Pregnancy: Amino acid concentrations in human embryological fluids

The concentrations of amino acids in samples of coelomic fluid(n = 15), amniotic fluid (n = 9) and maternal serum (n = 15)obtained from normal pregnancies between 7 and 12 weeks of gestationwere measured using reversed-phase chromato-graphy with pre-columnderivatization. The total molar concentration of the 18 aminoacids measured was 2.3 times higher in coelomic fluid than inmaternal serum. All amino acids except serine and tryptophanwere present in significantly higher concentrations in coelomicfluid than in maternal serum. Significant correlations betweenmaternal serum and coelomic fluid were only found for proline,tyrosine and tryptophan, suggesting that levels of the otheramino acids are mainly influenced by placental synthesis anddo not directly depend on maternal amino acid metabolism. Levelsof all amino acids were significantly higher in coelomic fluidcompared to amniotic fluid. Compared to maternal serum, theamniotic fluid contained significantly higher levels of arginine,lysine, alanine and tyrosine and lower levels of serine, glutamineand tryptophan. The total molar amino acid concentration decreasedsignificantly with gestational age in both coelomic fluid andmaternal serum. These results suggest that amino acids accumulatein coelomic fluid to support the metabolism of the secondaryyolk sac, and that the exocoelomic cavity is the reservoir formost nutrients needed by the embryo and early fetus in the firsttrimester of human pregnancy.     

In-vivo study of diazepam transfer across the first trimester human placenta

Diazepam transfer by the first trimester human placenta wasinvestigated at pregnancy termination between 6 and 12 weeksof gestation. Fetal fluid samples were obtained from the exocoelomicand amniotic cavities of 65 pregnancies between 8 and 25 minfollowing the i.v. administration of 0.1 mg/kg diazepam to themother. Diazepam was detected in one-third of coelomic fluidsamples and two-thirds of amniotic fluid samples. Maternal serumand urine diazepam concentrations correlated negatively andpositively respectively, with time from drug injection to sampling.Individual diazepam concentrations were low on the fetal side,and the corresponding concentrations were independent of maternalserum concentrations and the time from drug injection to sampling.Amniotic fluid diazepam content increased significantly withadvancing gestational age. A multiple regression analysis showedthat the diazepam content of the coelomic fluid was not influencedby maternal serum diazepam concentration, the time from druginjection to sampling or gestational age, whereas only gestationalage contributed to the diazepam content of amniotic fluid. Thesedata demonstrate that the placental transfer of diazepam occursfrom week 6 of gestation, indicate a preferential transfer ofthis drug to the amniotic cavity and suggest that diazepam mayaccumulate in fetal circulation and tissues during organo-genesis.     

Relationship between protein concentrations in embryological fluids and maternal serum and yolk sacsize during human early pregnancy

Coelomic fluid (n = 57), amniotic fluid (n = 61) and maternalserum (n=81) were obtained from normal pregnancies between 7.7and 13.9 weeks and assayed for total protein, -fetoprotein (FP),albumin and pre-albumin. The mean concentration of total proteinin matched samples was 18 times higher in maternal serum thanin the coelomic fluid and 54 times higher in the coelomic fluidthan in amniotic fluid. The concentrations of total protein,albumin and pre-albumin decreased and that of FP increased inmaternal serum with advancing gestation. The yolk sac volumeand the concentrations of total protein in the coelomic andamniotic fluids increased with gestational age. No differencewas found for the crown–rump length, yolk sac volume andprotein concentration in the coelomic fluid between two groupspresenting with low and high maternal serum pre-albumin concentrations.Before 11 weeks gestation, significant correlation was onlyfound between yolk sac volume and coelomic fluid concentrationof pre-albumin as evaluated by both electrophoresis and immunonephelometry.These results suggest that during the first trimester of normalpregnancy, the placental metabolism and transfer rate of proteinsis not directly influenced by the concentrations of proteinin the maternal circulation and that the transfer of proteinsthrough the amniotic membrane is limited. These results alsoindicate that during that period the secondary yolk sac maycontribute to the protein content of the exocoelomic cavity,and that the embryo and its yolk sac and subsequently the fetusare the main source of the proteins present in the amnioticfluid.     

Materno-fetal immunoglobulin transfer and passive immunity during the first trimester of human pregnancy

Passive transfer of immunity from the mother to the first trimesterfetus is of particular interest because of the reported highincidence of serious fetal sequelae due to congenital infection.We have examined the relationship between maternal serum, coelomicfluid and amniotic fluid concentrations of immunoglobulin (Ig)and complement. Ig fractions G (IgG), A (IgA), and M (IgM) andcomplement factors 3 (C3) and 4 (C4) were measured in 34 normalpregnancies between 6 and 12 weeks of gestation. The concentrationsof specific antibodies for Toxoplasma gondii, cytomegalovirus(CMV) and rubella were also measured on 21 matched samples fromthe same study group. IgG and IgA concentrations were detectedin all coelomic fluid samples whereas IgG was only measurablein two amniotic fluid samples. IgG and IgA concentrations wererespectively 28 and 128 times lower in coelomic fluid than inmaternal serum and probably reflect fetal serum concentrations.Significant positive linear correlations were found betweengestational age and the coelomic concentrations of IgG (P equals;0.001) and IgA (P = 0.014). There was no obvious associationbetween immunoglobulin concentrations in coelomic fluid andmaternal serum suggesting increasing active transport acrossthe placenta with advancing gestation. IgM, C3 and C4 were notdetected in coelomic or amniotic fluid samples. Specific antibodieswere found in 13 out of 63 samples of coelomic fluid and in32 out of 63 samples of maternal serum. They were found in coelomicfluid only if they were present in maternal serum. These resultssuggest that maternal IgG and IgA are potentially availableto the embryo as early as the 6th week of gestation. The presencein the coelomic fluid of the IgG fraction, both total and infectiousagent-specific transferred via the placenta, indicates thatthey may provide limited protection against congenital infectionin the first trimester.     

Genes or placenta as modulator of fetal growth: evidence from the insulin-like growth factor axis in twins with discordant growth

To determine whether fetal growth is regulated by placental and/or fetal factors, we measured maternal and fetal concentrations of insulin-like growth factor-I (IGF-I), IGF-II and insulin-like growth factor binding protein-1 (IGFBP-1) (total and non-phosphorylated) in dichorionic (DC) and monochorionic (MC) twins with (DC, n = 13; MC, n = 12) or without (DC, n = 13; MC, n = 12) discordant birth weight. In the discordant MC pregnancy, growth-restricted (IUGR) twins had lower IGF-II concentrations (P < 0.001) but similar IGF-I concentrations compared to the appropriate for gestational age(AGA) co-twin. The differences in IGF-II concentrations showed a positive association with percentage birth weight discordance (r = 0.60; P < 0.05) in MC twins. In contrast, IUGR DC twins had lower IGF-I concentrations (P < 0.05) but similar IGF-II concentrations compared to the AGA co-twins. There was a positive correlation between IGF-I concentrations and birth weight (r = 0.47; P < 0.05) in DC twins. Total IGFBP-1 concentrations were higher in both MC and DC IUGR twins (P < 0.05) compared to AGA twins. A negative association was found between total IGFBP-1 concentrations and birthweight of both MC (r = 0.47; P < 0.05) and DC (r = 0.58; P < 0.01) twins. No such differences in IGF concentrations were found between concordant MC and DC twin pairs. The maternal IGF concentrations were comparable between the MC and DC groups. These data suggest that growth discordances of twins exposed to the same maternal environment may be due to variations in either IGF-I or IGF-II/IGFBP-1, depending upon the functioning of the placenta.     

Hepatocyte growth factor concentration in the early second-trimester amniotic fluid does not predict fetal growth at birth.

The purpose of this study was to evaluate whether hepatocyte growth factor (HGF) concentrations in the early second-trimester amniotic fluid predict fetal growth at birth. HGF and insulin-like growth factor-I (IGF-I) concentrations in the early second-trimester amniotic fluid were measured in 12 pregnancies with small for gestational age (SGA) infants, 84 pregnancies with appropriate for gestational age (AGA) infants, and eight pregnancies with large for gestational age (LGA) infants. HGF concentrations were measured from the early second-trimester amniotic fluid samples using an enzyme-linked immunosorbent assay. IGF-I concentrations were measured from the early second-trimester amniotic fluid samples using an immunoradiometric assay. Maternal age in AGA group (34.2 +/- 5.5 years) was significantly lower than in SGA (37.9 +/- 3.0 years) and LGA (37.6 +/- 3.3 years) groups (P < 0.05). There were no significant differences for parity or gestational age at amniocentesis among the groups. There were significant differences for birth age, birth weight, neonatal height, and placental weight among the groups (P < 0.05). HGF concentrations in SGA, AGA and LGA groups were 16.9 +/- 6.6, 16.7 +/- 9.0 and 20.2 +/- 14.8 ng/ml respectively (not significant). There was no correlation between amniotic fluid HGF concentrations and birth weight, height or placental weight. There were also no significant differences for amniotic fluid IGF-I concentrations among the three groups. These results suggest that differences in HGF concentrations in the early second-trimester amniotic fluid do not predict fetal growth at birth. Further study is needed to clarify the role of high HGF concentrations in early second-trimester amniotic fluid during pregnancy.     

Molecular interactions during pregnancy: Distribution of interleukin-6 in maternal and embryonic tissues during the first trimester

Interleukin-6 (IL-6) distribution was investigated in coelomicfluid, amniotic fluid, maternal serum, decidua and placentalvillous tissue collected from 16 normal pregnancies between7 and 12 weeks of gestation. IL-6 levels in fluids and tissueswere measured by an immunoenzymatic assay and mouse monoclonalantibodies specific for IL-6 were used to localize immunohistochemicallyIL-6 in decidual and placental tissue. IL-6 was detected inall samples of coelomic and amniotic fluids and in most extractsof placental and decidual tissues. IL-6 concentration was significantly(P < 0.005) higher in the coelomic fluid than in amnioticfluid and was positively correlated with gestational age. Immunostainingfor IL-6 was present in both syncytiotrophoblast and extra-villoustrophoblast. IL-6 was in significantly (P < 0.001) higherconcentration in decidual than in placental tissues. These dataindicate that IL-6 is normally present in coelomic and amnioticfluids of early pregnancy and that IL-6 concentrations mainlyresult from the accumulation of trophoblastic-derived IL-6.During the first trimester IL-6 could play a role in tissueremodelling associated with placentation but also in the haematopoiesisfunction of the secondary yolk sac and in the generation ofnew vessels in placental villous tissue. coelomic fluid/early placenta/first trimester/interleukin-6/pregnancy     

Comparison of placental isoferritin levels in maternal serum and coelomic and amniotic fluids during first trimester human gestation

Placental isoferritin (PLF) is present in serum at high concentrations throughout normal gestation. The current study compared PLF concentrations in first trimester maternal serum with those of amniotic and coelomic fluids in 25 healthy, pregnant women. Use of a specific enzyme-linked immunosorbent assay indicated concentrations high (22 +/- 3 U/ml) in maternal serum, whereas significantly lower values were detected in both coelomic and amniotic fluids (5 +/- 2 and 3 +/- 2 U/ml, respectively, P < 0.005). It was further demonstrated that when pure PLF was added to amniotic and coelomic fluids, the PLF levels remained low in each compartment, suggesting that bioproducts produced inside the gestational sac may interfere with the PLF immunoassay.      

Follistatin and activin A in extra-embryonic coelomic and amniotic fluids and maternal serum in early pregnancy

Follistatin is a specific binding protein which controls bioavailability of activins and inhibins which have an important role in fetal development. In the first trimester of pregnancy bioactive dimeric inhibins are found at high concentrations in the extra- embryonic coelomic fluid, but the distribution of follistatin and activins is not known. We have used recently developed immunoassays for follistatin, activin A and activin AB to determine their presence in the intrauterine compartments during early pregnancy. Follistatin was present in highest concentrations in the extra-embryonic coelomic fluid (11.72 +/- 1.70 ng/ml; median +/- SEM), with less in maternal serum (6.35 +/- 4.58) and lowest amounts in amniotic fluid (0.97 +/- 0.52). Follistatin concentrations in extra-embryonic coelomic fluid were highly correlated with both dimeric inhibin isoforms. Activin A was present in only barely detectable amounts in some samples of extra- embryonic coelomic fluid (41% of samples) and maternal serum (26%) and was undetectable in all amniotic fluid samples. Activin AB was undetectable in all compartments. The presence of follistatin in the amniotic and extra-embryonic coelomic fluids may regulate the availability of bioactive activins and inhibins which are released into the intrauterine compartments during the development of the fetus and placenta in early pregnancy.      

Biologic tests for the diagnosis of amniotic fluid embolism

Amniotic fluid embolism is a rare, unpredictable and often lethal complication of pregnancy and childbirth. Because of its variable presentation, an early biologic test would help to establish the diagnosis. We investigated in maternal serum 4 components of amniotic fluid, i.e., alpha-fetoprotein (AFP), l'insuline like growth factor binding protein-1 (IGFBP-1), fetal fibronectin (fFN) and placental alpha1-microglobulin (PAMG-1). On the 6 cesareans controls involved, none of the makers increased after membranes section. PAMG-1 is unsuitable because its detection is always positive or doubtful even in the baseline. On the 7 cases suspected of amniotic fluid embolism, no detectable increase in any of those markers was noted in 3 cases, which is not in favour of this diagnosis. In the remaining cases, IGFBP-1 and fFN became detectable, confirming histological evidences of amniotic fluid embolism in 2 cases. The follow up of those markers in maternal blood confirmed the suspicion of amniotic fluid embolism at 21 wg in one case of ongoing pregnancy. These preliminary results point out the potential interest to assay maternal serum AFP, IGFBP-1 and fFN to confirm amniotic fluid embolism using rapid laboratory tests.     

A study on placental transfer of diclofenac in first trimester of human pregnancy

Diclofenac is a commonly used non-steroidal anti-inflammatory drug in women of reproductive age. It has teratogenic effects in animals. The aim of this study was to investigate the placental transfer of diclofenac in the first trimester of human pregnancy. Thirty patients undergoing surgical termination of pregnancy between 8 and 12 weeks gestation were given two doses of diclofenac before the procedure. Corresponding samples of maternal serum, amniotic fluid, coelomic fluid and fetal tissue were analysed by high-performance liquid chromatography. Diclofenac was detectable in all fetal tissue samples, with a concentration similar to that found in maternal venous samples. However, diclofenac was detectable in only 56.7 and 23.3% of the coelomic and amniotic fluid samples respectively, and the highest concentration attained was 80 and 5% of the maternal concentration respectively. In summary, we confirmed that diclofenac crosses the human placenta readily during the first trimester. Further studies are required to investigate the potential teratogenic effect of diclofenac in human embryos.     

Human placental growth hormone is increased in maternal serum at 20 weeks of gestation in pregnancies with large-for-gestational-age babies

To investigate the relationship between maternal serum concentrations of placental growth hormone (GH-V), insulin-like growth factor (IGF)-1 and 2, IGF binding proteins (IGFBP)-1 and 3 and birth weight in appropriate-for-gestational-age (AGA), large-for-gestational-age (LGA) and small-for-gestational-age (SGA) cases in a nested case-control study. Maternal serum samples were selected from the Screening for Pregnancy Endpoints (SCOPE) biobank in Auckland, New Zealand. Serum hormone concentrations were determined by ELISA. We found that maternal serum GH-V concentrations at 20 weeks of gestation in LGA pregnancies were significantly higher than in AGA and SGA pregnancies. Maternal GH-V concentrations were positively correlated to birth weights and customized birth weight centiles, while IGFBP-1 concentrations were inversely related to birth weights and customized birth weight centiles. Our findings suggest that maternal serum GH-V and IGFBP-1 concentrations at 20 weeks' gestation are associated with fetal growth.     

An in-vivo study on placental transfer of naproxen in early human pregnancy

BACKGROUND: Naproxen is one of the most common non-steroidal anti-inflammatory drugs used by women of reproductive age. Naproxen is known to be teratogenic in animals. The aim of this study was to investigate the placental transfer of naproxen in the first trimester of human pregnancy, and to determine the amount of the drug in different embryonic compartments. METHODS: Twenty-eight patients who requested surgical termination of pregnancy in the first trimester were given two oral 500 mg doses of naproxen before the surgical procedure. Four biological samples, maternal venous blood, coelomic fluid, amniotic fluid and fetal tissue, were collected from each patient for drug analyses by high performance liquid chromatography. RESULTS: Naproxen was detected in all samples. The mean (+/- SD) concentrations were 69.5 +/- 12.2 microg/ml, 6.4 +/- 2.4 microg/g, 1.85 +/- 1.03 microg/ml and 0.14 +/- 0.11 microg/ml in maternal serum, fetal tissue, coelomic fluid and amniotic fluid respectively. The mean amniotic fluid/maternal drug ratio and fetal/maternal drug ratio were 0.002 (range 0.0005-0.0064) and 0.092 (range 0.022-0.155) respectively. There was a positive correlation between the fetal drug concentration (r = 0.59, P = 0.001), amniotic fluid drug concentration (r = 0.47, P = 0.013), amniotic fluid/maternal ratio (r = 0.536, P = 0.003) and fetal/maternal ratio (r = 0.72, P < 0.001) with advancing gestational age. CONCLUSIONS: Although naproxen can cross the placenta readily in the first trimester of human pregnancy, only a small amount was present in fetal tissues. Since there is no information on whether this small amount of naproxen would be teratogenic or not, women of reproductive age who are taking naproxen regularly should be warned of the possible fetal side-effects.     

Local fetal signal is not required for maintaining IGFBP gene expression in the human decidua: evidence from extrauterine pregnancies

Insulin-like growth factor-II (IGF-II) from the invading extravillous cytotrophoblasts (EVTs) and insulin-like growth factor binding proteins (IGFBPs) from the maternal decidua interact at the feto-maternal interface and regulate implantation and placentation. To determine whether a local stimulus from the fetus is important in the regulation of IGFBP gene expression in the human decidua, we compared the expression of IGFBP genes in intra- and extrauterine (tubal) pregnancies. The expression of IGF-II and IGFBP-1 to IGFBP-6 mRNAs was determined by in-situ hybridization in the Fallopian tubes of extrauterine pregnancies and concurrent decidua (n = 6), and in the placentae and Fallopian tubes of intrauterine pregnancies (n = 6). All six IGFBP mRNAs were identified in the decidualized endometrium and decidualized Fallopian tubes of intra- and extrauterine pregnancies, with IGFBP-1 mRNA being the predominant mRNA. IGFBP-4 was the second most predominant mRNA and was slightly more abundant in the decidua of extrauterine pregnancies than of intrauterine pregnancies. IGF-II mRNA was expressed mainly in cells of fetal origin. The fact that the IGFBP mRNAs were expressed similarly in both intra- and extrauterine pregnancies indicates that the local physical stimulus from an implanting fetus is not necessary to induce or maintain decidual IGFBP gene expression.     

Role of the IGF system in trophoblast invasion and pre-eclampsia

Insulin-like growth factor-II (IGF-II) and IGF binding protein-1 (IGFBP-1) appear to play an important role in paracrine interactions at the maternal-fetal interface in human pregnancy. Patterns of expression of IGF-II and IGFBP-1 at the decidual-trophoblast interface suggest paracrine interactions occur between the IGF-II-expressing invading cytotrophoblast and maternal decidua-derived IGFBP-1. Autocrine/paracrine actions of trophoblast-derived IGF-II may be important in invasion, and for both trophoblast and decidual function. The actions of IGFBP-1 in binding IGF, and as an integrin ligand, suggest it may have multiple roles in the interactions between the invading trophoblast and the maternal decidua. Abundant decidual IGFBP-1 may interact with the IGF-II-expressing, protease-secreting trophoblast to modulate invasion. In-vitro studies of trophoblast-decidual cell interactions in invasion, and clinical observations in a gestational disorder with shallow placental invasion such as pre-eclampsia, have provided new insights into the possible role(s) of IGFBP-1 in trophoblast invasion. The precise mechanisms underlying IGF and IGFBP-1 action at the decidual-trophoblast interface remain to be elucidated. The potential predictive value of serum IGFBP-1 concentrations in pre-eclampsia also remains to be established.     

Hepcidin and iron species distribution inside the first-trimester human gestational sac

We have investigated factors affecting iron distribution in the first-trimester gestational sac, by the measurement of transferrin, non-transferrin-bound iron (NTBI) and pro-hepcidin (Hep) in maternal serum, coelomic fluid (CF) and amniotic fluid (AF) and by immunostaining for Hep in villous and secondary yolk sac biopsies. These samples were obtained from 15 first-trimester pregnancies at 8-11 weeks gestation. Transferrin concentrations were significantly lower in fetal (0.56 mg/ml) than maternal serum (1.71 mg/ml), with very low concentrations in CF and AF (0.09 mg/ml). In contrast, transferrin saturations were significantly higher in fetal (77%) than maternal serum (33%). NTBI was present in fetal serum, CF and AF, presumably as a consequence of low transferrin concentrations in these compartments. Pro-Hep was present at lower levels in fetal (140.0 ± 11.1) than maternal serum (206.2 ± 9.2) and at low concentrations in CF (19.4 ± 3.1) and AF (21.8 ± 5.2). Immunostaining with Hep antibody was found in the syncytiotrophoblast of first-trimester placenta as well as in mesothelial and endodermal layers of the secondary yolk sac at 10 weeks. The presence of Hep in syncytiotrophoblast cells of first-trimester placenta as well as in mesothelial and endodermal layers of the secondary yolk sac suggest a key regulatory role for this protein in iron transfer to the first-trimester fetus. The low transferrin concentrations and the presence of NTBI in CF and AF suggest that transferrin-independent iron transfer is important in early gestation.     

Insulin-like growth factors and their binding proteins in human follicular fluid.

Increasing evidence suggests that insulin-like growth factors I and II (IGF-I, IGF-II) have a regulatory role in animal granulosa cells. This study was undertaken to investigate the presence of IGF-I and IGF-II, as well as that of their binding proteins (BP), IGFBP-1 and IGFBP-3 in human serum and follicular fluid (FF). Preovulatory FF was obtained from 51 patients undergoing in-vitro fertilization. The IGFBP-1 level was found to be significantly higher (P less than 0.01) in FF than in serum, whereas IGF-I and IGFBP-3 values remained markedly lower (P less than 0.01) in FF. Serum IGF-II levels were slightly but not significantly elevated compared to values obtained in the FF of patients. A positive correlation (P less than 0.001) between individual serum and FF levels was observed only for IGF-I. When a group of poor responders was compared to patients with normal stimulation characteristics, no significant difference was found in either IGF or IGFBP levels in the FF. It is concluded that IGFBP-1 is produced locally, whereas the serum may possibly be the major source of IGF-I. No clear conclusions can be drawn regarding the source of FF IGF-II and IGFBP-3. Neither the absolute level nor the relationship of IGFs to their transport proteins could explain the poor response to ovarian stimulation.     

Posttranslational modifications of decidual IGFBP-1 by steroid hormones in vitro

Insulin-like growth factor binding protein-1 (IGFBP-1) appears to regulate insulin-like growth factors (IGFs; IGF-I and IGF-II) biological activity within the local environment of human placenta by modulating IGFs interaction with their receptors. Considering that posttranslational modifications of IGFBP-1 such as phosphorylation and proteolysis affect its affinity for IGFs, this study was undertaken to identify the role of estrogen and progesterone in this regard. The conditioned media of steroid hormone-treated decidual cells were evaluated using different approaches using sodium dodecyl sulphate-polyacrylamide gel electrophoresis (SDS-PAGE) and non-denaturing PAGE following immunoblotting as well as zymographys that contained gelatin and IGFBP-1 as substrates. Our results demonstrated that medroxy progesterone acetate (MPA) treatment increased both phosphorylated and non-phosphorylated decidual-secreted IGFBP-1, whereas 17beta-estradiol (E2) treatment attenuated its phosphorylated forms. Furthermore, the results of zymography revealed that steroid hormones regulated the activity of decidual-secreted matrix metalloproteinases (MMP)-2 and -9, in which E2 treatment up-regulated the MMP-9 activity. Finally, it was demonstrated in our study that decidual-secreted MMP-9 was capable of degrading human amniotic fluid-derived IGFBP-1. In conclusion, our data implicate steroid hormones in the control of IGF system activities at the embryo-maternal interface, at least in part, through their effects on the post-translation changes of decidual-secreted IGFBP-1 such as its phosphorylation and/or proteolysis.     

Determination of protein pattern in embryonic cavities of human early pregnancies: a means to understand materno-embryonic exchanges.

Exocoelomic and amniotic fluids were obtained by selective puncture under ultrasound guidance in normal human pregnancies between 5 and 13 weeks of gestation. Evaluation of the protein patterns in the exocoelomic fluid showed qualitative and quantitative changes with advancing gestation. During the second month of gestation, three electrophoretic bands were found with mobility compatible with albumin, alpha 1-globulin and beta-globulin and composed of at least eight proteins including: pre-albumin, albumin, alpha-fetoprotein (alpha-FP), alpha 1-protease inhibitor, haptoglobin, ceruloplasmin, transferrin and immunoglobulin-G, as revealed by immunoblotting. Protein patterns obtained between 9 and 13 weeks were comparable in exocoelomic fluid and in maternal serum except for the presence of alpha-FP in the alpha 1-globulin band. At the same gestational age, protein electrophoresis of amniotic fluid revealed four bands corresponding to albumin, alpha-FP, haptoglobin and transferrin. Creatinine levels were significantly lower (P less than 0.01) in amniotic fluid than in exocoelomic fluid, and alpha-FP levels were similar in both exocoelomic and amniotic fluids. These results suggest that the exocoelomic fluid is a transudate of the maternal serum except for the presence of high levels of alpha-FP, that amniotic and exocoelomic cavities are separated by a non-permeable membrane and that the secondary yolk sac plays an important role in early protein synthesis and transfer.