lukaszewicz, W. Latoszewska, H. Snochowska, Plasma GHRH, CRH, ACTH, betaendorphin, human placental lactogen, GH and cortisol concentrations at the third trimester of pregnancy, Acta Endocrinol. (Copenh.) 120 (6) (1989) 785–789. [6] F.A. Patel, J.W. Funder, J.R. Challis, Mechanism of cortisol/ progesterone antagonism in the regulation of 15-hydroxyprostaglandin dehydrogenase activity and messenger ribonucleic acid levels in human chorion and placental trophoblast cells at term, J. Clin. Endocrinol. Metab. 88 (6) (2003) 2922–2933. [7] M. McLean, R. Smith, Corticotrophin-releasing hormone and human parturition, Reproduction 121 (4) (2001) 493–501. [8] J.C. Condon, P. Jeyasuria, J.M. Faust, C.R. Mendelson, Surfactant protein secreted by the maturing mouse fetal lung acts as a hormone that signals the initiation of parturition, Proc. Natl. Acad. Sci. U.S.A. 101 (14) (2004) 4978–4983. [9] A.B. Brussaard, J. Wossink, J.C. Lodder, K.S. Kits, Progesteronemetabolite prevents protein kinase C-dependent modulation of gamma-aminobutyric acid type A receptors in oxytocin neurons, Proc. Natl. Acad. Sci. U.S.A. 97 (7) (2000) 3625–3630. [10] A.B. Brussaard, K.S. Kits, R.E. Baker, W.P. Willems, J.W. LeytingVermeulen, P. Voorn, A.B. Smit, R.J. Bicknell, A.E. Herbison, Plasticity in fast synaptic inhibition of adult oxytocin neurons caused by switch in GABA(A) receptor subunit expression, Neuron 19 (5) (1997) 1103–1114. [11] M.D. Majewska, Steroid regulation of the GABAA receptor: ligand binding, chloride transport and behaviour, Ciba Found Symp. 153 (1990) 83–97 (discussion 97–106).
M. Hill et al. / Journal of Steroid Biochemistry & Molecular Biology 105 (2007) 166–175 [12] M. Park-Chung, F.S. Wu, R.H. Purdy, A.A. Malayev, T.T. Gibbs, D.H. Farb, Distinct sites for inverse modulation of N-methyl-daspartate receptors by sulfated steroids, Mol. Pharmacol. 52 (6) (1997) 1113–1123. [13] M. Park-Chung, A. Malayev, R.H. Purdy, T.T. Gibbs, D.H. Farb, Sulfated and unsulfated steroids modulate gamma-aminobutyric acid A receptor function through distinct sites, Brain Res. 830 (1) (1999) 72–87. [14] S.M. Todorovic, S. Pathirathna, B.C. Brimelow, M.M. Jagodic, S.H. Ko, X. Jiang, K.R. Nilsson, C.F. Zorumski, D.F. Covey, V. Jevtovic-Todorovic, 5Beta-reduced neuroactive steroids are novel voltage-dependent blockers of T-type Ca2+ channels in rat sensory neurons in vitro and potent peripheral analgesics in vivo, Mol. Pharmacol. 66 (5) (2004) 1223–1235. [15] C.D. Putnam, D.W. Brann, R.C. Kolbeck, V.B. Mahesh, Inhibition of uterine contractility by progesterone and progesterone metabolites: mediation by progesterone and gamma amino butyric acid A receptor systems, Biol. Reprod. 45 (2) (1991) 266–272. [16] B.F. Mitchell, J.M. Mitchell, J. Chowdhury, M. Tougas, S.M. Engelen, N. Senff, I. Heijnen, J.T. Moore, B. Goodwin, S. Wong, S.T. Davidge, Metabolites of progesterone and the pregnane X receptor: a novel pathway regulating uterine contractility in pregnancy? Am. J. Obstet. Gynecol. 192 (4) (2005) 1304–1313 (discussion 1303–1305). [17] L. Milewich, N.F. Gant, B.E. Schwarz, G.T. Chen, P.C. MacDonald, 5 alpha-reductase activity in human placenta, Am. J. Obstet. Gynecol. 133 (6) (1979) 611–617. [18] R.A. Dombroski, M.L. Casey, P.C. MacDonald, 5-Alpha-dihydroprogesterone formation in human placenta from 5alpha-pregnan3beta/alpha-ol-20-ones and 5-pregnan-3beta-yl-20-one sulfate, J. Steroid Biochem. Mol. Biol. 63 (1–3) (1997) 155–163. [19] H. Mickan, J. Zander, Pregnanolones, pregnenolone and progesterone in the human feto-placental circulation at term of pregnancy, J. Steroid Biochem. 11 (4) (1979) 1461–1466. [20] M. Hill, A. Parizek, M. Bicikova, H. Havlikova, J. Klak, T. Fait, D. Cibula, R. Hampl, A. Cegan, J. Sulcova, L. Starka, Neuroactive steroids, their precursors, and polar conjugates during parturition and postpartum in maternal and umbilical blood. 1. Identification and simultaneous determination of pregnanolone isomers, J. Steroid Biochem. Mol. Biol. 75 (4–5) (2000) 237–244. [21] M. Hill, P. Popov, H. Havlikova, L. Kancheva, J. Vrbikova, R. Kancheva, V. Pouzar, I. Cerny, L. Starka, Altered profiles of serum neuroactive steroids in premenopausal women treated for alcohol addiction, Steroids 70 (8) (2005) 515–524. [22] A.R. Genazzani, F. Petraglia, F. Bernardi, E. Casarosa, C. Salvestroni, A. Tonetti, R.E. Nappi, S. Luisi, M. Palumbo, R.H. Purdy, M. Luisi, Circulating levels of allopregnanolone in humans: gender, age, and endocrine influences, J. Clin. Endocrinol. Metab. 83 (6) (1998) 2099–2103. [23] S. Luisi, F. Petraglia, C. Benedetto, R.E. Nappi, F. Bernardi, M. Fadalti, F.M. Reis, M. Luisi, A.R. Genazzani, Serum allopregnanolone levels in pregnant women: changes during pregnancy, at delivery, and in hypertensive patients, J. Clin. Endocrinol. Metab. 85 (7) (2000) 2429– 2433.
175
[24] B.E. Pearson Murphy, C.M. Allison, Determination of progesterone and some of its neuroactive ring A-reduced metabolites in human serum, J. Steroid Biochem. Mol. Biol. 74 (3) (2000) 137–142. [25] B.E. Pearson Murphy, S.I. Steinberg, F.Y. Hu, C.M. Allison, Neuroactive ring A-reduced metabolites of progesterone in human plasma during pregnancy: elevated levels of 5 alpha-dihydroprogesterone in depressed patients during the latter half of pregnancy, J. Clin. Endocrinol. Metab. 86 (12) (2001) 5981–5987. [26] M. Hill, M. Bicikova, A. Parizek, H. Havlikova, J. Klak, T. Fajt, M. Meloun, D. Cibula, A. Cegan, J. Sulcova, R. Hampl, L. Starka, Neuroactive steroids, their precursors and polar conjugates during parturition and postpartum in maternal blood. 2. Time profiles of pregnanolone isomers, J. Steroid Biochem. Mol. Biol. 78 (1) (2001) 51–57. [27] L. Dehennin, P. Lafarge, P. Dailly, D. Bailloux, J.P. Lafarge, Combined profile of androgen glucuro- and sulfoconjugates in postcompetition urine of sportsmen: a simple screening procedure using gas chromatography-mass spectrometry, J. Chromatogr. B: Biomed. Appl. 687 (1) (1996) 85–91. [28] H. Havlikova, M. Hill, L. Kancheva, J. Vrbikova, V. Pouzar, I. Cerny, R. Kancheva, L. Starka, Serum profiles of free and conjugated neuroactive pregnanolone isomers in non-pregnant women of fertile age, J. Clin. Endocrinol. Metab. (2006). [29] P. Lundgren, J. Stromberg, T. Backstrom, M. Wang, Allopregnanolonestimulated GABA-mediated chloride ion flux is inhibited by 3beta-hydroxy-5alpha-pregnan-20-one (isoallopregnanolone), Brain Res. 982 (1) (2003) 45–53. [30] C.E. Weaver, M.B. Land, R.H. Purdy, K.G. Richards, T.T. Gibbs, D.H. Farb, Geometry and charge determine pharmacological effects of steroids on N-methyl-d-aspartate receptor-induced Ca(2+) accumulation and cell death, J. Pharmacol. Exp. Ther. 293 (3) (2000) 747–754. [31] A. Malayev, T.T. Gibbs, D.H. Farb, Inhibition of the NMDA response by pregnenolone sulphate reveals subtype selective modulation of NMDA receptors by sulphated steroids, Br. J. Pharmacol. 135 (4) (2002) 901–909. [32] M.D. Wang, T. Backstrom, S. Landgren, The inhibitory effects of allopregnanolone and pregnanolone on the population spike, evoked in the rat hippocampal CA1 stratum pyramidale in vitro, can be blocked selectively by epiallopregnanolone, Acta Physiol. Scand. 169 (4) (2000) 333–341. [33] J.J. Koksma, R.E. van Kesteren, T.W. Rosahl, R. Zwart, A.B. Smit, H. Luddens, A.B. Brussaard, Oxytocin regulates neurosteroid modulation of GABA(A) receptors in supraoptic nucleus around parturition, J. Neurosci. 23 (3) (2003) 788–797. [34] S.E. Gilbert Evans, L.E. Ross, E.M. Sellers, R.H. Purdy, M.K. Romach, 3Alpha-reduced neuroactive steroids and their precursors during pregnancy and the postpartum period, Gynecol. Endocrinol. 21 (5) (2005) 268–279. [35] I. Oren, S.J. Fleishman, A. Kessel, N. Ben-Tal, Free diffusion of steroid hormones across biomembranes: a simplex search with implicit solvent model calculations, Biophys. J. 87 (2) (2004) 768–779. [36] P.M. Sheehan, G.E. Rice, E.K. Moses, S.P. Brennecke, 5Beta-dihydroprogesterone and steroid 5 beta-reductase decrease in association with human parturition at term, Mol. Hum. Reprod. 11 (7) (2005) 495–501.