Estradiol, farnesoid X receptor, and altered metabolism in pregnancy

Normal pregnancy is characterized by a gradual metabolic shift that results in women who have normal levels of cholesterol, bile acids, and glucose at the time of conception developing relative hypercholesterolemia, hypercholanemia, and insulin resistance in the third trimester.1 This phenomenon is an important physiological response to support the nutritional demands of the developing fetus and fulfill the maternal requirement for lipids that are used to produce pregnancy hormones. Metabolic alterations during pregnancy are important in the context of disease, as the gestational increases in lipids, bile acids, and glucose can result in the development of pregnancy-specific disorders, including intrahepatic cholestasis of pregnancy (ICP) and gestational diabetes mellitus. The most profound increases in lipid and bile acid levels occur in the third trimester, when reproductive hormone concentrations are at their highest during normal pregnancy. In this issue, Song et al.2 demonstrate that the transcription of the bile salt export pump (BSEP; ABCB11) is inhibited during mouse pregnancy, indicative of a cholestatic phenotype, and they show that this is mediated by the direct inhibition of the bile acid receptor farnesoid X receptor (FXR) by 17β-estradiol-activated estrogen receptor α (ERα). To study the dynamics of BSEP during pregnancy, Song et al. injected a BSEP promoter-luciferase plasmid into mice and assayed the resultant luciferase activity using an in vivo imaging system. BSEP promoter activity decreased 2 days before parturition (gestational day 19), corroborating previous findings that showed a reduction of BSEP gene expression at gestational day 18 relative to nonpregnant control mice.3 In the context of normal physiology, the observation that pregnancy mediates a reduction of BSEP expression is noteworthy, as the authors propose that this reduction results in an accumulation of hepatic bile acid levels and thus feedback inhibition of bile acid synthesis and increased cholesterol levels. Indeed, hepatic bile acid levels have been shown to be increased at gestational day 18.3 The cumulative effect of pregnancy-mediated reduction of BSEP expression could be a gradual development of hypercholanemia, as occurs in pregnant mice and humans. Moreover, if a woman has an inherited mutation in BSEP4 that reduces the transporter's activity, the phenomenon described by Song et al. could explain the onset of ICP with increasing 17β-estradiol levels in genetically predisposed pregnant women. The etiology of ICP has genetic, environmental, and endocrine components. A small proportion of ICP cases have mutations with known functional effects in BSEP and multidrug resistance protein 3 (MDR3) (ABCB4).4 Common genetic variation around the BSEP and MDR3 loci is also implicated in disease susceptibility.5 Single nucleotide polymorphisms (SNPs) in NR1H4 (FXR), the primary hepatic bile acid activated transcription factor, have also been reported in a small proportion of ICP cases.6 Although women with these genetic variants more commonly develop ICP, they are usually well outside pregnancy. There is an established body of research demonstrating that 17β-estradiol, its metabolite estradiol-17β-D-glucuronide, and ethinylestradiol can cause cholestasis in rodent models. Ethinylestradiol administration reduces murine Bsep, Ntcp, Oatp1, Oatp2, and Oatp4 mRNA and protein expression,7-9 whereas estradiol-17β-D-glucuronide has been shown to alter the subcellular localization of Bsep and Mrp2.10 Finally, there is in vitro evidence showing that 17β-estradiol-activated ERα can inhibit FXR activity.3 Any alterations in FXR activity are likely to influence bile acid and cholesterol homeostasis. Song et al. have provided in vitro and in vivo evidence of an ERα-FXR-BSEP endocrine axis in pregnancy. They demonstrate that physiological concentrations of 17β-estradiol suppress BSEP expression and have used an ERα–/– mouse model in which the 17β-estradiol-mediated reduction in BSEP expression is lost, indicating that the influence of 17β-estradiol on BSEP expression is ERα-mediated. Although they did not study hepatic bile acid levels in the ERα–/– mouse model ± 17β-estradiol, one can speculate that the bile acid concentrations would be higher in the 17β-estradiol administered wild-type mice due to an ERα-mediated reduction of BSEP expression. Using a combination of molecular techniques, Song et al. provide additional mechanistic insights by demonstrating that the activated ERα ligand binding domain interacts directly with a liganded FXR. Progesterone metabolites also contribute to pregnancy-mediated dyslipidemia and hypercholanemia, as well as in the etiology of ICP. There are several reports showing that levels of serum and urinary mono- and disulfated progesterone metabolites are markedly raised in normal pregnancy and in women with ICP.11 Sulfated progesterone metabolites at levels found in normal pregnancy and ICP can: (1) competitively inhibit NTCP-mediated bile acid uptake into hepatocytes12; (2) reduce BSEP-mediated efflux of bile acids13; and (3) act as a partial agonist of FXR, thereby reducing its activity and thus inhibit the induction of bile acid-sensitive hepatic genes consistent with cholestasis.14 Thus, it is likely that the enhanced susceptibility to hypercholanemia that is observed in pregnancy is a consequence of the combined effect of reproductive hormones that signal to antagonize bile acid metabolism and transport pathways (Fig. 1). The influence of reproductive hormones on hepatic bile acid homeostasis. In a nonpregnant state, (1) bile acid (BA)-activated FXR induces the transcription of BSEP, which subsequently effluxes BA into the canaliculus. During gestation hepatic bile acid levels rise because pregnancy levels of: 17β-estradiol (E2) activates ERα, potentiating (2) an interaction with the FXR-ligand binding domain, which inhibits the transcription of BSEP2; sulfated progesterone metabolites (P4S) increase, which act as (3) a partial FXR agonist, reducing BA-mediated FXR transcription,14 inhibits (4) NTCP-mediated uptake of BA,12 and (5) reduces BSEP-mediated BA efflux13; estradiol-17β-D-glucuronide (E2 17G) mediates (6) the internalization of transporters such as MRP2, reducing canalicular efflux of BA.10. The metabolic pathways that influence bile acid and lipid metabolism are intrinsically linked and any perturbation of hepatic bile acid levels will have an indirect influence on cholesterol concentration. Song et al. show that pregnancy levels of 17β-estradiol can reduce the expression and function of BSEP, which effluxes 80% of the hepatic bile acids across the hepatocyte canalicular membrane into the bile canaliculi.15 A gestational reduction in overall BSEP levels will likely: (1) increase hepatic bile acid concentrations and result in a homeostatic decrease in bile acid synthesis, thus indirectly increasing the hepatic cholesterol levels and contributing to gestational dyslipidemia; and (2) decrease the bile acid-driven secretion of cholesterol leading to an accumulation of oxysterols in the liver. It is plausible that other transporters, which are under the influence of FXR, are impacted by the pregnancy ERα-FXR axis, to contribute to the gestational dyslipidemia. Indeed, FXR-mediated induction of Mdr2/MRP3 has been shown to be inhibited by sulfated progesterone metabolites.14 An improved understanding of the endocrine signals that cause ICP to develop in susceptible women is of clinical importance. The severity of hypercholanemia as a consequence of elevated 17β-estradiol or progesterone sulfates will vary in women with specific genetic variants of relevance to bile acid/lipid pathways. Disease severity is of relevance to pregnancy outcome. There are accumulating data demonstrating increased rates of adverse fetal or neonatal outcome if the maternal serum bile acid level rises higher than 40 μmol/L.16, 17 There are also reports of higher total cholesterol, LDL cholesterol, and triglyceride levels in women with ICP that have higher serum bile acid levels.18 Gestational diabetes mellitus is also commoner in women with ICP.19 Thus, raised serum bile acids may signal alterations in lipid and glucose metabolism that have an additional impact over the influence of reproductive hormones on lipid and bile acid homeostasis. The clinical relevance of this work extends beyond ICP. Reproductive hormones are an important cause of drug-induced liver injury (DILI), and the mechanistic insights from studies of 17β-estradiol and progesterone interactions with hepatic nuclear receptors and biliary transporters are likely to be of relevance to other causes of DILI. Pregnancy is associated with increased rates of cholelithiasis and women with ICP have a higher prevalence of gallstones than noncholestatic controls.20 MDR3 variants are also associated with susceptibility to cholelithiasis and combined oral contraceptive pill-induced cholestasis.21 Although there are no current studies implicating BSEP mutations in oral contraceptive-related hepatic impairment, the work of Song et al. provides an impetus for this association to be pursued. This work is also of relevance to cardiovascular disease risk in women. In the Framingham Heart Study and the NHEFS the rate of coronary heart disease was higher in women with six or more pregnancies.22 As cardiovascular disease is the leading cause of death in women,23 it is hoped that future work will focus further on the impact of reproductive hormone signals on metabolic disease risks not only in the context of pregnancy but also to provide insights into lifelong disease prevention. Shadi Abu-Hayyeh, Ph.D. Catherine Williamson, M.D.