![[Feedback]](/icons/banner/feedbackACT.gif){kind=icon}
![[For Subscribers]](/icons/banner/subscriberACT.gif){kind=icon}
![[Archive]](/icons/banner/archiveACT.gif){kind=icon}
![[Search]](/icons/banner/searchACT.gif){kind=icon}
![[Contents]](/icons/banner/tocACT.gif){kind=icon}
|
|
|
|
|
||
Department of Pharmacology and Clinical Pharmacology (A.C.C., J.A.K., J.W.P., M.D.T.), the Liggins Institute (A.C.C., J.A.K., P.E.v.Z., M.D.M.), and Department of Obstetrics and Gynecology (M.D.M.), the University of Auckland, Auckland, New Zealand
(Received December 30, 2003; accepted April 19, 2004)
 |
Abstract |
|---|
 Top Abstract Materials and MethodsResultsDiscussionReferences |
|---|
).
In adults, approximately 75% of a dose of AZT is glucuronidated and excreted in urine; 14% is excreted unchanged and 10% is metabolized via another path (cytochrome P450 and NADPH-dependent cytochrome-P450 reductase). Only around 1% of a dose is metabolized to the active triphosphorylated metabolite (Blum et al., 1988). Peak concentrations of AZT and AZT-glucuronide (AZT-G) in plasma are reached 0.5 and 0.8 h, respectively, after oral dosing. The pharmacokinetic parameters for the drug include an oral bioavailability of 63%, an apparent volume of distribution of 98 liters, and a hepatic clearance of approximately 62 l/h (Holford and Benet, 1995). The plasma half-lives of both parent drug and glucuronide metabolite are approximately 1 h (Blum et al., 1988). Previous researchers have reported either that there is no appreciable glucuronidation of AZT by the placenta (Bawdon et al., 1992), or that the glucuronide metabolite (AZT-G) is present, but below the level of sensitivity of the assay used (Liebes et al., 1990). Others have reported undetectable levels of AZT-G but detectable levels of an unidentified, polar metabolite (Schenker et al., 1990).
The pharmacokinetics and pharmacodynamics of AZT show massive interindividual variability of up to 50-fold (Mazzei et al., 1990; Marchbanks et al., 1995; Fletcher and Balfour, 1996; Pacifici et al., 1996; Rodman et al., 1999). Furthermore, in vitro human liver data indicate a positive skew to the frequency distribution of uridine diphosphate glucuronosyltransferase (UGT) activities in the human population (Pacifici et al., 1996). Since glucuronidation is mainly responsible for clearance of AZT, the skew in UGT activities may be the explanation for the wide variations in AZT serum concentrations observed.
Metabolism and clearance of AZT is performed by the UGT family (Veal and Back, 1995). AZT is conjugated to the glucuronide moiety of the cofactor uridine diphosphate glucuronic acid (UDPGA), and the resulting polar metabolite is excreted in urine. The isoform UGT2B7 is thought to be the major isoform responsible for this reaction (Barbier et al., 2000). Two UDPGA transporters have been identified and cloned. One is designated Frc and transfers UDPGA, UDP-N-acetylglucosamine, and potentially UDP-xylose from the cytoplasm into the lumen of the endoplasmic reticulum (Goto et al., 2001; Selva et al., 2001). The other (hUGTrel7) transports UDPGA and N-acetylgalactosamine (Muraoka et al., 2001). The significance of these two transporters in the placenta is 2-fold. First, the Frc transporter has a demonstrated role in embryogenesis and development (Goto et al., 2001; Selva et al., 2001); and second, both transporters have a demonstrated role in supplying cofactor for glucuronidation.
Although the regulation of UGT enzyme expression and activity is not fully understood, delivery of UDPGA to the active site has been postulated to be a rate-limiting factor for UGT reaction rates (Guéraud and Paris, 1998). The specific aim of the present study was to assess the metabolism and transport of AZT and the disposition of the glucuronide metabolite in the human placenta using a comprehensive approach including the use of subcellular fractions, cell culture, and placental perfusion. The transport characteristics of the UDPGA cofactor were also studied with the aim of determining the rate and direction of transfer in the placenta to aid understanding of UGT enzyme regulation in this organ. In addition, in light of recent literature (Goto et al., 2001; Selva et al., 2001), we speculated that an understanding of the placental transport and disposition of UDPGA may have important implications for placental pharmacology and fetal development.
|
Materials and Methods |
|---|
|
TopAbstractMaterials and MethodsResultsDiscussionReferences |
|---|
Perfusion Experiments. Tissue collection. Term placental tissue was obtained with informed consent from women undergoing elective Caesarian section at term due to malpresentation or previous Caesarian section. First trimester tissue was obtained from women undergoing elective termination of pregnancy within the first trimester (12 weeks). Tissue was collected under approval from the Auckland Ethics Committee. Mothers who smoked were specifically excluded from the study due to the possibility of vascular abnormalities (Pirani and MacGillivray, 1978; Cnattingius et al., 1993). Placentas were obtained within 5 min of delivery and transported immediately from the delivery suite to the laboratory.
Perfusion system. Peripheral lobules from term placentas (n = 5) were perfused using separate, recirculating maternal and fetal circuits. The procedure was based on the method of Cannell et al. (1988) with some modifications. Briefly, a suitable artery/vein pair was selected, cannulated, and excised. Fetal perfusion was commenced with the maternal aspect uppermost and two infusion needles pushed 0.5 cm into the decidua to provide the maternal circulation. Perfusate was made from [modified] M199 supplemented with heparin (25,000 IU/ml), gentamicin (100 mg/l), glucose (2g/l), L-glutamine (100 mg/l), sodium bicarbonate (2.9 g/l), and dextran (29 g/l fetal and 7.5 g/l maternal perfusate). The maternal circuit flow rate was 18 ml/min and in the fetal circuit, 3 ml/min. Equilibration was performed for 20 min to assess physical integrity of tissue and circuits (fetal arterial pressure <40 mm Hg and leakage of perfusate from fetal to maternal circuit <2 ml/h); then, the maternal and fetal reservoirs were replaced with fresh medium containing bolus doses of AZT (2.67 ng/ml, 10 µM) and 14C-UDPGA (3 µCi/ml); "Time 0" samples were taken immediately from all circuits. Metabolic viability was assessed with samples collected hourly from maternal and fetal arteries and analyzed for oxygen and carbon dioxide saturation, electrolytes, and glucose consumption (Bayer 865 Blood Gas Analyzer; Bayer Corp.-Diagnostics Div., Tarrytown, NY), ß-human chorionic gonadotropin (ß-hCG) production (Elecsys 2010 Diagnostic Analyzer; Roche Diagnostics, Indianapolis, IN), and lactate production (Hitachi 917; Roche Diagnostics). Biochemical validation assays were performed by the Diagnostic Specimen Laboratory and National Testing Centre at Greenlane Hospital, Auckland, New Zealand. Viability was assessed according to the criteria described by Cannell et al. (1988), such that establishment of both maternal and fetal circuits occurred within 30 min of delivery, pressure in the fetal circuit was 
40 mm Hg, leakage of perfusate from fetal to maternal circuit was <2 ml/h, and tissue was blanched as perfusion proceeded. AZT transfer and metabolism was assessed in five placentas with samples (300–400 µl) collected from the maternal artery, maternal vein, fetal artery, and fetal vein at 0, 5, 10, 20, and 30 min and then at 1, 1.5, 2, 2.5, 3, 3.5, and 4 h, snap-frozen in liquid nitrogen, and stored at -80°C until analysis. The ability of the placenta to transfer the cofactor UDPGA was assessed in four placentas by addition of radiolabeled 14C-UDPGA (3 µCi/ml, 10 µM) to the maternal reservoir, and samples (100–200 µl) were taken and stored as above.
Radioactive determination of 14C-UDPGA concentrations. Samples were analyzed for the transfer of 14C-UDPGA in a Wallac Rack-beta (PerkinElmer Wallac, Turku, Finland) scintillation counter. Samples were thawed on ice, and an aliquot (100 µl) of each sample was added to scintillation tubes with 3.5 ml of Starscint scintillant and shaken to mix phases. A standard curve was constructed with known amounts of radioactive cofactor from 0.05 to 10 µM (r2 = 0.998) and radioactive counts were transformed to micromolar amounts of cofactor using Prism 3.0 software (GraphPad Software Inc., San Diego, CA).
Microsomal Metabolism of AZT. Human placental or liver microsomes [0.5 mg final concentration, prepared as previously described (Collier et al., 2002)] were preincubated for 2 min at 37°C in 0.1 M Tris buffer containing 5 mM MgCl2 and 0.05% Brij 58, and 0 to 1000 µM AZT. The reaction was initiated by addition of UDPGA (10 mM) such that the final volume was 187 µl and incubations proceeded for 20 min. The reaction was stopped with ice-cold perchloric acid (24%, 10 µl), the internal standard 4-MUG (100 µM) was added, and tubes were centrifuged at 1500g for 10 min. Supernatant was transferred to a new tube; then, 2 M KOH (1 µl) was added to raise pH, centrifuged at 10,000g and snap-frozen in liquid nitrogen, and then stored at -80°C until assay.
Metabolism of AZT in the Human Placental Cell Line JEG-3 and Primary Human Placental Explant Culture. Human placental JEG-3 cells were cultured as previously described (Keelan et al., 1999) in Ham's F12 medium containing 1% penicillin/streptomycin/glutamine supplemented with either 10% FBS or 0.1% BGG (serum-free). Medium was removed and treatments (0–100 µM AZT) were added to the plates; then cells were returned to a 37°C 95% air/5% CO2 incubator for 4, 12, and 24 h. At each time point, medium (1 ml) was collected and immediately frozen at -20°C for later chromatographic analysis. Primary placental explants from first trimester and term human placenta were cultured as previously described (Blumenstein et al., 2001). After 24 h in culture, medium was removed, treatments (0–100 µM AZT in Dulbecco's modified Eagle's medium containing 1% penicillin/streptomycin/glutamine with 1% FBS) were added to the cultures, and the plates were returned to a 37°C 95% air/5% CO2 incubator for 24 h. Medium from each plate (1 ml) was then withdrawn and frozen at –80°C until analysis.
Analysis of AZT and AZT-G by High Pressure Liquid Chromatography. AZT and AZT-G detection was performed under isocratic conditions on a Waters Alliance 2690 HPLC machine (Waters, Milford, MA) using a µBondapak C18 column (5 µm x 30 cm) at a flow rate of 1 ml/min. The mobile phase consisted of 1:9 acetonitrile/ammonium formate (450 mM, pH 3.5) filtered through a 2-µm filter under vacuum. Detection was performed with a Waters Alliance 996 photodiode array detector (Waters) and a signal channel extracted at 267 nm. Prior to analysis, all samples were thawed and an aliquot (187 µl) was added to a micro tube containing 10 µl of ice-cold perchloric acid. Internal standard (4-MUG, 100 µM) was added, and the perfusate sample was centrifuged at 10,000g for 20 min. Supernatant was then transferred into a separate tube containing 2 M KOH (1 µl), 150 µl was added to an HPLC vial, and vials were queued and kept at 4°C until injection, whereupon 100 µl of sample was injected onto the column and eluted for 20 min. Quantitation was performed by determining the ratio of the area under the concentration-time curve (AUC) for parent drug or the metabolite to that of the internal standard and using the standard curve to establish sample concentration with Millennium 32 software (version 3.2, Waters). Standard curves were generated over the range 0.5 to 50 µM, using five concentrations and a sample blank. Peak confirmation was performed by spectral comparison at 267 nm using a diode array detector, and by liquid chromatography/mass spectrometry (Agilent 1100; Agilent Technologies, Palo Alto, CA). Liquid chromatographic conditions were as described above for HPLC, and MS detection was performed with a full mass scan (100–1000 nm) using negative electrospray at 350°C and a flow rate 12 l/min.
Under the HPLC conditions described above, AZT-G, the internal standard 4-MUG, and AZT eluted at 6.5, 9.6, and 15.3 min, respectively. The limit of sensitivity for this assay was 0.05 µM for AZT and AZT-G (26 ng/ml and 50 ng/ml, respectively); however, the limit of quantitation for both AZT and AZT-G was 0.5 µM (CV 9.9 and 6.8%, respectively) due to limitation in the standard curve range. Average r2 for the standard curves was 0.997 ± 0.006 for AZT-G and 0.994 ± 0.001 for AZT (mean ± S.D., n = 6). Extraction efficiency was performed by comparing standards prepared in microsomal assay buffer, cell culture media, and perfusion media to standard curves in blank (phosphate-buffered saline) matrix. Extraction efficiency for AZT (104 ± 3%, mean ± S.D.) and AZT-G (86 ± 1%, mean ± S.D.) was consistent across all matrices tested. Diode array spectral comparison demonstrated that the products derived from both placental and liver incubations had the same spectra as pure, synthetic standards (data not shown). Using liquid chromatography/mass spectrometry, a full mass scan (100–1000 m/z) detected AZT (266 m/z) and AZT-G (442 m/z) from both human liver and placental incubations (data not shown).
Data Transformation and Statistical Analysis. AZT transfer was analyzed using eq. 1: CL (maternal-fetal) = [(CFV – CFA) x QF]/CMA, where CFV is the concentration in the fetal vein, CFA is the concentration in the fetal artery, QF is the flow rate in the fetal circuit, CMA is the average concentration in the maternal artery, and CL is clearance. For transfer in the opposite direction, the inverse equation was used, where maternal arterial and venous concentration and flow rate were divided by the fetal concentration (Cannell et al., 1988). This equation is appropriate for compounds that reach equilibrium in the perfusion model. AZT-G and UDPGA transfer were assessed with eq. 2: CL = (K x V)/C. Here, K is the slope of drug appearance in the recipient reservoir, V is the recipient perfusate volume, and C is the mean drug concentration in the donor circuit. This equation is appropriate for compounds that do not reach equilibrium in the model (Schenker et al., 1987). All graphical and statistical analyses were performed using GraphPad Prism 3.0.
|
Results |
|---|
|
TopAbstractMaterials and MethodsResultsDiscussionReferences |
|---|
; Cannell et al., 1988). Although the ß-hCG production we observed was within one standard deviation of that reported for viable, oxygenated perfusion (Cannell et al., 1988), the standard deviations in previously published data were much larger than our figures. This is consistent with the findings of Boal et al. (1997), who demonstrated that with 4 h of perfusion, AZT caused a significant reduction of ß-hCG secretion compared with nonexposed placentas.
|
The conversion of AZT to its glucuronide by JEG-3 cells cultured with either 10% FBS or 0.1% BGG are presented in Fig. 1. JEG-3 cells glucuronidated AZT in vitro when cultured with 10% FBS (Fig. 1, A, C, and E), but when cultured in serum-free medium (supplemented with 0.1% BGG), glucuronidation was almost completely abolished (Fig. 1, B, D, and F). The conversion of AZT to its glucuronide in primary placental explant cultures was variable (0–3 µM) and did not differ significantly between first trimester and term placental cultures at any treatment concentration (1, 10, or 100 µM AZT; Fig. 1).
|
The enzyme kinetics of AZT glucuronidation were studied in triplicate in samples prepared by pooling microsomes from n = 12 term placentas and n = 16 first trimester placentas. Results were fitted to both hyperbolic and two-site binding models and assessed with an F test. Michaelis-Menten kinetics proved to be most appropriate in both cases. There was no significant difference in kinetic parameters between first trimester and term placental microsomes: Vmax 1st trimester = 15.8 ± 1.7 pmol/min/mg protein, Km = 9.7 ± 2.0 µM, whereas Vmax term = 19.97 ± 0.8 pmol/min/mg protein, Km = 10.6 ± 1.2 µM (Fig. 2).
|
In the perfusion model, AZT was very rapidly transferred across the placental barrier (Fig. 3). Peak AZT concentrations were observed in the maternal circuit at 5 min, whereas measurable AZT concentrations appeared in the fetal circuit as early as 5 min but consistently at 10 min after the introduction of the drug to the system. No significant difference in AZT concentrations between maternal artery and maternal vein or fetal artery and fetal vein at any time point was observed (Fig. 3, A and B). The concentrations of AZT in the maternal and fetal circuits reached equilibrium after approximately 3 h of perfusion.
|
There was no significant difference in the transport of the parent drug AZT in the fetal direction compared with the maternal direction (P = 0.68, t test; Table 2). After 4 h, when the perfusion was terminated, 51 ± 0.27% (mean ± S.E.M., n = 5) of the dose of AZT administered had entered the fetal circulation, and the perfusion was at equilibrium.
|
AZT-G was measurable in the maternal circuit within 5 min of the introduction of AZT in two of five placentas tested. However, in the fetal circuit, AZT-G first appeared at 10 min in one perfusion and was not consistently observed until 30 min of dual, recirculating perfusion. AZT-G accumulated over time in the maternal artery circuit until 90 min, at which time equilibrium was reached (Fig. 3C). In the maternal venous circuit, significant time-dependent accumulation was observed until 30 min, at which time equilibrium was reached. No significant time-dependent increase in AZT-G concentration was observed in either of the fetal circuits (Fig. 3D). There was no significant difference between maternal artery and maternal vein concentrations of AZT-G at any given time point or between fetal artery and fetal vein concentrations at any given time point (Fig. 3C).
Mean transport of AZT-G was significantly higher in the fetal-maternal versus the maternal-fetal direction (P < 0.05, t test; Table 2). Preferential transport of the AZT-G metabolite out of the fetal compartment was the net result of the difference in transport rate. Upon termination of perfusion, 2.0 ± 0.9% (mean ± S.E.M., n = 5) of the original dose of AZT had been converted to the glucuronide form. Furthermore, at this time, the concentration of AZT-G in the maternal circuit was 2-fold that in the fetal circulation (0.22 ± 0.06 µM versus 0.11 ± 0.05 µM in the maternal and fetal circuits, respectively).
The transplacental transfer of UDPGA in the human placenta in the perfusion model was rapid and shifted toward the fetal circuit. Concentrations in the maternal circuit began to fall at 5 min post-UDPGA addition, and transfer was almost complete at 4 h (Fig. 4A). A significant decline in concentrations of 14C-UDPGA in the maternal circuit was observed with time (P < 0.001 by ANOVA for both maternal artery and maternal vein to 240 min), whereas a significant increase was recorded in the fetal circuit concentrations of the cofactor with time (P < 0.01 fetal artery and P < 0.05 fetal vein by ANOVA, to 240 min; Fig. 4B). A peak in UDPGA concentration was observed at 60 min in the maternal vein. A corresponding increase in UDPGA concentration in the maternal artery was also observed at the next time point (90 min). After this time, the net flow of UDPGA was toward the fetal compartment. Thus, the rate of UDPGA transport across the placenta was significantly greater toward the fetal compartment (P < 0.05, t test; Table 2).
|
|
Discussion |
|---|
|
TopAbstractMaterials and MethodsResultsDiscussionReferences |
|---|
; Schenker et al., 1990; Boal et al., 1997; Oliviero et al., 1999).
Interestingly, our data suggest that AZT at high concentrations may decrease its own metabolism in the human placenta. In these studies, AZT glucuronidation was significantly decreased in JEG-3 cells and in term placental explant cultures cultured in the presence of 100 µM AZT. A decrease in glucuronidation is not unexpected becuase AZT is cytotoxic; however, the magnitude of the decrease in metabolism suggests that other mechanisms may also play a role. Serum was also shown to affect AZT glucuronidation in vitro. Serum withdrawal almost completely eradicated glucuronidation in both JEG-3 cells and primary cultures. This may be due to serum-responsive elements in the UGT gene structure or, alternatively, to the absence of essential cellular survival factors which are present in serum. Although these studies alone are not strong enough to support these statements, previous research from our laboratory has demonstrated decreased UGT activity in placentas exposed to AZT, an effect which is amplified in the absence of serum (Collier et al., 2003). However, it is unlikely that protein binding effects would cause changes in glucuronidation since AZT is only 25% protein bound in human plasma (Holford and Benet, 1995); these results may also be explained by saturation of UGT activity in the higher concentration ranges, and in vitro data presented by Barbier et al. (2000) would indicate that this is at least partly the case.
The enzyme kinetic profile of AZT in placental microsomes did not differ significantly between term and first trimester placenta. The derived Km values suggest that a high-affinity, low-capacity enzyme is responsible for glucuronidation of AZT. Both Michaelis-Menten and two-binding-site models were compared for goodness of fit, but the Michaelis-Menten model proved to be most appropriate (F test). Furthermore, Eadie-Hofstee plots were linear, suggesting the involvement of only one enzyme (or a dominant isoform). These data support the single-isoform paradigm of Barbier et al. (2000), who demonstrated that only UGT2B7 glucuronidated AZT in UGT-transfected HEK297 cells. The Km data derived from our studies are lower than those published, although this may be a function of the different assays used to assess UGT activity. The involvement of a low-capacity, high-affinity UGT may also explain the failure of previous researchers to quantify glucuronidation of AZT in the human placenta. Studies of glucuronidation performed in vitro have previously been of limited value due to consistent underestimation of the actual rates of glucuronidation in vivo (Miners et al., 2000).
The quantification of AZT glucuronidation by the human placenta, although contradictory to other studies, is extremely plausible in light of the low level (2%) of metabolism observed and the high sensitivity of the assay used. Studies by previous researchers may have furnished negative results due either to the short duration of perfusion (1 or 2 h; Liebes et al., 1990, 1992; Bawdon et al., 1992) compared with our experimental design (4 h), the lower concentration of AZT used by others (commonly 3 µM; Liebes et al., 1990; Schenker et al., 1990), or an open maternal circuit for perfusion which would not allow for cumulative levels of the metabolite to be measured when AZT-G is preferentially transferred in the maternal direction (Schenker et al., 1990; Bawdon et al., 1992). Alternatively, it has been demonstrated that the enzyme ß-glucuronidase is present and active in the human placenta (Collier et al., 2002b). Hence, it is possible that regardless of methodological approach, low or undetectable rates of glucuronidation may be due to cleavage of the glucuronidated product back to parent drug. On a whole-organism level, a 2% contribution by the placenta to glucuronidation is unlikely to be important in terms of drug clearance. Other effects of AZT on the placenta including apoptosis and alterations in the expression and activities of other metabolizing enzymes have been described and may be important in maintaining the structural and functional integrity of the placental barrier (Collier et al., 2003).
Transport and clearance characteristics for the AZT-G metabolite showed significantly greater clearance from the fetal compartment compared with the maternal compartment. The transport characteristics of AZT-G (detectable in the maternal circulation before the fetal circulation, 2-fold higher concentrations in the maternal circuit after 4 h of perfusion, and slow rate of transfer) were characteristic of the elimination of polar compounds by means of extracellular pores and/or paracellular pathways. The maternal-fetal transfer rate observed for the metabolite AZT-G in these studies (0.2 ± 0.009 ml/h/g placenta) is also similar to that reported by Liebes et al. (1992) in studies of transfer with synthetic AZT-G (0.29 ± 0.06 ml/h/g placenta), demonstrating that the metabolite-produced and synthetically derived forms are dispositionally similar. We do not know why this apparent equilibrium in production of AZT-G was reached. It is possible that UGT activity is saturated by the concentration of AZT used in these studies, or that the placenta loses enzyme activity over the course of the perfusion (although this is unlikely since metabolic viability was demonstrated).
Although transport of UDPGA across the human placenta showed a significant preference toward the fetal compartment and was almost complete after 4 h of perfusion, a small transient peak in the maternal circuit was observed at 60 min (Fig. 4). Although the reason for this is unknown, more than one UDPGA transporter exists and transfer occurs at multiple intracellular sites. Also, nucleotide sugar transporters are promiscuous for many biological sugars, and it is possible that multiple biological transporters are able to convey UDPGA (Kawakita et al., 1998). In the human body, the Km for UDP-sugar transporters is around 10 µM, meaning that the concentrations of UDPGA used herein fall within physiological range for multiple sugar transporters (Kawakita et al., 1998). Furthermore, if UDPGA was initially transported out of the fetal compartment by a protein for which it is a minor substrate, it is conceivable that competitive inhibition of this protein occurred (Ki values for UDP-sugar transporters are 1–5 µM; Kawakita et al., 1998), resulting in net transfer toward the fetus by the principal transporter. Two UDP-sugar transporters, Frc and hUGTrel7, have been identified and characterized thus far; one (Frc) is integrally associated with embryonic development (Goto et al., 2001; Selva et al., 2001), and both were postulated to be involved with catalysis of glucuronidation (Muraoka et al., 2001). Complete transport of UDPGA into the fetal compartment, as observed here, is consistent with both the cofactor and its transporter being important for embryogenesis and morphogenesis. In addition, the extremely fast rate of transport of UDPGA across the placenta demonstrates that the rate-limiting step for glucuronidation in vitro is not supply of the cofactor. It has been shown previously that glucuronidation in the placenta is not constrained by constitutive levels of intracellular UDPGA, since it is present at high concentrations, approximating those in the human kidney (Cappiello et al., 2000), and it is more likely that low levels of enzyme expression and turnover are responsible for low activity. Despite the argument that UDPGA is not rate-limiting for glucuronidation, the absolute amount present in tissues is not a reliable indicator of UGT activity. For instance, the fetal liver was also reported to contain UDPGA even though UGTs are inactive in the fetal liver until after birth (Onishi et al., 1979; Kawade and Onishi, 1981; Coughtrie et al., 1988). It has been demonstrated that fetal relative to adult organ concentrations of UDPGA are very low (approximately 20 and 60% of the adult liver and kidney, respectively) and suggested that this may be a limiting factor for fetal glucuronidation (Cappiello et al., 2000). Therefore, the ability of the placenta to transfer UDPGA to the fetus, as demonstrated here, would imply a role for the placenta in facilitating enzyme activity and development. As a caveat to these results, we did not confirm that the 14C-UDPGA in the circuits remained intact. Although there would not be active esterases or phosphatases in sterile M199, the we acknowledge that the placenta does contain, at the very least, active carboxyl esterases (Yan et al., 1999) and that residual blood contamination (of which there is very little) may contain phosphatases and esterases, thus resulting in UDPGA breakdown upon exposure to the syncytium. The amount of activity of these enzymes and their specificity for UDPGA is unknown at this time.
In conclusion, although we have demonstrated the ability of the placenta to glucuronidate AZT in low amounts (approximately 2% of a dose), our results support previous reports that the human placenta does not provide a barrier to the transport of AZT across the maternal-fetal interface and that transport is bidirectional. This indicates that the fetus would be exposed to AZT in concentrations approximating those in the maternal serum and that therapeutic failure of the drug is not caused by placental obstruction of drug transfer. The finding that AZT-G is preferentially transferred out of the fetal compartment at rates that are significantly higher than transfer into the compartment is supportive of protective clearance in the placenta whereby polar metabolites are eliminated. It also confirms our earlier assertion, based on the activity localization and expression of the UGT enzymes, that glucuronidation in the placenta is primarily fetoprotective (Collier et al., 2002a,b). Finally, we have described the preferential transplacental transport of UDPGA into the fetal compartment, a novel finding which may have important pharmacological and developmental implications. We anticipate with great interest further work defining the cellular location and functional characteristics of both hUGTrel7 and Frc in the human placenta and their role in embryogenesis.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
ABBREVIATIONS: AZT, 3'azido-3'-deoxythymidine; AZT-G, 3'azido-3'-deoxythymidine glucuronide; UDPGA, uridine diphosphate glucuronosyltransferase; UGT, uridine diphosphate glucuronosyltransferase; BGG, bovine gamma globulin; 4-MUG, 4-methylumbelliferone ß-D-glucuronide; ß-hCG, ß-human chorionic gonadotropin; FBS, fetal bovine serum; HPLC, high performance liquid chromatography; ANOVA, analysis of variance; CL, clearance.
Address correspondence to: Abby Collier, University of Nevada, Reno, MS 199, Reno, NV 89557. E-mail: abbyC{at}unr.edu
|
References |
|---|
|
TopAbstractMaterials and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  N. Antony, J. J. Bass, C. D. McMahon, and M. D. Mitchell Myostatin regulates glucose uptake in BeWo cells Am J Physiol Endocrinol Metab, November 1, 2007; 293(5): E1296 - E1302. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
 |
 |
 |
|
 |
 |
 |
