Genotoxins and the initiation of sporadic breast cancer

Francis L.Martin 0 0 Institute of Cancer Research, Haddow Laboratories , Cotswold Road, Sutton, Surrey SM2 5NG, UK Breast cancer is the most frequently diagnosed female malignancy world-wide. The aetiology of the majority of cases remains obscure and the only genotoxin as yet known to initiate breast cancer is ionizing radiation. High penetrance susceptibility genes probably account for no more than 5-10% of cases. The breast, which consists of 70-90% adipose tissue, has a unique morphological structure. Dispersed within it are the functional elements that are lined with cancer-susceptible epithelial cells. Numerous dietary and/or environmental fat-soluble compounds are known to be rodent mammary carcinogens. Extracts of lipid obtained following collagenase digestion of elective reduction mammoplasty tissues from UK resident women showed activity in short-term genotoxicity assays in 40% of cases. More active lipid extracts tended to come from donors whose human mammary epithelial cells (HMECs) exhibited pre-existing DNA single-strand breaks (SSBs). Lipid extracts also induced morphological transformation of mammalian cells in vitro. To increase cohort size, extracts of UK breast milk were examined for genotoxicity and similar activity profiles were observed. Viable cells, a large percentage of which were HMECs, were recovered from breast milk and examined for pre-existing SSBs and for the ability of the donor's own milk extract to induce SSBs. Again, donors whose untreated cells contained the most SSBs tended to yield genotoxic breast milk extracts. Breast milk cells were also able to activate rodent mammary carcinogens to DNA-damaging species. These studies provide good evidence for in vivo exposure of HMECs to genotoxic agents years before the peak in occurrence of sporadic breast cancer. Work is in progress to characterize these agents and to determine their possible role in breast cancer aetiology. - Scale of the problem and risk factors Breast cancer is the most frequent malignancy occurring in women world wide, accounting for 20% of all female cancers (Higginson et al., 1992). Currently, one in every 89 women in Western Europe and North America will develop the disease by the age of 85. Aetiological risk factors that could account for the majority of cases, which are post-menopausal and sporadic, remain obscure and genetic predisposition probably accounts for only some 510% of breast cancer cases (Higginson et al., 1992). The observation that the mutational spectrum in the p53 gene in human mammary tumours differs from that attributable solely to background endogenous processes points to environmental factors playing a role in the aetiology of breast cancer (Biggs et al., 1993). Breast cancer incidence varies widely across the globe, with a much higher incidence in Western countries than in, for example, the Far East (Higginson et al., 1992). Migration studies have shown that breast cancer incidence in JapaneseAmerican women increases to match the resident incidence by the second or third generation (Ziegler et al., 1993). There is evidence linking alcohol intake and breast cancer incidence amongst individuals lacking both the GSTM1 and GSTT1 genes (Park et al., 2000). Intensity of alcohol intake rather than duration may also be a risk factor (Bowlin et al., 1997). Such epidemiological investigations into the relationship of breast cancer risk to geneenvironment interactions are still in their infancy (Williams and Phillips, 2000). For instance, observation of an increased risk of breast cancer from smoking in slow NAT2 acetylators (Ambrosone et al., 1996) has not been verified (Hunter et al., 1997; Delfino et al., 2000). The only environmental exposure proven to induce breast cancer is ionizing radiation (Tokunaga et al., 1987). Total cumulative exposure to oestrogen is also seen as a risk factor: thus nulliparity, late age at first pregnancy, early menarche and late menopause may play a role in breast cancer risk (Feigelson and Henderson, 1996). However, oestrogens probably act as tumour and growth promoters rather than as complete carcinogens (Higginson et al., 1992). In simple terms, the breast consists of 7090% adipose tissue, in which are dispersed the functional elements that are lined with epithelial cells from which breast cancers commonly arise. CarcinogenDNA adducts in many tissues, including the breast, have been observed in vivo in different human populations (Perera et al., 1995; Perera, 2000). Human exposures to carcinogens via the diet and the environment include exposure to polycyclic aromatic hydrocarbons (PAHs), nitroPAHs and heterocyclic aromatic amines (HAAs, food mutagens), many of which are fat soluble and capable of inducing mammary tumours in rodents (El-Bayoumy, 1992). This lends credence to the proposition (Beer and Billingham, 1978) that human mammary lipid could act as a reservoir or depot for mutagenic/genotoxic agents capable of inducing DNA damage in adjacent epithelial cells in vivo. These epithelial cells express CYP1B1 and CYP1A1, which are involved in PAH and HAA activation (Larsen et al., 1998; Williams et al., 1998), as well as N-acetyltransferases and sulphotransferases, that further metabolically activate hydroxylated derivatives of these compounds (Sadrieh et al., 1996; Williams et al., 2000a). human mammary epithelial cells (HMECs) can thus activate HAAs and PAHs to DNA-binding species (Grover et al., 1980; Pfau et al., 1992; Carmichael et al., 1996). It has also been hypothesized that peroxidases, including lactoperoxidase, synthesized in HMECs, and myeloperoxidase, present in neutrophils, may play a key role in the initial activation of aromatic amines to DNA-binding species (Josephy, 1996). A large amount of lipid can be obtained from essentially healthy women following collagenase digestion of tissues removed at elective reduction mammoplasty (Easty et al., 1980). Such tissues can be obtained from young women compared with the usual age of subjects from whom tumouradjacent breast tissue is obtained (Li et al., 1996; Perera et al., 1995). An advantage of using reduction mammoplasty tissue is that it can be examined for DNA damage and for the presence of agents capable of inducing genotoxic events many years before the age-related peak in sporadic breast cancer incidence. Breast milk is a natural, lipid-containing secretion of the breast that is more readily available than elective reduction mammoplasty tissues and that can be obtained non-invasively from a larger cohort of individuals. There is also evidence that body lipid, as opposed to dietary lipid, turns over to provide lipid for secretion in breast milk (Villalprando and del Prado, 1999), bringing with it, one could reasonably surmise, any contaminating genotoxins. The detection of low (p.p.b.) concentrations of aromatic amines, one of which is known to be a rodent mammary carcinogen, has been reported in breast milk (DeBruin et al., 1999). The presence of (...truncated)