Päivi Myllynen

Drug transporters in the human blood-placental barrier.

Studies on the increasing number of transporters found in the placental barrier are gaining momentum, because of their tissue‐specific expression, significance in physiology and disease, and the possible utilization of the emerging knowledge in pharmacology. In the placenta, both syncytiotrophoblast and fetal capillary endothelium express transporters. Fetal exposure is determined by the net effect of combination of transporters, their nature and localization in relation to placental cells and their substrate specificity. Although the significance of placental transporters on human fetal drug exposure is almost an unstudied field so far, their potential use to design drugs that do not cross the placenta is already being pursued. It is thus of interest to review the existing knowledge of human placental transporters. Transporters in all groups which take part in drug transport are found in human placenta. Especially, ATP‐binding cassette transporters ABCG2/breast cancer resistance protein, ABCB1/P‐glycoprotein and ABCC2/MRP2 are all expressed at the apical surface of syncytiotrophoblast facing maternal blood and are putatively important protective proteins both for placental tissue and the fetus, because they are efflux transporters and their substrates include many drugs and also environmental chemicals. Such protective effect has been shown in animals, but these results cannot be directly extrapolated to humans due to interspecies differences in placental structure and function. Experimental models utilizing human placental tissue, especially human placental perfusion, offer valuable possibilities, which have been insufficiently studied so far.

The fate and effects of xenobiotics in human placenta

During past decades, knowledge on placental drug metabolism and mechanisms of placental transfer has increased significantly. Most pharmaceutical drugs administered during pregnancy cross the placenta to some extent. The important properties determining the placental transfer by passive diffusion are molecular weight, pKa, lipid solubility and protein binding. In addition to passive diffusion, compounds may cross the placenta via active transfer, facilitated diffusion, phagocytosis and pinocytosis. This review gives an update of efflux transporter proteins and xenobiotic-metabolizing enzymes that modify the fate and effects of drugs in the placenta.

Developmental expression of drug metabolizing enzymes and transporter proteins in human placenta and fetal tissues.

Transporter proteins and xenobiotic metabolizing enzymes have a crucial role in the fate of xenobiotics in human body. The expression in human placenta and fetal tissues of the proteins most commonly participating in pharmaco/toxicokinetics is reviewed. In case human data are not available, relevant animal data are included. Among transporter proteins ABC transporters, monoamine transporters and organic anion transporters are pharmacologically and toxicologically of main interest. From xenobiotic enzymes, both CYP enzymes and transferases are expressed in fetal liver already during pregnancy. In the placenta, the variety of enzymes is much more restricted. During development dynamic changes occur in both xenobiotic metabolizing enzymes and drug transporters. Although the knowledge has increased substantially over the past years it is apparent from the literature that there are uncharacterized areas, especially regarding developmental expression patterns and regulation of transporters in fetal tissues and placenta. Knowledge about tissue-specific distribution and functional significance will aid our understanding of the differences in drug response and risks for adverse events during fetal development.

Aflatoxin B1 transfer and metabolism in human placenta

Aflatoxin B1 (AFB1), a common dietary contaminant, is a major risk factor of hepatocellular carcinoma (HCC). Early onset of HCC in some countries in Africa and South-East Asia indicates the importance of early life exposure. Placenta is the primary route for various compounds, both nutrients and toxins, from the mother to the fetal circulation. Furthermore, placenta contains enzymes for xenobiotic metabolism. AFB1, AFB1-metabolites, and AFB1-albumin adducts have been detected in cord blood of babies after maternal exposure during pregnancy. However, the role that the placenta plays in the transfer and metabolism of AFB1 is not clear. In this study, placental transfer and metabolism of AFB1 were investigated in human placental perfusions and in in vitro studies. Eight human placentas were perfused with 0.5 or 5microM AFB1 for 2-4 h. In vitro incubations with placental microsomal and cytosolic proteins from eight additional placentas were also conducted. Our results from placental perfusions provide the first direct evidence of the actual transfer of AFB1 and its metabolism to aflatoxicol (AFL) by human placenta. In vitro incubations with placental cytosolic fraction confirmed the capacity of human placenta to form AFL. AFL was the only metabolite detected in both perfusions and in vitro incubations. Since AFL is less mutagenic, but putatively as carcinogenic as AFB1, the formation of AFL may not protect the fetus from the toxicity of AFB1.

ABCG2/BCRP decreases the transfer of a food-born chemical carcinogen, 2-amino-1-methyl-6-phenylimidazo[4,5-b]pyridine (PhIP) in perfused term human placenta.

We have studied the role of ATP binding cassette (ABC) transporters in fetal exposure to carcinogens using 2-amino-1-methyl-6-phenylimidazo[4,5-b]pyridine (PhIP) a known substrate for ABC transporters as a model compound. In perfusion of human term placenta, transfer of (14)C-PhIP (2 microM) through the placenta resulted in fetal-to-maternal concentration ratio (FM ratio) of 0.72+/-0.09 at 6 h. The specific ABCG2 inhibitor KO143 increased the transfer of (14)C-PhIP from maternal to fetal circulation (FM ratio 0.90+/-0.08 at 6 h, p<0.05) while the ABCC1/ABCC2 inhibitor probenecid had no effect (FM ratio at 6 h 0.75+/-0.10, p=0.84). There was a negative correlation between the expression of ABCG2 protein in perfused tissue and the FM ratio of (14)C-PhIP (R=-0.81, p<0.01) at the end of the perfusion. The expression of ABCC2 protein did not correlate with FM ratio of PhIP (R: -0.11, p=0.76). In addition, PhIP induced the expression of ABC transporters in BeWo cells at mRNA level. In conclusion, our data indicates that ABCG2 decreases placental transfer of (14)C-PhIP in perfused human placenta. Also, PhIP may modify ABC transporter expression in choriocarcinoma cells.

Placental transfer and metabolism: an overview of the experimental models utilizing human placental tissue.

Over the decades several ex vivo and in vitro models which utilize delivered human placenta have been developed to study various placental functions. The use of models originating from human placenta to study transplacental transfer and related mechanisms is an attractive option because human placenta is relatively easily available for experimental studies. After delivery placenta has served its purpose and is usually disposed of. The purpose of this review is to give an overview of the use of human placental models for the studies on human placental transfer and related mechanisms such as transporter functions and xenobiotic metabolism. Human placental perfusion, the most commonly used continuous cell lines, primary cells and tissue culture, as well as subcellular fractions are briefly introduced and their major advantages and disadvantages are discussed.

Transplacental transfer of acrylamide and glycidamide are comparable to that of antipyrine in perfused human placenta

Most drugs can penetrate the placenta but there are only a few studies on placental transfer of environmental toxic compounds. In this study, we used dual recirculating human placental perfusion to determine the transfer rate through the placenta of a neurotoxic and carcinogenic compound found in food, acrylamide and its genotoxic metabolite glycidamide. Putative acrylamide metabolism into glycidamide during the 4-h perfusions and acrylamide-derived DNA adducts in placental DNA after perfusions were also analyzed. Placentas were collected immediately after delivery and kept physiologically functional as confirmed by antipyrine kinetics, glucose consumption and leak from fetal to maternal circulation. Acrylamide (5 or 10 microg/ml) or glycidamide (5 microg/ml), both with antipyrine (100 microg/ml), was added to maternal circulation. Acrylamide and glycidamide were analyzed in the perfusion medium by liquid chromatography/mass spectrometry. Acrylamide and glycidamide crossed the placenta from maternal to fetal circulation with similar kinetics to antipyrine, suggesting fetal exposure if the mother is exposed. The concentrations in maternal and fetal circulations equilibrated within 2h for both studied compounds and with both concentrations. Acrylamide metabolism into glycidamide was not detected during the 4-h perfusions. Moreover, DNA adducts were undetectable in the placentas after perfusions. However, fetuses may be exposed to glycidamide after maternal metabolism. Although not found in placental tissue after 4h of perfusion, it is possible that glycidamide adducts are formed in fetal DNA.

Meta-analysis of data from human ex vivo placental perfusion studies on genotoxic and immunotoxic agents within the integrated European project NewGeneris

In the E.U. integrated project NewGeneris, we studied placental transport of thirteen immunotoxic and genotoxic agents in three ex vivo placental perfusion laboratories. In the present publication, all placental perfusion data have been re-analyzed and normalized to make them directly comparable and rankable. Antipyrine transfer data differed significantly between the studies and laboratories, and therefore normalization of data was necessary. An antipyrine normalization factor was introduced making the variance significantly smaller within and between the studies using the same compound but performed in different laboratories. Non-normalized (regular) and normalized data showed a good correlation. The compounds were ranked according to their transplacental transfer rate using either antipyrine normalized AUC120 or transfer index (TI120(%)). Normalization generated a division of compounds in slow, medium and high transfer rate groups. The transfer rate differed slightly depending on the parameter used. However, compounds with passage similar to antipyrine which goes through the placenta by passive diffusion, and good recovery in media (no accumulation in the tissue or adherence to equipment) were highly ranked no matter which parameter was used. Antipyrine normalization resulted in the following ranking order of compounds according to AUC(120NORM) values: NDMA ≥ EtOH ≥ BPA ≥ IQ ≥AA ≥ GA ≥ PCB180 ≥ PhIP ≥ AFB1 > DON ≥ BP ≥ PCB52 ≥ TCDD. As the variance in all parameters within a study decreased after antipyrine normalization, we conclude that this normalization approach at least partially corrects the bias caused by the small methodological differences between studies.

Placental transfer and DNA binding of benzo(a)pyrene in human placental perfusion

Benzo(a)pyrene (BP) is the best studied polycyclic aromatic hydrocarbon, classified as carcinogenic to humans. The carcinogenic metabolite, benzo(a)pyrene-7,8-dihydrodiol-9,10-epoxide (BPDE), binds covalently to DNA. The key enzyme in this metabolic reaction is CYP1A1, which has also been found in placenta and human trophoblastic cells. By using human placental perfusion we confirmed that BP added to the maternal circulation in concentrations of 0.1 and 1 microM reaches fetal compartment but somewhat slower than the freely diffusible reference substance antipyrine. A well-known P-glycoprotein (ABCB1/P-gp) antagonist verapamil did not affect the transfer more than it did in the case of antipyrine, indicating that ABCB1/P-gp does not have a role in BP transfer. In one of the two placentas perfused for 6 h with the higher concentration of BP (1 microM) BPDE specific DNA adducts were found in placental tissue after the perfusion, but not before. The ability of human trophoblastic cells to activate BP to BPDE-DNA adducts was confirmed in human trophoblastic BeWo cells. This study shows that maternal exposure to BP leads to the exposure of the fetus to BP and/or its metabolites and that placenta itself can activate BP to DNA adducts.

Microsomal metabolism of carbamazepine and oxcarbazepine in liver and placenta

Metabolism of both carbamazepine (CBZ) and oxcarbaze-pine (OCBZ) were catalyzed by human liver microsomes and microsomes from livers of CBZ-induced or non-induced C57BL/6 mice. Human placental microsomes metabolized only OCBZ. Mouse liver microsomes metabolized CBZ to carbamazepine-10,11-epoxide (CBZ-E), 10- hydroxy-10,11-dihydro-carbamazepine (10-OH-CBZ), 3- hydroxy-carbamazepine (3-OH-CBZ), 10,11-trans-dihydroxy-10,11-dihydro-carbamazepine (10,11-D) and to an unidentified metabolite. CBZ-pretreatment of mice increased both ethoxyresorufin O-deethylase activity in the liver and the amount of CBZ-E in microsomal incubations regardless of the age of mice. Human liver microsomes catalyzed the formation of CBZ to 9-hydroxymethyl-10-carbamoyl acridan (9-AC) in addition to CBZ-E, 3-OH-CBZ and 10-OH-CBZ. OCBZ was metabolized to its active metabolite in all incubations. An unknown metabolite was also present in some of the incubations. Human liver microsomes catalyzed only minute covalent binding of CBZ and OCBZ to DNA. Binding of OCBZ was, however, one order of magnitude greater than binding of CBZ. Human placental micro-somes from the mothers on CBZ therapy did not catalyze CBZ metabolism. The same microsomes catalyzed OCBZ metabolism to 10-OH-CBZ and to an unknown metabolite. These results indicate autoinduction in CBZ metabolism in mouse liver. Due to the higher binding of OCBZ than CBZ to DNA in vitro, further studies on the potential mutagenicity of OCBZ may be warranted.