British Journal of Cancer

Fitzpatrick, R; Davies, L

doi: 10.1038/bjc.1998.254pmid: 9635826

Telang, NT; Arcuri, F; Granata, OM; Bradlow, HL; Osborne, MP; Castagnetta, L

doi: 10.1038/bjc.1998.255pmid: 9635827

Targeted overexpression of the c-myc oncogene induces neoplastic transformation in immortalized, non-tumorigenic mouse mammary epithelial cells (MMEC). Experiments in the present study were conducted to examine whether cellular transformation induced by c-myc oncogene is associated with altered metabolism of 17beta-oestradiol (E2). The parental, MMEC and the stable c-myc transfectant (MMEC/myc3) cell lines were compared for major oestrogen metabolic pathways, namely E2 and E1 interconversion, and C2- and C16alpha-hydroxylation by both high-pressure liquid chromatography (HPLC) analysis and the 3H release assay using specifically labelled [C2-3H]E2 or [C16alpha-3H]E2. The reductive conversion of E1 to E2 was about 14-fold and 12-fold higher than the oxidative conversion of E2 to E1 in MMEC and MMEC/myc3 cells respectively. However, in MMEC/myc3 cells, both reductive and oxidative reactions were decreased by about 32% and 12% relative to those seen in the parental MMEC cells (P = 0.0028). The extent of C16alpha-hydroxylation was increased by 164.3% (P < 0.001), with a concomitant 48.4% decrease (P < 0.001) in C2-hydroxylation in MMEC/myc3 cells; this resulted in a fourfold increase in the C16alpha/C2 hydroxylation ratio in this cell line. Thus, a persistent c-myc expression, leading to aberrant hyperproliferation in vitro and tumorigenesis in vivo, is associated with an altered oestrogen metabolism. However, it remains unclear whether this represents a result of oncogene expression/activation or is rather a consequence of phenotypic transformation of the cells.

Ponchel, F; Milner, J

doi: 10.1038/bjc.1998.256pmid: 9635828

p53 is activated in response to DNA damage and functions in the maintenance of genetic integrity. Loss of p53 function because of mutation of the p53 gene is associated with over half all human cancers. Certain human p53 mutants are conformationally flexible in vitro and are temperature sensitive, with partial or complete recovery of wild-type (wt) properties at 32 degrees C. We have now tested the functional capacities of selected p53 mutants in vivo, by transfection into established human cell lines. Unexpectedly, we found that wt p53 can be temperature sensitive for transactivation of a co-transfected target gene in vivo. Flexible mutants retained varying degrees of functional capacity in transfected cells, and the recipient cell line appeared to be a significant determinant of both wt and mutant p53 function; importantly, two p53 null cell lines commonly used to study p53 function (Saos-2 and Hep3B) differed markedly in this latter respect. We also show that the p53 mutant V272M, which exhibits sequence-specific DNA binding in vitro, is nonetheless defective for transactivation and is unable to induce apoptosis in vivo. The valine 272 residue may thus be crucial for properties (other than sequence-specific DNA binding) that are important for p53 function(s) in vivo.

Burger, H; Nooter, K; Boersma, AWM; Kortland, CJ; Stoter, G

doi: 10.1038/bjc.1998.257pmid: 9635829

We examined the sensitivity for cisplatin-induced apoptosis in a panel of four testicular germ cell tumour (TGCT) cell lines and monitored the cellular expression of the apoptosis-related proteins p53, Bcl-2 and Bax. Three of four TGCT cell lines (NT2, NCCIT and S2) were hypersensitive for cisplatin-induced apoptosis, while the TGCT cell line 2102 EP appeared to be resistant for cisplatin-induced apoptosis, even at relatively high drug concentrations (12.5 microM). For all four cell lines, the induction of apoptosis by cisplatin correlated with drug sensitivity in the MTT assay. The differences in chemosensitivity and induction of apoptosis could not be attributed to differences in cellular platinum accumulation, DNA platination or platinum-DNA adduct removal. We next analysed the relationship between p53 status and cisplatin-induced up-regulation of p53, and the susceptibility to cisplatin-induced apoptosis. Wild-type p53 containing NT2 and 2102 EP cells showed p53 up-regulation upon drug treatment, and NCCIT (mutant p53) and S2 (no p53 protein) cells did not. Consistently, the increase in wild-type p53 protein in NT2 and 2102 EP cells led to an increase in mRNA level of the p53 downstream gene p21/WAF/CIP, whereas mutant p53-containing NCCIT cells and p53-non-expressing S2 cells could not transactivate this p53-responsive gene. As NT2, NCCIT and S2 were readily triggered into apoptosis, while 2102 EP cells failed to undergo cisplatin-induced apoptosis, our data suggest that the presence of wild-type and/or transactivation-competent p53 might not be an absolute prerequisite for efficient induction of apoptosis in TGCT cell lines. Also endogenous levels of Bcl-2 and Bax expression did not correlate with cisplatin-induced apoptosis. In addition, the endogenous Bcl-2 and Bax expression was not affected by cisplatin treatment. The present study suggests that, at least in our panel of TGCT cell lines, hypersensitivity for cisplatin-induced apoptosis might not be necessarily correlated with the presence of wild-type p53 and is probably not associated with Bcl-2 and Bax expression.

Takagi, Y; Osada, H; Kuroishi, T; Mitsudomi, T; Kondo, M; Niimi, T; Saji, S; Gazdar, AF; Takahashi, T; Minna, JD; Takahashi, T

doi: 10.1038/bjc.1998.258pmid: 9635830

Accumulating evidence suggests that the p53 gene is a good target for molecular epidemiological studies. We previously reported an association between the presence of p53 mutations and lifetime cigarette consumption. Although over 675 p53 mutations have been reported in lung cancers in the literature thus far, very little is known about the nature of such changes in lung cancers in the absence of a smoking background. In the present study, we therefore analysed 69 non-small-cell lung cancer specimens from individuals without any history of active smoking and identified p53 mutations in 26% of the cases. Statistical analysis of the present cohort of non-smokers also showed absence of significant relationship between p53 mutations and age, sex, histological type or disease stage. Comparison of mutational spectra between the present results in non-smokers and previously reported mutations in smokers clearly demonstrated G:C to T:A transversions to be significantly less frequent in non-smokers than in smokers (OR 5.35, 95% CI 1.77-16.12). Interestingly, G:C to C:G and G:C to A:T mutations were also observed in tumours of non-smokers at similar frequencies to G:C to T:A mutations, suggesting that these mutations can occur relatively frequently in the absence of active smoking. This study is, to our knowledge, the largest so far analysing a well-defined cohort of non-smokers in a single laboratory.

Sørlie, T; Martel-Planche, G; Hainaut, P; Lewalter, J; Holm, R; Børresen-Dale, A-L; Montesano, R

doi: 10.1038/bjc.1998.259pmid: 9635831

Exposure to aromatic amines is considered a major risk factor for the development of bladder cancer. In this study, we have analysed the pattern of point mutations in several tumour genes in 21 cases of bladder cancer arising among western European workers exposed to aromatic amines in an attempt to determine whether this exposure may be associated with a unique spectrum of mutations. Of the four genes analysed (p53, p16MTS1, p21WAF1 and H-ras), only p53 showed a high frequency of mutations (in 8 out of 21 cases, 38%). Two mutations were found in p16, one in H-ras and none in p21 exon 3. All mutations were at G:C base pairs, mostly at non-CpG residues. This spectrum of mutations, which is highly suggestive of an involvement of exogenous carcinogens, is however identical to the spectrum of p53 mutations detected in bladder cancers of the general population. In exposed workers, p53 mutations were associated with tumour grade and with high occupational and tobacco exposure. Taken together, our data suggest that the same carcinogens may be responsible for the development of bladder cancers in workers exposed to aromatic amines and in the general population.

Kozin, SV; Gerweck, LE

doi: 10.1038/bjc.1998.260pmid: 9635832

Theory suggests that the transmembrane pH gradient may be a major determinant of the distribution of lipophilic weak electrolytes across the cell membrane. The present study evaluates the extent to which this factor contributes to pH-dependent changes in the cytotoxicity of two such chemotherapeutic drugs: chlorambucil and mitoxantrone. Experiments were performed with two cell types of the same origin but exhibiting different pH gradients at the same extracellular pH (pHe): CHO cells cultured under normal physiological conditions (pH 7.4) and acid-adapted cells obtained by culturing under low pH conditions (6.8). Over the pHe range examined (6.0-7.6), the difference between intracellular pH (pHi) and pHe increased with decreasing pHe. Acid-adapted cells were more resistant to acute changes in pHi than normal cells, resulting in substantially larger gradients in these cells. Drug cell survival curves were performed at pHe values of 6.4, 6.8 and 7.4. The cytotoxicity of chlorambucil, a weak acid, increased with decreasing pHe, and low pH-adapted cells were more sensitive than normal cells at the same pHe. In contrast, for the weak base, mitoxantrone, cytotoxicity increased with pHe and was more pronounced in normal cells. As predicted by the theory, the cytotoxicity of both drugs changed exponentially as a function of the pH gradient, regardless of cell type. For mitoxantrone, the rate of such change in cytotoxicity with the gradient was approximately two times greater than for chlorambucil. This difference is probably due to the presence of two equally ionizable crucial groups on mitoxantrone vs one group on chlorambucil. It is concluded that the cellular pH gradient plays a major role in the pH-dependent modulation of cytotoxicity in these weak electrolytes. The data obtained also suggest that a pronounced differential cytotoxicity may be expected in vivo in tumour vs normal tissue. In comparison with normal cells at a pHe of 7.4 (a model of cells in normal tissues), acid-adapted cells at a pHe of 6.8 (a model of cells distal from supplying blood vessels in tumours) were more sensitive to chlorambucil, with a dose-modifying factor of approximately 6, and were more resistant to mitoxantrone by a factor of 14.

Ranson, M; Andronicos, NM; O'Mullane, MJ; Baker, MS

doi: 10.1038/bjc.1998.261pmid: 9635833

Overexpression of urokinase-type plasminogen activator and its receptor correlates with metastatic capacity in breast cancer. In this study we show that the urokinase/urokinase receptor-overexpressing, metastatic human breast cancer cell line MDA-MB-231 (1) bound significantly more cell-surface plasminogen in a lysine-dependent manner and (2) was capable of generating large amounts of plasmin compared with the non-metastatic cell lines MCF-7 and T-47D. In addition, distinct plasminogen binding proteins were detected in the plasma membranes of the cell lines, suggesting heterogeneity of binding proteins. Plasminogen binding was analysed using a combination of dual-colour fluorescence flow cytometry and ligand histochemistry (for comparative and cellular localization of ligand binding), and fluorimetry (for Scatchard analysis). Apart from revealing the greater plasminogen binding capacity of MDA-MB-231 cells, flow cytometry and histochemistry also revealed that, in all three cell lines, non-viable or permeabilized cells bound significantly more plasminogen in a lysine-dependent manner than viable or non-permeabilized cells. Viable MDA-MB-231 cells bound plasminogen with moderate affinity and high capacity (Kd = 1.8 microM, receptor sites per cell 5.0 x 10(7). Our results indicate that differences in cell surface-specific plasminogen binding capacity between cell lines may not be detectable with binding techniques that cannot distinguish between viable and non-viable cells.

Shamash, J; Salam, AH; Davies, DC; Williams, A; Joel, S; Lister, TA

doi: 10.1038/bjc.1998.262pmid: 9635834

The flux of calcium forms an important intracellular messenger system. The bcl-2 oncoprotein is thought to make cells resistant to a variety of insults, including cytotoxic drugs, by the suppression of apoptosis, which appears to involve the repartitioning of intracellular calcium. Three drugs that affect calcium pathways and may influence this repartitioning, i.e. dantrolene, azumolene (a water-soluble dantrolene analogue) and nimodipine, were studied in cell culture, using both a transformed follicle centre lymphoma cell line and primary culture of lymphoma cells in vitro in a manner that resulted in a growth pattern closely resembling that of the malignancy in vivo. Dantrolene and azumolene were potent inducers of cell death in both systems reducing the viable cell count by 70-90% in comparison with normal controls. Nimodipine, in comparison, appeared to have no significant effect. These results obtained in an in vitro setting suggest that further evaluation of dantrolene and azumolene for the treatment of low-grade non-Hodgkin's lymphoma is warranted.

Gasparian, AV; Laktionov, KK; Belialova, MS-O; Pirogova, NA; Tatosyan, AG; Zborovskaya, IB

doi: 10.1038/bjc.1998.263pmid: 9635835

The mapping of allelic loss on the short arm of chromosome 1 has been performed in non-small-cell lung cancer. We used a set of 11 microsatellite loci spanning 1p to examine the frequency of allelic imbalance in a panel of 58 tumours. Fifty-one of 58 (87.9%) cases have shown somatic allelic loss at one or more loci tested. The two shortest regions of the overlap (SRO) of the deletions have been identified: SRO 1 at 1p13.1 and SRO 2 at 1p32-pter. Allelic losses at these regions have been compared among adenocarcinoma and squamous cell carcinoma and no difference has been found. In contrast to SRO 1, deletions at SRO 2 significantly correlated with advanced stage of the disease as well as post-operative metastasizing and relapse. These data may suggest that SRO 1 and SRO 2 can harbour tumour-supressor genes (TSGs) involved in different stages of NSCLC development. SRO 2 is still quite large and its refined mapping should help attempts to clone and identify the putative TSG(s). Microsatellite instability (replication errors) affecting only 6 (10.3%) of 58 tumour samples is an infrequent genetic alteration at the loci tested.