L-asparaginase: inhibition of early mitosis in regenerating rat liver.

L-Asparaginase in agouti serum and in extracts from Escherichia coli inhibits the early wave of mitosis occurring in rat liver approximately 30 hours after hepatectomy, but even with continued treatment of the animal the later wave at 50 hours is not inhibited. This result differs from the permanent inhibition of growth which asparaginase causes in various tumors.

Fluorescent banding patterns of rat chromosomes in normal cells and primary hepatocellular carcinomas.

Quinacrine staining permits identification of rat chromosomes in metaphase that were formerly classified only in groups within the karyotype. This technique defined two types of abnormal chromosomes in cells of rat hepatomas.

Overexpression of the multidrug resistance gene mdr3 in spontaneous and chemically induced mouse hepatocellular carcinomas.

Overexpression of a family of plasma membrane glycoproteins, known as P-glycoproteins, is commonly associated with multidrug resistance in animal cells. In rodents, three multidrug resistance (mdr or pgp) genes have been identified, but only two can confer the multidrug resistance phenotype upon transfection into animal cells. Using the RNase protection method, we demonstrated that the levels of three mdr gene transcripts differ among mouse tissues, confirming a previous report that the expression of these genes is tissue specific (J.M. Croop, M. Raymond, D. Huber, A. DeVault, R. J. Arceci, P. Gros, and D. E. Housman, Mol. Cell. Biol. 9:1346-1350, 1989). The levels of mdr transcripts were determined for mouse liver tumors spontaneously arising in both C3H/HeN and transgenic animals containing the hepatitis B virus envelope gene and for tumors induced by two different carcinogenic regimens in C57BL/6N and B6C3-F1 mice. The mdr3 gene was overexpressed in all 22 tumors tested. Our results demonstrate that overexpression of the mdr3 gene in mouse liver tumors does not require exposure of the animals to carcinogenic agents and suggest that its overexpression is associated with a general pathway of hepatic tumor development. The overexpression of the mdr3 gene, which is the homolog of human mdr1 gene, in hepatocellular carcinomas may be responsible for the poor response of these tumors to cancer chemotherapeutic agents.

Thioacetamide hepatocarcinogenesis.

Thioacetamide (TAA) was administered to inbred male ACI rats for 1 year to determine its hepatocarcinogenicity. The carcinogen was fed at a level of 0.035% in a semipurified diet so that other components could be manipulated. The following groups were used: group I, TAA alone; group II, TAA with riboflavin content at 0.002% (one-third of the daily requirement); group III, TAA plus a supplement of 0.05% crude bovine thyroid; and group IV, TAA, 0.002% riboflavin, and thyroid supplement. All groups demonstrated cirrhosis, neoplastic nodules, and cholangiofibromas to varying degrees. Group IV was invariably the most advanced in all lesions, while group II demonstrated the next most advanced picture of cholangiofibromas. These lesions persisted for 1 year after cessation of the diet. Primary hepatocarcinomas (PHC) were produced by groups I, III, and IV, with the highest number being in group IV. Group II evoked only primary cholangiocarcinomas (PCC). A number of both types of carcinomas transplanted successfully. Thus TAA appears to be a hepatocarcinogen capable of inducing either PHC or PCC, with the tumor type evoked somewhat dependent on the associated diet.

Fibronectin production by human mammary cells.

Human mammary cells were examined for the presence of the high-molecular-weight surface glycoprotein fibronectin. Early passage mammary epithelial cell and fibroblast cultures from both carcinomas and normal tissues were tested for the presence of cell-associated fibronectin by immunofluorescence microscopy and for the synthesis and secretion of fibronectin by specific immunoprecipitation of metabolically labeled protein. In vivo frozen sections of primary carcinomas and normal tissues were tested for the localization of fibronectin by immunofluorescence microscopy. In contrast to the extensive fibrillar networks of fibronectin found in the fibroblast cultures, the epithelial cell cultures from both tissue sources displayed a pattern of cell-associated fibronectin characterized by powdery, punctate staining. However, the cultured epithelial cells, as well as the fibroblasts, secreted large quantities of fibronectin into the medium. Putative myoepithelial cells also displayed extensive fibrillar networks of fibronectin. The difference in cell-associated fibronectin distribution between the epithelial cells and the fibroblasts and putative myoepithelial cells provided a simple means of quantitating stromal and myoepithelial cell contamination of the mammary epithelial cells in culture. In vivo, normal tissues showed fibronectin primarily localized in the basement membrane surrounding the epithelial cells and in the stroma. Most primary carcinomas displayed powdery, punctate staining on the epithelial cells in addition to the fibronectin present in the surrounding stroma.

Morphological classification of mouse liver tumors based on biological characteristics.

Examination of three strains of inbred mice suggested that specific morphological types of hepatic tumors are the result of genetic predisposition while other types of tumors are associated with exposure to chemical carcinogens. A simple classification system is proposed. Additional studies indicated that a putative promoting agent, such as phenobarbital, increased the incidence of those tumor types usually associated with spontaneous appearance and only in strains with a genetic predisposition to spontaneous tumorigenesis. Carcinogens induce an increase in all types of tumors.

Inhibition of spontaneous hepatocarcinogenesis in C3H/HeN mice by transplanted hepatocellular carcinomas.

Cell suspensions of tumor fragments derived from spontaneous or chemically induced primary hepatocellular carcinomas obtained from inbred C3H/HeN mice were transplanted into young male mice of the same strain. Transplantable hepatocellular carcinomas were excised as soon as they were detected, and all recipient mice were killed at one year of age. In control C3H/HeN mice, the incidence of primary hepatocellular carcinomas was 41% (41 of 100). In mice in which there was no growth of transplantable carcinomas, whether originally given injections of tumor cell suspensions or fragments, the overall incidence of primary hepatocellular carcinomas was 49% (35 of 72) with one to six tumors per liver at time of sacrifice. Transplantable hepatocellular carcinomas were established only in mice that had received tumor fragments. In these mice, from which established transplantable hepatocellular carcinomas had been excised, the overall incidence of primary hepatocellular carcinomas was 12% (3 of 25) with one tumor being found in each of three livers. The time of appearance of excision of transplantable hepatocellular carcinoma did not affect this decrease in primary hepatocellular carcinoma incidence.

Patterns of spontaneous metastasis of transplantable hepatocellular carcinomas.

A number of transplantable rat hepatocellular carcinomas of varied phenotype were examined for their ability to metastasize. A striking diversity of pattern was observed, ranging from presentation in almost every organ to none at all. No relationship between metastatic capability and growth rate, tumor size, chromosome composition, or other functional characteristics was noted. Interestingly, several of the tumors demonstrated a lymphatic to vascular route of spread, very similar to that of human tumors.

Lack of proportionality between rate of cell division and induction of tumors in carcinogen-exposed regenerating livers.

Although a single pulse of dimethylnitrosamine administered during the regenerative response of liver subsequent to 70% hepatectomy resulted in more primary hepatocellular carcinomas in treated livers than in controls, the response was not proportionate to the level of cell division. Further, the use of 55-g male Sprague-Dawley rats that displayed an extremely active regenerative response did not significantly shorten the lag period before the appearance of tumors. Of additional interest was the finding that the post-S period might be even more susceptible to dimethylnitrosamine than the S phase was. These results support the suggestion that many aspects of the interaction between carcinogens and dividing cells and the requirement for subsequent events for the development remain unclear.

Alteration of hepatocytes by subcarcinogenic exposure to n-2-fluorenylacetamide.

These experiments examined the effects of a single, subcarcinogenic dose of dimethylnitrosamine or N-hydroxyfluorenylacetamide when administered after a subcarcinogenic dietary regimen of N-2-fluorenylacetamide. Control rats that received either carcinogen diet alone or a single dose of carcinogen demonstrated neither hepatic nodules nor hepatocellular carcinomas. Those animals that received dimethylnitrosamine subsequent to carcinogen diet demonstrated many persistent hepatic nodules and 100% hepato-cellular carcinomas. These data support the concept that the nodules produced by subcarcinogenic ingestion of N-2-fluorenylacetamide are not composed simply of normal hepatocytes undergoing compensatory regeneration but consist of cells that have been altered by the carcinogen. One manifestation of this alteration is an increased susceptibility to further carcinogenic evolution.

Tumor phenotype and susceptibility to progression as an expression of subpopulations of initiated murine cells.

Currently, it is conceived that a number of events, or hits, are required for the induction of tumors by chemical agents. The first phase of this sequence, initiation, is considered to result from at least one event in the genetic apparatus. Analyses of this sequence, however, usually give little consideration to the nature of the target cell or to the characteristics of the resultant tumors. Vesselinovitch et al. (Cancer Res., 38: 2003-2010, 1978) have reported that a single, small pulse of carcinogen can induce early and numerous liver tumors when administered neonatally to mice with a genetic predisposition to hepatotumorigenesis. In the current study, the nonpredisposed strain C57BL/6N was also shown to be highly susceptible to diethylnitrosamine during the neonatal period. C57BL/6N demonstrated large numbers of two of the three types of liver tumors seen in livers of genetically predisposed mice, one of which required the additional stimulus of dietary phenobarbital for growth. Tumors of more malignant phenotype were demonstrated only in genetically predisposed mice (C57BL/6N X C3H/HeN F1) that received one dose of carcinogen. These findings suggest that the phenotype of a tumor that results from a pulse of a chemical carcinogen may depend upon the target cell. The initiated cells that result from this hit may vary from those that demonstrate very little progression in cell type and may or may not require exogenous enhancement of growth to those that can progress very rapidly to fully malignant behavior. The latter might arise from a hit in a genetically initiated cell, the result of which is a more rapid progression in tumor type.

Sequential phenotypic and biochemical alterations during chemical hepatocarcinogenesis.

Chronic exposure to chemical carcinogens induces in the target tissue a series of complex morphological and biochemical alterations that precede the appearance of overt cancer. Three types of experiments are described: (a) exposure of livers that had received subcarcinogenic doses of N-2-fluorenylacetamide to a subcarcinogenic dose of dimethylnitrosamine resulted in a 100% yield of neoplastic nodules and hepatocellular carcinoma: (b) neither normal hepatocytes nor those obtained from neoplastic nodules were agglutinated by any of the lectins tested. This finding was also true for slowly growing cells from carcinomas, while those of rapidly growing carcinomas were agglutinated by several lectins; (c) analysis of nonhistone proteins isolated from neoplastic nodules demonstrated the appearance of many new species in euchromatin when compared with normal liver. Carcinomas demonstrated an even greater number of new species and they were demonstrated in heterochromatin as well.

Role of the AKR gene locus AKv-1 in susceptibility to chemical induction of thymic lymphomas.

Various strains of mice demonstrate widely differing susceptibility to chemical induction of thymic lymphomas, in both timing and incidence. In AKR mice tumors appear very early and at high incidence after a single dose of N-methyl-N-nitrosourea, while in other strains they appear later and at lower incidences. In an attempt to determine the potential role of AKR ecotropic murine leukemia virus loci in this process, congenic mice of NFS/N background, into which the highly productive ecotropic murine leukemia virus loci AKv-1 or AKv-2 has been transferred, were challenged with N-methyl-N-nitrosourea. Although they had a lower incidence of thymic lymphomas than did the parental donor AKR, the NS.AKv-1 mice had a tumor incidence twice that of NFS/N or NS.AKv-2. However, no difference in timing was noted, and these three strains demonstrated tumor appearance much later than that of AKR/N. It is suggested that the presence of the AKv-1 loci, or a gene of the closely associated genomic region, increases the number of target cells that are susceptible to N-methyl-N-nitrosourea.

Progression of tumor histiotype during mouse hepatocarcinogenesis associated with the viable yellow (Avy) gene.

Increasing attention has been focused recently upon those factors in carcinogenesis that are responsible for the proliferation of initiated cells and the increasingly aberrant phenotype that they progressively manifest. The agouti locus allele Avy (viable yellow) has been shown to be associated with conditions which favor promotion of cells that have been initiated by a wide variety of causes, in many organs, but has not been previously associated with tumor progression in those systems. In the current study, the presence of the Avy gene in a strain of mice not normally predisposed to hepatocarcinogenesis, C57BL/6N was, for the first time, associated not only with much earlier appearance, but with progression of the histiotype of hepatic tumors, following neonatal administration of diethylnitrosamine. At 52 weeks, 28 C57BL/6N mice demonstrated 7 mouse liver tumors 0.5 cm or greater in diameter, all of more benign histiotype, without associated metastasis. The 31 C57BL/6N-Avy demonstrated 194 mouse liver tumors at that time, 22% of which were of malignant histiotype, 19% of which were associated with metastasis. This system would appear to offer the possibility of identifying the underlying mechanisms for components of the carcinogenic process. In addition, the C57BL/6N-Avy mouse appears to offer advantages as a test animal in bioassay procedures that use the liver as a target organ. Thus, it represents a mouse with little or no spontaneous predisposition to hepatocarcinogenesis, with a predicted short lag period toward response to hepatocarcinogens.

Heme enzyme patterns in rat liver nodules and tumors.

Chemically induced rat hepatocyte nodules and carcinomas have a reduced capacity to oxidize drugs. The reduction in monoxygenase activity results largely from the partial loss of cytochrome P-450, a heme-containing terminal electron acceptor. To determine whether the cytochrome P-450 deficit was indicative of an altered heme metabolism, we quantitated four heme-containing proteins in normal rat liver and in rat liver nodules and cancers induced by 2-acetylaminofluorene or diethyl-nitrosamine: cytochrome P-450; cytochrome bs; catalase (EC 1.11.1.6); and tryptophan 2,3-dioxygenase (EC 1.13.11.11). The amounts of these components in nodules were 45%, 88%, 50%, and 59% of normal liver, respectively; in 2-acetylaminofluorene-induced cancers, 65%, 74%, 64%, and 65%, respectively; and in diethylnitrosamine-induced cancers, 40%, 69%, 56%, and 52%. delta-Aminolevulinic acid synthase (EC 2.3.1.37), the rate-limiting enzyme in the heme synthetic pathway, and heme oxygenase (EC 1.14.99.3), a degradative enzyme, were also quantitated. The amounts of these enzymes in nodules were 95% and 138% of normal liver, respectively, whereas in 2-acetylaminofluorene-induced cancers, they were 47% and 233%, and in diethylnitrosamine-induced cancers, they were 50% and 175%. These data indicate that four nonmitochondrial liver hemoproteins were diminished to about the same extent in hepatic nodules and cancers. Nodules and cancers also demonstrated an increased capacity for heme degradation, while cancers also demonstrated a decreased capacity for heme synthesis. Thus, the resistance of nodules and tumors to P-450-activated cytotoxic agents may ultimately result from a disturbance in heme metabolism.

Heme enzyme patterns in genetically and chemically induced mouse liver tumors.

Chemically induced rat hepatocyte nodules and hepatomas have repeatedly been shown to be deficient in Phase I drug-metabolizing enzymes. Some of these reduced activities are attributable to a diminution of the heme-containing terminal electron acceptor, cytochrome P-450. We recently demonstrated that spontaneous mouse liver tumors exhibit the same deficiency. Therefore, chemically induced and spontaneous liver tumors share common metabolic alterations which are likely to represent intrinsic characteristics of the tumorigenic process and are independent of its etiology. To determine whether the cytochrome P-450 deficit was the result of an altered heme metabolism, we quantitated four heme-containing proteins in normal mouse liver, spontaneous mouse liver tumors, and those induced by a single injection of diethylnitrosamine: cytochrome P-450; cytochrome b5; tryptophan 2,3-dioxygenase (EC 1.13.11.11); and catalase (EC 1.11.1.6). The amounts of these components in spontaneous tumors relative to normal liver were 0.35, 0.68, 0.76, and 0.51, respectively. Similar values were obtained with chemically induced tumors. The enzymes delta-aminolevulinic acid synthase (EC 2.3.1.37), the rate-limiting enzyme in the heme synthetic pathway, and heme oxygenase (EC 1.14.99.3), a degradative enzyme, were also quantitated. The amounts of these enzymes in spontaneous tumor relative to liver were 0.49 and 1.51, respectively. Again, similar values were observed for the chemically induced tumors. Alteration of the latter two enzyme activities may be sufficient for the altered hemoprotein patterns seen in mouse liver tumors. Further, this pattern of metabolic alteration is common to both chemically induced and spontaneous tumors. Thus, tumor resistance to cytotoxic agents activated by the monooxygenase system is not necessarily induced by exposure to these agents, nor as a result of selection.

alpha-Fetoprotein levels and hepatic alterations during chemical carcinogenesis in C57BL/6N mice.

Hepatocellular carcinomas were induced by administration of acetylaminofluorene or chlordane to C57BL/6N mice. Lesions which closely resembled the neoplastic nodule described as a putative premalignant lesion in rats were evident. alpha-Fetoprotein elevations were noted only in the presence of the malignant lesions for both carcinogens. In this regard, the responses of these mice were similar to those seen during spontaneous hepatocarcinogenesis.

Progressive DNA damage in hepatic nodules during 2-acetylaminofluorene carcinogenesis.

The method of alkaline elution was used to detect DNA alteration in rat liver throughout the carcinogenic sequence that resulted from exposure to a standard four-cycle feeding regimen of 2-acetylaminofluorene (AFF). At the end of 3 weeks of AAF feeding, DNA from aliquots of whole liver demonstrated a small but significant degree of damage. By the end of the fourth AAF feeding, the liver exhibited numerous nodules that could be dissected free from surrounding tissue. DNA from these putative premalignant lesions showed approximately 136% more damage than that seen at the end of 3 weeks of AAF feeding. Two to 4 months following the cessation of AAF, DNA from persistent nodules was examined for eivdence of alteration. Despite the prolonged absence of exposure to AAF, DNA damage was found to have progressed and was as much as 320% greater than that seen at the end of the first cycle. While the persistence of DNA damage during chronic exposure to a carcinogen has been reported previously, the observations that the DNA of a focal lesion which is putatively premalignant was damaged and, further, that this alteration progressed in the absence of a carcinogen, are unique.

Diminished hepatic binding protein for desialylated glycoproteins during chemical hepatocarcinogenesis.

The lectin hepatic binding protein has a specific binding capacity for desialylated serum glycoproteins and is limited to hepatocyte membranes. This binding capacity was reduced by approximately 60% in the neoplastic nodules which resulted from exposure of rat livers to N-2-acetylaminofluorene. The binding capacity of the primary hepatocellular carcinomas which resulted from this regimen was reduced by 95%. The loss of binding capacity was found to be proportional to the decreased concentration of immunologically detectable lectin in the altered tissues.

Heme synthesis in normal mouse liver and mouse liver tumors.

Hepatic cancers from mice and rats demonstrate decreased levels of delta-aminolevulinic acid synthase, the rate-limiting enzyme in the heme synthetic pathway, and increased heme oxygenase, the heme-catabolizing enzyme. These findings suggest that diminution of P-450, b5, and catalase in these lesions may result from a heme supply that is limited by decreased heme synthesis and increased heme catabolism. Heme synthesis was measured in mouse liver tumors (MLT) and adjacent tumor-free lobes (BKG) by administering the radiolabeled heme precursors 55FeCl3 and [2-14C]glycine and subsequently extracting the heme for determination of specific activity. Despite reduced delta-aminolevulinic acid synthase activity in MLT, both tissues incorporated [2-14C]glycine into heme at similar rates. At early time points, heme extracted from MLT contained less 55Fe than that from BKG. This was attributed to the findings that MLT took up 55Fe at a slower rate than BKG and had larger iron stores than BKG. The amount of heme per milligram of protein was also similar in both tissues. These findings militate against the hypothesis that diminished hemoprotein levels in MLT result from limited availability of heme. It is probable, therefore, that decreased hemoprotein levels in hepatic tumors are linked to a general program of dedifferentiation associated with the cancer phenotype. Diminution of hemoprotein in MLT may result in a relatively increased intracellular heme pool. delta-Aminolevulinic acid synthase and heme oxygenase are, respectively, negatively and positively regulated by heme. Thus, their alteration in MLT may be due to the regulatory influences of the heme pool.