Kenneth D. Tew

Cytotoxic properties of estramustine unrelated to alkylating and steroid constituents

Using cultured HeLa S3 cells an ID50 of 2.5 micrograms/ml was found after a twenty-four-hour incubation with estradiol-17 beta- 3N -bis-(2-chloroethyl) carbamate (estramustine). Similar ID90 values were found in two Walker 256 rat carcinoma cell lines which were either sensitive or resistant to nitrogen mustards. Alkaline elution methodology revealed the complete absence of DNA strand breaks or cross-links in cells receiving up to 10 micrograms/ml estramustine for twenty-four hours. Nuclear uptake was 1.34 per cent of the available drug, one third of which was hydrophobically associated with the protein/phospholipid components of the nuclear matrix. In the human prostatic cell lines DU145 and PC3 , estramustine caused a drastic dose-dependent increase in the mitotic index. This increase resulted from an arrest of cells in metaphase, with highly contracted disoriented chromosomes present. Rapid reverse of the arrest on removal of drug resulted in cell death. Neither nor-nitrogen mustard nor estradiol demonstrated antimitotic properties. The lack of macromolecular alkylation together with the observed antimitotic effects predict a mechanism of action for estramustine which is distinct from either of its constituent components.

Acquired drug resistance is accompanied by modification in the karyotype and nuclear matrix of a rat carcinoma cell line

A Walker 256 breast carcinoma cell line (WR) exhibiting a greater than 20-fold resistance to alkylating agents has been selected from a parent cell line (WS). Karyotypic heterogeneity was apparent, with a number of differences evident between WR and WS cells. The modal chromosome number for WS is 62; for WR, 54; double minutes were found only in WR, whereas spontaneous chromosomal aberrations were present in approx. 40% of the WS cells. No similar aberrations were observed in WR. Using SDS-gel electrophoresis and subsequent silver staining, differences in the profile of nuclear matrix proteins in WR and WS were observed. A diffuse band at approx. 70 kD in the WS was absent in WR cells. This protein was phosphorylated, together with a number of the other major matrix polypeptides. Levels of phosphorylated matrix proteins were approximately equivalent in both WR and WS cell lines, but matrix protein phosphorylation levels were approx. 2-fold higher than corresponding values for bulk nuclear proteins. Selective pressure of drug exposure has resulted in enhanced genetic stability in WR cells and observed karyotype differences are accompanied by modifications in the structural proteins of the nuclear matrix. Whether the observed differences are the cause or result of drug resistance remains to be established.

Probes to study the effect of methyl nitrosourea on ADP-ribosylation and chromatin structure at the subunit level

Abstract Treatment of HeLa cells with MNU results in a significant activation of the chromatin modifying enzyme, poly(ADP-Rib) polymerase, in nuclei and at the nucleosome level of chromatin. This appeared to be due to a direct effect of MNU on chromatin structural elements, initiating this study on the effect of MNU-induced carbamoylation at the subunit level of chromatin, the core nucleosome. Core nucleosomes were prepared by micrococcal nuclease digestion of HeLa nuclei isolated from cells treated with [ carbonyl - 14 C]MNU Trypsin digestion of core nucleosomes containing carbamoylated proteins showed that 80% of the 14 C-label was covalently associated with these proteins in a domain not sensitive to trypsin digestion. When core nucleosomes containing ADP-ribosylated histones were prepared from HeLa nuclei previously incubated with [ 3 H] NAD, nearly 80% of the covalently bound ADP-Rib was proteolytically removed by trypsin. These data suggest that lysines distal to the NH 2 -termini of histones were most susceptible to carbamoylation, while amino acid residues near the NH 2 -termini were subject to ADP-ribosylation. DNA of carbonyl-modified nucleosomes, phosphorylated ( 32 P) at the 5′-end, was digested 2–3 times faster with DNAase I than DNA in control nucleosomes. DNAase I generated the 10-base repeat pattern in these modified particles typical of native nucleosomes. Mapping the DNAase I cleavage sites revealed that sites 100, 90, 70, 60 and 50 nucleotides from the 5′-end of the nuclosomal DNA are made more susceptible to DNAase I as a result of cellular treatment with MNU. Such data suggest the carbamoylation of amino acids distal to the NH 2 -termini of histones may have a significant effect on histone-histone interactions with concomitant alteration of DNA-histone interactions.

A comparative analysis of drug-induced DNA effects in a nitrogen mustard resistant cell line expressing sensitivity to nitrosoureas

In the Walker 256 rat mammary carcinoma cell line, WR, resistance to nitrogen mustards (NM) is accompanied by collateral sensitivity to chloroethylnitrosoureas (CENUs). DNA-interstrand cross-links, DNA-protein cross-links, and sister chromatid exchange (SCE) induction were assayed in WR and the parent cell line (WS) after treatment with nitrogen mustard (HN2), phosphoramide mustard (PM), chlorozotocin (CLZ) and 1-(2-chloroethyl)-3-cyclohexyl-1-nitrosourea (CCNU). Treatment of cells with HN2 caused extensive levels of cross-links, approximately 50% of which were DNA-interstrand, equal in both WR and WS, whereas PM caused no detectable cross-links in either cell line. CLZ induced low levels of DNA-interstrand cross-links, similar in WR and WS, but no DNA-interstrand cross-links could be detected in either cell line after treatment with CCNU. Both CLZ and CCNU induced low levels of DNA-protein cross-links in both cell lines, though higher in WR than WS. There was no difference in the rate of removal of HN2-induced DNA-interstrand or DNA-protein cross-links or total CLZ-induced cross-links by the two cell lines, suggesting that differential repair was not relevant to the expression of resistance. Both HN2 and PM caused more SCEs in WS than in WR, whereas CLZ and CCNU induced more SCEs in WR. Thus, NM-induced SCEs were related to cell killing but not cross-linking, whilst CENU-induced SCEs were related to cell killing and DNA-protein but not DNA-interstrand cross-links. Furthermore, the collateral sensitivity of WR cells to CENUs was not due to the differential induction of DNA-interstrand cross-links or repair of total cross-links, or repair of total cross-links, although higher levels of DNA-protein cross-links occurred in WR, and these may be either a cause or a consequence of increased susceptibility of these cells to CENUs. Presumably NMs and CENUs have several distinct and separate macromolecular targets which result in differential cell killing. It is concluded that a range of lesions occurred after treatment of WR and WS cells with either NMs or CENUs and that, in these cell lines, there is no simple correlation between drug-induced cross-linking, SCE induction and cytotoxicity.

Correlation of nitrosourea murine bone marrow toxicity with deoxyribonucleic acid alkylation and chromatin binding sites

All of the clinically available nitrosourea antitumor agents produce serious treatment-limiting bone marrow toxicity. A reduction in this toxicity can be achieved by attaching the chloroethylnitrosourea cytotoxic group to C2 (chlorozotocin) or C1 (1-(2-chloroethyl)-3-(beta-D-glucopyranosyl)-1-nitrosourea, GANU) of glucose. Both glucose analogs are less myelotoxic in mice than 1-(2-chloroethyl)-3-cyclohepyl-1-nitrosourea (CCNU) or 1-(4-amino-2-methylpyrimidin-5-yl)methyl-3-(2-chloroethyl)-3-nitrosourea (ACNU), while retaining comparable antitumor activity against the murine L1210 leukemia. To define the nuclear mechanisms for this reduced myelotoxicity, alkylation of L1210 and murine bone marrow DNA was quantitated. With the use of the endonuclease micrococcal nuclease and DNase I, the sites of alkylation within the chromatin substructure were determined. Experiments were performed on L1210 leukemia or bone marrow cells that had been incubated in vitro for 2 hr with 0.1 mM [14C]chloroethyl drug. The quantitative alkylation of DNA by GANU was 1.3-fold greater in L1210, as compared to bone marrow, cells. This ratio of DNA alkylation is comparable to the 1.3 ratio we previously reported for chlorozotocin [L. C. Panasci, D. Green and P. S. Schein, J. clin. Invest. 64, 1103 (1979)]. In contrast, the ratio of alkylation (L1210:bone marrow DNA) for the myelotoxic ACNU was 0.66, similar to 0.59 for CCNU. Nuclease digestion experiments demonstrated that chlorozotocin and GANU preferentially alkylated internucleosomal linker regions of bone marrow chromatin, while nucleosome core particles were the preferred targets of CCNU and ACNU. The reduced myelotoxicity of chlorozotocin and GANU may be correlated with the advantageous ratio of L1210:bone marrow DNA alkylation and preferential alkylation of internucleosomal regions of bone marrow chromatin.

Mechanism of action of 2-haloethylnitrosoureas on deoxyribonucleic acid: Pathways of aqueous decomposition and pharmacological characteristics of new anticancer disulfide-linked nitrosoureas

We have examined the pharmacological characteristics of three dinitrosated isomers of N,N-bis[N(2-chloroethyl)-N-carbamoyl]cystamine [CNCC-(D), 1C1G1325] differing in the relative positions of the nitroso substituents [CNCC-(C), (1,1 dinitroso); CNCC-(S), 3,3 dinitroso); and CNCC-(M), (1,3-dinitroso)] and which were designed to be subject to preferential bioreductive activation in hypoxic tumors. The decomposition products of the isomers formed under physiological conditions [both in the absence and in the presence of dithiothreitol (DDT)] were identified and quantified. For example, CNCC-(S) in phosphate buffer, pH 7.0, and 37 degrees gave rise to 2-chloroethylisocyanate, bis(2-chloroethyl)urea and bis(2-hydroxyethyl)disulfide, whereas in the presence of DTT it afforded 2-chloroethylisocyanate, bis(2-chloroethyl)urea, bis(2-hydroxyethyl) disulfide, thiirane and 2-mercaptoethanol. Control aqueous decomposition profiles were performed with two known metabolites of CNCC, namely 3-(2-chloroethyl)-1-(2-thioethyl)-1-nitrosourea and 3-(2-chloroethyl)-1-(2-methylthioethyl)-1-nitrosourea. CNCC-(C) caused 20% interstrand cross-linking of lambda-DNA in 2 hr, whereas in the presence of DTT the extent of cross-linking increased to 38% in the same time period. In contrast, isomer (S) showed no detectable cross-linking in 7 hr. This thiol potentiation of cross-linking which is observed with other 2-chloroethylnitrosoureas is explained by nucleophilic attack at the carbonyl group and subsequent stereoelectronically controlled decomposition of the tetrahedral intermediate. The relative extents of carbamoylating activity of the CNCC isomers were obtained using a [14C]-lysine assay which showed (S) approximately equal to (M) greater than (C). Inhibition of glutathione reductase for both Walker 256 resistant (WR) and Walker 256 sensitive (WS) strains showed that isomer (S) inactivated the enzyme more effectively than isomer (C) in accord with the carbamoylating activity results. The higher carbamoylators (S) and (M) also showed greater effects on the intracellular thiol pools in both WR and WS cells indicative of sulfhydryl conjugation and efflux and/or inhibition of the GSH metabolic enzymes. In vitro cytotoxicity studies with human DU 145 prostatic carcinoma cells showed the isomer cytotoxicity was (M) greater than (C) greater than (S) over a 24-hr incubation period. The reduced cytotoxic potential of CNCC-(S) in both the Walker 256 cells and in the human prostatic carcinoma cells may be a function of an interaction between GSH and the drug thereby protecting other more critical nucleophilic targets within the nucleus.

Biochemical and cytotoxic properties of the isomeric forms of N,N'-bis[N-2-chloroethyl)-N-nitrosocarbamoyl] cystamine.

Three isomeric forms of a cystamine-containing chloroethylnitrosourea, N,N-bis[N-(2-chloroethyl)-N-nitrosocarbamoyl]cystamine (CNCC), have been identified and separated by high pressure liquid chromatography. Isomer S, 3,3-bis[N-(2-chloroethyl)-N-nitrosocarbamoyl] ethyl disulfide, was significantly less cytotoxic than isomer C, 1,1-bis [N-(2-chloroethyl)-N-nitrosocarbamoyl] ethyl disulfide, or isomer M, 1,3-bis[N-(2-chloroethyl)-N-nitrosocarbamoyl] ethyl disulfide, in either a human Namalva lymphoblastoid or a rat Walker 256 carcinoma cell line. Both isomers S and C inhibited DNA synthesis at a 50 microM concentration. A structural analysis of the isomeric forms suggested that bioreduction of the disulfide bond would permit both isomers to produce isocyanate moieties which would carbamoylate intracellular proteins and depress nucleic acid synthesis. The reduced cytotoxic potential of isomer S is consistent with a prolongation in the half-life of production of alkylating carbonium species that lack the capacity to cross-link macromolecules. Overall, the relative position of the NH group within each of the nitrosourea isomers appears critical to the biological properties of the drug.

Importance of Phospholipids to Nitrosourea Interactions with the Nuclear Envelope and Associated Nucleic Acids

Nuclease digestion of triton isolated HeLa cell nuclei has resulted in nuclear envelopes containing less than 10% of the nuclear macromolecule. Alkylation and carbamoylation of the nuclear envelope (NE) fraction by chloroethylnitrosoureas (HNU) was disproportionately high (approximately 35% of total nuclear alkylation and approximately 55% carbamoylation) suggesting that the nucleophilic species in the envelope were preferential targets for drug binding. Digestion of the envelope fraction with phospholipase C (PLC), followed by buoyant density gradient separation in Percoll (70%):0.15 M NaCl (30%), demonstrated that some envelope:nucleic acid attachments were destabilized as a result of phospholipid cleavage. In addition, the buoyant density of HNU-modified NE macromolecules was slightly altered by PLC digestion, suggesting that cleavage of polar phospholipids did not completely destabilize the chloroethylated products from the nuclear envelope.