Mechanisms of nitric oxide-induced cytotoxicity in normal human hepatocytes.

Chronic exposure of hepatocytes to reactive nitrogen species (RNS) following liver injury and inflammation leads not only to functional and morphological alterations in the liver but also to degenerative liver diseases and hepatocellular carcinoma. Previously, we showed that S-nitroso-N-acetylpenicillamine-amine (SNAP), which generates nitric oxide, and 3-morpholinosydnonimine (Sin-1), which generates equal molar concentrations of superoxide and nitric oxide resulting in peroxynitrite production, exhibited different levels of cytotoxicity to normal human hepatocytes in culture. The aim of the present study was to elucidate some of the molecular and cellular pathways leading to hepatocyte cell death induced by RNS. Following treatment of the hepatocytes with SNAP or Sin-1, gene-specific DNA damage was measured in mtDNA and a hprt gene fragment using a quantitative Southern blot analysis. Both agents induced dose-dependent increases in DNA damage that was alkaline labile, but not sensitive to both formamidopyrimidine-DNA glycosylase (fpg) and endonuclease III, which recognize 8-oxoguanine, thymine glycol, and other oxidized pyrimidines. DNA damage was two- to fivefold greater in mtDNA than in the hprt gene fragment. There was a persistent and marked increase in DNA damage posttreatment that appeared to arise from the disruption of electron transport in the mitochondria, generating reactive species that saturated the repair system. DNA damage induced by Sin-1 and SNAP led to cell-cycle arrest in the S-phase, growth inhibition, and apoptosis. The data support the hypothesis that the functional and morphological changes observed in liver following chronic exposure to RNS are, in part, the result of persistent mitochondrial and nuclear DNA damage.

Modulation of Ki67, p53 and RARbeta expression in normal, premalignant and malignant human oral epithelial cells by chemopreventive agents.

BACKGROUND: Aberrant expression of Ki67, p53 and RARbeta are characteristic of many tumor types including those of the oral cavity. Chemopreventive agents may act by modulating their expression to more normal levels. MATERIALS AND METHODS: The effects of 21 chemopreventive agents on the expression of Ki67, p53 and RARbeta were determined using a human in vitro model of normal, premalignant and malignant oral epithelial cell lines. RESULTS: Ki67 and mutant p53 (mtp53) were overexpressed in both the premalignant and malignant cell lines, whereas expression of RARbeta was high in the normal, low in the premalignant and not detectable in the malignant cell lines. Most of the agents selectively inhibited the expression of Ki67 in the premalignant and malignant cell lines. Eight of the 21 agents increased, while four agents decreased, the levels of mtp53 protein in the premalignant cell line. In the malignant cell line, five of the agents increased, while ten agents decreased mtp53 protein levels. The agents increased RARbeta expression to near normal levels in the premalignant cell line. CONCLUSION: The data suggest that the suppression of Ki67 and mtp53 are good indicators of the effectiveness of agents in premalignant and malignant oral cells, whereas the enhancement of RARbeta is a measure of effectiveness in premalignant oral cells.

Identification and isolation of human epithelial cell colonies that express specific gene products.

A new and simple method for the identification and isolation of human cell clones expressing specific gene products is presented. This technique is analogous to the colony blotting techniques described for prokaryotes. In this technique, human epithelial cells are grown in tissue-culture dishes, and cells from the colonies on the dish are transferred in situ to a high-density cationized quaternary amine-charged nylon membrane. The membrane can then be processed multiple times, using antibody and/or nucleic acid probes, for the identification of those cells producing the desired protein, mRNA or DNA sequence. Once identified, the desired cell colony is isolated for expansion, direct DNA sequencing and/or cell function. We demonstrate the potential of the technique by identifying and isolating colonies of replicating normal human liver hepatocytes producing albumin and keratin.

The establishment and continuous subculturing of normal human adult hepatocytes: expression of differentiated liver functions.

The use of normal adult liver hepatocytes in cell culture for biochemical, toxicological and pharmacological studies has been greatly limited owing to the loss of replicative capacity and differentiated liver function. This is contrary to the ability of the liver to regenerate following injury in vivo. This suggests that liver "stem" or "transitional" hepatocytes exist that upon proper stimulus divide and differentiate into mature hepatocytes. In this study we report the establishment and culture of hepatocytes from normal human adult liver, which: (1) possess replicative capacity sufficient to subpassage 12-15 times (27-37 cumulative population doublings); (2) can be cryopreserved for subsequent use without loss of replicative capacity; and (3) upon differentiation in culture synthesize albumin and keratin 18 and metabolize benzo[a]pyrene. The ability of these cells to divide or express differentiated functions appears to be due to a number of cellular, biochemical and physical characteristics that are present during the primary establishment and subsequent growth phases of the cell cultures. Disassociation of cells from excess liver tissue was best achieved by combining the mechanical action of the Stomacher with very low amounts of proteolytic enzymes and EGTA. The cell lines appeared to grow best when established and subpassaged in an mALPHA medium supplemented with insulin, hydrocortisone, transferrin, epithelial growth factor and fetal bovine serum (prescreened for human hepatocyte cell growth). The seeding density and cell-cell contact in culture appeared to be important for both cell division and expression of liver function. When cells were seeded at a low density and subpassaged before confluency, the cells continued to divide. Albumin and keratin 18 synthesis occurred primarily in tightly packed cell clusters. When cells were seeded at a high density, near confluency, albumin and keratin 18 synthesis occurred uniformly in all of the cells of the culture and the culture metabolized benzo[a]pyrene to water-soluble metabolites, which covalently bound to cellular DNA. This appearance of liver functions was consistent with the "transition" of hepatocytes to a terminally differentiated state. Nonhepatic markers, i.e., alpha-fetoprotein, factor VIII and gamma-glutamyl transpeptidase activity were not expressed in cells cultured at either low or high density. Thus, the data presented here indicate that normal human adult liver hepatocytes, once established in culture, can be subpassaged to a high number of population doublings, cryopreserved for later use, and modulated to express differentiated liver functions.

An enzyme-linked immunosorbent assay (ELISA) for the detection and quantitation of the tumor marker 1-methylinosine in human urine.

A highly sensitive enzyme-linked immunoassay (ELISA) was developed to detect and quantify the tumor marker, 1-methylinosine (m1I), in human urine. The rabbit antisera was highly specific for m1I with negligible or no inhibition by other nucleosides excreted into urine. Using the competitive ELISA, nanogram amounts of m1I were easily measured directly in urine. The assay agreed with our previous hplc analysis of m1I in urine for identifying those individuals with chronic myelogenous leukemia. Thus, this assay should greatly facilitate the quantitation of m1I as a tumor marker.

Characteristics of long-term human epithelial cell cultures derived from normal human fetal kidney.

Epithelial cell cultures were prepared from normal human fetal kidney and established in long-term culture. The growth characteristics and production of keratin, and alkaline phosphatase (AP) and gamma-glutamyl transpeptidase (GGT) activities were compared in a modified minimal essential medium (mMEM), D-valine-containing modified alpha-MEM (mALPHA) and L-valine mALPHA. The mean number of cumulative population doublings (CPDL) was significantly (P less than 0.001) enhanced with the L-valine mALPHA (40.8 CPDL) over that achievable in mMEM (14.2 CPDL) or D-valine mALPHA (18.3 CPDL) media. In all three media, greater than 95% of the cells in culture produced keratin throughout the life span of these cultures. Surface-associated fibronectin was absent in these cell cultures. AP and GGT activities increased as a function of subpassage and time in culture, with the greatest activity in the L-valine mALPHA. The expression of these renal cell-associated functions suggests that these cells in culture are proximal tubule epithelial cells. The conditions and procedures described in this paper can provide a human kidney epithelial cell culture system for studying human renal function, metabolism, cytotoxicity, genotoxicity, and transformation.

Contact insensitivity of a subpopulation of normal human fetal kidney epithelial cells and of human carcinoma cell lines.

Early passage normal human fetal kidney epithelial cells were inoculated on top of a confluent monolayer of X-ray lethally irradiated human fibroblasts to determine the colony-forming ability of these epithelial cells. The results indicate that the great majority of the epithelial cells did not have the clonogenic ability on the fibroblast cell mat, although they were capable of colony formation on plastic surface without the cell mat. A small subpopulation of these epithelial cells, however, was able to proliferate on the cell mat. These contact-insensitive fetal epithelial cells were found to be deficient in gap junction-mediated intercellular communication, to contain keratin and gamma-glutamyl transpeptidase but not fibronectin. These contact-insensitive cells appear to have greater proliferative potential than the parental cell population and to exist transiently in early passage but not in late passage culture. The ability of proliferation on cell mat was found to be shared by 22 different human carcinoma cell lines that were tested. This unique clonogenic ability of normal contact-insensitive and human carcinoma cells on the cell mat could provide a selection method for presumptive normal stem and tumor cells and for an assay for screening potential antitumor drugs and assessing the efficacy of chemotherapeutic drugs against a given tumor.

Growth-promoting effect of 12-O-tetradecanoylphorbol-13-acetate on cultured normal human fetal kidney epithelial cells.

Normal human epithelial cells have been reported to be sensitive to growth inhibition by 12-O-tetradecanoylphorbol-13-acetate in contrast to neoplastic counterparts. Our studies with normal human fetal kidney epithelial cells, however, show that 12-O-tetradecanoylphorbol-13-acetate may increase cell density and promote the growth of these cells at early passages. The result, together with three previous reports showing similar effects in normal human melanocytes, prostatic epithelial cells, and an unidentified cell type in human epidermal cell culture, indicates that human cells may exhibit divergent responses to 12-O-tetradecanoyl-phorbol-13-acetate for induction of cellular differentiation or proliferation.

A method for isolating large numbers of viable disaggregated cells from various human tissues for cell culture establishment.

A method is described for the isolation of large numbers of viable disaggregated cells from human tissues. This method combined the mechanical action of a Stomacher Model 80 Lab Blender, 0.1 mg/ml trypsin or 0.5 mg/ml collagenase, and 0.1 mM [ethylene bis(oxyethylenenitrolo)]-tetraacetic acid (EGTA). Tissue (0.2 to 1.0 g) obtained from human fetal intestine, kidney, liver, lung, and skin were separately minced into approximately 1-mm3 pieces. The pieces were placed in a sterile bag containing 60 ml of calcium- magnesium-free phosphate buffered saline, the appropriate enzyme (0.1 mg/ml trypsin or 0.5 mg/ml collagenase) plus 0.1 mM EGTA, and 0.1% methylcellulose. The bag was then placed into the blender and mixed at a low speed for 3 to 20 min at room temperature. After a single cell suspension was observed by phase contrast microscopy, 10 ml of bovine calf serum was added to the cell suspension to inactivate the proteolytic enzymes. At this time 130 ml of cold Hanks' balanced salts solution containing 5% bovine calf serum was added and the entire cell suspension passed through a tissue sieve (100 mesh, 140 micron) and the cells collected by centrifugation. These cells were then resuspended into the appropriate culture medium. In comparison to other methods for establishment of cell cultures from human tissues, the method described requires shorter incubation times with relatively low concentrations of proteolytic enzymes, and yields two- to three-fold greater number of cells per tissue with 86 to 93% viability. Also, depending on the cell type, 50 to 75% of the isolated cells attached to the culture vessel within 24 h. Variation of the time and concentration of digestive enzymes can be used to select different cell types for culture.

Repair of O6-methylguanine in human fetal brain and skin cells in culture.

The repair of O6-methylguanine was measured in cell cultures derived from human fetal brain and skin. Cells derived from nine different fetal specimens were treated in culture with N-methyl-N-nitrosourea (MNU). The brain cell cultures used were a mixture of glial and neuronal cell types, while the skin cell cultures were predominantly fibroblasts. The amount of O6-methylguanine initially induced and remaining in cellular DNA was quantitated as a function of time (0-4 h) by h.p.l.c. analysis of the acid hydrolyzates of in vitro alkylated cellular DNA. Very little (less than 10%) of the 7-methylguanine was lost from the DNA of both cell cultures 4 h post-treatment. Approximately 50% of the O6-methylguanine induced in cellular DNA by MNU was lost within 0.7 h in both the human fetal brain and skin cells in culture. Within 4 h, 80% of this methylated guanine was lost. The kinetics for the removal of O6-methylguanine in the brain and skin cells appeared biphasic. A rapid initial phase was followed by a gradual slower phase of repair. These studies indicate that cells derived from human fetal brain and skin exhibit the same degree of proficiency for the repair of the potential pro-carcinogenic O6-methylguanine lesion.

Quantitation of O6-ethyldeoxyguanosine in ENU alkylated DNA by polyclonal and monoclonal antibodies.

Rabbit polyclonal and mouse monoclonal antibodies were developed against O6-ethylguanosine conjugated with keyhole limpet hemocyanin. Radioimmunoassay (RIA) and a modified enzyme-linked immunosorbant assay (ELISA) were established for the determination of antigen-antibody binding and quantitation of potentially mutagenic O6-ethyldeoxyguanosine (O6-EtdGuo) in DNA treated with ethylnitrosourea (ENU) in vitro and in vivo. Optimum as well as reproducible antibody binding could be observed with conjugate concentrations at 0.1 ng/well immobilized by overnight drying at 37 degrees C. RIA was several-fold more sensitive than ELISA in detecting inhibition with O6-EtdGuo requiring 0.1 pmol for the 50% inhibition of tracer-antibody binding. In competitive inhibition assays with polyclonal and monoclonal antibodies, a linear dose resonse relation was obtained with DNA hydrolyzates alkylated in vitro with increasing concentrations of ENU. Significantly lower modification levels, e.g., 16.0 fmol O6-EtdGuo, at an O6-EtdGuo/dGuo molar ratio of 2.6 X 10(-7) in a hydrolyzate of 80 micrograms rat liver DNA ethylated in vivo with 10 micrograms ENU/g body weight was determined immunologically. The rate of elimination of O6-EtdGuo determined in human fetal kidney epithelial cells treated with 0.65 mM ENU showed that 50% of initial O6-EtdGuo was removed within 1 h followed by a slow phase of repair with 23% remaining at 8 h post-treatment.

DNA repair following ultraviolet and N-ethyl-N-nitrosourea treatment of cells cultured from human fetal brain, intestine, kidney, liver, and skin.

DNA excision repair was measured in cell cultures derived from human fetal brain, intestine, kidney, liver, and skin following ultraviolet (UV) irradiation and N-ethyl-N-nitrosourea (ENU) treatment. Cells in early passages were exposed to 5 or 10 J of UV radiation per sq m or to 25 microM to 3.5 mM ENU. DNA excision repair was determined by (a) scintillation counting and autoradiography to measure unscheduled DNA synthesis (UDS) and (b) the UV-endonuclease-sensitive site assay to measure pyrimidine dimers directly. The level of UDS following treatment of these cell cultures with UV was both time and dose dependent. UDS also increased with increasing doses of ENU up to 350 microM but decreased at doses greater than 500 microM. Cells derived from human fetal brain, kidney, and liver appeared to exhibit lower (50 to 80%) levels of UDS following UV irradiation or ENU treatment than did cells cultured from human fetal skin or intestine. The loss of UV-endonuclease-sensitive sites assayed in skin, liver, and kidney cells over a 24-hr period confirmed the differences observed by UDS in these cells. Skin cells removed 50% of the initial pyrimidine dimers from their DNA within an 8-hr period and 65 to 86% in 24 hr. Kidney and liver cells, on the other hand, removed only 28 and 32% of the initial dimers, respectively, over a 24-hr period. The data suggest differential excision repair responses following UV irradiation and ENU treatment of cells derived from different human fetal organs.