| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Original Studies |
Royal Postgraduate Medical School, Institute of Obstetrics and Gynaecology, Queen Charlotte’s and Chelsea Hospital, Goldhawk Road, London W6 0XG United Kingdom
Address all correspondence and requests for reprints to: Rekha Bajoria, Royal Postgraduate Medical School, Institute of Obstetrics and Gynaecology, Queen Charlotte’s and Chelsea Hospital, Goldhawk Road, London W6 0XG United Kingdom. E-mail: rbajoria{at}rpms.ac.uk
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The transmembrane transfer of radiolabeled and commercial TRH were comparable across both preterm and term placenta. When transport was studied from the maternal to fetal side, the maternal concentration of TRH declined rapidly from 100% at time 0 to 19.31 ± 2.26% at 8 h with a concomitant increase in the fetal concentration from undetectable to a maximum of 2.56 ± 0.38% with a fetomaternal ratio of 0.16 ± 0.01. Transfer of TRH from the fetal to maternal compartment was similar to that of maternal to fetal. Chromatography of maternal and fetal media showed that TRH was metabolized by the membrane into small molecular weight fragments. Treatment of the membrane with p-HMSA increased TRH transport from the maternal to fetal compartment to 18.12 ± 0.91 (P < 0.001) with an fetomaternal ratio of 0.35 ± 0.02 (P < 0.001). Although transmembrane transfer of TRH from the fetal to maternal side was also increased by p-HMSA, levels achieved were less than that from maternal to fetal (12.26 ± 1.50%; P < 0.05). These results suggest that the human fetal membrane acts as an enzymatic barrier to the bidirectional transfer of TRH from 24 weeks gestation.
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Accordingly, we hypothesized that TRH, a tripeptide with a molecular weight of 330 Da, crosses the chorioamniotic membrane, from which it would then be transferred to the fetal circulation from amniotic cavity by fetal swallowing. Oral administration of TRH has been shown in nonpregnant adults to produce a sizable thyrotrophic response, greater in fact than that obtained following intravenous injection (15, 16, 17). In pregnancy, intraamniotic administration of thyroid hormones to correct fetal thyroid deficiency state is effective in maintaining fetal levels, indicating that at least in late gestation, fetal swallowing can account for transfer of thyroid hormones from amniotic fluid to the fetal circulation (18). To elucidate alternative pathways of transfer of TRH from the maternal to fetal circulation, we studied the transfer and metabolism of TRH across the preterm and term human fetal membranes.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Membrane preparation
Intact human fetal membranes (amnion, chorion, and decidua) were obtained from uncomplicated normal term (37–42 weeks gestation) or preterm pregnancies (24–32 weeks gestation) immediately after vaginal or cesarean delivery. The reflected portion of the membranes were collected by trimming it away from the site of rupture and approximately 1 cm from the chorionic plate. Membranes were washed in PBS containing 1% penicillin-streptomycin-gentamicin to remove blood. Intact membrane was mounted over the end of a 15-mm wide hollow glass cylinder and held in place with silicon O-rings (13). Excess membranes around the rim of the glass cylinder were trimmed. Silicon rubber washers were used to suspend the experimental system in a 12-well tissue culture plate (ICN Biomedicals, Costa Mesa, CA) in such a manner that membranes were completely immersed in the culture media without touching the bottom of the plate. Culture media was comprised of TC-199 containing sodium bicarbonate, 10% ITS (insulin 12.5 mg, transferrin 12.5 mg, selenium 12.6 µL, linoleic acid 11.1 µL, and BSA 2.5 g in 20 mL), 2 mmol glutamine, and 1% penicillin-streptomycin. Cylinders were placed so that the choriodecidua (maternal) side faced downward. Culture media (1.5 mL) was added to each side of the membrane to produce a two-compartmental experimental system. Membranes were incubated overnight at 37 C under 95% O2 and 5% CO2. The tissue culture media was removed, and membrane preparations reincubated with fresh media containing the study drug. At each time point, cell viability of the membrane was determined by typan blue negative staining (12). Heparin and creatinine were used as within experimental controls to normalize the transfer rate of TRH for membrane-related variables and thus reduce interexperimental variability.
Experiment protocol
One hundred nanograms synthetic TRH (n = 4) or 125I-labeled TRH (100,000 cpm; n = 4), 650 U heparin, and 300 µg creatinine were added to either the maternal or fetal compartments. In three further experiments, transfer across the preterm membrane was studied by adding 125I-labeled TRH. Spontaneous degradation of TRH at 37 C for 24 h was determined by incubating 125I-labeled TRH (100,000 cpm/mL) in culture media alone with no membrane present.
To elucidate the enzyme(s) responsible for metabolism of TRH by the fetal membranes, eight experiments were undertaken with 200 µM of the dipeptidase enzyme inhibitor p-HMSA added to the maternal or fetal compartment 30 min before addition of unlabeled (n = 4) or 125I-labeled TRH (n = 4). Similarly, four sets of preterm membranes were incubated with 200 µm p-HMSA before addition of 125I-labeled TRH.
Membranes were incubated in triplicate for 0.5–8 h. Bidirectional transfer of TRH across the membrane was determined by taking samples at 0.5, 1, 2, 4, 6, and 8 h. Uptake of TRH was measured by homogenizing the membrane in an ultraturrax high-speed homogenizer (Citenco Ltd., Borehamwood, Herts, UK) in 20 mL PBS buffer at room temperature. Ten-milliliter aliquots of the homogenized tissue were centrifuged at 3000 x g for 15 min, and TRH concentration determined in the supernatant.
Transmission electron microscopy
Samples of membrane at time 0 and 12 h post-TRH incubation were fixed for 1–2 h in 3% glutaraldehyde in 0.1 M cacodylate buffer, pH 7.2, washed in fresh buffer, and postfixed in 1% osmium tetroxide. Samples were dehydrated through a graded series of alcohol and embedded in Araldite epoxy resin (Taab Laboratories Equipment, Reading, Berkshire, UK). Ultrathin sections were cut perpendicular to the plane of the membrane using a diamond knife. Sections were stained in uranyl acetate and lead citrate before examination in an Hitachi HU 12A transmission electron microscope (Hitachi, Wokingham, Berkshire, UK) operated at 50 KV.
Analytical methods
Column chromatography on Sephadex G-10 was used in the radiolabeled experiments to separate TRH from its metabolites in maternal and fetal culture media. One milliliter of maternal, fetal, and placental homogenate was applied to the column calibrated with 0.05 M phosphate buffer. The void volume was 5 ml. The column was eluted with phosphate buffer, and 100-µL fractions collected. Similarly, 1 mL 125I-labeled TRH in phosphate buffer containing 15,000 cpm was also applied to the column, and 100-µL fractions collected. The amount of radioactivity of each fraction was determined.
The amount of radioactivity at time 0 of incubation was considered as
the reference value. At each sample point, residual TRH was expressed
as a percentage of the reference. In 125I-labeled TRH
experiments, a known volume of all the samples were counted for 2 min
on a 
-counter. Uptake of 125I-labeled TRH by the
membrane was determined by counting the membrane disk at each sample
point for 2 min.
The concentration of unlabeled TRH was quantified by a sensitive and specific immunoassay (3), with a lower limit of detection of 10 pg in a 300 µL ethanol extract. Cross-reaction of the TRH analog His-pro-NH2, p-Glu-pro, p-Glu-His, p-Glu-His-Pro with anti-TRH antibody was consistently <0.0001%. Mean recovery of nonradioactive TRH from serum and tissue extract over the concentration range 10–5000 pg was >95%. Mean coefficients of variation for different TRH concentrations in sera and tissue extract were 2.5–12.7%. Mean intraassay coefficient of variation of TRH was 10%.
The concentration of heparin was measured colorimetrically (19) with a coefficient of variation of 10–14%. The lower limit of detection was 0.1 IU/mL. Creatinine concentration was determined by colorimetric assay (20), with a coefficient of variation of 7–12%. The sensitivity of the assay was 0.1 mg/mL.
Data analysis
Values were expressed as mean ± SEM as a percentage of initial dose added unless otherwise indicated. Data between two groups as a function of time were compared by two-way ANOVA. One-way ANOVA was used to compare blocked variables between groups. P values of <0.05 were considered significant. Maternal (MAUC) and fetal (FAUC) area under the curve concentration of TRH was calculated by the trapezoidal rule (19). The transfer rate of TRH was determined by fitting a weighted line for the standard deviations of means of the fetal concentration of TRH between 30–480 min. The transfer rate of TRH was calculated from the slope of this line.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
) was characterized by a highly
microvillous surface, greatly infolded lateral plasma membranes with
variable amounts of dilatation, and a prominent basement membrane.
Numerous cytoplasmic vesicles, bundles of tonofilaments, and lipid
droplets were present in the cytoplasm throughout the duration of the
experiment. The structure of the chorionic trophoblast epithelial cells
(Fig. 1) was not adversely affected by the experimental procedure. The
plasma membranes and cytomembrane system of the cells remained intact,
and the mitochondria were also unaffected.
|
Maternal to fetal.
When TRH was added to the maternal
compartment, the concentration of TRH declined to 25.45 ± 0.39
ng/mL with a MAUC of 386 ± 3 ng min/mL (Fig. 2). Fetal levels increased to 1.86
± 0.04 ng/mL (slope = 0.004; r = 0.99) with a fetal/maternal
(F/M) ratio of 0.08 ± 0.01 at 8 h (Fig. 3). The FAUC and FAUC/MAUC ratio were
8 ± 1 ng min/mL and 0.02 ± 0.01, respectively. The transfer
rate of TRH was 2 pg/min. The transfer of heparin and creatinine was
1.01 ± 0.02 and 18.32 ± 1.04% of initial dose,
respectively.
|
|
), and MAUC levels
(P < 0.05), fetal concentration (P <
0.001; slope = 0.04; r = 0.98), FAUC (P <
0.01), F/M ratios (P < 0.01), and transfer rate (40
pg/min) were increased (Fig. 3). The transfer of heparin (1.11 ±
0.01%) and creatinine (18.73 ± 1.24%) was comparable with
experiments without p-HMSA.
Fetal to maternal transfer.
When TRH was added to the fetal
compartment, fetal levels at 8 h (52.13 ± 0.29 ng/mL) (Fig. 2), FAUC (510 ± 10 ng min/mL), and FAUC/MAUC ratios (0.02 ±
0.01) were also higher than maternal values in the maternofetal
transfer experiments. However, maternal TRH concentration at 8 h
(1.78 ± 0.05 ng/mL; slope = 0.003; r = 0.97), M/F ratio
(0.03 ± 0.01) (Fig. 3), and MAUC (8.83 ± 0.25 ng min/mL)
were comparable with fetal levels in maternal-fetal transfer group. The
transfer rate of TRH was also similar (4.6 pg/min). The transfer of
heparin (1.34 ± 0.03%) and creatinine (14.36 ± 1.14%) was
comparable with the maternofetal transfer experiments.
However, when the membranes were pretreated with p-HMSA, fetal
(slope = 0.02; r = 0.91; P < 0.001) (Fig. 2)
and maternal levels of TRH (P < 0.01), FAUC
(P < 0.01), MAUC (P < 0.05),
MAUC/FAUC ratio (P < 0.01), and M/F ratio
(P < 0.001) (Fig. 3B) were significantly higher than
in the untreated group. The rate of transfer of TRH increased by 5-fold
(20 pg/min). Transfer of heparin and creatinine in these experiments
was 1.33 ± 0.07% and 15.64 ± 1.20% initial dose,
respectively.
Transfer of radiolabeled TRH
There was no difference in the mean maternal and fetal
concentrations in experiments with term (n = 4) or preterm
membranes (n = 4) (Table 1), and
therefore data from the two groups were pooled as shown in Fig. 4.
|
|
and Table 1) fetal concentration
(2.56 ± 0.38%; slope = 0.004; r = 0.99), FAUC,
FAUC/MAUC, F/M ratio (Fig. 3C) were similar to the commercial group of
experiments. The transfer rate of TRH was also comparable (1.16 ±
0.01 pg/min). Membrane uptake of TRH was 5.02 ± 0.75%. The
transfer of heparin and creatinine in these experiments was 1.05
± 0.01% and 20.13 ± 0.80% of initial dose, respectively (Fig. 4).
When membranes were pretreated with 200 µM p-HMSA, total
radioactivities in the maternal compartment, fetal, MAUC, FAUC, and F/M
ratio were comparable with untreated group of experiments (Table 1 and
Fig. 5). However, concentration of intact
TRH in the maternal (P < 0.01) fetal compartment
(slope 0.03; r = 0.98; P < 0.001) was higher with
a transfer rate of 30 pg/min (Fig. 4). Similarly, F/M ratio
(P < 0.001), MAUC (P < 0.001), FAUC
(P < 0.001), and FAUC/MAUC ratio (P <
0.001) of intact TRH were higher than in the untreated group (Table 1).
The transfer of heparin (1.12 ± 0.04%) and creatinine
(21.73 ± 2.14%) was comparable with control experiments.
Pretreatment with p-HMSA was associated with significant reduction in
the TRH metabolites (P < 0.001) (Fig. 5)
|
) and FAUC
(P < 0.05) were higher than the maternal levels in the
materno-fetal transfer experiments, values were essentially similar to
those with commercial TRH (Fig. 5). Net transmembrane transfer of
intact TRH (slope = 0.004; r = 0.96), M/F ratio (Fig. 3),
MAUC, MAUC/FAUC ratio, and transfer rate of TRH (1.4 pg/min) were
comparable with those of materno-fetal transfer of
125I-labeled TRH and commercial TRH group (Table 1). The
transfer of heparin (1.06 ± 0.03%) and creatinine (17.14 ±
1.31% dose) was similar between groups.
When membranes were incubated with p-HMSA, the total radioactivity of
TRH in the maternal, fetal compartment as well as F/M ratio, MAUC, and
FAUC were comparable with the untreated group. The fetal concentration
of intact TRH (slope = 0.02; r = 0.99; P <
0.01) (Fig. 4), FAUC (P < 0.05 ), maternal
concentration (slope = 0.02; r = 0.99; P <
0.001), M/F ratio (P < 0.01) (Fig. 3), MAUC
(P < 0.001), and MAUC/FAUC ratio (P <
0.05) were all higher than in the untreated group of experiments (Fig. 5). The transfer rate of TRH was also increased by 6-fold (12 pg/min).
However, these values were comparable with those obtained from
nonlabeled TRH groups of experiment. The transfer of heparin (1.15
± 0.05% dose) and creatinine (19.22 ± 2.64% dose) were similar
to the control group. These results suggest that pretreatment with
p-HMSA increases transfer of intact TRH by minimizing its
metabolism.
Metabolism of TRH.
The chromatogram of the maternal
samples in materno-fetal transfer experiments (Fig. 6) showed multiple peaks. The prominent
peak, occurring in fractions 40–50 after the void volume was because
of TRH as it corresponded to the intact TRH peak. Subsequent multiple
peaks were caused by fragments of TRH. The concentrations of intact TRH
and total small molecular weight TRH fragments were 33.53 ±
1.20% and 48.84 ± 1.71% respectively. Although a chromatogram
of the fetal samples showed a number of small peaks, the distinct
single peak in fractions 40–50 was attributed to intact TRH (1.84
± 0.41%), suggesting that TRH is predominantly present in the fetal
compartment in the intact form. This was further evident from the
chromatogram of the fetal and maternal samples from feto-maternal
transfer experiments, which showed multiple peaks present in the
maternal compartment (5.52 ± 0.51% caused by TRH fragments and
0.63 ± 0.14% caused by intact TRH) compared with a distinct
single peak in the fetal samples. However, when membranes were
incubated with p-HMSA, a single peak attributed to intact TRH
predominated in all samples in both maternal to fetal and fetal to
maternal experiments (Fig. 6).
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
We investigated the possibility of an alternative pathway apart from the placenta by which TRH might enter into the fetal compartment. We used an in vitro dual chamber model of fetal membrane containing both amnion and chorion. Although not extensively used to study drugs or hormone transfer across the chorioamniotic membranes, culture conditions to preserve membranous cellular integrity have been optimized in our laboratory (7, 12, 13, 14). Furthermore, various investigators have shown similar systems capable of endogenous production of interleukins, PGs, and PRL (21, 22, 23, 24, 25). We have also recently shown that creatinine, antipyrine, and TSH cross the membrane, and the rate of transfer of these drugs depends on their molecular weight and lipid solubility (10). In this study, we further established the ultrastructural viability of this system to study transmembrane transfer of small molecules. Organelles in both amnion and chorion were preserved with no evidence of intracellular damage or cellular oedema during 12 h of culture under experimental conditions. Furthermore, mitochondria in the chorion showed well-organized cristae, indicating that membrane cells remained viable after 12 h incubation. We also used creatinine and heparin as a internal markers to exclude spurious increases in transfer of TRH from either mechanical discontinuity of the membranes or the presence of nonviable cells. We included only experiments in which heparin transfer at 8 h was <1.5%, and the cultured membrane disk stained negatively with trypan blue.
Using this system, we studied transfer of natural TRH both from maternal to fetal and fetal to maternal sides. Our data suggest that TRH, despite its low molecular weight (330 Da) failed to cross the membranes in significant quantity. In the same experiments, transmembrane transfer of creatinine (molecular weight <113 Da) was markedly higher than TRH, which excludes inadequate contact of the membranes with the culture fluid as an explanation for minimal transfer. Although the pharmacological characteristics of TRH such as its lipid solubility, degree of ionization (pKa 6.8), and amphoteric nature would otherwise suggest free transfer across the membrane (26), the rate of transfer we found was comparable with that of the hydrophilic polar macromolecule, heparin.
Metabolism by the membrane might also explain minimal transmembrane passage of TRH. It is well established that both human amnion and chorion can convert T4 by the inner ring deiodinase enzyme to reverse T3 (11). Cytokines like interleukin-1 and PGE2 are also known to be degraded by membranes (12, 13, 14). Therefore, the minimal transmembrane passage of TRH could be caused by its degradation into smaller molecular weight fragments. We used an RIA with minimal cross-reactivity to known metabolites to quantitate TRH and found a total recovery of TRH of <50%. This suggests metabolism by the membrane as the likely cause of poor passage of TRH across the membrane.
We also used 125I-labeled TRH to study metabolism of TRH by the membrane because it has the added advantage of allowing quantitation of net transfer of TRH over and above endogenous levels. Radiolabeling TRH did not alter its biological properties, because transfer of intact 125I-labeled TRH was comparable with the synthetic TRH molecules. The maternal and fetal chromatograms indicate that TRH was degraded into smaller molecular weight fragments. Although these metabolites were predominantly found in the maternal compartment, small proportions were also present in the fetal compartment. It is unlikely that fragments of TRH represent spontaneous degradation because >95% of TRH was present in intact form when incubated at 37 C for 8 h in culture media in the absence of membranes. Although no attempt was made to characterize the various fragments of TRH, we speculate that dipeptides containing His-pro or His-Pro-NH2 were the major degradation products. Because the His-molecule of TRH was labeled with 125I, we were unable to exclude formation of non-His- containing TRH metabolites by the membrane.
An attempt was made to identify the enzyme(s) responsible for TRH degradation by the amnion and chorion. TRH is known to be metabolized by nonspecific cytosolic enzymes such as dipeptidase and pyroglutamyl peptidase I and by specific membrane-bound enzymes such as pyroglutamyl peptidase II (27, 28). To elucidate the enzyme responsible, we studied the effect of p-HMSA, a known dipeptidase and pyroglutamyl peptidase I inhibitor, on transmembrane passage of TRH. The increased bidirectional transfer of TRH in the presence of p-HMSA is unlikely to be caused by membrane damage because transfer of heparin and creatinine were similar to the control group. This is further substantiated by the marked reduction in TRH metabolites on maternal and fetal chromatograms. Hence, our data suggest that minimal transmembrane transfer of TRH is most likely caused by its metabolism by dipeptidase/type-1 pyroglutamyl peptidase. Furthermore, because we failed to find any difference in transfer of TRH between preterm and term membranes, it indicates that TRH degrading enzymes are present in fetal membrane from 24 weeks gestation.
Although we made no attempt to determine the precise location of the enzymes, some inferences can be drawn. Despite comparable transfer rates in either direction, metabolites were predominantly found in the maternal compartment. This, together with the higher fetal concentration of intact TRH in the fetal-maternal transfer than maternal levels in materno-fetal group, suggests that dipeptidases are predominantly present on the chorion rather than amnion.
In conclusion, our data indicate that human fetal membranes, like term placentae, prevent free passage of TRH because of metabolism by type 1 peptidases. Furthermore, the membranes form a composite enzymatic barrier to TRH from as early as 24 weeks, and thus transmembraneous passage is not the mechanism by which maternally administered TRH crosses to the fetus. This, together with our recent observation that TRH does not cross the human placenta from 24 weeks (2, 3, 4), suggests that maternally administered TRH more likely stimulates the fetal pituitary-thyroid axis indirectly via secretion of TSH-like substances by the placenta. Further studies are needed to elucidate the mechanism by which maternal TRH administration elicits a fetal thyrotrophic response.
|
Footnotes |
|---|
Received January 30, 1997.
Revised April 10, 1997.
Revised June 5, 1997.
Accepted June 17, 1997.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
) are expressed as
nanograms per milliliter1.
