Search

In this study we show that benzo[a]pyrene (B[a]P) and the cyclopenta polycyclic aromatic hydrocarbons (CP-PAH) cyclopenta[c,d]pyrene (CPP), benz[j]aceanthrylene (B[j]A) and benz[l]aceanthrylene (B[l]A) induce apoptosis in Hepa1c1c7 cells, as measured by fluorescence microscopy and flow cytometry. The compounds induced formation of the active form of caspase-3, cleavage of its intracellular substrate, poly(ADP-ribose)polymerase (PARP), and DNA fragmentation. B[j]A was found to be the most potent in inducing apoptosis, followed by B[a]P, CPP and B[l]A. All compounds increased expression of CYP1A1 with relative potencies B[j]A > B[a]P >> CPP > B[l]A, corresponding well with their relative apoptotic responses. alpha-Naphthoflavone (alphaNF), an inhibitor of CYP1A1, reduced the induced apoptosis. B[a]P and CP-PAH exposure also resulted in an accumulation of the tumour suppressor protein p53. No changes were observed in the protein levels of Bax and Bcl-2, whereas the anti-apoptotic Bcl-xl protein was down-regulated, as judged by western blot analysis. Fluorescence microscopic analysis revealed a translocation of p53 to the nucleus and of Bax to the mitochondria. Furthermore, caspase-8 was activated and Bid cleaved. Interestingly, the levels of anti-apoptotic phospho-Bad (Ser155 and Ser112) had a biphasic increase after B[a]P or CPP treatment. Whereas alphaNF markedly reduced the activation of B[a]P to reactive metabolites, as measured by covalent binding to macromolecules, it did not inhibit the up-regulation of phospho-Bad. Neither of the compounds triggered apoptosis in primary cultures of rat lung cells (Clara cells, type 2 cells and lung alveolar macrophages), possibly due to a lack of CYP1A1 induction. In conclusion, B[a]P and the CP-PAH induced apoptotic as well as anti-apoptotic signals in Hepa1c1c7 cells.

The genotoxic effects of the environmental contaminants benz[j]aceanthrylene (B[j]A), benz[l]aceanthrylene (B[l]A) and benzo[a]pyrene (B[a]P), and the metabolism of radiolabelled B[j]A, were studied using rat lung microsomes and various types of isolated rat lung cells from control and Aroclor 1254 (PCB) treated animals. All three compounds (10 or 20 microg/plate) resulted in low, but detectable, levels of His+ revertants in the Salmonella assay when plated with control lung microsomes. The two cyclopenta polycyclic aromatic hydrocarbons (CP-PAH) B[j]A and B[l]A, gave increased levels of revertants when plated with microsomes from PCB-treated animals. Clara cells, type 2 cells and alveolar macrophages isolated from control rats were exposed to B[j]A, B[l]A or B[a]P (30 microg/ml, 1 h), but neither of the cell types showed any DNA damage when measured by alkaline filter elution. However, both B[j]A and B[l]A (30 microg/ml, 2 h) caused DNA adducts in all three cell types, measured by the 32P-post-labelling technique, whereas no B[a]P adducts were detected (30 microg/ml, 2 h). The total DNA adduct levels in Clara cells, type 2 cells and macrophages exposed to B[j]A were 0.085 +/- 0.033, 0.053 +/- 0.001 and 0.170 +/- 0.030 fmol/microg DNA, respectively, whereas the total levels in cells exposed to B[l]A were 0.140 +/- 0.070, 0.140 +/- 0.030 and 0.220 +/- 0.080 fmol/microg DNA, respectively. Cells exposed to B[j]A revealed only one adduct which corresponds with the B[j]A-1,2-oxide DNA adduct. Judged from high performance liquid chromatography (HPLC) analysis using radiolabelled B[j]A (30 microg/ml, 30 min), the major metabolite formed in control microsomes was B[j]A-1,2-diol. Thus, oxidation at the cyclopenta ring appears to be the most important activation pathway for B[j]A with control rat lung cells. Exposure of lung cells to CP-PAH (30 microg/ml, 2 h) isolated from PCB pretreated rats resulted in slightly increased DNA adduct levels in Clara cells and macrophages when compared to cells isolated from control rats. Furthermore, the adduct pattern had shifted, and no apparent B[j]A-1,2-oxide adduct could be detected on the thin layer chromatography (TLC) plate. In contrast, the major metabolite formed with microsomes from PCB-treated animals was still the B[j]A-1,2-diol.

Autophagy, measured as the sequestration of an endogenous cytosolic enzyme (LDH), showed a progressive rate reduction during diethylnitrosamine-induced rat liver carcinogenesis. In primary hepatocellular carcinomas the autophagic activity was only one-fourth of that seen in normal hepatocytes. Reduced autophagy was also observed in peritumorous hepatocytes and in cells from preneoplastic liver, and a complete suppression of autophagic protein degradation was seen in normal hepatocytes treated with ascitic fluid from an ascites hepatoma, suggesting that tumour cells and their precursors may produce autophagy-suppressive factors with an autocrine and paracrine action. In cells from the transplantable rat ascites hepatoma, Yoshida AH-130, autophagic activity was negligible during active (logarithmic) growth, but increased to approximately 0.4%/h at high cell density, i.e. in stationary phase. In contrast to normal hepatocytes, autophagy in the AH-130 cells was not inhibited by ascitic fluid. The hepatoma cells would thus appear to have lost some aspects of autophagy regulation while retaining others. However, even the highest rate of hepatoma cell autophagy was only one-tenth of the maximal activity seen in normal hepatocytes, confirming the hypothesis that reduced autophagy may be an important aspect of growth deregulation in liver cancer.

Benz[j]aceanthrylene (B[j]A) and benz[l]aceanthrylene (B[l]A), two isomeric cyclopenta polycyclic aromatic hydrocarbons (CP-PAH) structurally related to 3-methylcholanthrene, were studied with respect to their genotoxic effects in isolated liver and lung cells. Both compounds were found to cause DNA adducts measured by the 32P-postlabelling technique. The level of DNA-adducts in rat hepatocytes exposed to 30 micrograms/ml B[l]A and B[j]A for 4 h were 46.5 +/- 22.0 and 8.3 +/- 5.1 fmol/micrograms DNA respectively. Using butanol extractions, the major and one of the minor B[j]A adducts co-chromatographed with B[j]A-1,2-oxide adducts of 2'-deoxyadenosine and 2'-deoxyguanosine. Thus, oxidation at the cyclopenta-ring of B[j]A appears to be an important activation pathway. In hepatocytes, 3-30 micrograms/ml of B[j]A and B[l]A induced DNA damage and repair measured both as increased alkaline elution of DNA and as increased incorporation of [3H]TdR in the DNA. B[l]A was somewhat more potent than B[j]A in inducing DNA repair. Reactive CP-PAH intermediates formed in the hepatocytes caused mutations in Salmonella typhimurium TA98 upon co-incubation. DNA adducts were also observed in isolated rabbit lung cells exposed to 30 micrograms/ml B[l]A or B[j]A for 2 h. A total of 14.5 +/- 6.9, 2.9 +/- 2.1 and 0.2 +/- 0.6 fmol B[l]A adducts/micrograms DNA were observed in Clara cells, type II pneumocytes and alveolar macrophages respectively. The main B[l]A adduct observed in the liver cells was not found in the lung cells. On the other hand, the levels of B[j]A adducts in the lung cells were in the range 4-14% of that found in liver cells, and no major differences between the various lung cells were observed. Neither B[l]A nor B[j]A induced DNA damage measured by alkaline elution in the lung cells, indicating that these adducts are not alkali labile.

Hepatocellular carcinomas from rats of different strains, subjected to a variety of carcinogenic treatment regimens in different laboratories (initiation by diethylnitrosamine or dimethylhydrazine, promotion by phenobarbital, 2-acetylaminofluorene, nafenopin, orotic acid or deoxycholic acid, growth stimulation by partial hepatectomy or necrogenic CCl4 treatment), were all found to be predominantly diploid by flow cytometric analysis, in contrast to normal liver tissue in which polyploid nuclei were predominant. A switch from polyploidization to diploid growth would thus seem to be a common property of malignant liver tumours. Benign neoplastic liver nodules were likewise predominantly diploid, with the exception of nodules induced by long-term deoxycholic acid treatment in Fischer rats. In addition to containing a majority of polyploid cells, the latter nodules failed to progress to the carcinoma stage.

Orotic acid (OA), a promoter of liver carcinogenesis, inhibited proliferation of primary hepatocytes in culture as monitored by labelling index, mitotic index and total DNA content. The mitoinhibitory effect of OA was seen even in the presence of a strong mitogen such as epidermal growth factor (EGF). The growth inhibitory effect of OA was not due to cell killing. Upon exposure to OA the hepatocytes exhibited an increase in the ratio of uridine nucleotides to adenosine nucleotides, and as this ratio increased the response of hepatocytes to proliferate in the presence or absence of EGF decreased. Washing the hepatocytes free of added OA resulted in a gradual decrease in the ratio of uridine nucleotides to adenosine nucleotides, paralleled by an increase in hepatocytic proliferation. Adenine, an agent that inhibits the metabolism of OA to uridine nucleotides, not only inhibited the increase in the ratio of uridine nucleotides to adenosine nucleotides but also counteracted the OA-induced mitoinhibitory effect. These results, together with our earlier observations, suggest that an imbalance in nucleotide pools composed of an increase in uridine nucleotides and a decrease in adenosine nucleotides appears to be important for OA-induced mitoinhibition.

Primary rat hepatocellular tumours, induced by a combination of diethylnitrosamine and 2-acetylaminofluorene, were examined for the presence of neuroendocrine peptides by immunocytochemical methods. Two-thirds of the tumours showed positive immunostaining for either neuron-specific enolase (NSE), protein S-100 or bombesin. NSE was commonly observed both in hepatocarcinomas and in neoplastic nodules, whereas protein S-100 was more frequently seen in carcinomas (49% positive) than in nodules (13% positive). Bombesin, previously shown to function as an autocrine growth factor in small-cell carcinoma of the lung, was present in neurosecretory granules in 13% of the nodules and 29% of the carcinomas. Normal, preneoplastic and peritumorous liver tissue, including the frequent atypical foci present in the latter two categories, was uniformly negative for all neuroendocrine markers. The foci, like the nodules and carcinomas, generally stained positively for the liver tumour marker glutathione S-transferase type P (GSTP). The results suggest that dysdifferentiation of altered hepatocytes in a neuroendocrine direction may be a common, late event in liver carcinogenesis which could possibly contribute to tumour formation, e.g. by establishing autocrine or paracrine circuits.

Hepatocytes in the adult rat are normally mostly tetraploid, but sequential carcinogen treatment (diethylnitrosamine/2-acetylaminofluorene) induces the development of a predominantly diploid hepatocyte population. By two-parametric flow cytometry (simultaneous measurement of DNA and protein content within the same cell) it could be shown that diploid hepatocytes from carcinogen-treated rats had only half the size (protein content) of the tetraploid cells from the same liver, but the same size as diploid hepatocytes from normal rats.

DNA content was measured by flow cytometry in isolated nuclei from 71 neoplastic nodules and 15 hepatocellular carcinomas isolated from rat liver at various times after treatment with an initiation--promotion regimen employing diethylnitrosamine and 2-acetylaminofluorene. Nodules and carcinomas contained mostly diploid nuclei as compared with both surrounding and normal hepatocytes which were predominantly polyploid. There appears to be a positive correlation between the degree of diploidy in nodules and their rate of proliferation. No aneuploid populations were identified in any neoplasm despite good peak resolution. These results show that an alteration in proliferation pattern from normal polyploidizing growth to diploid--diploid divisional growth is a consistent characteristic throughout the carcinogenic process in our experimental model.

2-Acetylaminofluorene (AAF), given in the diet at 0.02% for 4 weeks, is an effective promoter of liver carcinogenesis initiated by partial hepatectomy (PH) plus diethylnitrosamine (DEN) in the inbred rat strain Wistar Kyoto. AAF promotes the early (6 week) appearance of phenotypically altered (gamma-glutamyltranspeptidase-positive) cells as well as the later appearance of neoplastic nodules (2-4 months) and hepatocarcinomas (4-8 months). Promotion does not seem to involve selective cytotoxicity (selection of AAF-resistant hepatocytes), since neither AAF alone nor DEN + AAF has any inhibitory effect on overall liver growth.

The sequential treatment of young Wistar rats with two different carcinogens (diethylnitrosamine - plus partial hepatectomy - as an initiator, and 2-acetylaminofluorene as a cytotoxic selection pressure) induces the appearance of foci and nodules of liver cells which are phenotypically altered. By means of an algorithm which takes into account binuclearity as well as cell-to-cell aggregation it is possible to compute cellular ploidy distributions from flow-cytometric analysis of either hepatocyte suspensions or suspensions of hepatocytic nuclei. Cell suspensions isolated from carcinogen-treated rats can be shown to contain, already after 8 weeks, approximately 70% small, diploid hepatocytes, whereas suspensions from normal or partially hepatectomized control livers contain only approximately 10% diploid cells (the remainder being mostly tetraploid). Isolated nodules, i.e., expanding clones of proliferating cells, believed to be neoplastic precursor lesions, contained almost only diploid cells. These observations suggest that the selective outgrowth of a population of small, diploid hepatocytes may be a significant early step in the development of liver cancer.