alpha-Fetoprotein production and erythropoiesis in cell cultures of human fetal liver.

alpha-Fetoprotein and the synthesis of heme associated with hemoglobin were measured simultaneously in short-term cultures of human fetal liver cells to correlate the relationship of alpha-fetoprotein to erythroid cell function. Both synthetic processes decreased exponentially during the first 5 days of culture. The use of media supplemented with different batches of fetal calf serum and porcine portal vein serum indicated that the optimal conditions for the production of alpha-fetoprotein were different from those required for the synthesis of heme associated with hemoglobin. Moreover, the alpha-fetoprotein-producing cells could be separated from erythroid cells after velocity sedimentation in Ficoll gradients. Although it is well known that erythropoiesis and alpha-fetoprotein production occur simultaneously during ontogenesis, alpha-fetoprotein itself (0.01-100 micron g/ml) did not stimulate heme synthesis in liver erythroid cells. Erythropoietin did not stimulate alpha-fetoprotein production. It is concluded that there is no cause-effect relationship between alpha-fetoprotein production and erythroid cell fuction in human fetal liver cells and that the two processes occur independently in different cell types.     

Investigation of the content of alpha-fetoprotein and serum albumin in the vitreous body of the eye of human embryos

The content of serum albumin and alpha-fetoprotein in the vitreous body of the eyes of human fetuses from the 16th through the 24th week was investigated. It was detected that albumin and alpha-fetoprotein in the vitreous body of human eyes are presented in equal molar concentrations in the 16th week. There is 1.5-fold increased concentration of alpha-fetoprotein in comparison to albumin during the 17th week. After the 17th week, there was a reduction in the concentration of both proteins. It was reported that cyanine dye, used for detection of albumin, does not interact with alpha-fetoprotein.     

Determination of the distribution of fatty acids and diethylstilbestrol between serum albumin and alpha-fetoprotein by concanavalin A affinity chromatography

The distribution of fatty acids and diethylstilbestrol between serum albumin and alpha-fetoprotein was measured in vitro by a new method based on the separation of the two proteins by virtue of the binding specificity of concanavalin A for the carbohydrate moiety of alpha-fetoprotein. Human and bovine proteins were investigated. It was found that palmitate and oleate were distributed almost equally between albumin and alpha-fetoprotein, while docosahexaenoate and diethylstilbestrol bound preferentially to alpha-fetoprotein even at an albumin: alpha-fetoprotein ratio of 10:1. The results confirm the binding specificity of alpha-fetoprotein for polyunsaturated fatty acids and also show that alpha-fetoprotein binds diethylstilbestrol much more strongly than albumin does. This suggests that alpha-fetoprotein may play a role in the fetal uptake of diethylstilbestrol.     

Isolation and characterization of alpha-fetoprotein from the mouse hepatoma BW7756.

Purification of alpha-fetoprotein from mouse hepatoma BW7756 extracts was performed using ammonium sulfate precipitations, gel filtration, ion-exchange chromatography and isoelectric focusing. These procedures produced a 5.6% yield of alpha-fetoprotein with 96% purity. Polyacrylamide slab gel electrophoresis, extended agarose electrophoresis and immunoelectrophoresis demonstrated that mouse hepatoma alpha-fetoprotein migrated at pH 8.6 as a rapid alpha1, or postalbumin globulin. Crossed antibody electrophoresis of the agarose zone containing alpha-fetoprotein failed to demonstrate microheterogeneity. Molecular weight analysis of the mouse hepatoma alpha-fetoprotein on a calibrated Sephadex G-200 column yielded a value of 72 000-73 000 for the native protein. Sodium dodecyl sulfate gel electrophoresis subsequently demonstrated a single polypeptide chain with a molecular weight of 72 000. Amino acid analysis showed the alpha-fetoprotein to be an acidic protein dominated by hydrophobic residues. The total carbohydrate content was 5.5%, and 3 mol of sialic acid were detected per mol of alpha-fetoprotein. Although neutral sugars were the principal class present, galactosamine was the most abundant single sugar detected.     

Changes in the rhythms of the alpha-fetoprotein content of the blood serum of mice with hepatoma 22A and of tumor cell proliferation depending on the time of cyclophosphane administration

A study was made of the content of alpha-fetoprotein (alpha-FP) in the blood serum and of proliferative processes in hepatoma 22A-bearing mice after injection of cyclophosphamide in a dose of 150 mg/kg at varying time of 4-day tumor growth. The data obtained indicate that the time of appearance of alpha-FP in the blood, the pattern of rhythmic fluctuations in protein content and in the number of DNA-synthesizing and mitotic cells of the tumor depend on the time of the cytostatic injection during the day. Besides, cyclophosphamide exerts a more powerful inhibitory effect on the content of alpha-FP in the blood serum and less powerful on the proliferative processes in hepatoma 22A.     

Immunochemical purification and characterization of ovine alpha-fetoprotein.

Ovine alpha-fetoprotein was successfully isolated from fetal sheep serum by using rabbit anti-ovine alpha-fetoprotein linked to an agarose immunoadsorbent column. Antibody used in this affinity chromatography column was produced by immunizing a rabbit with highly purified alpha-fetoprotein-antibody complex to yield a monospecific antiserum to ovine alpha-fetoprotein. Following affinity chromatography, alpha-fetoprotein was further purified by preparative polyacrylamide disc gel electrophoresis ultimately yielding a 105-fold purification. The purified alpha-fetoprotein was homogeneous on analytical polyacrylamide disc gel electrophoresis. Ovine alpha-fetoprotein was found to be immunochemically related to human alpha-fetoprotein and to exhibit a molecular weight and amino acid composition similar to other mammalian alpha-fetoproteins.     

Binding of bilirubin by bovine and human alpha-fetoprotein.

alpha-Fetoprotein, a fetal protein associated with certain tumors, was found to bind bilirubin. Addition of human or bovine alpha-fetoprotein to bilirubin solutions enhanced the light absorbance of bilirubin and shifted its maximum. Bovine alpha-fetoprotein caused a marked shift towards shorter wavelengths, while human alpha-fetoprotein gave a slight red shift. The spectral changes were used to study the characteristics of the binding of bilirubin by bovine alpha-fetoprotein. These studies indicated the presence of one binding site/molecule of alpha-fetoprotein with an association constant of about 1.1 . 10(6) M-1. A difference between the spectral changes brought about by alpha-fetoprotein and albumin allowed comparison of their relative affinities for bilirubin. The spectrum approximated the average between the spectra induced by the two proteins when the ratio of bovine alpha-fetoprotein to bovine albumin was 6.3 : 1, and of the human proteins 21 : 1, respectively. These results show that alpha-fetoprotein from two species binds bilirubin with an affinity somewhat lower than that of albumin. Binding of bilirubin by alpha-fetoprotein is in agreement with the recent demonstration of structural homology between alpha-fetoprotein and albumin. Whether alpha-fetoprotein plays a role in the metabolism of bilirubin or other degradation products of heme remains to be investigated.     

Binding of the chymotrypsin substrate, tryptophan methyl ester, by rat alpha-fetoprotein

We studied how tryptophan methyl ester and related compounds inhibit binding of estrone to rat alpha-fetoprotein and find that: (a) like chymotrypsin, alpha-fetoprotein binds tryptophan esters with higher affinity than tryptophan or its amides; (b) the affinity of alpha-fetoprotein for tryptophan methyl ester is 3.7 . 10(-4) M, which is close to the affinity of chymotrypsin (10(-4) M); (c) alpha-fetoprotein binding of tryptophan methyl ester is stereoselective and pH dependent. All of these observations suggest that there is a specific interaction between alpha-fetoprotein and the chymotrypsin substrate, tryptophan methyl ester, and that rat alpha-fetoprotein contains a site with some structural similarities to the catalytic site in chymotrypsin. Since we also find that tryptophan methyl ester is a competitive inhibitor of estrone binding to alpha-fetoprotein, it is possible that the protease substrate binding site on alpha-fetoprotein is spatially close to the estrone binding site.     

Alpha-fetoprotein and albumin in experimentally-induced exencephaly in the rat.

Alpha-fetoprotein and albumin were quantified in the sera and amniotic fluids from control, Vitamin A-treated non-exencephalic and Vitamin A-treated exencephalic rat fetuses. Exencephaly was associated with amniotic fluid alpha-fetoprotein concentrations which were significantly elevated over those of Vitamin A-treated non-exencephalic and of untreated fetuses. Amniotic fluid albumin concentrations also were higher in the exencephalic fetuses than in the non-exencephalic fetuses. Serum alpha-fetoprotein and albumin concentrations were lower in the exencephalic than in the non-exencephalic fetuses. The results are cosistent with simple diffusion across a defective barrier as the cause of elevated amniotic fluid alpha-fetoprotein concentrations in the presence of open neural tube defects. This experimental model of neural tube defects result in changes in amniotic fluid alpha-fetoprotein similar to those changes found in human amniotic fluid alpha-fetoprotein concentrations in the presence of neural tube defects.     

Carbohydrate structure of the concanavalin A molecular variants of alpha-fetoprotein

The structure of the carbohydrate units of alpha-fetoprotein from fetal calf serum has been studied. Glycopeptides and oligosaccharides were prepared from alpha-fetoprotein by protease treatment and hydrazinolysis, respectively, and subjected to carbohydrate and amino acid analysis. Two N-glycosidic glycans are present in each alpha-fetoprotein molecule. These were separated into concanavalin A (ConA)-reactive and nonreactive species on ConA-Sepharose. Methylation analysis, Smith degradation, sequential exoglycosidase treatments, and sizing suggested that the major, ConA-nonreactive fraction is composed of triantennary and the minor, ConA-reactive fraction, of biantennary complex-type ConA-reactive and -nonreactive fractions of intact alpha-fetoprotein, respectively, and refractionated on Con A-Sepharose. The results indicate that 75% of alpha-fetoprotein molecules contain two triantennary complex-type glycans, 20% contain one triantennary and one biantennary glycan, and 5% contain two biantennary glycans. The last two molecular variants are bound to ConA. These results explain, at least in part, the previously found heterogeneity of alpha-fetoprotein with respect to charge and molecular size, and provide a biochemical basis for the differing reactivities toward ConA of alpha-fetoprotein from the yolk sac, fetal liver, and various tumors.     

DNaseI sensitivity of the rat albumin and alpha-fetoprotein genes.

We have analyzed the DNaseI sensitivity of chromatin from the rat albumin and alpha-fetoprotein genes in the fetal liver (which synthesizes albumin and alpha-fetoprotein), adult liver (which synthesizes albumin), fetal yolk sac (which synthesizes alpha-fetoprotein), and adult kidney (which synthesizes neither). Active genes were much more sensitive than their kidney counterparts, and the adult liver alpha-fetoprotein and fetal yolk sac albumin genes showed intermediate levels of sensitivity. Sensitivity was analyzed as a function of the extent of DNaseI digestion. Rate constants were calculated for the degradation of individual DNA hybridization bands and normalized to the intrinsic rate constants of the same bands degraded in purified DNA. This enabled us to eliminate the inconsistencies that otherwise result from comparing chromatin sensitivity of different DNA sequences, or chromatin sensitivity in different nuclear environments.     

Synthesis of DNA and alpha-fetoprotein in the monolayer culture of mouse hepatocytes

Relation between processes of proliferation and synthesis of the embryonal serum protein alpha-fetoprotein (AFP), the influence on these processes of polyelectrolyte dextran sulfate (DS) and dimethyl-sulfoxide (DMSO) has been studied in the monolayer culture of mouse hepatocytes. In control cultures the correlations between the time of appearance and the level of DNA and AFP synthesis were observed. DS and DMSO were found to inhibit both processes. Cell proliferation could be reestablished by addition of epidermal growth factor. In case of the influence of DMSO, it wasn't followed by the induction of AFP synthesis. This the processes of DNA and AFP synthesis in monolayer cultures of mouse hepatocytes can be separated. The elongated incubation of hepatocytes with collagenase during their obtaining, abolished the effects of DS. This shows that surface components of hepatocytes, lost upon enzyme degradation, may be involved in the mechanism of DS effect.     

Detection and measurement of fetomaternal haemorrhage: serum alpha-fetoprotein and the Kleihauer technique.

A raised maternal serum alpha-fetoprotein concentration was taken as an indicator of fetomaternal haemorrhage due to amniocentesis and was used to calculate the volume of the fetal bleed. The alpha-fetoprotein concentration seemed to be a more sensitive and reliable indicator than the established Kleihauer technique and may have further applications in antenatal testing.     

alpha-Fetoprotein synthesis in transformed fetal rat liver cells

We have reported that transformed fetal liver cells produced a variant alpha-fetoprotein of 65K that differed from the mature alpha-fetoprotein of 69K and 73K in the polypeptide backbone. In the present study, we demonstrated that the biosynthetic pathway of the variant alpha-fetoprotein differed from that of the mature alpha-fetoprotein. The 65K variant was synthesized first as a preprotein of 49.5K which was processed to a polypeptide of 59K in the presence of microsomal membranes. The latter was the precursor of the variant alpha-fetoprotein found in cells and medium of transformed fetal liver cells. The 65K alpha-fetoprotein was encoded by a mRNA of 16S while mature AFP was encoded by a mRNA of 20S.     

Ultrastructural immunoenzymatic study of alpha-fetoprotein-producing cells in the human fetal liver.

Cells producing alpha-fetoprotein in human fetal liver have been studied with specific horseradish peroxidase labeled immunoglobulins. Under light microscopy, the alpha-fetoprotein is strictly localized in the cytoplasm of certain hepatocytes, distributed randomly in the hepatic lobule. Ultrastructural examination of thesame cells shows that the alpha-fetoprotein is present within the cytoplasm. Ultrastructural differences are described in hepatocytes according to whether or not the cell is producing alpha-fetoprotein at the time of sampling. These observations lead to the hypothesis that alpha-fetoprotein may correspond to a particular functional state of the hepatocyte in human fetal liver.     

Amino acid sequence homology between rat alpha-fetoprotein and albumin at the COOH-terminal regions

The nucleotide sequence of the 3'-terminal untranslated region and a portion of the coding region of rat alpha-fetoprotein mRNA has been determined from a cloned double-stranded cDNA. the amino acid sequence of the COOH-terminal portion of alpha-fetoprotein was inferred from the nucleotide sequence and compared to the amino acid sequence of the corresponding portion of rat, bovine, and human albumin. A striking homology in amino acid sequence between alpha-fetoprotein and albumin was observed. These results confirm earlier suggestions that the two proteins are closely related in structure and probably arose from a common ancestral gene.