Liver carcinogenesis by diethylnitrosamine in the rat.

Diethylnitrosamine was continuously administered to rats at a dose rate of low toxicity. Ninety-two percent of the animals died with multicentrical hepatocellular carcinomata within a narrow and highly reproducible time interval. Discontinuing the carcinogen during the experiment resulted in a prolonged median time until death, a reduced tumor yield, and a lessened slope of the dose-response curve. Partial hepatectomy after discontinuation of the drug did not change either tumor yield or time of death. The obtained dose-response relationships support the concept that carcinogenic effects of single doses are irreversible and cumulative. Daily, low-dose, total-body x-irradiation had no significant effect on the response of rat liver to the carcinogen.

Neoplastic transformation of fetal rat brain cells in culture after exposure to ethylnitrosourea in vivo.

A single, transplacental pulse of N-ethyl-N-nitrosourea (ENU; 75 mug/g body wt) to the fetal (18th day of gestation) BD IX rat led to death with malignant tumors of the central and peripheral nervous system after a median time of approximately 195 days. In contrast to untreated control cells, dissociated brain cells transferred to long-term cell culture 20-90 hours after the ENU pulse became tumorigenic after approximately 200 days, as assayed by reimplantation into baby BD IX rats. This was preceded by a characteristic sequence of phenotypic alterations (termed "stages I-IV"). During early primary culture (stage I), both ENU and control cultures exhibited stationary glia-like cells on a growing layer of epithelioid (possibly glia precursors) and few fibroblast-like cells. Stage II (approximately 10th-40th day) was characterized by a constant proportion of glia-like cells in the ENU cultures and by their disappearance in the controls. During stage III (approximately 40th-100th day), slowly proliferating glia-like cells in the ENU cultures formed "piled-up" foci. They could then be removed from the underlying cell layer and cultured separately. Transition to stage IV (approximately 100th-200th day) was marked by proliferation of morphologically altered cells, which subsequently acquired the capacity of form colonies in semisolid agar and finally became tumorigenic (stage V). This system may represent a model for the study of malignant transformation.

Elimination of O6-ethylguanine from the DNA of brain, liver, and other rat tissues exposed to ethylnitrosourea at different stages of prenatal development.

The magnitude of the neurooncogenic effect of N-ethyl-N-nitrosourea (EtNU) in the BD IX rat is strongly dependent on the developmental stage of the nervous system at the time of carcinogen exposure, with a maximum during late prenatal and early postnatal development. Both with increasing postnatal age and in the direction of early embryonic development (prior to Postnatal Day 15), the yield of neuroectodermal tumors in the brain and peripheral nervous system declines sharply. Using a competitive radioimmunoassay for O6-ethyldeoxyguanosine (O6-EtdGuo), we have ascertained that the initial degree of DNA ethylation in BD IX rat tissues (including brain) is independent of the developmental stage at the time of transplacental (i.v.) exposure to a constant single dose of EtNU over a time range from Prenatal Day 11 to a postnatal age of 102 days. O6-EtdGuo is highly persistent in the DNA of peri- and postnatal rat brain but enzymatically removed from the DNA of other tissues, notably liver. The present analyses by radioimmunoassay indicate that O6-EtdGuo is equally persistent in the DNA of prenatal BD IX rats exposed to EtNU (50 micrograms/g body weight) on the 11th, 13th, or 16th day of gestation but removed enzymatically from other prenatal tissues. The rate of removal from the DNA of liver (Prenatal Day 16) is higher than the corresponding rate in 10-day-old (postnatal) BD IX rats. On Prenatal Day 11 to 12 (when a neurooncogenic effect first became apparent after transplacental exposure of BD IX rats to EtNU; S. Ivankovic and H. Druckrey, Z. Krebsforsch., 71: 320-360, 1968), the number of cells per brain is approximately 2 X 10(5). When a limited number of experimental animals are used, and regardless of the incapacity of neural precursor cells to remove O6-EtdGuo from their DNA, this target population size may be incompatible with the manifestation of a rare event such as malignant transformation.

Immunological quantification by high-affinity antibodies of O6-ethyldeoxyguanosine in DNA exposed to N-ethyl-N-nitrosourea.

Three immunological methods [radioimmunoassay (RIA), enzyme-linked immunosorbent assay, and radioimmunosorbent technique] were established for quantification of the potentially mutagenic O6-ethyldeoxyguanosine (O6-EtdGuo) in DNA treated with the carcinogen ethylnitrosourea in vivo or in vitro. To obtain high-affinity antibodies for specific detection of low levels of O6-EtdGuo in small amounts of DNA (cells), different schemes were applied for immunization of rabbits with the hapten O6-ethylguanosine coupled to various carrier proteins (rat serum albumin, bovine serum albumin, keyhold limpet hemocyanin). Low-dose immunization with the hapten-keyhold limpet hemocyanin conjugate resulted in antibodies with an affinity constant of 1 to 2 X 10(10) liters/mol and very low levels of cross-reactivity with normal as well as other alkylated DNA components. The RIA (the most sensitive of the three assays) detects 0.05 pmol of O6-EtdGuo at 50% inhibition of tracer (O6-ethyl[8,5'-3H]-3'-deoxyguanosine)-antibody binding. This permits quantification by RIA of O6-EtdGuo at an O6-EtdGuo:2'-deoxyguanosine molar ratio of approximately 3 X 10(-7) in a hydrolysate of 100 micrograms of ethylated DNA. By chromatographic separation of O6-EtdGuo prior to the RIA, this value can be lowered to less than 5 X 10(-8).

Tumor-selective modification of cellular microenvironment in vivo: effect of glucose infusion on the pH in normal and malignant rat tissues.

The pH distributions in transplanted neural (TV1A, BT1A) and hepatic (HV1A3) tumors and in brain and kidney of BDIX rats were analyzed as a function of serum glucose concentration (SGC), tumor size, and tissue architecture. Tissue damage during pH measurements in vivo could be minimized by the use of pH microelectrodes with tip diameters of less than or equal to 10 micrometers. In normoglycemic rats, the pH in TV1A tumors was only slightly lower than in brain or kidney. However, at 6 hr after the induction of hyperglycemia by continuous i.v. infusion of glucose, the average pH in TV1A tumors had fallen to 6.7 at an SGC of 27 mM and to 6.1 at an SGC of 50 mM. A similar glucose-mediated pH reduction was observed in BT1A and HV1A3 tumors. No significant increase in tissue acidity occurred in brain and kidney. The pH in tumors had reached its minimum at 2 hr after the onset of high-dose glucose infusion (SGC, 50 mM) and could be maintained at this level in hyperglycemic rats for at least 48 hr. In hyperglycemic hosts, an increased retention of acidic metabolites in the tumor tissue with decreasing vascular density was reflected by a tumor size (age)-dependent pH reduction and a higher degree of intratumoral pH variation. In partially necrotic tumors, pH values as low as 5.2 were recorded. Oral administration of NaHCO3 to tumor-bearing rats had no effect on the average pH in TV1A tumors.

Differential formation of O6-ethylguanine in the DNA of rat brain chromatin fibers of different folding levels exposed to N-ethyl-N-nitrosourea in vitro.

Extended (histone H1-depleted), 11-nm-thick chromatin fibers and condensed 25- to 35-nm-thick chromatin fibers, representing the first and second level of DNA folding in chromatin, respectively, as well as nucleosome core particles, were isolated from fetal rat brain cells and briefly exposed to N-ethyl-N-nitrosourea (EtNU) in vitro. The O6-ethyl-2'-deoxyguanosine (O6-EtdGuo):2'-deoxyguanosine (dGuo) molar ration in DNA enzymatically hydrolyzed to 2'-deoxynucleosides was determined by competitive radioimmunoassay using an anti-O6-EtdGuo monoclonal antibody with an affinity constant for O6-EtdGuo of approximately 2 X 10(10) liters/mol. In comparison to naked DNA (O6-EtdGuo:dGuo relative value, 1.0) the O6 atom of guanine proved to be increasingly protected from ethylation by EtNU in the DNA of the histone H1-depleted chromatin fibers (relative value, approximately 0.6) and of the condensed chromatin fibers (relative value, approximately 0.4). From the O6-EtdGuo:dGuo relative values obtained for the DNA of nucleosome core particles (approximately 0.5) and of H1-depleted chromatin fibers (approximately 0.6), it follows that the accessibility of the O6 atom of guanine to the electrophilic ethyldiazonium ion generated from EtNU in internucleosomal DNA (protein free) is about twice that found in nucleosomal DNA. The overall O6-EtdGuo:dGuo molar ratio in the DNA of H5 rat hepatoma cells exposed to EtNU in vitro was similar to that of the DNA of the condensed 25- to 35-nm chromatin fibers. Since the bulk of genomic DNA is organized in the form of these chromatin fibers, the overall degree of intercellular O6-EtdGuo formation appears to be mainly determined by this folding level of DNA, and not, or only to a low degree, by the DNA folding levels of higher-order chromatin structures such as those found in large loops or domains of chromatin fibers within the cell nucleus.

pH distributions in transplanted neural tumors and normal tissues of BDIX rats as measured with pH microelectrodes.

The distribution of pH values was measured in transplanted neuroectodermal TV1A tumors and in brain and kidney of BDIX rats in vivo. Tissue damage during pH measurements could be minimized by the use of Hinke-type pH glass microelectrodes with maximum diameters of the pH-sensitive tips of less than or equal to 10 micrometers (sensitivity, 58 to 60 mV/pH unit at 37 degrees; response time (95%), less than or equal to 3 sec; drift, less than or equal to 0.01 pH unit/hr). The advantages and limitations of this technique are discussed in relation to other methods for the analysis of extracellular pH. In tumors weighing 1.0 to 2.5 g, pH values ranged from 6.8 to 7.1 (mean, 7.0). The pH distribution in tumors weighing 4 to 6 g was shifted to slightly lower values, with an average pH of 6.9 (range, 6.7 to 7.1). No marked pH differences were found between the tumors and normal tissues. The pH values measured in brain and kidney ranged from 6.6 to 7.3 (mean, 7.0) and 6.7 to 7.3 (mean, 7.1), respectively. Within single tumors, local pH variations in the range of 0.2 to 0.3 pH unit were observed. The local pH values measured in certain tumor areas are, however, sufficiently low to partially inhibit proliferation and colony formation in cultured malignant cells.

Electrophysiological properties of ethylnitrosourea-induced, neoplastic neurogenic rat cell lines cultured in vitro and in vivo.

A comparative analysis was performed on the electrophysiological properties of 11 neoplastic neurogenic cell culture lines and five other cell lines of different origin (HV1C, rat bile duct carcinoma; BICR/M1RK, rat mammary tumor; HeLa, human cervix carcinoma; 3T3, mouse embryo; REe, rat embryo). Neurogenic lines were derived either from N-ethyl-N-nitrosourea-induced neoplasms of the nervous system or from cultured fetal rat brain cells that had undergone neoplastic transformation in vitro after exposure to N-ethyl-N-nitrosourea in vivo. Electrical membrane excitability was lacking in all neurogenic cells analyzed. Their membrane potential and input resistance values were similar to those of the nonneurogenic lines. Intercellular ionic coupling was consistently observed between cells of a fibroblastoid shape or cells bearing multiple cytoplasmic processes (i.e., all neurogenic lines HV1C, BICR/M1RK, and 3T3). Epithelioid cells (i.e., HeLa, REe, an NV1C subpopulation, and a GV1C1 variant) showed no such intercellular communication. In vivo monolayer cultures on glass coverslips were obtained by a modified i.p. diffusion chamber technique. Under these conditions, the cells (with the exception of a glioma-derived cell line) retained the morphological appearance and electrophysiological properties observed in vitro.

DNA excision repair profiles of normal and leukemic human lymphocytes: functional analysis at the single-cell level.

Recent evidence has linked cellular DNA repair capacity to the chemosensitivity of cancer cells to alkylating agents. Using single-cell gel electrophoresis ("comet assay"), we have analyzed the induction and differential processing of DNA damage in human lymphocytes derived from healthy donors and from patients with chronic lymphatic leukemia (CLL) after exposure to N-ethyl-N-nitrosourea in vitro. The extent of comet formation in lymphocytes after N-ethyl-N-nitrosourea exposure appears to depend predominantly on the processing of DNA repair intermediates, because strand breaks in plasmid DNA were not induced by ethylation before the addition of nuclear proteins. Although the initial level of a specific alkylation product (O6-ethylguanine) in nuclear DNA was uniform, different dose-response curves were obtained for the comet size in individual cell samples immediately after exposure, with small intercellular variation. The individual kinetics of DNA repair varied significantly between specimens derived from both healthy individuals and CLL patients; for the DNA repair half-time (t1/2), large difference was found. Pretreatment of cells with methoxyamine as a DNA repair modifier blocking the base excision repair pathway revealed a quite similar extent of base excision repair-independent DNA incision in almost all normal lymphocyte samples. In contrast, this portion varied relatively and absolutely to a great extent among individual samples of CLL lymphocytes, suggesting a loss of stringent control of DNA repair processes in these cells. The comet assay can thus be used to gain information about interindividual variation in the efficiency of different DNA repair processes in small samples of normal cells and their malignant counterparts.

Suppression of ethylnitrosourea-induced schwannoma development involves elimination of neu/erbB-2 mutant premalignant cells in the resistant BDIV rat strain.

Contrary to the response of rats of the highly sensitive inbred strain BDIX, BDIV rats are resistant to the induction of malignant schwannomas by exposure to the alkylating N-nitroso carcinogen N-ethyl-N-nitrosourea (EtNU). In BDIX rats, a point mutation at nucleotide 2012 in the transmembrane region of the neu/erbB-2 gene has proved to be a very early marker of initiated Schwann precursor cells with an elevated risk of malignant transformation, and is diagnostic of the resulting schwannomas. To gain insight into the cellular and molecular mechanisms responsible for the resistance of the BDIV strain, comparative quantitative neu mutation analyses combined with histomorphological studies were performed on the trigeminal nerves of EtNU-treated BDIV and BDIX rats as well as on their (BDIX x BDIV) F1 progeny. It was found that neu-mutant Schwann cells are initially present at comparable frequency in the trigeminal nerves of both resistant and sensitive animals. Contrasting with the progressive multiplication of mutant Schwann cells in BDIX trigeminal nerves, however, the numbers of mutant cells began to decrease during the intermediary phase of the carcinogenic process in BDIV animals, and premalignant neu-mutant cells were no longer detectable by the time BDIX rats developed full-blown trigeminal schwannomas. The resistance of BDIV rats thus involves the elimination of initiated neu-mutant Schwann cells during the postinitiation period of EtNU-induced schwannomagenesis via mechanisms that remain to be clarified.

Capacity of individual chronic lymphatic leukemia lymphocytes and leukemic blast cells for repair of O6-ethylguanine in DNA: relation to chemosensitivity in vitro and treatment outcome.

The elimination kinetics of the alkylation product O6-ethylguanine (O6eGua) from nuclear DNA were determined in individual lymphocytes or blast cells isolated from 27 patients with chronic lymphatic leukemia (CLL) and 26 patients with de novo acute myeloid leukemia (AML). A monoclonal antibody-based immunocytological assay was used for quantification of O6eGua in DNA of individual cells after pulse exposure of cells to N-ethyl-N-nitrosourea (EtNU). In cell specimens from a given patient, no major subpopulations with significantly different capacities for repair of O6eGua were observed. The time required to remove 50% of induced O6eGua residues varied interindividually between 0.5 and 8.4 h in CLL lymphocytes and between 0.8 and 6.3 h in leukemic blast cells. An inverse relationship was found between the rate of removal of O6eGua from DNA and the chemosensitivity of cells to EtNU, 1,3-bis(2-chloroethyl)-1-nitrosourea or mafosfamide in vitro. High rates of O6eGua repair and pronounced resistance to mafosfamide, 1,3-bis(2-chloroethyl)-1-nitrosourea, and EtNU in vitro were found in samples from 8 CLL patients nonresponsive to chemotherapy with alkylating agents. In AML patients treated with anthracyclines and 1-beta-D-arabinofuranosylcytosine, no relation was found between DNA repair capacity and treatment outcome. However, increased P-glycoprotein expression was observed between specimens derived from AML patients who had failed to reach complete remission (n = 12) after chemotherapy versus responsive patients (n = 14). DNA repair rate was not related to chemosensitivity to Adriamycin and 1-beta-D-arabinofuranosylcytosine in vitro, nor were cellular glutathione content, glutathione S-transferases activity, or P-glycoprotein expression.

Highly sensitive, specific detection of O6-methylguanine, O4-methylthymine, and O4-ethylthymine by the combination of high-performance liquid chromatography prefractionation, 32P postlabeling, and immunoprecipitation.

A highly sensitive and specific method for the detection of O6-methylguanine (O6-meG), O4-methylthymine (O4-meT), and O4-ethylthymine (O4-etT) has been established by combining prefractionation by high-performance liquid chromatography (HPLC), 32P postlabeling, and immunoprecipitation by monoclonal antibodies (PREPI method). DNA was enzymatically hydrolyzed to 2'-deoxynucleoside-3'-monophosphates (3'-dNps). Each alkyl 3'dNp was separated by reverse-phase HPLC, radiolabeled at the 5' position with [gamma-32P]ATP and polynucleotide kinase. After removing 3'-phosphate for better recognition by the antibodies, the resulting alkyl nucleotides were further fractionated by HPLC and finally precipitated specifically with respective antibodies. The detection limits were 1 fmol for all the alkyl nucleotides analyzed, so that one adduct in 10(8) of its normal counterpart nucleotide can be determined using approximately 100 micrograms (O4-meT and O4-etT) or approximately 150 micrograms (O6-meG) of DNA, i.e., 3-5 x 10(7) cells corresponding to approximately 10 ml of peripheral blood or a few hundred milligrams of tissue. By the use of the PREPI method, three leukocyte and three liver DNA samples from Japanese living in the Tokyo area were analyzed with respect to O-alkyl adduct content. O6-meG was detected in all three of the leukocyte samples (O6-meG:G molar ratio, 1.1 x, 0.8 x, and 1.6 x 10(-8) as molar ratios to guanine). Neither O4-meT nor O4-etT was detected (detection limit, O4-alkylT:thymine molar ratio less than 0.5 x 10(-8)). Among the liver samples analyzed, two cases showed positive O6-meG values (4.2 x and 1.1 x 10(-7) O6-meG:guanine molar ratios). Contrary to the leukocyte DNA, O4-meT (3.9 x, 4.3 x, and 7.5 x 10(-8) as O4-meT:thymine) and O4-etT (1.9 x, 4.9 x, and 8.7 x 10(-8) as O4-etT:thymine) were detected in all the liver samples. These results indicate the validity of the PREPI method for molecular epidemiological studies on DNA alkylation products.

pH in human tumor xenografts and transplanted rat tumors: effect of insulin, inorganic phosphate, and m-iodobenzylguanidine.

Various strategies to improve the therapeutic index of anticancer agents aim at inducing, by stimulation of aerobic glycolysis, temporary pH differences between malignant and normal tissues which can be exploited to activate cytotoxic agents selectively in tumors. We have investigated whether the pH reduction induced by glucose, the "drug" commonly used to increase lactic acid production in malignant tissues, can be augmented by pharmacological manipulation of tumor cell glycolysis. At normal plasma glucose concentration (6 +/- 1 mM), inorganic phosphate, a modifier of hexokinase and phosphofructokinase activity, had no effect on pH in two transplanted rat tumors and a human tumor xenograft line (average pH, 6.80; range, 6.65-6.95). When plasma glucose concentration was raised to 30 +/- 3 mM by i.v. infusion of glucose, inorganic phosphate reduced the pH in those tumors which exhibited only a moderate pH response to glucose per se (mean pH, 6.60) to an average value of 6.20 (range, 6.05-6.35). In the same setting, insulin, continuously infused at dose rates up to 600 milliunits/kg body weight/min, did not result in acidification of tumor tissue exceeding that induced by glucose alone. However, the H+ ion activity in both transplanted rat tumors and human tumor xenografts was increased by m-iodobenzylguanidine (MIBG), an inhibitor of mitochondrial respiration. For example, at normoglycemia, MIBG reduced the mean pH in a human mesothelioma xenograft from 6.90 to 6.70. This pH value was further reduced to 6.20 by simultaneous low-dose i.v. glucose infusion (plasma glucose concentration, 14 +/- 3 mM). The acidosis induced by inorganic phosphate and MIBG was tumor specific. Normal tissues of tumor-bearing hosts were only marginally sensitive to hyperphosphatemia or MIBG administration. These results indicate that the known stimulatory effect of exogenous glucose on lactic acid production in malignant tumors in vivo can be further accentuated or, as in the case of MIBG, partially replaced by pharmacological manipulation of aerobic glycolysis using clinically established drugs.

Proton-mediated liberation of aldophosphamide from a nontoxic prodrug: a strategy for tumor-selective activation of cytocidal drugs.

Based on the findings that the pH in malignant tumors can be preferentially decreased by stimulation of their aerobic glycolysis, acid-sensible prodrugs, which are nearly nontoxic at physiological pH, were synthesized. At lower pH, however, these compounds are cleaved with liberation of a cytotoxic species. The prototypic drug compound 2-hexenopyranoside of aldophosphamide was prepared, which releases aldophosphamide by acid-catalyzed hydrolysis. Exposure of cultured M1R rat mammary carcinoma cells to this agent at pH 7.4 only resulted in slight toxicity. However, when drug treatment was performed at pH 6.2, the mean pH in malignant tumors of hyperglycemic hosts, the colony-forming fraction of M1R cells decreased to 0.05 and 0.0001 of controls treated at pH 7.4 after exposure for 24 h and 48 h, respectively. The synthesis of the 2-hexenopyranoside of aldophosphamide is described in detail.

Ethylnitrosourea-induced development of malignant schwannomas in the rat: two distinct loci on chromosome of 10 involved in tumor susceptibility and oncogenesis.

Inbred rodent strains with differing sensitivity to experimental tumor induction provide model systems for the detection of genes that either are responsible for cancer predisposition or modify the process of carcinogenesis. Rats of the inbred BD strains differ in their susceptibility to the induction of neural tumors by N-ethyl-N-nitrosourea (EtNU). Newborn BDIX rats that are exposed to EtNU (80 microg/g body weight; injected s.c.) develop malignant schwannomas predominantly of the trigeminal nerves with an incidence >85%, whereas BDIV rats are entirely resistant. A T:A-->A:T transversion mutation at nucleotide 2012 of the neu (erbB-2) gene on chromosome 10, presumably the initial event in EtNU-induced schwannoma development, is later followed by loss of the wild-type neu allele. Genetic crosses between BDIX and BDIV rats served: (a) to investigate the inheritance of susceptibility; (b) to obtain animals informative for the mapping of losses of heterozygosity (LOH) in tumors with polymorphic simple sequence length polymorphisms (SSLPs); and (c) to localize genes associated with schwannoma susceptibility by linkage analysis with SSLPs. Schwannoma development was strongly suppressed in F1 animals (20% incidence). All of the F1 schwannomas displayed LOH on chromosome 10, with a consensus region on the telomeric tip encompassing D10Rat3, D10Mgh16 and D10Rat2 but excluding neu. A strong bias toward losing the BDIV alleles suggests the involvement of a BDIV-specific tumor suppressor gene(s). Targeted linkage analysis with chromosome 10 SSLPs in F2 intercross and backcross animals localized schwannoma susceptibility to a region around D10Wox23, 30 cM centromeric to the tip. Ninety-four % of F1 tumors exhibited additional LOH at this region. Two distinct loci on chromosome 10 may thus be connected with susceptibility to the induction and development of schwannomas in rats exposed to EtNU.

Immunoanalytical detection of O4-ethylthymine in liver DNA of individuals with or without malignant tumors.

The detection and quantitation of carcinogen-DNA adducts in human cells are the key parameters in the molecular dosimetry of human exposure to environmental carcinogens. For investigating the possible relevance of alkylating N-nitroso compounds as causative agents in human carcinogenesis, we have quantitated O4-ethyl-2'-deoxythymidine (O4-EtdThd) in human liver DNA obtained from 33 autopsy specimens, i.e., 13 cases with primary liver cancer (LC), 8 with cancers other than liver cancer (OC), and 12 with noncancerous diseases (NC). None of the cases analyzed had a history of known occupational exposure to ethylating agents. The detection limit for O4-EtdThd was 3 X 10(-8) as a O4-EtdThd/dThd molar ratio in DNA, which was attained by the combination of prefractionation of DNA hydrolysates (= 20 mg of DNA/sample) by high performance liquid chromatography and competitive radioimmunoassay using anti-(O4-EtdThd) monoclonal antibody ER-01. Except for one case in each group, O4-EtdThd [or, alternatively, (an) unidentified structural modification(s) of DNA recognized by monoclonal antibody ER-01] was detected at mean (+/- SD) O4-EtdThd/dThd molar ratios of 39.9 +/- 40.2 x 10(-8), 53.5 +/- 74.0 X 10(-8), and 11.7 +/- 6.5 X 10(-8), respectively, in LC, OC, and NC. The difference of the O4-EtdThd content in DNA between LC and NC, or between LC + OC and NC, was statistically significant at P less than 0.05. These results suggest that humans are exposed to ethylating agents in vivo and that a premutagenic DNA lesion (O4-EtdThd) eventually accumulates in DNA, possibly to a biologically significant extent.

Hydrogen ion-mediated enhancement of cytotoxicity of bis-chloroethylating drugs in rat mammary carcinoma cells in vitro.

Aerobic glycolysis, a metabolic characteristic of malignant cells, can be exploited to increase the concentration of lactic acid selectivity in tumor tissues in vivo by systemic administration of glucose (E. Jähde and M. F. Rajewsky, Cancer Res., 42: 1505-1512, 1982). To investigate whether a more acidic microenvironment can enhance the effectiveness of cytocidal drugs, we have analyzed the colony-forming capacity of M1R rat mammary carcinoma cells exposed to bis-chloroethylating agents in culture as a function of extracellular pH (pHe). At pHe 6.2 the cytotoxicity of 4-hydroperoxycyclophosphamide, as measured by inhibition of colony formation, was potentiated by a factor of approximately 200 as compared to pHe 7.4. Similar results were obtained with mafosfamide, nitrogen mustard, nornitrogen mustard, melphalan, and chlorambucil; not, however, with ifosfamide. As indicated by experiments using the ionophor nigericin for rapid equilibration of pHe and intracellular pH (pHi; measured with pH-sensitive microelectrodes), modulation of drug action by varying pHe primarily resulted from the concomitant decrease in pHi. The acidic microenvironment enhanced cytotoxicity most effectively during the phase of cellular drug uptake and monofunctional alkylation of DNA. DNA cross-link formation appeared to be less affected by pH, and lowering of pHe during the phase of cross-link removal was only marginally effective.

Reduction of intratumoral pH by the mitochondrial inhibitor m-iodobenzylguanidine and moderate hyperglycemia.

The interstitial pH of RIF-1 tumors was selectively lowered by i.p. administration of the mitochondrial inhibitor meta-iodobenzylguanidine (MIBG; 40-100 mg/kg), supported by sustained moderate hyperglycemia (plasma glucose concentration, 14 mM) in rats or by a single i.p. bolus injection of glucose (1.5 g/kg) in mice. Responses were evaluated in a multicenter study by pH measurements with semimicroelectrodes and 31P magnetic resonance spectroscopy, by biochemical analysis of tissue and plasma levels of glucose and lactate, and by positron emission tomography analysis of 2-[18F]fluoro-2-deoxy-D-glucose uptake. In both schedules, treatment with MIBG and glucose reduced the mean intratumoral pH as recorded with semimicroelectrodes to 6.2. In the mouse model, treatment with MIBG plus glucose was accompanied by a 2-3-fold stimulation of 2-[18F]fluoro-2-deoxy-D-glucose uptake and a corresponding increase in tumor glucose content. Responses were maximal in male mice with tumors of 0.2-0.8 g. 31P magnetic resonance spectroscopy analysis revealed no changes in intracellular pH or metabolic status, indicating that only extracellular pH was affected. MIBG was synergistic with bolus or continuous glucose administrations by a dual mechanism. The drug reduced by up to 5-fold the amount of glucose required for effective reduction of intratumoral pH and promoted the availability of (extra) glucose to tumor tissue in a stress-related, sympathomimetic response. Moreover, by converting oxic tumor cells into functionally hypoxic cells, combined treatment resulted in a more homogeneous decrease in intratumoral pH which included better perfused peripheral tumor areas. The effects of combined treatment on tumor glucose metabolism could be monitored noninvasively by 2-[18F]fluoro-2-deoxy-D-glucose positron emission tomography analysis.

Glial cell-specific differences in repair of O6-methylguanine.

Normal and malignant cells of the oligodendrocyte lineage show increased sensitivity to alkylating agents compared to astrocytes. One of the most mutagenic DNA lesions formed following exposure to alkylating agents is O6-alkylguanine. To determine whether the increased sensitivity to nitrosoureas seen in oligodendrocytes is due to decreased repair capacity for O6-alkylguanine, removal of this lesion from DNA was assessed in primary cultures of rat oligodendrocytes, astrocytes, and microglia. Glial cells were exposed to 1 mM N-methyl-N-nitrosourea for 1 h and allowed 8 or 24 h for repair. Repair was evaluated using an immunoslot blot technique and a monoclonal antibody which recognizes O6-methylguanine (O6MeGua). Astrocytes removed O6MeGua more efficiently (approximately 80% in 24 h) than either oligodendrocytes (approximately 20%) or microglia (approximately 4%). Determination of O6-alkylguanine-DNA-alkyltransferase (AT) activity revealed that astrocytes contain 0.4 pmol/mg protein, which is average by comparison to other cell types. Both oligodendrocytes and microglia exhibited very low levels of AT (oligodendrocytes, 0.08; microglia, 0.01 pmol/mg protein). These data are the first to show that within different populations of glial cells, O6MeGua adduct removal is substantially reduced in both oligodendrocytes and microglia. Rapid removal of O6MeGua in astrocytes coupled with persistence of this mutagenic lesion in oligodendrocytes following exposure of the developing central nervous system to nitrosoureas could contribute to the observed formation of oligodendrogliomas. Inefficient removal of O6MeGua in oligodendrogliomas might also account for their response to chemotherapeutic regimens involving alkylating agents such as procarbazine, lomustine, and carmustine. The lack of repair of O6MeGua in microglia suggests that primary lymphomas of the central nervous system might be sensitive to treatment with alkylating drugs whose toxicity depends on repair of this adduct.

Induction of the common fragile site FRA3B does not affect FHIT expression.

A long-standing issue concerns the extent to which fragile sites predispose to cancer-associated chromosomal rearrangements. The FHIT gene at chromosome 3p14.2 spans the most common fragile site, FRA3B, in the human genome. Although the FHIT gene is altered in many human cancers, its status as a tumor suppressor gene has remained controversial, particularly since functional studies provided contradictory results. It had been suggested the FHIT alterations result from FRA3B induction promoted by the interference of carcinogens with DNA replication. Here we investigated the effect of FRA3B induction on FHIT expression. Common fragile sites were induced by treatment with aphidicolin and scored cytogenetically. FHIT transcription was analysed by RT--PCR and RNase protection analysis. Unexpectedly, FHIT transcription proceeded unchanged after fragile site induction. Aberrant FHIT transcripts lacking one or more exons were not observed. Moreover, Western blots revealed that the levels of FHIT prior to and following fragile site induction was unchanged, whereas p53 was found at elevated levels after induction. FRA3B induction thus has no direct effect on FHIT transcription and translation.