Does placenta prevent dangerous substances from passing from the mother to the embryo?

3.3 Translocation Into the Brain of Offspring Through the Placental Barrier

Placental barrier, composed of both maternal and fetal tissue, is another internal barrier that can protect the development of embryo (Chu et al., 2010). It could protect the fetus from being affected by harmful substances in maternal blood circulation, whereas the fetus could get nutrients and oxygen from the mother via the placenta. However, a great number of studies (Di Bona et al., 2014; Semmler-Behnke et al., 2014) have already revealed that after pregnant mice/rats were exposed to exogenous substances, such as NPs, those substances could be detected in the brain of the fetus, and then they can perturb the homeostasis of brain or even lead to neuronal death. Those harmful impacts on fetus brain are related with psychiatric disorders such as autism, schizophrenia, and depression in their later life (O’Connor and Paley, 2009; Ekblad et al., 2010). As a consequence, those findings suggested that placental barrier plays an important part in fetal growth and development.

In a study by Mohammadipour et al. (2014), pregnant Wistar rats were administered TiO2 NPs intragastrically daily from gestational day 2 to 21. The TiO2 NPs concentration in the hippocampus of 1-day-old neonates, determined by ICP-MS, was significantly upregulated as compared with that in the control group. Yamashita et al. (2011) also detected silica (70 nm) and TiO2 (35 nm) NPs in the placenta and fetal brain after the pregnant mice were injected intravenously with these NPs, which led to pregnancy complication.

In another study (Takahashi et al., 2010), the authors discovered that when pregnant mice were treated with TiO2 NPs, the levels of dopamine and its metabolites were increased in some regions of the fetal brain on postnatal day 21. Another study (Umezawa et al., 2012) adopted microarray to assess gene expression changes in brains of male fetus and pups after pregnant mice were exposed to TiO2 NPs. Data showed that the gene expression related with the dopamine neuron system was altered. Shimizu et al. (2009) also used microarray in their study, and pregnant mice were administered TiO2 NPs. By analyzing the gene expression alternations in brain of male fetus and pups, data obtained revealed that the expression of genes related with oxidative stress, neurotransmitters, and psychiatric diseases was dysregulated. The neurobehavioral performance of the offspring might be moderately altered because of maternal exposure to TiO2 NPs (Hougaard et al., 2010). Similarly, Cui et al. (2014) discovered that when Sprague–Dawley rats were injected subcutaneously with TiO2 NPs, the antioxidant ability of pups’ brain was impaired. Although these researches did not measure the contents of TiO2 NPs in the brain directly, data collected indirectly demonstrated that the TiO2 NPs in the maternal circulation system would affect the development of the embryo. Moreover, TiO2 NPs could impair brain development and finally lead to brain disorders in later life.

Pathophysiology of Antibody Development and Transfer

The placental barrier offers certain protections to the fetus (Daniels and Reid, 2010). For a pregnant RhD-negative woman to be immunized due to RhD incompatibility, there must be direct contact of fetal red cells with the maternal circulation. This occurs most commonly in the setting of fetomaternal hemorrhage (FMH) prior to or at the time of delivery. Exposure to less than 1 ml of fetal blood is enough to induce maternal alloimmunization against this highly immunogenic antigen, and almost all deliveries will result in the passage of at least trace amounts of fetal erythrocytes into the maternal circulation (Finn et al., 1961). Trauma, miscarriage, and abortion can also result in varying degrees of FMH. Incompatible transfusions (an RhD-negative woman receiving an RhD-positive transfusion) can also result in alloimmunization. Exposure to a foreign antigen in the appropriate setting will initially result in host production of IgM antibody. In the setting of HDFN, this is not dangerous to the fetus because exposure usually occurs at delivery, and this type of antibody cannot cross the placenta. The rare case of HDFN in the first pregnancy typically results from unidentified prior sensitization (transfusion, missed abortion, trauma, procedure such as amniocentesis without prophylaxis), or a significant initial exposure such as in the setting of severe FMH that occurs prior to delivery.

With repeated exposure, an anamnestic response occurs, resulting in predominant production of IgG antibodies against the incompatible antigen, which typically is a high titer, sustained response. Once an IgG antibody develops, it can cross the placenta via Fc receptor-mediated transport. Although transfer of IgG can begin early in the second trimester, the rate of transfer increases as gestation progresses, becoming maximal in the third trimester. When the antibody is present in the fetal circulation, it is available to bind to fetal red cells resulting in opsonization. As these antibody-coated cells traverse the fetal circulation, they are recognized by Fc receptors and fetal reticuloendothelial macrophages engulf them, resulting in the characteristic findings of extravascular hemolysis. Severely affected patients can develop kernicterus or fetal hydrops. Erythrocyte hemolysis with RhD incompatibility rarely occurs through initiation of the complement cascade. Though IgM is a powerful inducer of complement, it cannot cross the placenta. Certain subclasses of IgG can fix complement in certain circumstances (especially subclasses IgG1 and IgG3), but this occurs very infrequently in HDFN because the antigen density is too low to result in complement fixation and the D antigen, unlike many other antigens, does not cap (move in the membrane), making complement activation less likely. Hemoglobinemia and hemoglobinuria are rarely found but can result in renal insufficiency, disseminated intravascular coagulation, or other complications.

Placental Transfer: the “Passive” Barrier Concept

The term placental barrier includes a somewhat false notion, because the placenta is not a true barrier. Instead, the placenta is the entry through which the fetus is exposed to chemicals. Passive diffusion is hereby a common phenomenon, primarily driven by the concentration gradient between the maternal and the fetal compartment, further modulated by maternal, fetal, and placental blood circulation. The rate and extent of drug transfer mainly relates to the physicochemical and structural characteristics of the specific compound, as well as to the physiologic characteristics of the maternal-placental-embryonic-fetal unit.26-32 The concept of the placenta as lipoid membrane is hereby useful to describe and predict the impact of physicochemical characteristics of a specific compound on its placental transfer. Most drugs cross the placental membranes by passive diffusion. This rate is governed mainly by physicochemical factors according to Fick's law26-30,32:

[20-1]Rate of diffusion =D×Δc×A/d

where A = area of exchange, d = membrane thickness, Δc = drug concentration gradient across the membrane (e.g., difference between maternal and fetal plasma drug concentrations), and D = diffusion constant for the drug. Such a definition immediately reflects the impact of gestational age (area of exchange, membrane thickness), dose, and maternal disease characteristics (area of exchange, membrane thickness) or treatment modalities (e.g., prenatal lung maturation affects placental growth).6,33 From this equation, it may be predicted that a larger area of placental exchange (A), consisting of membranes with limited thickness (d), favors placental transfer of drugs. A, d, and Δc can be determined in a model; however, D is far more difficult to predict because it results from the interactions between the membrane and the molecule. The resistance within the tissue layers interposed between the maternal and fetal circulations (compartments) limits the diffusion, which is significant for hydrophilic molecules. In the human placenta, two layers contribute to this diffusional resistance: the trophoblast and the endothelium. Hydrophilic molecules either have to pass through these layers (i.e., the membrane hypothesis) or find their way through water-filled channels that extend through the trophoblast and communicate with the intracellular channels of the endothelial layer (i.e., the aqueous pores hypothesis). Rapid placental transfer is therefore related to better lipid solubility and low ionization and protein binding of drugs with a molecular weight (MW) of <500 daltons.

The permeability of lipid-soluble substances is much higher. For these substances, the placental flux rate is limited mainly by availability of drug at the area of exchange, which is ultimately determined by blood flow.34 Therefore the initial maternal-fetal concentration gradient, Δc, is dependent on uterine and umbilical blood flow. This mathematical concept has been applied to predict placental barrier permeability using the quantitative structure-activity relationship (QSAR) method.6 Such a QSAR hereby integrates the chemical structures of molecules and the available biologic properties to estimate—in this setting—placental permeability.6

A number of studies have noted that as placental thickness and the number of placental layers decreased and the area of exchange increased during gestation, increased placental transfer occurred.26,28-32 However, in vivo, the placenta barrier cannot fully be described by only anatomic parameters such as area or thickness. Thornburg and Faber35 found that in rabbit placenta, the fetal endothelium, which is not markedly altered during pregnancy, is the layer defining the transfer rate of many drugs; area of exchange (A) and membrane thickness (d) apparently are of secondary importance. Maternal-to-fetal transfer occurs across the placental barrier, made up of both the syncytiotrophoblast on the maternal side and the endothelial cell layer on the fetal side.36 The diffusion of drugs across the hemotrichorial placenta (mouse, rat) often is faster than across the hemomonochorial placenta (monkey, human). Interestingly, Elad and associates recently described the contribution of the human fetal endothelial cell layer on overall placental transfer of nutrients.36

In the later sections of this chapter, the impact of lipid solubility, ionization, protein binding, MW, and stereoselectivity (passive barrier aspects) on placental drug transfer will be discussed. This will be preceded by a discussion on PK models of maternal-fetal drug disposition and followed by aspects related to the active regulator function of the placenta (selective transfer mechanisms, placental drug metabolism).

Placental Barrier

The term “placental barrier” is a widely held false notion, as the placenta is not a true barrier for the transfer of most drugs and toxicants from mother to fetus. The placenta has been characterized as “a lipid membrane that permits bidirectional transfer of substances between maternal and fetal compartments” rather than as a barrier. In humans the placental barrier consists of the trophoblastic epithelium, covering the villi, the chorionic connective tissue, and the fetal capillary endothelium. The average thickness of the barrier varies from the first trimester (20–30 μm) to the third trimester (2–4 μm) (Wloch et al., 2009). At term the average exchange area is approximately 11 m2 and the placental blood flow rate is approximately 450 mL/min (Pacifici and Nottoli, 1995).

The two most important factors involved in transplacental transfer of toxicants are (1) physicochemical properties of the chemical and (2) type of placenta. Any chemical with a molecular weight (MW) <1000 readily crosses the placenta, and most drugs of use and abuse, pesticides, metals, mycotoxins, plant alkaloids, and other xenobiotics have an MW <1000. Hence, these chemicals are not restricted from reaching the fetus. However, the placenta poses a limited permeability barrier to chemicals with an MW >1000 (Gupta, 2009; DeSesso et al., 2012). Metabolic processes in the mother or the placenta can biotransform high-MW chemicals into low-MW chemicals, thus allowing them to cross the placental barrier. Chemical properties, such as lipophilicity, polarity, and degree of ionization, can also affect the placental barrier. Giaginis et al. (2009) described the applicability of a quantitative structure–activity relationship (QSAR) for modeling drugs/chemicals that pass through the human placental barrier.

The other factor that predominantly influences the transplacental transfer of chemicals is the type of placenta. From the limited literature available, it appears that the placental barrier is partial and selective to some xenobiotics and is recognized in the simpler choriovitalline type of placenta present in rodents, and also in the chorioallantoic type present in higher mammals. In general, the more complex multilayered placenta of higher animals can make it more difficult for xenobiotics to gain access to the fetus.

So far, diffusion (simple or facilitated) has been proven to be the only mechanism by which drugs and toxicants cross the human placenta, although several animal studies have suggested a role for active transport or pinocytosis. The rate and extent of transfer differ for various compounds (Welsch, 1982). The rate of diffusion is determined by the maternal–fetal drug gradient, uterine and umbilical blood flow, MW of drug/toxicant, protein binding, lipid solubility, and degree of ionization. These factors also determine the time required for maternal–fetal equilibrium.

The evidence for placental transfer of chemicals stems either from direct detection of a chemical residue or its metabolite in the placenta, umbilical cord blood, and embryo/fetus or from the specific biochemical and morphological changes induced by a chemical toxicant in the placenta/fetus. Placental transfer of various classes of pesticides is described in our previous publications (Gupta, 1995, 2009, 2012; Pelkonen et al., 2006). Similar information on placental transfer of metals, mycotoxins, and drugs has been reported elsewhere (Eisenmann and Miller, 1996; Pelkonen et al., 2006; Gupta, 2009, 2012).

In essence, the anatomical placental barrier for most toxicants/drugs (including charged molecules such as d-tubocurarine, highly ionized salicylates, pesticides, metals, mycotoxins, and narcotics) is a widely held false assumption, as they cross the placenta and reach the embryo/fetus and, thereby, produce a variety of toxicological and teratogenic effects. Finally, it is noteworthy that since the mid-1990s, another kind of placental barrier (i.e., metabolic barrier) has been recognized by physicians, toxicologists, pharmacologists, and others. Human term placenta, by having a significant butyrylcholinesterase (BuChE) activity, metabolizes cocaine and thereby serves as a metabolic barrier to protect the conceptus (Simone et al., 1994).

Effects on Pregnancy and the Newborn

Arecoline crosses the placental barrier. Adverse pregnancy outcome has been reported from a number of studies. Controlling for tobacco and alcohol use a large Taiwanese cohort study found that mothers who chewed areca through pregnancy have poorer birth outcomes, including lower birth weight, reduced birth length, and a lower ratio of male to female offspring. Another study from Papua New Guinea has confirmed the association of areca nut usage and low birth weight. Of interest this later study reported that one-third of women indicated that a reason they had continued to chew through pregnancy was to prevent morning sickness. In general babies appear healthy in most perinatal respects. There has been single case report of suspected neonatal withdrawal.

Pregnancy and Lactation

Oxycodone readily penetrates the placental barrier (Kokki, Franco, et al., 2012) and newborn infants, whose mothers have been given oxycodone, should be observed closely in order to detect signs and symptoms of opioid-related central nervous system (CNS) adverse effects.

Oxycodone is distributed into breast milk in low concentrations. This is a concern because long-lasting maternal use of opioids during breastfeeding can cause infant CNS depression. Once the mother’s milk comes in, it is recommended to limit the maternal intake of oxycodone to a few days, and a maximum oxycodone dosage of 30 mg daily is suggested (Lam et al., 2012).

Pre-Clinical Research

Although lepirudin crosses the placental barrier in laboratory animals, there is no evidence of harm to the fetus Product Information Refludan (2006). However, administration of lepirudin at 30 mg/kg/day (1.2 times the recommended maximum human total daily dose) to pregnant rats during organogenesis and perinatal-postnatal periods, suggested an increased maternal mortality due to undetermined causes Product Information Refludan (2006).

Lepirudin is not genotoxic in the Ames test, the Chinese hamster cell (V79/HGPRT) forward mutation test, the A549 human cell line unscheduled DNA synthesis (UDS) test, the Chinese hamster V79 cell chromosome aberration test, or the mouse micronucleus test. An effect on fertility and reproductive performance of male and female rats is not observed with lepirudin at i.v. doses up to 30 mg/kg/day Product Information Refludan (2006).

Lepirudin causes bleeding in animal toxicity studies. Antibodies against hirudin that appeared in several monkeys treated with lepirudin resulted in a prolongation of the terminal half-life and an increase of AUC (area under the curve) plasma values of lepirudin Product Information Refludan (2006).

Ochratoxin A in fetus and milk

Ochratoxin A crosses the placental barrier. A study in Poland showed that ochratoxin A crosses the human placenta to fetal blood and that it is excreted into breast milk (Postupolski et al., 2006). Maternal blood samples were collected immediately before labor, cord blood was obtained when the umbilical cord was transected, and breast milk was obtained 3–4 days postpartum. The mean ochratoxin A levels were 1.14 ng/mL (median 1.00 ng/mL) and 1.96 ng/mL (median 1.83 ng/mL) in maternal serum and neonatal serum, respectively. The ratio of maternal serum to fetal serum was 1.96 and the values ranged from 0.6 to 4.0. Ochratoxin A crosses the placental barrier in horses (Minervini et al., 2013). The mares were on commercial feedstuffs and 83% had detectable levels (50 pg/mL) of ochratoxin A in their serum. The serum levels of ochratoxin A for mares ranged from 69.5 to 252.6 pg/mL. Fifty percent of the foals were serum positive for ochratoxin A and the values ranged from 69.5 to 252.6 pg/mL. The maternal/fetal ratio for serum from foals positive for ochratoxin A ranged from 0.8 to 2.1. In the species studied, the level of ochratoxin A in maternal serum is not a good indicator of those of ochratoxin A in fetal serum at the time of birth. Finding ochratoxin A in the neonate before nursing and in milk are biomarkers of maternal to fetal and maternal to neonatal exposures.

Ochratoxin A is transferred from plasma to milk and the mammary gland is an excretory route for ochratoxin A. Since ochratoxin A is bound to serum proteins it is likely that both simple diffusion and transport mechanisms are involved in the shift of ochratoxin A from blood to milk. The transfer rate of ochratoxin A from blood to milk may change with the percentage of ochratoxin A that is bound to serum proteins. The milk/blood ratio in rats has been shown to be 0.4 to 0.7 after the animals were administered 50 μg ochratoxin A five days a week for 21 days, and similar results were observed in a single oral dose study (Breitholtz-Emanuelsson et al., 1993a, Hallen et al., 1998). The reported levels of ochratoxin A in breast milk ranges from nondetectable to 6600 ng/L (Breitholtz-Emanuelsson et al., 1993b; Galvano et al., 2008; Duarte et al., 2011). An Egyptian study showed that 32/50 mothers (72%) had serum and milk, levels of ochratoxin A (Hassan et al., 2006). The infants of these mothers also had serum levels of ochratoxin A. An Italian study found that maternal serum levels of ochratoxin A were not significantly correlated with the levels of ochratoxin A in breast milk and the serum/milk ratios were not linear (Biasucci et al., 2011). The duration of exposure may increase the levels of ochratoxin A in breast milk and this phenomenon is considered linked to binding of ochratoxin A to serum proteins. Dietary habits of the mother influence the levels of ochratoxin A in breast milk, with consumption of foods high in ochratoxin A resulting in the higher levels in milk (Miraglia et al., 1995; Skaug et al.; 2001; Galvano et al., 2008). The mean maternal serum/breast milk ratio at day 3–4 of lactation was 0.0058 with a low value of 0.005 and high value of 0.012 (Postupolski et al., 2006). Ochratoxin A was identified in the breast milk of a mother living in a water-damaged home (Thrasher et al., 2012). Estimates of the dietary levels of exposure to ochratoxin A were not provided. Studies have shown that ochratoxin A is not excreted in milk from ruminants given low doses of ochratoxin A (Battacone et al., 2010). However, field sampling of milk has revealed the presence of ochratoxin A in milk sampled from tanker trucks in different parts of Sweden, in milk from market basket sampling (China), and in samples of cows’ milk taken from bulk tanks (France), suggesting that dietary levels higher than those used in experimental studies could be occurring (Breitholtz-Emanuelsson et al., 1993b; Boudra et al., 2007; Chen et al., 2012). These findings are consistent with a study in sheep that showed ochratoxin A is not completely metabolized to ochratoxin α in the rumen (Blank et al., 2003).

Ochratoxin A is transferred to eggs (Frye and Chu, 1977). A recent study in hens fed a diet containing 5 ppm ochratoxin A had ochratoxin A appearing in the eggs after 5 days of feeding the contaminated diet and the level of ochratoxin A peaked in the eggs at 7.4±1.03 ng/g on day 21 (Zahoor et al., 2012). Laying hens fed a diet containing 2 ppm ochratoxin A for 3 weeks did not have detectable levels (0.5 μg/kg) of ochratoxin A in their eggs (Denli et al., 2008). Other studies have shown that poultry diets with less than 2 ppm do not result in detectable levels being transferred to eggs (Battacone et al., 2010).

Coumarin embryopathy

VKAs readily cross the placental barrier and can reach the fetus. The teratogenic risk associated with the use of VKA during pregnancy continues to be of importance because maintaining long-term anticoagulation is essential in women with heart valve replacement. Substitution with LMWH in sufficient doses during the first trimester of pregnancy and prior to delivery improves fetal outcome but increases maternal morbidity and mortality (McLintock 2011, 2013, Abildgaard 2009, Vitale 1999). However, recent studies on the use of warfarin during pregnancy have shown that both maternal and fetal outcomes are greatly improved if low-dose warfarin (≤5 mg/d) is used throughout pregnancy and replaced with LMWH close to delivery (McLintock 2013, De Santo 2012, Malik 2012, Geelani 2005).

The embryotoxicity of VKA, particularly that of warfarin, is well-known. Warfarin has been found to produce a characteristic pattern of malformations in the children of women who took this drug during pregnancy. Common features of this pattern of malformations, collectively called coumarin embryopathy or fetal warfarin syndrome, include nasal hypoplasia, stippled epiphyses, and growth retardation (Hall 1980). In a review of 63 case reports of coumarin embryopathy published after 1955, van Driel (2002a) found that anomalies of the skeleton were the most predominant feature, occurring in 51 (81%) of the 63 cases. Midfacial hypoplasia, that included a small upward pointing nose with indentations between the tip of the nose and nares, depressed nasal bridge, defective development of the nasal septum, micrognathia, a prominent forehead, and a flattened appearance of the face, was described in 47 of the cases. Stippling in the epiphyseal regions (chondrodysplasia punctata) was described in 32 (51%) of the 63 cases, mostly along the axial skeleton, at the proximal femora and in the calcanei. Limb hypoplasia, primarily involving the distal digits, may be found in up to one-third of children with coumarin embryopathy (Pauli 1993). Other anomalies reported and summarized by van Driel (2002a) were CNS abnormalities, disturbances of eye and ear development, abnormal heart development, asplenia syndrome, kidney agenesis, cleft lip, jaw and palate and pulmonary hypoplasia. Minor physical anomalies reported were lowset or poorly developed ears, a high-arched palate, hypertelorism, antimongoloid palpebral fissures, and widely spaced nipples. Hepatopathy lasting up to 4 months of age in addition to features typical of coumarin embryopathy were described in a premature infant whose mother had been treated with phenprocoumon up until 24 weeks of pregnancy (Hetzel 2006). It is likely that the liver dysfunction observed in this infant resulted from a toxic effect of phenprocoumon on the fetus similar to that which occasionally occurs with the drug in adults.

A Placental and Mammary Transfer

Methylmercury was found to cross the placental barrier readily with a direct toxic effect on the fetus. It was noted that the blood–mercury of the affected infants was significantly higher than that of the mother (Amin-Zaki et al., 1974). Kuhnert et al. (1981) compared the mercury levels in human maternal blood and cord blood and reported that there was 30% more methylmercury in the erythrocytes of fetal blood than in the erythrocytes of the maternal blood. Other study further indicated that transfer of mercury from maternal blood to fetal blood far exceeded the reverse (Reynolds and Pitkin, 1975; Hamada et al., 1997). When the brain–mercury was compared, it was also found that the concentration of mercury in the fetal brain was at least twice that of their mothers (Amin-Zaki et al., 1974). This information strongly endorsed the concept that the fetus serves as a “mercury trap” in pregnant mothers, resulting in higher tissue concentrations of mercury than in the maternal tissues.

Aside from placental transfer, the developing infant can also be exposed to mercury via the mother’s milk (Amin-Zaki et al., 1981; Grandjean et al., 1994; Kacew, 1996). Amin-Zaki et al. (1974) reported that when the mother received substantial exposure to methylmercury, suckling infants may acquire blood–mercury levels in excess of 200 ppb, which is considered to be minimal toxic level for adults. Studies by Grandjean et al. (1994) also demonstrated that hair–mercury of infants increased with the length of the nursing period. Further-more, increasing time interval from weaning to hair sampling did not decrease in mercury concentration, suggesting that a slow or absence of elimination of methylmercury during this time period (first year) of life.