Cell-free studies of developmental changes in synthesis of alpha-foetoprotein and albumin in the mouse liver

Expression of alpha-fetoprotein and prostate-specific antigen genes in several tissues and detection of mRNAs in normal circulating blood by reverse transcriptase-polymerase chain reaction

BACKGROUND

alpha-Fetoprotein (AFP) and prostate-specific antigen (PSA) in serum are widely used as tumor markers in the evaluation of prognosis and management of patients with hepatocellular carcinoma and prostate cancer, respectively. To establish the molecular diagnosis of cancer, reverse transcriptase polymerase chain reaction (RT-PCR) for AFP and PSA was used to identify circulating cancer cells in the blood of cancer patients. Here, we examined the tissue-specificity of AFP and PSA and tested whether AFP and PSA are suitable targets in the detection of certain cancer cells by RT-PCR using peripheral blood samples.

METHODS

Tissue specificity of AFP and PSA was analyzed by Northern blotting and RT-PCR. Probes for AFP and PSA were hybridized with poly A+ RNAs from 50 human tissues. RT-PCR for AFP and PSA mRNA was performed using several cancerous tissues and normal tissues and peripheral blood cells from seven healthy volunteers.

RESULTS

Broad expression of AFP was observed in several tissues and a large amount of AFP mRNA was found in fetal liver. PSA was expressed in prostate, salivary gland, pancreas and uterus. By RT-PCR, AFP and PSA mRNA were detected in several tumors, including salivary pleomorphic adenoma, hilar bile duct carcinoma, pancreatic carcinoma, transitional cell carcinoma of urinary bladder and thyroid papillary carcinoma. Furthermore, AFP and PSA mRNAs were frequently detected by RT-PCR, even in peripheral blood cells from healthy volunteers.

CONCLUSIONS

Neither AFP nor PSA showed tissue-specific expression. AFP and PSA mRNA were detected in several diseased and non-diseased tissues and normal circulating blood by RT-PCR.

Hormonal therapy for hepatocellular carcinoma

Hepatocellular carcinoma is not only the leading cause of male cancer death in Taiwan, but also one of the most common cancers in the world. The survival of hepatocellular carcinoma patients is very low, mainly due to the lack of effective treatments. Radiation and chemotherapies in general are not satisfactory: surgery itself is the most effective treatment for hepatocellular carcinoma but only on small resectable tumors. The overall prognosis is still poor. Previously, we have found that the level of glucocorticoid receptor and its mRNA in hepatocellular carcinoma was significantly higher than that of the adjacent liver tissue. This correlated well with the elevated serum alpha-fetoprotein levels in patients with hepatocellular carcinoma. Recently, a female hormone, progesterone, has been found to inhibit the expression of alpha-fetoprotein in hepatoma cells. In addition, progesterone has been used to treat a few hepatocellular carcinoma patients with promising responses. These results together with our hypothesis that the expression of alpha-fetoprotein is regulated by glucocorticoid receptor complex in proliferating hepatocellular carcinoma cells lead to the conclusion that steroid hormones and/or their antagonists may interfere with the function of glucocorticoid receptors in tumors, consequently regulate tumor growth. The potential of hormonal therapy for treatment of hepatocellular carcinoma is worthy of further investigation.

Fine structure and function of hepatocytes during development

This paper reviews the fine structure and function of hepatocytes during fetal and postnatal development. Bile canaliculi develop to a mature appearance during perinatal and early postnatal periods, while bile secretory function is immature at birth and develops during the postnatal period. The rough endoplasmic reticulum is prominent and remains unchanged in amount during development, and the Golgi complex is large from early stages of fetal life. The smooth endoplasmic reticulum (SER) appears shortly before birth and increases in quantity to the adult level after birth. In mouse hepatocytes, Sv (area per unit cytoplasmic volume) of SER increases in perivenular cells between 1 and 10 days of age, although it remains low in periportal cells. Similarly, Sv of total ER increases in both periportal and perivenular cells between 1 and 5 days of age and then becomes greater in perivenular than periportal cells. This suggests that the postnatal increase in the drug-metabolizing capacity occurs predominantly in perivenular hepatocytes. SER proliferates after phenobarbital (PB) administration in both perivenular and periportal cells in 3-, 5-, and 10-day-old mice, and predominantly in perivenular cells in 20-day-old and adult mice. Thus the conspicuous proliferation of SER in perivenular hepatocytes after PB administration, characteristic of adult liver, becomes manifest during postnatal development. In mouse hepatocytes, Vv (volume per unit cytoplasmic volume) of mitochondrial matrix and peroxisomes and Sv of mitochondrial inner membrane and cristae increase in both periportal and perivenular cells between birth and 10 days of age. Then, Vv of mitochondrial matrix remains unchanged in periportal cells but decreases in perivenular cells. In general, the process of postnatal hepatocyte differentiation appears to include several phases of development; cell organelles develop during the early postnatal period, subsequently the cells undergo both functional and structural heterogeneity, and the late postnatal period after weaning is the time for a marked increase in cell size.

Liver-specific expression of the mouse alpha-fetoprotein gene is mediated by cis-acting DNA elements

We have identified cis-acting regulatory elements in the 5' flanking region of the mouse alpha-fetoprotein (Afp) gene, using the expression of the bacterial gene for chloramphenicol acetyltransferase (CAT) in a transient expression assay. Tissue-specific enhancer activity was determined by transfection of mouse hepatoma (BWTG3) and fibroblast cells (C127, NIH 3T3) with various DNA fragments linked to the CAT gene. A 5.4-kilobase restriction fragment was shown to have characteristics typical of enhancers, including the ability to function independent of orientation and position and the ability to enhance transcription from a heterologous promoter. The enhancer activity was greatest in the hepatoma cells, which express Afp. By deletion analysis, it was demonstrated that enhancer activity is present in several subfragments, indicating the presence of more than one element in this fragment. An additional regulatory element within 950 base pairs of the Afp transcription initiation site has been identified and shown to confer tissue-specific expression on the CAT gene. This fragment, which lacks enhancer activity, contains the Afp promoter region and mediates the tissue-specific expression of the CAT gene when driven by nonspecific viral enhancers. We conclude from our studies that there are several types of regulatory elements in the 5' flanking region of the Afp gene that help mediate tissue-specific expression.

Oncodevelopmental and hormonal regulation of alpha 1-fetoprotein gene expression

The main features of the oncodevelopmental biology of alpha 1-fetoprotein (AFP) are reviewed. Progress made in the molecular biology of AFP gene regulation is discussed and we present our recent data on the mechanisms of AFP suppression by glucocorticoid hormones. The relationship between AFP gene transcription and cell replication is examined, and it is suggested that the degree of methylation of the AFP gene (or of co-methylated regulatory DNA sequences) conditions its response to hormones.

Developmentally regulated mRNAs in mouse liver

The mouse liver contains a group of 10--12 different tissue-specific mRNAs, each present at an average concentration of 12,000--15,000 copies per cell [Hastie, N. D. & Bishop, J. O. (1976) Cell 9, 761--774]. We have determined, by translation in vitro, that these mRNAs are developmentally regulated in the liver. We have also used specific cloned probes to quantitate the developmental time course of expression of five different abundant liver mRNAs. We have found that there are at least three periods during liver development when specific abundant mRNAs are first detectable: prior to 14 days postconception, at birth, and during the onset of sexual maturity. These results indicate that all the members of this mRNA group are not under common developmental regulation. One of the abundant liver mRNAs (p54 mRNA) increases more than 1000-fold in the liver 1 day before birth. We discuss factors that may be involved in the developmental regulation of expression of the genes encoding these mRNAs.

Immunoprecipitation assay of alpha-fetoprotein synthesis by cultured mouse hepatoma cells treated with estrogens and glucocorticoids

This investigation was to study the biosynthesis of 3H-labeled alpha-fetoprotein (AFP) by cultured mouse hepatoma (HEPA-2) cells. Both the function and regulation of this oncodevelopmental gene are unknown. However, evidence indicates that mechanisms controlling the expression of AFP involve aspects of both normal embryonic development and neoplastic transformation. the secretion of AFP was analyzed during different phases of the growth cycle to provide information on AFP production using standard culture conditions. The highest rate of secretion occurred during the stationary phase, followed by the late logarithmic and early logarithmic phases of growth, respectively. The production of AFP was then determined following the addition of glucocorticoids and estrogens in an attempt to understand hormonal factors that may be involved. Studies utilizing estradiol-17 beta indicated that the secretion of AFP did not appear to be sensitive to this steroid even though sucrose density gradient analysis of HEPA-2 cytosol, for estrogenic receptors, revealed competitive binding moieties on the 8S and 4S regions of the gradient. In contrast, the secretion of the total complement of proteins, including AFP, was significantly stimulated by the glucocorticoids, dexamethasone and corticosterone. Analysis of HEPA-2 cytosol for glucocorticoid receptors revealed binding components in the 7S and 3-4S regions of the gradient. The 3H-dexamethasone binding appeared to be stereospecific since nonlabeled dexamethasone, but not nonlabeled estradiol-17 beta, effectively displaced the bound radioactivity. The glucocorticoid-binding component in HEPA-2 therefore displayed characteristics reported for glucocorticoid receptors in normal liver and other hepatomas.

Structure of the alpha-fetoprotein gene in the mouse

The mouse alpha-fetoprotein mRNA is the product of a single-copy gene whose mRNA coding sequences are represented discontinuously in the genome. Several EcoRI genomic fragments which contain portions of the alpha-fetoprotein gene have been cloned using the EK2 vector lambda gt WES . lambda B. In addition, a mouse genomic library has been screened to obtain a 15.75-kilobase segment of DNA that includes more than 85% of the alpha-fetoprotein coding sequence. Analyses by restriction endonuclease mapping and electron microscopy showed that the mRNA sequence is interrupted by at least 11 intervening sequences which occupy 90% of the cloned DNA.

Plasma diethylstilboestrol binding proteins of rat, mouse and man in the course of development: relations with the binding of estradiol

High diethylstilboestrol (DES) binding has been demonstrated in fetal and adult sera from man, rat and mouse by equilibrium dialysis and electrophoretic techniques. In the adults of the three species and in the human fetus only albumin shows an elevated binding capacity for DES. By contrast, in the case of rat and mouse embryos there are two proteins, namely albumin and alpha-fetoprotein, which afford major and quantatively similar contributions to the binding. Human alpha-fetoprotein does not bind DES. These phenomena are analysed in relation to the estrogen binding characteristics of the alpha-fetoproteins of the three species.