Loss of Placental P-Glycoprotein May Be a Common Pathophysiologic Pathway for First Trimester Spontaneous Miscarriages

Introduction

Early pregnancy loss is the most common complication in pregnancy and is often diagnosed before the onset of symptoms. ¹ Successful maintenance of early pregnancy depends on an endocrinologic and metabolic equilibrium resulting from physiological changes occurring at the materno–fetal interface ensuring successful implantation and maintenance. ¹ Recently, the role of adenosine triphosphate (ATP)-binding transmembrane transporters has been implicated in peri-implantation events and fetal/placental development. ²

P-glycoprotein (P-gp), the highly conserved mammalian ATP-binding cassette transmembrane multidrug protein transporter ³ encoded by the ABCB1(MDR1) gene, ⁴ is broadly expressed in normal human tissues ⁵ and primarily functions as an efflux pump protecting cells from toxic xenobiotics and endogenous metabolites. ⁶ Arceci et al. and Axiotis et al. showed high levels of P-gp expression in human and murine secretory and gestational endometrium regulated by estrogen and progesterone and suggested that P-gp may be important in implantation, utero-placental transport of substrates, and in early embryo–endometrial interactions.^7,8 Subsequent studies demonstrated that P-gp is localized both in the apical surface of the syncytial microvillous membrane and in the syncytium facing membrane of the cytotrophoblast of the first trimester placenta and that P-gp expression is greatly reduced at term. ⁹ This differential expression in early versus late pregnancy and the localization of P-gp in both syncytio and cytotrophoblast strongly suggest a pivotal role in utero-placental transport of substances across the uterine epithelium during early pregnancy as well as in the protection of the embryo from toxic substances during implantation and early embryogenesis.^2,10,11

We believe that understanding utero-placental P-gp pathophysiology may provide insight into early pregnancy loss. The aim of this study was to investigate P-gp expression in first trimester miscarriages and compare the level of expression to first trimester elective terminations. We hypothesized that P-gp is reduced in early pregnancy loss and that P-gp dysregulation may be involved in early pregnancy pathologies.

Material and Methods

Results

Age range was from 18 to 43 years (mean = 33.6 years) in the spontaneous abortion group and from 21 to 41 years (mean = 32.5) in the elective termination group. Gestational age of the placental tissue from spontaneous abortion group ranged from 6 weeks to 18 weeks (mean = 11 weeks) and that of the elective termination group ranged between 8 and 13 weeks (mean = 9.7 weeks). There was no difference between the age and gestational age between the 2 groups (P = .629 and P = .222, respectively). Beta human chorionic gonadotropin (β-HCG) values of the spontaneous abortion group ranged from undetectable to 58 681 mU/mL. Most patients from elective termination group had only a positive qualitative urine β-HCG test and quantitative β-HCG values were not available. One patient from the spontaneous abortion group had a history of recurrent first trimester abortions. Ultrasonogram did not show any abnormalities other than absent fetal heart rate in the spontaneous abortion group and showed normal fetal heart rate and no other abnormalities in the elective termination group. None of the patients from both groups had any evidence of infection during pregnancy.

Immunohistochemistry for P-gp showed 90% of controls (elective termination group) (18/20) to be positive for staining with both JSB-1 and C219. Ten percent (2/20) showed loss of P-gp expression; 75% of first trimester spontaneous abortion cases (27/36) showed total absence of P-gp (score 0), 19% (7/36) showed rare villi with discontinuous immunostaining (score 0), and 6% (2/36) showed positive immunostaining with P-gp (score 1+ and 2 +, one each). Altogether, 94% (34/36) of chorionic villi in first trimester spontaneous abortions showed loss of P-gp staining in the apical microvillous border of syncytiotrophoblasts and cytotrophoblasts (Figure 1). Tissue stained with C219 showed basolateral membrane staining, in addition to the apical membrane of the syncytiotrophoblasts that was shown by JSB-1 antibody, resulting in a diffuse smudgy pattern. Only apical membranous staining of syncytiotrophoblasts was taken into consideration for scoring. Although the JSB-1 stain was much darker, crisper, and easier to interpret than C219, there was no discrepancy between the staining score obtained from JSB-1 and C219. Our results were statistically significant with a P value of <.0001 (Table 1). OPEN IN VIEWER

OPEN IN VIEWER

Table 1.

Summary of P-gp Immunostain Score in the Placental Tissues of Elective and Spontaneous Abortion.

Mode of Abortion	Number of Cases With P-gp Immunostain Semiquantitative score				P
	0	1+	2+	3+
Elective (N = 20)	2	3	12	3	Fisher’s exact test P < .0001
Spontaneous (N = 36)	34	1	1	0

Karyotype was not available for any of the cases of elective termination. Seventeen out of 36 cases of spontaneous abortion had karyotyping done, 14 cases had normal karyotype, and 3 had complex karyotype (with more than 3 abnormalities). All 3 cases with complex karyotype demonstrated complete absence of P-gp. Of the remaining 14 cases, 1 had positive staining for P-gp (2+) and the remainder showed complete absence of P-gp staining. The difference was statistically significant when the staining pattern of these cases of spontaneous abortion with normal karyotype were compared with elective termination, with a P value of <.0001 (Table 2).

OPEN IN VIEWER

Table 2.

P-gp Immunostain Score in the Placental Tissues of Spontaneous Abortion With Normal Karyotype and Elective Termination.

Mode of Abortion	Number of Cases With P-gp Immunostain Semiquantitative Score				P
	0	1+	2+	3+
Elective (N = 20)	2	3	12	3	Fisher’s exact test P < .0001
Spontaneous (N = 14)	13	0	1	0

Discussion

To the best of our knowledge, this is the first demonstration of loss of P-gp in first trimester miscarriage. This finding in conjunction with the known high expression of P-gp in normal first trimester placental tissues suggests an important role of P-gp in the maintenance of early pregnancy. For instance, they efflux steroid hormones, drugs, environmental toxins and also transport cholesterol and fat-soluble vitamins functioning as a gatekeeper at the feto–maternal interface. They are also substrates for various cytokines and chemokines. ²

Our study of first trimester fetal tissue from elective termination of pregnancy concurred with previous investigations showing high P-gp expression and localization to the apical surface of the microvillous membrane of the syncytium and the syncytium facing membrane of the cytotrophoblast.^9,10,12 This localization suggests that P-gp has both a role in fetal protection and maternal–fetal substrate transport. Furthermore, gestation-dependent expression of P-gp, high levels during first trimester and reduced levels at term, also suggests the potential importance of P-gp in early fetal protection from toxic substances. This has been established by previous studies showing increasing fetal digoxin accumulation with advancing pregnancy in murine models. ¹³ Reduced P-gp levels in the later stages of pregnancy may be essential to allow exposure to maternal endogenous factors including corticosteroids that play a significant role in fetal maturity. ²

P-gp expression in early pregnancy is regulated by multiple factors such as steroid hormones, medication, cytokines, hypoxia, ¹⁰ and bacterial lipopolysaccharide ¹¹ either independently or synergistically. Studies have shown that drugs that are substrates for P-gp and those that inhibit P-gp posed increased risk of teratogenicity. ¹⁴ Bacterial and viral infections have also shown to downregulate or impair P-gp function by inducing secretion of cytokines such as interleukin (IL)-1β, IL-6, and tumor necrosis factor-α. ¹⁵ Therefore, we selected our cases and controls carefully to eliminate any possible confounding factors such as medications, drug abuse, infections, exogenous hormones, and chronic illness.

In our study, we did not have the karyotype of all cases. It is therefore possible that more cases had abnormal karyotype. Nevertheless, spontaneous miscarriage could either be a consequence of decrease in P-gp or P-gp dysregulation which may represent a common pathophysiologic pathway for early pregnancy loss independent of etiology. Of the 36 cases from the spontaneous abortion group, only 2 cases showed positive staining for P-gp, whereas in the elective termination group, 18/20 cases had positive staining for P-gp. Therefore, we propose that dysregulation of P-gp is a common pathophysiologic pathway in abortions, which leads to dysfunction of the efflux pumps, exposing the fetus to various toxic insults and subsequent abortion. Although it is possible that the degenerative changes in specimens from spontaneous abortion may have an impact on their staining characteristics unlike the specimens from elective termination of pregnancy which are usually fresh, we believe that the statistical significance in our study is too high to be attributed to such a possible confounding factor. Drugs that induce or activate P-gp limiting the toxicity caused by its substrates may be of therapeutic interest. P-gp activators, a new class of compounds with the ability to immediately increase P-gp activity without waiting for increased protein expression, may be of particular therapeutic interest because they immediately promote the P-gp efflux pump. ¹⁶ However, the fact that the distribution of P-gp is not limited only to placental tissues but also found functional in other organ systems pose a great challenge. Our study provides a foundation and impetus to further study of the role of P-gp in the maintenance of early pregnancy.