Mouse hepatoblastomas: a histologic, ultrastructural, and immunohistochemical study

Congenital ascites due to hepatoblastoma with extensive peritoneal implantation metastases in a premature equine fetus

A premature dead equine fetus with excessive fluctuating distension of the abdomen was delivered by extraction. Post-mortem examination revealed ascites and a solitary, irregular, bulging, multinodular, firm, yellow mass of 25 cm in diameter in the right liver lobe. Extensive peritoneal implantation metastases were present. The masses were composed of polygonal embryonal cells arranged in sheets and nests. Based on the immunohistochemical expression of Ki67, low molecular weight cytokeratin and alpha-1 fetoprotein, a diagnosis of hepatoblastoma with peritoneal implantation metastases was made.

Proliferative and nonproliferative lesions of the rat and mouse hepatobiliary system

The INHAND Project (International Harmonization of Nomenclature and Diagnostic Criteria for Lesions in Rats and Mice) is a joint initiative of the Societies of Toxicologic Pathology from Europe (ESTP), Great Britain (BSTP), Japan (JSTP) and North America (STP) to develop an internationally-accepted nomenclature for proliferative and non-proliferative lesions in laboratory animals. The purpose of this publication is to provide a standardized nomenclature and differential diagnosis for classifying microscopic lesions observed in the hepatobiliary system of laboratory rats and mice, with color microphotographs illustrating examples of some lesions. The standardized nomenclature presented in this document is also available for society members electronically on the internet (http://goreni.org). Sources of material included histopathology databases from government, academia, and industrial laboratories throughout the world. Content includes spontaneous and aging lesions as well as lesions induced by exposure to test materials. A widely accepted and utilized international harmonization of nomenclature for lesions of the hepatobiliary system in laboratory animals will decrease confusion among regulatory and scientific research organizations in different countries and provide a common language to increase and enrich international exchanges of information among toxicologists and pathologists.

Hepatoblastoma in a cat

Hepatoblastomas are neoplasms that originate from putative pluripotential stem cells of the liver. A hepatic mass from an 8-year-old Abyssinian cat was composed of cords and sheets of neoplastic cells, with scattered rosettes and small ductal structures. Most neoplastic cells had a pale eosinophilic cytoplasm and a round to ovoid nucleus. The tumor also had short spindle cells with an oval nucleus. Immunohistochemically, neoplastic cells were weakly positive for embryonic hepatocellular markers, such as alpha-fetoprotein and cytokeratin (CK) 8/18, but negative for the hepatocellular marker Hepatocyte Paraffin 1. The cells were also positive for CD56/neural cell adhesion molecule and for the biliary epithelial markers CK 7, CK 8/18, CK CAM5.2, and vimentin, but negative for CK 20. Some neoplastic cells expressed neuroectodermal or neuroendocrine markers, such as protein gene product 9.5 and synaptophysin, but were negative for chromogranin A and not argyrophilic by the Grimelius technique. The cat died soon after the biopsy without clinical improvement.

Liver cancer stem cells

In an effort to review the evidence that liver cancer stem cells exist, two fundamental questions must be addressed. First, do hepatocellular carcinomas (HCC) arise from liver stem cells? Second, do HCCs contain cells that possess properties of cancer stem cells? For many years the finding of preneoplastic nodules in the liver during experimental induction of HCCs by chemicals was interpreted to support the hypothesis that HCC arose by dedifferentiation of mature liver cells. More recently, recognition of the role of small oval cells in the carcinogenic process led to a new hypothesis that HCC arises by maturation arrest of liver stem cells. Analysis of the cells in HCC supports the presence of cells with stem-cell properties (ie, immortality, transplantability, and resistance to therapy). However, definitive markers for these putative cancer stem cells have not yet been found and a liver cancer stem cell has not been isolated.

Evaluation of gene expression related to hepatic cell maturation and differentiation in a chemically induced mouse hepatoblastoma cell line

The MHB-2 cell line, established from a mouse hepatoblastoma (HB), was subjected to the reverse transcriptase-polymerase chain reaction (RT-PCR) for evaluation of gene expression related to cell differentiation. RNAs for c-kit, CD34, thy-1, albumin, cytokeratin (CK) 8, 18 and 19 could be detected, but expression of alpha-fetoprotein, glucose-6-phosphatase, tyrosine aminotransferase and CK7 was not observed. MHB-2 cells were positive for CK8/18 but negative for c-kit, CD34, thy-1 and albumin on protein level. Immunohistochemical staining of the HB in vivo revealed diffusely expressed c-kit. Thy-1-positive HB cells were sparsely observed, but the tumor was negative for CD34 and rarely positive for CK8/18. By in situ hybridization, the HB was positive for CK18 but negative for CK19. Slight expression of albumin, but the lack of immature hepatocytic marker suggested some heterogeneous hepatocyte or an undifferentiated cell from other origin. Furthermore, positive expression of CK19 as well as CK8 and CK18 in culture strongly suggested the differentiation into a biliary lineage or the bidirectional state. In conclusion, the present study indicated the mouse HB to have de-differentiated, bipotent, or biliary-like cell characteristics, and considering the histological difference between HB and biliary tumors, it suggests the mouse HB cells are closely like some sort of hepatic undifferentiated cells.

Three equine cases of mixed hepatoblastoma with teratoid features

Hepatoblastoma was diagnosed in 3 Thoroughbreds at the University of Kentucky Livestock Disease Diagnostic Center (LDDC) since 1997. Case #1 involved a fetus with a well-demarcated, multilobulated, solitary mass that extended from the left liver lobe. Case #2 was observed in a neonate with a primary hepatic mass and multiple metastases in the skin, brain, meninges, and stylohyoid bone. Case #3 was a solitary hepatic mass incidentally discovered in a neonate at necropsy. Microscopically, the masses were similarly composed of sheets and cords of fetal and embryonal epithelial cells that frequently formed sinusoid-like structures. Intermixed with the neoplastic epithelial cells were variable amounts of hemorrhage, necrosis, osteoid, and bone. Immunohistochemically, the epithelial cells stained variably positive for alpha- fetoprotein, frequently positive for vimentin, and occasionally positive for cytokeratin. All 3 cases were diagnosed as mixed hepatoblastoma with teratoid features.

Embryonic stem cell-derived embryoid bodies in three-dimensional culture system form hepatocyte-like cells in vitro and in vivo

Pluripotent embryonic stem (ES) cells can be a source of hepatocytes for bioartificial livers or transplantation. In this study, embryoid bodies (EBs) were formed from ES cells cultured in polypropylene conical tubes. The EBs were then inserted into a collagen scaffold three-dimensional culture system and stimulated with exogenous growth factors and hormones to induce hepatic histogenesis. The EB-derived cells expressed liver-specific genes, and albumin-positive cells formed cord-like structures that were not present in two-dimensional monolayer culture systems. However, these albumin- positive cells were not cytokeratin 18 positive. Electron microscopy showed immature hepatocyte- like cells having tight junctions, rough endoplasmic reticulum, and intercellular canaliculi. The scaffold including EB-derived hepatocyte-like cells was transplanted into the median lobes of partially hepatectomized nude mice. After 7 and 14 days, cells positive for both albumin and cytokeratin 18 appeared in the transplant and formed clustered aggregates. Thus the collagen scaffold three-dimensional culture system and the liver regeneration environment induced hepatocyte-like cells and hepatic lobule-like aggregates from EBs. Therefore, differentiating EBs in the scaffold culture system may be useful in developing bioartificial livers, secondary livers, and as pharmaceutical models.

Immunohistochemical characterization of hepatoblastomas in B6C3F1 mice treated with diethylnitrosamine and sodium phenobarbital

Hepatoblastomas (HBs) were induced in B6C3F1 male mice by diethylnitrosamine (DEN) and sodium phenobarbital (PB). Six-week-old mice received a single intraperitoneal dose of DEN followed by a continuous treatment with PB in diet at a concentration of 0 (group 1) or 500 (group 2) ppm for 50 weeks. HBs were observed in 13 of 21 (62%) group 2 mice, with typical histologic features as reported previously, while no such tumors were observed in group 1. Seven of 13 (54%) HBs were found in and/or adjacent to hepatocellular adenomas (HCAs) or hepatocellular carcinomas (HCCs). Immunohistochemically, all HBs were positive for S-100 protein but negative for keratin, alpha-fetoprotein (AFP), albumin (ALB) and vimentin, while HCC cells occasionally reacted positively for AFP with a mosaic pattern. HCC and HCA cells were occasionally positive for ALB. Non-neoplastic hepatocytes and normal bile ducts were positively stained for ALB and keratin/S-100 protein, respectively. S-100 protein is known to be expressed in many mesenchymal tissues and neoplasms including neuroectodermal elements but negative in cells of the hepatic lineage. Thus, the present immunohistochemical results suggested that mesenchymal differentiation occurs in mouse HB cells as observed in human HBs, one of the most frequent infant liver tumors in humans. Although the susceptibility of mouse HBs to PB-promotion suggests a hepatocytic histogenesis, the present immunohistochemical results support the hypothesis that the mouse HB is derived from pluripotent endodermal stem-like cells.

Greater expression of transforming growth factor alpha and proliferating cell nuclear antigen staining in mouse hepatoblastomas than hepatocellular carcinomas induced by a diethylnitrosamine-sodium phenobarbital regimen

Transforming growth factor alpha (TGF-alpha) is a potent stimulator of normal hepatocyte proliferation, considered to have relationship to the liver regeneration or carcinogenesis. In this study, we investigated immunohistochemically the association between expression of TGF-alpha and cell proliferation activity in mouse hepatoblastomas (HBs) and hepatocellular carcinomas (HCCs) induced in B6C3F1 mice by diethylnitrosamine and sodium phenobarbital. The TGF-alpha-positive rate in HBs (29.2%) was significantly higher than that in HCCs (12.7%). Likewise, the proliferating cell nuclear antigen-positive rate (22.2%) was higher than the HCC value (14.5%). On the individual data for both TGF-alpha and PCNA, most of the HBs showed higher positive rates than HCCs. In HBs, TGF-alpha was localized only in the nuclei, whereas some HCC cells stained positive both in their nuclei and cytoplasm (0.6%). These results suggest expression of TGF-alpha and its localization might be linked to cell proliferation and play a role in malignant progression of mouse HBs.

Two-year toxicity and carcinogenicity study of methyleugenol in F344/N rats and B6C3F(1) mice

Methyleugenol (MEG) was tested for toxicity/carcinogenicity in a 2-yr carcinogenesis bioassay because of its widespread use in a variety of foods, beverages, and cosmetics as well as its structural resemblance to the known carcinogen safrole. F344/N rats and B6C3F(1) mice (50 animals/sex/dose group) were given MEG suspended in 0.5% methylcellulose by gavage at doses of 37, 75, or 150 mg/kg/day for 2 yr. Control groups (60 rats/sex and 50 mice/sex) received only the vehicle. A stop-exposure group of 60 rats/sex received 300 mg/kg/day by gavage for 53 weeks followed by the vehicle only for the remaining 52 weeks of the study. A special study group (10 animals/sex/species/dose group) were used for toxicokinetic studies. All male rats given 150 and 300 mg/kg/day died before the end of the study; survival of female rats given 150 mg/kg/day and all treated female mice was decreased. Mean body weights of treated male and female rats and mice were decreased when compared to control. Area under the curve results indicated that greater than dose proportional increases in plasma MEG occurred for male 150 and 300 mg/kg/day group rats (6 and 12 month) and male 150 mg/kg/day mice (12 month). Target organs included the liver, glandular stomach, forestomach (female rats) and kidney, mammary gland, and subcutaneous tissue (male rats). Liver neoplasms occurred in all dose groups of rats and mice and included hepatoadenoma, hepatocarcinoma, hepatocholangioma (rats only), hepatocholangiocarcinoma, and hepatoblastoma (mice only). Nonneoplastic liver lesions included eosinophilic and mixed cell foci (rats only), hypertrophy, oval cell hyperplasia, cystic degeneration (rats only), and bile duct hyperplasia. Mice also exhibited necrosis, hematopoietic cell proliferation, and hemosiderin pigmentation. Glandular stomach lesions in rats and mice included benign and malignant neuroendocrine tumors, neuroendocrine cell hyperplasia, and atrophy and in mice included glandular ectasia/chronic active inflammation. In female rats, the forestomach showed a positive trend in the incidences of squamous cell papilloma or carcinoma (combined). Male rats also exhibited kidney (renal tubule hyperplasia, nephropathy, and adenomacarcinoma), mammary gland (fibroadenoma), and subcutaneous tissue (fibroma, fibrosarcoma) lesions. Male rats also exhibited malignant mesotheliomas and splenic fibrosis. These data demonstrate that MEG is a multisite, multispecies carcinogen.

Establishment and characterization of a cell line from a chemically-induced mouse hepatoblastoma

We established a cell line (MHB-2) from a hepatoblastoma (HB) induced by diethylnitrosamine (DEN) and sodium phenobarbital (PB) in male B6C3F1 mice and examined the biological characteristics of MHB-2. MHB-2 cells grew as monolayers in culture and showed a spindle or polygonal shape. Immunohistochemically, the original tumor cells and MHB-2 cells were negative for keratin, alpha-fetoprotein and albumin. Electron microscopically, MHB-2 cells had irregular-shaped nuclei with prominent nucleoli, abundant free ribosomes, myelinosomes, desmosomes and surface microvilli. Growth of this cell line was significantly accelerated by hepatocyte growth factor (HGF) and expression of its receptor c-met was confirmed by the reverse transcription-polymerase chain reaction (RT-PCR). MHB-2, however, was not found to be tumorigenic when transplanted into the subcutaneous tissue of syngeneic, nude or scid mice. To our knowledge, this is the first report on the establishment of a cell line derived from a mouse HB. MHB-2 would be useful for further studies to clarify the biological characteristics of mouse HB.

Hepatoblastoma in a dog

A hepatoblastoma was found in a 13-year-old female Maltese dog. Histologically, the tumor showed a wide trabecular pattern and was frequently accompanied with vascular lake formation. Tumor cells were positive for cytokeratin and neuron specific enolase, but negative for chromogranin. Electronmicroscopically, tumor cells were accompanied with continuous basement membrane and had poorly developed desmosomes. Sinusoidal endothelia had fenestration and were surrounded by myofibroblast-like cells. To the best of our knowledge, this paper is the first report of morphological studies on canine hepatoblastoma.

In situ hybridization demonstration of albumin mRNA in B6C3F1 murine liver and hepatocellular neoplasms

In situ hybridization was used to detect albumin mRNA in normal liver and hepatocellular neoplasms in 20 male B6C3F1 mice between 17 and 24 months of age. Positive signals for albumin were observed consistently in the cytoplasm of hepatocytes in normal liver, particularly in periportal areas. No signals were observed in other cells, such as Kupffer's cells, mesenchymal cells, or bile duct epithelium. Of hepatocellular adenomas, 11/11 (100%) stained positively for albumin mRNA, whereas 14/15 (93%) of primary hepatocellular carcinomas showed positive expression. Albumin mRNA was also detected in extrahepatic metastases of hepatocellular carcinoma, including 9/15 (60%) of pulmonary neoplasms and 5/12 (42%) of metastases at other sites. The pulmonary metastases of hepatocellular carcinoma frequently exhibited a glandular, papillary, or sarcomatous histologic appearance. The presence of albumin in these tumors, lacking characteristic hepatocellular phenotype, is a potential determinant of hepatic lineage. We conclude that in situ hybridization for albumin mRNA in mice is a useful tool in the differential diagnosis of hepatocellular carcinoma, particularly in the case of pulmonary metastasis. This technique may also enable recognition of hepatocyte differentiation in glandular structures with phenotypic features of biliary cells, as seen in mixed hepatocellular-cholangial neoplasms.

Further evidence for promoter-dependent development of hepatoblastoma in the mouse

In previous studies, we found that male D2B6F1 mice fed phenobarbital (PB) for 53 weeks following N-nitrosodiethylamine (NDEA) initiation developed a high (70-80%) incidence of malignant hepatoblastomas. A very low (3.3%) incidence of such tumors occurred in the absence of promoter treatment in NDEA-initiated mice observed for 60 weeks, although nearly 50% of these animals developed hepatocellular lesions. To investigate whether hepatocellular lesions in NDEA-initiated mice or spontaneous hepatocellular lesions promoted by PB in mice given PB but no NDEA, progress to hepatoblastomas later in life, mice exposed to NDEA alone and PB alone were maintained for 110 weeks. Hepatocellular tumors (adenomas and carcinomas) occurred in almost all (97%) mice given NDEA alone. However, only 10% of NDEA-treated mice developed hepatoblastomas. Thus, despite its ability to induce hepatocellular neoplasms, NDEA treatment alone was rarely sufficient to induce hepatoblastomas in these mice. In contrast, PB treatment in the absence of NDEA initiation promoted the development of spontaneously occurring hepatocellular lesions, a significant number (37%) of which progressed to hepatoblastomas. Our observations clearly show that in this animal model the development of hepatoblastoma from its precursor cells (hepatocellular adenoma and carcinoma cells) occurs predominantly in the presence of promoting agents such as PB.

Dissimilar frequency of hepatoblastomas and hepatic cystadenomas and adenocarcinomas arising in hepatocellular neoplasms of D2B6F1 mice initiated with N-nitrosodiethylamine and subsequently given Aroclor-1254, dichlorodiphenyltrichloroethane, or phenobarbital

Aroclor-1254 (Ar-1254) and dichlorodiphenyltrichloroethane (DDT) were compared to phenobarbital (PB) for their ability to promote hepatocellular proliferative lesions to hepatocellular adenomas and carcinomas and to hepatoblastomas in D2B6F1 male mice initiated with N-nitrosodiethylamine (NDEA). Hepatocellular neoplasms developed in all mice given NDEA and were more numerous in mice fed promoters. Multiplicities decreased in the order Ar-1254 > PB > DDT, indicating that Ar-1254 was more potent than either PB or DDT at the dosage levels used. PB was the most effective of the 3 agents in stimulating the evolution of hepatocellular neoplasms to hepatoblastoma. The incidence of hepatoblastomas in the NDEA.PB group was 72% but was only 27% in NDEA-initiated, DDT-promoted mice and 33% in low-dose and only 9% in high-dose Ar-1254-promoted mice. In contrast, lesions resembling benign and malignant cholangiocellular neoplasms were frequently found within hepatocellular tumors in Ar-1254-promoted mice but not in mice fed PB or DDT, either alone or after NDEA. Some cystic glandular structures in Ar-1254-promoted mice contained mucous cells, argentaffin cells, and Paneth cells and thus constituted intestinal metaplasia. Hepatoblastoma and intestinal metaplasia/cholangiocellular tumor morphology appear to constitute different patterns of genetic programming induced by certain promoters in expanding clones of initiated hepatocytes, on favorable genetic backgrounds such as that of D2B6F1 male mice.

The stem cells of the liver ? a selective review

The current status of the much-debated question of the still-hypothetical stem cells of the liver is reviewed, with an emphasis on their role in hepatocarcinogenesis. The widely held view of the primacy of the hepatocyte, notably of the mononuclear diploid type, in this process--the "hepatocytic theory"--has been compared with variants of the "stem cell hypothesis" based on the "non-parenchymal epithelial cells" of the liver--the "oval" or biliary ductular cells, the "nondescript periductular" cells and the "primitive" bipotential epithelial cells. An attempt has been made to concentrate mainly on the more recent publications, in an effort to balance the conflicting opinions expressed by comparing results obtained by the newer procedures currently in use. Despite some interesting and relevant findings it appears that the evidence in favour of the stem-cell hypothesis is still circumstantial and that the hepatocytic theory has not been invalidated. Presumably the question of the hepatic stem cells will be answered when the riddle of hepatocarcinogenesis has been solved.

Transplantable mouse hepatoblastoma: histologic, ultrastructural and immunohistochemical study

Hepatoblastoma found adjacent to a liver cell adenoma in an aged (CBA x C57Bl/6)F1 male mouse was transplanted to the syngeneic host and passed through 30 generations. Histologically, tumours that grew on transplantation retained principal morphological features of the primary tumour. Transplanted tumours were negative for AFP and for antigen A6. This antigen in the normal mouse liver is found in epithelial cells lining bile ducts and ductules including the terminal Hering canals. Some of the Hering cells were consistently A6 negative under normal conditions, and the suggestion is made that these A6 negative cells might be the cells of hepatoblastoma origin.

Applications of immunohistochemistry in rodent tumor pathology

Immunohistochemistry can serve as a valuable adjunct to rodent tumor pathology. Specific antigens may be localized to cells and tissues in normal organs, preneoplastic lesions, and benign and malignant tumors. The immunoreactivity of polyclonal and monoclonal antibodies to these antigens provide a more accurate basis for tumor diagnosis and aid in understanding pathogenesis. Ultimately, the application of more precise understanding of tumor histogenesis and diagnosis will lead to more accurate interpretations of tumor incidence data for safety assessment in toxicology.