Anticarcinogenesis and Radiation Protection

Cancer is a multi-step process, involving three stages that can be operationally distinguished: initiation, promotion and progression. Initiation involves an irreversible alteration of the cellular DNA that permits the transformation of the cell to a non-malignant state. Promotion produces conditions that allow the initiated cell to become malignant, and progression is the growth of the malignant cell to a tumor.

Oxygen may act not only as an oxidant to initiate free radical reactions, but it can also act as a substrate for the propagation of these reactions. The spontaneous reaction of molecular oxygen with radicals is commonly referred to as autoxidation. Autoxidation is responsible for the deterioration of many manufactured plastics and rubber goods. Rancidity and spoilage of foodstuffs is a direct result of the autoxidation of fats, which are most susceptible to air oxidation and present, to a large extent, in virtually all foods.

Free radicals are generated by ionizing radiations (e.g., X rays, γ rays, electrons, neutrons, and Rn α-particles),1 and the biological effects they cause are in some ways a consequence of their reactions.2 Free radicals may also be generated by other processes and agents and numerous chemicals generate them under physiological conditions. Cancer promoters3 and antineoplastic drugs4 cause DNA strand breaks via free radical reactions. Metabolic processes are also perceived as possible sources of free radicals.5–7 The repair of free radicals and the elimination (or reduction) of their biological effects by radioprotectors (mainly sulfhydryls) is a well-established defense mechanism in biological systems8. Although the interaction of anticarcinogens (primarily antioxidants) with free radicals is feasible, this is a poorly understood process.8,9

The reactivity of some thiyl radicals with molecular oxygen were measured and reported much before the intracellular thiols began to arouse the wave of interest that has characterized the last decade or so (1,2). Cellular thiols, almost entirely consisting of glutathione (GSH), are considered at the present time as one of the most important system capable of protecting cells against free radicals formed during oxidative metabolism or from exposure to drugs or ionizing radiation.

Ionizing radiation can be lethal to mammalian cells. For those cells which are irradiated and survive, irradiation can also be mutagenic or, in the case of cells in animal tissues, carcinogenic. The biological effects of ionizing radiation are generally believed to originate in free radical reactions. In particular, a radical competition model has been proposed to account for the “oxygen effect” on radiation lethality — the so-called “oxygen fixation” hypothesis (Figure 1) (1–5).

Although carotenoid pigments have been implicated as anti-carcinogenic compounds for several years, based on both epidemiological evidence (1) as well as experiments in animals (2,3), the exact mechanism whereby this widely distributed class of componds functions is still poorly understood. What appears to be important however, is the fact that many of the effects of carotenoids invivo and invitro can be observed with pigments that donot function as precursors of vitamin A (retinol). For example, beta-carotene may exert its biological effects merely by functioning as a precursor of retinal and retinol. On the other hand, there are carotenoid pigments, such as canthaxanthin (4,4’-diketo-beta-carotene) which also exhibit anti-carcinogenic properties and cannot be converted to retinol (Figure 1). Under these circumstances, we must look at the properties of the intact molecules in order to understand their functions.

Antioxidants are known to have an anticarcinogenic effect1,2. Although they are capable of reducing the incidence of tumors induced by chemical carcinogens, their effect on radiation-induced tumors has not been clearly demonstrated3,4. The mechanism of the anticarcinogenic action of antioxidants is poorly understood, and despite numerous mechanistic studies of autoxidation processes and antioxidants in model systems, detailed reaction mechanisms under physiological conditions are still elusive. However, inhibition of free radical processes may be an important part of this activity.

The mouse skin tumor promoter phorbol-12-myristate-13-acetate (PMA) is a potent inflammatory agent that not only elicits the infiltration of neutrophils, monocytes and lymphocytes to the site of application (1) but also induces human neutrophils and monocytes, but not lymphocytes, to release clastogenic factor(s) (CF) (2). CF are characterized as lipophilic (ethylacetate extractable) substances of relatively low molecular weight (<10 kDa) whose formation is mediated by superoxide anions and whose activity is inhibited by Cn superoxide dismutase (SOD), free-radical scavengers and antagonists of arachidonic acid (AA) metabolism (3). A correlation between CF formation and lipid peroxidation has been demonstrated (4). These properties of CF are suggestive of free-radical chain reactions which can give rise to lipid peroxides and possibly account for the formation and/or action of CF.

Clastogenic factors (CF), also called chromosome breakage factors, were first described by radiobiologists in 1968 (for review see 1). It was noted that not only do therapeutically or accidentally irradiated persons show an increased frequency of chromosome breaks and rearrangements in their own cells, but the plasma of these persons contains chromosome damaging material able to induce chromosome damage after transfer into cell cultures. These CF were circulating in the blood stream of the irradiated persons even years after the irradiation event (2). Plasma from irradiated animals had a tumorigenic effect in rats (3), suggesting a role for CF in carcinogenesis.

Metastases represent perhaps the single most important hindrance to improved cancer patient survival. The formation of metastases is the culmination of a complex series of tumor cell-host interactions called the metastatic cascade (Figure 1; 1,2). During tumor progression, tumor cells become separated from the primary tumor mass and invade into the surrounding tissue. Occasionally, tumor cells will invade blood vessels or lymphatics (intravasation) and are readily disseminated throughout the host. A relatively small percentage of the circulating tumor cells will arrest at the vessel wall and invade into the surrounding tissue (extravasation) to establish a secondary tumor site (3–5). The interaction of the circulating tumor cells with host immune cells may result in destruction of the tumor cells, whereas interaction with host platelets and/or blood coagulation components may enhance the metastatic process. Evidence for the involvement of platelets in metastasis has been demonstrated by several investigators (for review see 6,7). Enhancement of tumor cell arrest and adhesion to the vessel wall by platelets has been postulated as the mechanism for this phenomenon, but its exact nature is undetermined. Because of the potent effects of arachidonic acid metabolites (i.e., prostacyclin and thromboxane A₂) on platelet function, we have hypothesized that tumor cells shift the balance between these metabolites in favor of platelet aggregation (8). We have previously proposed the use of modifiers of arachidonic acid metabolism as antimetastatic agents, although the efficacy of antimetastatic therapy with these modifiers has been inconsistent. Prostacyclin (PGI₂) has been recently shown to reduce the incidence of pulmonary.

In recent years, there has been considerable interest in the involvement of free radicals in carcinogenesis (1–4). Much of the research emphasis has focused on intermediates of oxygen reduction such as superoxide anion and hydroxyl radical (2). Organic analogs of superoxide anion and its protonated form, perhydroxyl radical, have received less attention but may be just as important biologically as reduced oxygen intermediates. Our laboratory is interested in the role of hydroperoxide-dependent oxidations in carcinogenesis and much of our research is concerned with the chemistry and biology of peroxyl radicals generated by metal-catalyzed fragmentations of unsaturated fatty acid hydroperoxides. We believe evidence exists to support the hypothesis that peroxyl free radicals contribute to metabolic activation of chemical carcinogens (e.g., polycyclic hydrocarbons) in the initiation phase of tumorigenesis; circumstantial evidence implicates peroxyl radicals in tumor promotion as well.

The fatty acid oxygenases, cyclooxygenase and lipoxygenase, require hydroperoxides to achieve maximal rates of reaction, and their reaction products are hydroperoxides that can stimulate a faster oxygenation reaction. The cyclooxygenase can be stimulated to half-maximal velocity by 20 picomoles of lipid hydroperoxide per ml (1), while it can generate nanomoles of the hydroperoxide, prostaglandin G₂ (PGG₂). Soybean lipoxygenase, and all other lipoxygenases, also react faster as the level of lipid hydroperoxide is amplified. Thus, intracellular amplification of lipid hydroperoxides by the fatty acid oxygenases may be closely related to high rates of formation of eicosanoids (prostaglandins and leukotrienes). This relationship has broader importance when overproduction of eicosanoids is recognized to be a common occurrence in many major pathophysiological events (2). The intracellular amplification by the fatty acid oxygenases of “triggering” levels of hydroperoxide (3) to higher levels that permit pathological rates of eicosanoid synthesis needs careful study. Intracellular levels of to 10−9to 10−7 M lipid hydroperoxide seem likely to be involved in this amplification.

Investigation of the phenomenon of xenobiotic-induced peroxisome proliferation in liver parenchymal cells continues to provide fascinating details about the differential regulation of peroxisomal enzymes, their role in the production of intrahepatic oxidative stress and the possible relationship between sustained oxidative stress and the eventual development of hepatocellular carcinomas (1–3), Hepatic peroxisome proliferation was first noted nearly twenty years ago in the livers of rats treated with the hypolipidemic drug Clofibrate (4,5). Subsequently, several structurally diverse chemicals have been identified as hepatic peroxisome proliferators (1,2,6). Long-term feeding studies demonstrated the hepatocarcinogenicity of several of these agents (2,7–9), although they are not genotoxic in short-term tests (10–13). On the basis of these observations it was proposed that peroxisome proliferators form a novel class of chemical carcinogens (8), An understanding of the mechanism of induction of peroxisome proliferation and the metabolic events that accompany persistent increase in the number of these organelles is necessary to resolve the question whether these nongenotoxic hepatocarcinogenic peroxisome proliferators pose the same carcinogenic risk to humans as those that are genotoxic.

Oxygen free radicals react with and cause damage to DNA. Hydroxyl free radicals (·OH) add to DNA to yield several products, one of which is 8-hydroxydeoxyguanine (1). Digestion of DNA to the nucleoside level yields 8-hydroxydeoxyguanosine (8-OHdG) which can be very sensitively quantitated (to 20 femtomoles) by HPLC-electrochemical detection (2,3). The potential significance of the 8-OHdG adduct in living systems has been demonstrated by its presence in the DNA of human granulocytes following TPA-stimulated oxidative damage (3). The amount of 8-OHdG present in the granulocyte DNA was decreased by the addition of superoxide dismutase and/or catalase, thus emphasizing that oxygen free radical intermediates are the responsible agents (3).

Active oxygen species generated by stimulated PMNs were found to be mutagenic and carcinogenic (1,2). We (3–5) and others (6,7) have shown that oxygen species produced by phagocytic cells are capable of modifying bases in DNA exposed to them. Those identified to date are 5-hydroxymethyl uracil (HMU) (3–5), thymine glycol (TG) (4–6) and 8-hydroxyguanine (BOHG) (7). All of them are known to be formed by the action of ionizing radiation (8–11). HMU and TG are excreted in the urine by humans and rats which points to these thymine derivatives as products of the normal biological processes (12), That these oxidized DNA base derivatives are potentially harmful to the well-being of humans and animals is shown by the existence of the repair enzymes HMU- and TG-glycosylases in mammalian cells (13–16). There is also evidence that 80HG might be repaired as well (11). However, when removal of abnormal bases from DNA is not complete or timely, they have the potential to exert their deleterious effects. For example, TG is thought to provide a replication block (17,18), whereas, HMU and particularly its deoxyribonucleoside HMdU are mutagenic (19). HMdU, which is known to be incorporated into cellular DNA, is also cytotoxic and cytostatic to a number of mammalian cells, and acts as a radiomimetic agent as it causes diarrhea and leukopenia, symptoms of acute radiation sickness (20–23). Thus PMNs, whose role is to protect against invading bacteria or other opsonized particles, are capable of imparting a radiation-like damage, certain types of which might be heritable.

It is well known that the carcinogenic potency of radiation and chemical carcinogens is closely correlated with the induction of DNA damage (1–3); however, the ultimate mechanisms involved are far from clear. In the case of high energy radiation, generation of hydroxyl radicals (OH·) probably plays a key role, either by direct attack of OH· on DNA, or indirectly via initiation of lipid peroxidation (4,5). Initiation of lipid peroxidation may also be a consequence of the generation of oxygen centered radicals in the oxidative stress associated with metabolism of some chemical carcinogens(6). Thus lipid peroxidation may be one of the mechanisms linking both radiation and metabolism of chemical carcinogens to DNA damage. This view is supported by reports (7–12) that products of lipid peroxidation can cause DNA damage in model systems and mutations in prokaryotes. However, actual data showing DNA damage to eukaryotic genes by lipid peroxidation is limited and indirect (5,13). In this communication, evidence is reported showing extensive damage to DNA of liver mitochondria occurring concomitantly with mitochondrial lipid peroxidation. Mitochondrial rather than nuclear DNA was studied since the former exists as a discrete circular molecule of single molecular weight, which facilitates detection of small degrees of DNA damage. The study of the relationship between lipid peroxidation and DNA damage in mitochondria is important in the assessment of the carcinogenic potency of oxidative stress, especially since many strong carcinogenic agents preferentially attack mitochondrial DNA (14).

Active oxygen species have been implicated in a wide variety of environmental and health effects, including aging, heart disease, and induction of cancers. One of these reactive species, singlet oxygen (1Δ_gO₂), the lowest energy electronically excited state of molecular oxygen, has evoked particular interest, because of its potential to react with a variety of biologically important substrates.and because it may be generated in living systems by the decomposition or interconversion of other active oxygen species, photosensitization reactions involving endogenous sensitizers, electron transport systems, and some enzyme systems (references cited in 1).

Radiation biologists have proposed that various reactive oxygen species (referred to hereafter as reactive oxygens) which are produced during reactions between radiation and the biological system, are largely responsible for the adverse biological effects of radiations (1). Recently, reactive oxygens have been implicated in the toxic action of numerous chemicals (2).

Simple monofunctional alkylating agents such as methylmethane sulphonate (MMS) and N-methyl-N’-nitro-N-nitrosoguanidine (MNNG) cause a variety of alkylated lesions in DNA, and these lesions can cause the induction of mutation and cell death in both human and bacterial cells. The repair of these DNA alkylation damages reduces cell killing and the induction of mutations and chromosome damage. In E. coli, the repair of 06-methylguanine (06MeG) and 04-methylth5anine (04MeT) specifically prevents these lesions from causing G:C to A:T and A:T to G:C transition mutations, because if left unrepaired these lesions mispair during replication (1–6). On the other hand, the repair of N3-methylpurines (N3MeA and N3MeG) and 02-methylpyrimidines (02MeC and 02MeT) in E. coli specifically prevents cell killing (7–9). The repair of DNA alkylation damage in mammalian cells is somewhat less well understood, and it is not yet clear which alkylated lesions cause mutation and which cause cell death. Like E. coli, mammalian cells can repair 06alkylG, N3alkylA, N3alkylG, 04alkylT, 02alkylT and 02alkylC lesions (10–21). However, because observations have been made with a variety of cell types the results have not always been consistent. For instance, some studies indicated that 04MeT is repaired by rat liver cells (13,20,21), while others did not (22–24). Not all cell lines are able to repair all the alkylated lesions; for instance, CHO and V79 cells (25,26) and some human tumor cell lines (27–29) are unable to repair 06MeG. The ability of these particular cell lines to repair alkylated pyrimidines has not yet been measured. Other human cells, however, have been shown to repair 06MeG (19,27–29), 04alkylT, 02alkylT, 02alkylC (11), N3alkylA, N7alkylG (11,14,15), N3alkylG (15,30) and N7alkylA (15).

Agents that damage DNA are known to cause both mutagenesis and carcinogenic effects to cells. One specific type of DNA damage results in the loss of a purine or pyrimidine base, resulting in the formation of a base-free sugar or so-called apurinic/apyrimidinic (AP) site. Agents responsible for the induction of AP sites include gamma radiation (1), alkylating agents (2), and several other environmental and chemical carcinogens (3). AP sites can also form spontaneously in DNA (4), and they often result from the action of specific DNA glycosylases, such as uracil-DNA-glycosylase, which act to remove unusual or damaged DNA bases from DNA (5). The repair of AP sites can result in the formation of a mutation at the AP site, and such mutations may ultimately be responsible for the induction of the carcinogenic process.

Reactive oxygen species such as hydrogen peroxide, superoxide radical, and hydroxyl radical are formed during aerobic respiration (1). These reactive oxygen species can react with vital cell components such as the cell membrane, protein, as well as DNA, potentially leading to mutagenesis and cell death (1). In Escherichiacoli. repair endonucleases. such as exonuclease III (2, 4) and endonuclease IV (3) appear to be involved in the repair of oxidative damages in DNA. Further, endonuclease III was shown to recognize oxidative base damages (4).

Free radicals are highly reactive chemical species that are generated by a number of different mechanisms, including x-rays, certain chemical carcinogens and even aerobic metabolism. In order to survive the damages caused by free radical species, cells have evolved a number of protective mechanisms (5). Since DNA is an important target for free radicals, the ability to repair radical-induced DNA damages should constitute a significant cellular survival response. For a number of years our laboratory has been involved in identifying, quantitating, and assessing the biological impact of X-ray-induced DNA damages (26, 27), many of which are produced by free radicals (2,10). Since the 5–6 double bond of thymine is particularly susceptible to free radical attack (5), enzymatic, chemical and immunological methods have been used to focus our efforts on radical-induced oxidation products of thymine (26, 27). Endonuclease III from E. coli has been used as a tool to probe x-irradiated DNA for thymine ring saturation (15) and fragmentation (16) products. Furthermore, both thymine glycols and urea residues (a fragmentation product of thymine hydroperoxides) act as replicative blocks invitro (4,9,12,25) and are lethal lesions invivo (12,23).

It has been shown by extensive product analysis that the saturation of the C5-C6 bond of thymine and the concomitant loss of planarity of the thymine ring are notable characteristics of the DNA lesions produced by ionizing radiation (1). Recently, we (2) and other laboratories (3–5) have found that thymine glycols present in a DNA template constitute replicative blocks to DNA synthesis invitro. Interestingly a base, presumably adenine, is incorporated opposite thymine glycol; however, DNA synthesis is arrested at the site. In order to obtain a more complete picture of the biological consequence of thymine C5-C6 bond saturation, we have focused on dihydrothymine. Although dihydrothymine is produced by ionizing radiation preferentially under anoxic conditions (6.7), it shares common structural characteristics with thymine glycol: the thymine rings of both compounds are no longer planar due to the saturation of C5-C6 bond and assume a half chair conformation with C5 and C6 significantly out of the plane of the other four ring atoms (8). Since it is difficult to selectively introduce dihydrothymine into DNA by ionizing radiation or chemical reagents, we have used the following alternative approaches. In the first approach, DNA containing dihydrothymine was prepared by invitro DNA synthesis catalyzed by DNA polymerase in the presence of dihydrothymidine 5’-triphosphate (DHdTTP). In the second approach, we developed an invivo system in which exogeneously added dihydro thymidine was incorporated into DNA. Thus we could obtain information not only about the coding property of dihydrothymine and the effect of dihydrothymine on the local structure of DNA, but also about the initial interactions of the molecule with DNA polymerases.

Free radicals and activated oxygen species are generated in a variety of ways. For example, most of the biological effects of ionizing radiations can be ascribed to the oxygen radicals they produce, such as superoxide anion (O₂$$\overline \bullet $$) and hydroxyl radical (·OH) (1). Such radicals are also produced deliberately by macrophages during inflammatory responses, evidently as a means of destroying invading cells (2). It seems likely that the normal reduction of O₂ to water by cytochrome C oxidase might inadvertently release intracellular activated oxygen. Indeed, bacteria devoid of superoxide dismutase, which destroys O₂$$\overline \bullet $$, display a strongly increased spontaneous mutation frequency in aerated but not in hypoxic culture (3). This apparent production of oxygen radicals under normal aerobic conditions underscores the need to understand the nature of the cellular defenses against oxidative damage. We are investigating the mechanisms that cells employ to avert oxidative damage and to correct such damage when it does occur.

5-Hydroxymethyluracil (HmUra) can be formed in DNA from thymidine by the action of ionizing radiation or by activated leukocytes (1–7). HmUra can be removed from DNA in vitro through the action of HmUra-DNA glycosylase (8,9) The repairability of HmUra suggests that it is deleterious to cells but the nature of its effects on cell function are uncertain. HmUra can also be introduced into cellular DNA as a result of the incorporation of the nucleoside, 5-hydroxymethy1–2’-deoxyuridine (HmdUrd) (10–12). HmUrd has been reported to be toxic to cells in culture and to animals (11,13,14) but the mechanism of this toxicity is unknown.

In response to DNA damage, the cell initiates a series of enzymatic reactions designed to repair its genetic structure and restore its normal function. In some cases, the metabolic consequences of these reactions can so drastically perturb cellular homeostasis that they lead to cell death. It is important to understand these biochemical processes in order to design rational approaches to alter their course.

Increased understanding of the tumor promotion phase of chemical carcinogenesis has resulted from the discovery of pure tumor promoting-agents and the development of animal models in which tumor promoters modify experimental carcinogenesis (1). Simultaneously, interest in this area of cancer research has been stimulated by epidemiological studies indicating that a promotion phase is important in the development of human cancer (2). Tumor promoters cause or allow the expression of the latent tumor phenotype induced in some cells by limited doses of carcinogens. Commonly, the model systems where tumor promotion plays an important role in neoplastic development are those involving epithelial tissues, in particular those composed of more than one cell type or cells in different states of maturation such as skin, breast and bronchus. As a classical model for carcinogenesis in a lining epithelium, mouse skin has provided most of the conceptual framework regarding the biology of tumor promotion (1,3).

Specific biochemical and genetic events that appear to be essential to promotion of transformation in JB6 cells have been identified.1 The two earliest events appear to be the activation of protein kinase C (PKC) (and the subsequent phosphorylation of its substrates) and an increase in oxidant stress. There is a significant body of evidence that implicates reactive oxygen, and the superoxide radical anion in particular, as an important mediator of tumor promotion.2 This evidence includes studies with nonphorbol tumor promoters as well as TPA, several tissues invivo, and a number of different model systems invitro. For example, copper(II)(3,5-diisopropylsalicylate)₂ (CuDIPS) is a potent inhibitor of phorbol ester-mediated tumor promotion in the epidermis of CD-I mice.3 CuDIPS is a lipophilic copper complex that catalyzes the dismutation of the superoxide anion radical at a rate similar to that of SOD, its biological counterpart Invivo.4 In invitro studies with JB6 P+ cells, SOD was observed to inhibit promotion of transformation by TPA. Other invitro studies have demonstrated that SOD will inhibit TPA-dependent radiogenic transformation of C3H10T1/2 cells.5,6

The evidence is convincing that oxidants and agents which induce a cellular prooxidant state can act as carcinogens, in particular as promoters and progressors (1–7). Bonafide oxidants with promotional activity include H₂O₂, superoxide, ozone, hyperbaric oxygen, peroxyacetic acid, chlorobenzoic acid, benzoyl-peroxide, decanoylperoxide, cumene-hydroperoxide, p-nitro-perbenzoic acid and periodate (8,9). Infiltrated phagocytes represent a major source of oxidants in inflamed tissues (10,11) and in several instances inflammation appears to be a prerequisite for promotion (12–14). Tumor promotion results in the clonal expansion of Initiated cells at the cost of the surrounding tissue. This can involve the modulation of the expression of growth- and differentiation-related genes and selective cytostatic effects on non-initiated cells. It is of obvious interest to elucidate the mechanism(s) by which oxidants exert their promotional activity. In addition, the question arises whether other classes of promoters which lack oxidizing properties might owe at least part of their promotional activity to the induction of a cellular prooxidant state (1). Support for this notion has been obtained for the phorbol-ester promoter phorbol-12-myristate-13-acetate (PMA). PMA induced an increase in the ratio of oxidized over reduced glutathione in mouse epidermal cells (5). Furthermore, several antioxidants, in particular CuZn-superoxide dismutase (SOD), inhibit cellular reactions induced by PMA as well as promotion invivo and invitro implicating the superoxide radical O₂ in its mechanism of action (1, 4, 9, 15).

Although the molecular mechanisms involved in carcinogenesis are poorly understood at this time, intermediary events in the development of cancer have been described through the selective actions of discrete chemical agents (1). These processes have been studied in various model systems and have been termed initiation, promotion and progression (2). Initiation appears to involve the modification of cellular DNA, resulting in genotypically altered cells, while promotion encompasses a continuum of events allowing for the selection and clonal expansion of the initiated cells (3). Finally, progression completes the conversion of pre-malignant cells to malignant cells. Substantial evidence supports the involvement of free radical species, especially those derived from molecular oxygen, in multiple aspects of these processes (4).

Our original suggestion in 1980 (1,2). that free radicals may be involved in tumor promotion was based on linkage of a number of puzzling observations about tumor promoters with newer insights into the biological and pathological role of free radicals. Of note was that the then known skin tumor promoters in the classic two-stage mouse skin tumor promotion system were all inflammatory agents. However, the lack of correlation among phorbol esters of tumor promoting ability with inflammation, as measured by standard techniques such as number of infiltrating inflammatoiy cells, seemed to preclude a role for inflammation in tumor promotion. Yet it was hard to ignore the many clinical observations, dating back to the nineteenth century, that inferentially seemed to point to chronic inflammation having some role in tumor development, observations ranging from the frequency of basal cell tumors on the bridge of the nose of eye glass wearers, to colorectal carcinoma in patients with ulcerative colitis. Our first approach to a possible role for free radicals in tumor promotion was to consider that the source of such free radicals in the mouse skin tumor initiation-promotion model might be derived from the activation of phagocytic cells, a known source of free radicals and active states of oxygen. Further, hepatic models of tumor promotion tend to feature compounds that are powerful inducers of cytochrome P-450 or of peroxisomes, both conceivably sources of free radicals or other active states of oxygen.

Repeated exposure of mice to UVB (280 to 320 nm) radiation results in the development of a state of tumor-specific suppression that renders the host Incapable of mounting an effective immunological rejection response to antigenic determinants expressed by UV-induced skin tumor cells (1,2). In this system, normal mice can reject transplanted UV-induced tumor cells, whereas UV-treated mice cannot. The objective of this research was to determine if UVB irradiation influences mouse skin tumorigcnesis induced by an initiation-promotion protocol.

The possible role of free radicals during experimental chemical carcinogenesis has been recently reviewed and speculated on by several authors (1–4). Much of the direct evidence for free-radical participation in carcinogenesis has come from the mouse skin model system.

Previous studies of the differential response of normal, initiated and tumor cells to TPA have been performed on cell lines derived from normal epidermis, epidermis which had been treated invivo with chemical carcinogen, or disaggregated tumor tissue. Thus, the cell lines were heterogeneous in cellular origin. In the present study, derivatives of a single cell lineage were examined for responsiveness to TPA at various stages of tumorigenic potential after chemical treatments invitro. The advantage of this approach is that differences in target cell response can be related to defined stages of transformation, rather than individual variation among cells of different parental origins.

Activated oncogenes, particularly of the ras gene family, have been detected in a variety of solid human tumors and interpreted as causal agents in the development of neoplasia (1, 2). Immortalized and aneuploid mesenchymal cells from the rodent NTH 3T3 cell line could be rendered tumorigenic by transfection oncogenes (3–5) while normal fibroblasts required two cooperating oncogenes or additional enhancers (6–9). So far, it has not been possible to malignantly transform human fibroblasts with oncogenes in vitro (10, 11).

The objective of the present work was to obtain chemopreventive compounds that can trap direct-acting carcinogens within the lumen of the gastrointestinal tract and thus prevent these carcinogens from attacking tissues of the host. Many direct-acting carcinogens are electrophiles (19, 36). One possible strategy for blocking their action is by trapping them with nucleophiles (electron donors). In the studies to be presented, emphasis has been placed on trapping direct-acting carcinogens in two sites, i.e., the stomach and the large intestine. To some extent different considerations pertain for each.

Indole-3-carbinol (I-3-C), a normal constituent of the human diet via cruciferous vegetables, was examined for its ability to protect mice against 24-hour N-nitrosodimethylamine (NDMA)-mediated hepatotoxicity. NDMA (20 mg/kg body weight) alone produced extensive hemorrhagic and centrolobular necrotic lesions, with a necrotic severity index of 3.0 ± 0.4 (scale of 0-5). Treatment with 50 mg/kg body weight of I-3-C by gavage, 1 hour prior to NDMA, substantially protected against hemorrhagic lesions. Furthermore, I-3-C lowered the NDMA-mediated tissue necrotic index to 1.5 ± 0.3, by reducing the extent of tissue necrosis rather than the severity in the necrotic region. Release of liver enzymes into the blood correlated with the histopathology; I-3-C reduced NDMA-mediated elevated activities of plasma alanine transaminase and ornithine transcarbamylase by 84% and 51.3%, repectively (Table 1). Plasma activities of these enzymes were used as indicators of hepatotoxicity. Although no changes in liver non-protein sulfhydryls were evident at 24 hours after NDMA, ascorbate levels were reduced to 40% of controls values. Treatment with I-3-C prior to NDMA prevented this decline in tissue ascorbate concentrations.

Ozone (O₃), a reactive species of oxygen, is an important natural constituent of the atmosphere (1). Background levels of ozone in the lower atmosphere may range up to 0.1 ppm and are modified by geographic elevation, solar radiation and climatic conditions (2). Since some ozone effects are radiomlmetic its actions may be enhanced in the presence of ionizing radiation from background and/or man-made sources (3,4).

It is well-known that endurance exercise training leads to an overall increase in the oxidative capacity of oxidative type (red) skeletal muscles through increased biogenesis of mitochondria (1, 2). While overall muscle mitochondrial content increases, the mitochondrial lipid/protein ratio and the specific activities of mitochondrial enz37mes remain constant (2). Besides their known physiological function as providers of ATP through oxygen reduction, mitochondria are also potential sites of oxygen free radical formation by erroneous, incomplete electron transfer to oxygen, most probably at the site of ubiquinone oxidoreduction (3). Mitochondrial free radical production is likely to increase with the massive increase in oxygen consumption during physical activity. This suggests that exercise should result in an increased demand on antioxidant systems to respond to the increased production of oxygen-derived and other free radicals.

Aggressive oxygen species (such as superoxide anion, hydroxyl radical, singlet oxygen and longer-lived reaction products (e.g. hydroperoxides, alkenals) have been implicated in cancerogenesis but it is very difficult to show their direct role in vivo. Indirect evidence can be obtained from the inspection of the body’s multilevel defense system against oxygen radicals which includes essential antioxidants, i.e. β-carotene and the vitamins A, C and E. Since the dietary supply of the latter can vary considerably cancerogenesis might, at least in part, be inversely related to the status of antioxidant vitamins. There is growing evidence in animals and humans in favour of this concept. In animals vitamin A deficiency results in metaplasia whereas experimentally induced tumors can be diminished by β-carotene, vitamins A, C and E. In the human many dietary surveys have convincingly shown that the intake of fresh fruits and leafy green-yellow vegetables as well as the calculated consumption of the above mentioned essential antioxidants is inversely related to the mortality from cancers (1–4). Dietary surveys have, however, inherent weaknesses and thus require confirmation by the measurement of plasma antioxidants in prospective studies.

Cancer is the second leading cause of mortality in the U.S., responsible for over 20% of the approximately two million total annual deaths (1). Treatment of cancer has provided and will continue to provide enormous benefits to some categories of affected individuals. At the same time, however, interventions that may prevent cancer in healthy individuals could, at least in theory, also afford great benefits to society as a whole. For example, a complete cure for acute leukemia, which accounts for several thousand of the approximately 400,000 total cancer deaths in the U.S. each year, would certainly be a remarkable breakthrough in medical research. In terms of public health impact, however, even a small reduction, on the order of 30%, in the development of epithelial cell cancers, which account for 90% of U.S. cancer deaths, due to dietary supplementation with micronutrients could conceivably prevent over 100,000 cancer deaths annually. With respect to known etiologies, cigarette smoking is the leading identified avoidable cause of cancer as well as mortality from all causes, accounting for about 30% of all deaths due to malignancy in the United States (2). Heavy alcohol consumption, which ranks second, is responsible for about 3% of cancer deaths. Recently, promising, but unproven, hypotheses have suggested that a substantial percentage of cancers may be preventable through dietary changes. Although diet has been postulated to account for as many as 35% of the annual deaths from cancer, which characteristics of diet, if any, increase or decrease the risk of cancer remains unclear (2).

Vitamin A, beta-carotene, and synthetic retinoids have attracted wide interest as premising chemopreventive agents, based on work in rodents and in cultured cells and organs, and low dietary intake or low serum levels of vitamin A or beta-carotene have been associated with high risk for lung cancer (1–4). Even if the data were stronger, observational epidemiological studies cannot establish whether or not changing the vitamin A status of specific individuals or groups of individuals would alter their subsequent cancer risk. The hypothesis that increasing retinoid intake in humans would reduce the risk of lung cancer must be tested in humans. For both practical and ethical reasons, such studies should be undertaken initially among those at highest risk.

There is a great variation in cancer incidence with diet, as has been recently reviewed (1,2). Epidemiological data suggest that environmental, specifically nutritional, factors play a major role in the etiology of cancer at many different sites (1–3). There are now several epidemiologic studies which suggest that components of vegetables might play a beneficial role in lowering the incidence of cancer (some examples of such studies are given in references 1-4). Although many compounds with anticarcinogenic potential are present in vegetables, it is possible that anticarcinogenic protease inhibitors contribute to the low cancer rates observed in certain human populations with high levels of vegetables in the diet. For example, the low cancer incidence rates in the Japanese and Seventh-Day adventists could be due to high levels of dietary protease inhibitors; it has been estimated that individuals in these populations ingest, on the average, more than 330 mg of protease inhibitors per day (3). There are, however, many other hypotheses which have been presented to explain the low cancer rates in these human populations. So many different variables are present in the diet that the effect of any specific anticarcinogenic agent cannot be distinguished in such epidemiologic studies; however, it is possible to distinguish the effects of specific anticarcinogenic agents in laboratory experiments. In this report, our own laboratory studies on anticarcinogenic protease inhibitors will be summarized and discussed. Although laboratory studies can give much information about the effects of potential chemopreventive agents, ultimately epidemiologic intervention studies must be performed to determine whether candidate chemopreventive agents are truly capable of preventing cancer in human populations. The current evidence that dietary protease inhibitors do have a role in lowering the cancer incidence in human populations has recently been reviewed (6 and 7).

Protease inhibitors have been shown to possess anticarcinogenic properties. Two synthetic protease inhibitors, tosyl phenylethyl chloro-methyl ketone (TPCK) and tosylarginine methylester (a competitive trypsin substrate), have been demonstrated to block the formation of skin tumors promoted with TPA in mice (1). These studies were repeated using the actinomycete-derived protease inhibitor leupeptin (2), thus establishing an in vivo basis for the involvement of proteases in the mechanism of carcinogenesis.

Over the last few years, working with a diet that is devoid of choline and in methionine (CD) we have observed that when this diet is fed to rats, they develop not only fatty liver, but also hepatocellular necrosis and cancer. In the course of our investigation as to the mechanism of cancer development by a dietary deficiency without any added carcinogen, we found that early nuclear lipid peroxidation (1), DNA alteration (2) and cell proliferation are very common features in the liver. Free radical generation and DNA alteration in a proliferating organ has been proposed as the initiating event in the development of liver cancer (see Fig. 1). However, the nature of free radical generated and the nature of DNA alteration are not known.

Higher plants contain an extensive array of biologically active chemicals, some of which are potent modifiers of chemical carcinogenesis (1). Specifically, some of these agents have been shown to be active in inhibiting the initiation stage of the carcinogenesis process (2). Members of the Allium genus, which include onions and garlic, are rich in sulfur-containing compounds. Among the major components of garlic oil are diallyl disulfide (DADS, 66%) and diallyl sulfide (DAS, 14%) (3). It has been previously shown that the latter agent, DAS, can inhibit 1,2-dimethylhydrazine-induced genotoxicity in the murine colonic epithelium (4) and cyclophosphamide-induced genotoxicity in the murine urothelium and hair follicles (5). Recently it has been shown that DAS can inhibit 1,2-dimethylhydrazine-induced colon tumorigenesis (6). The goal of this study was to determine the optimal time for the oral administration of DAS, prior to carcinogen treatment, and to determine the efficacy of other structurally analogous sulfur-containing compounds.

Metastatic inefficiency (1) presents a unique challenge for the investigation of local and systemic factors in physiologic anticarcinogenesis. It appears that a point has been reached whereby there is general agreement that the transfer of cancer cells from the primary tumor to secondary sites is a multi-step process (2) and likely also multifactorial since each step is affected by different factors (3,4). Purely mechanical factors seem to be insufficient to explain why 99.9% of circulating tumor cells are destroyed before they are able to produce a viable metastatic focus. Recent animal studies indicate that only specialized cancer cells are capable of initiating metastatic foci and that they are to a certain degree site specific (5). There is a need for further development of animal models, whereby experimental conditions are controlled, in order to explain the diversity of response to treatment (6).

Ionizing radiation is considered to cause its damage to cellular DNA via OH free ¡radicals and by direct ionization of the DNA (1). In attempting to simulate the OH portion of this damage we have used hydrogen peroxide (2). We showed that the probable mechanism of damage production by this agent was through OH radicals produced by its interaction with variable valency metal ions bound to DNA (in a Fenton type reaction). Previously it had been shown that hydrogen peroxide treatment causes single-strand breaks (SSB) (but not double-strand breaks, DSB) in cellular DNA (3). More recently the reaction of hydrogen peroxide with DNA was found to give the same range of base-damaged products as that formed by low-LET ionizing radiation (4). Therefore, the spectrum of all damage types and their relative yields produced in the individual moieties of the DNA (sugar and base) by OH radicals is the same for hydrogen peroxide as for low-LET ionizing radiation (5). When cells are treated with hydrogen peroxide at 0°C, killing does not occur until the number of SSB present (and the base damage concomitantly produced) is 2,600 times higher than that present when equivalent cell kill is induced by ionizing radiation (2). Clearly although the same damage types are formed by hydrogen peroxide their presence is inconsequential to the cell survival. We conclude that it is not the chemical structure of the OH radical-damaged sites per se which is responsible for the ionizing radiation-induced cell killing. It is the concentration of these damaged sites in local regions of the DNA (producing what we have called locally multiply damaged sites, LMDS) which is responsible for the effectiveness of ionizing radiation in inducing cell killing.

The perception of radiation hazards in society must be completely re-evaluated as a result of the recent realization that radon represents the largest source of natural background radiation to the U.S. population (1). Studies designed to identify a link between lung cancer and radon in the home, by seeking a correlation between lung cancer incidence and variations in regional radon levels, have produced results that are equivocal. In epidemiological studies, uraninum miners exposed for many years to high radon levels in the mines show an enhanced incidence of lung cancer (2–4). The perceived risk to the public from breathing lower levels of radon in the home are based on an extrapolation from the high concentrations experienced by the miners. However, extrapolating from the uranium miner data involves complications because most miners studied were also heavy cigarette smokers. The number of non-smoking miners is too small to separate the effects of radon from those of the cigarette smoke.

In vitro cell systems are well established in the study of mechanisms of neoplastic transformation by a variety of agents including chemicals, ultraviolet light and ionizing radiation (for recent reviews see references 1-4). However, with few exceptions, systems that are used for quantitative studies are rodent-derived cell lines, usually of fibroblastic origin. This situation exists for the very good reason that human cells in culture are extremely difficult to transform, and where this has been achieved the frequency of transformation is three to four orders of magnitude less than that typically seen for rodent cells (2). There is an obvious need for a human cell system(s) for use in in vitro investigations of neoplastic transformation.

Radiation is the most thoroughly studied environmental carcinogenic agent and cancer induction is the most significant late effect of radiation. Ionizing radiation creates a number of lesions in DNA including base modification, single and double strand breaks, and sugar damage. A significant proportion of the radiation-induced mutations in somatic cells are associated with chromosomal deletions and gene rearrangements. These alterations are the result of breakage of the chromosomes followed by loss of genetic material (deletions) or rejoining of broken chromosomes (rearrangements) (l). Radiation has also been found to be a point mutagen in certain model assay systems. The relationship of these geno-toxic effects of radiation to molecular mechanisms of radiation carcinogenesis is not yet understood. We have decided to address this issue using current molecular concepts involving oncogene activation applied to the biologically well-defined experimental model of rat skin carcinogenesis by ionizing radiation. Several types of tumors develop in rat skin following localized exposure to single or fractionated doses of ionizing radiation (2) which include squamous cell carcinomas, basal cell carcinomas, sarcomas, clear cell carcinomas, and sebaceous cell tumors. The histologic type of the tumor presumably reflects the cell type of origin.

The association between exposure to asbestos fibers and the development of bronchogenic carcinoma of the lung and diffuse mesothelioma of the pleura and peritoneum have been well established in both man (1,2) and experimental animals (3,4). Furthermore, cigarette smoking can enhance the incidence of lung cancers in asbestos workers in a synergistic fashion. (5). It is not known for certain whether specific chemicals or the high LET α-irradiation from deposited polonium is the important factor in carcinogenesis by cigarettes and its enhancement by asbestos. Previous studies from this laboratory have found that asbestos fibers, at a concentraction which itself was Ineffective in inducing oncogenic transformation in vitro did potentiate the oncogenicity of low LET ionizing radiation in a supra-additive fashion (6,7). Furthermore, leaching the fibers with acid to remove any possible chemical contaminants did not diminish their ability to enhance the oncogenic transforming potential of radiation (7). These results suggest that the mere physical presence of the fibers is crucial for the observed radiation-fiber interaction.

The glutathione redox cycle is the major pathway for providing cells with reducing equivalents for bioreduction of drug-or radiation-induced superoxide and peroxide (1).

Bis (3, 5-diisopropylsalicylato) copper II (CuDIPs), a low molecular weight lipid-soluble complex, has been suggested to be an effective, nontoxic radioprotectant in mice exposed to gamma radiation from a 60Co source (1). CuDIPs has also been shown to have antiinflammatory activity (2) and to catalyze the disproportionation of superoxide (3). In view of these observations, the present study examined the toxicity and radioprotective efficacy of CuDIPs, as well as those of diisopropylsalicylate (DIPs) and CuCl₂, in both male and female B6D2F1 mice.

Glutathione (L-γ-glutamyl-L-cysteinyl-glycine), a tripeptide present in substantial concentrations in virtually all mammalian cells, plays a variety of roles in catalysis, metabolism, transport, and in cellular protection. A review of glutathione metabolism was presented at the first International Conference on Radioprotectors and Anticarcinogens (1); see also (2–6). A summary of glutathione metabolism is presented in Fig. 1. Understanding of the biochemistry of glutathione has led to procedures by which cellular levels of glutathione may be decreased or increased; glutathione metabolism may be modulated in other ways by selective inhibition of certain enzymes.

N-acetylcysteine (NAC), a precursor of intracellular cysteine and glutathione (GSH)) (1), is extensively used in the treatment of patients suffering from respiratory diseases. Moreover, this synthetic aminothiol has been shown to possess a variety of antitoxic properties in humans, animals and in vitro test systems (2).

Aminothiols play a fundamental role in the physiology of the living cell. The most abundant and best studied compound of this class is glutathione (GSH). It is considered of paramount importance for many functions, such as amino acids transport, protein and DNA synthesis, membrane structure and activity (for thorough reviews see refs. 1 and 2). Moreover, GSH being a nucleophilic compound, it hampers alterations induced by reactive oxygen species and by endogenous or exogenous metabolites (for a review see ref. 3). Depletion of hepatic GSH levels affects a number of biochemical functions, as well as the efficiency of the cell response to physically or chemically induced injuries (3,4).

Glutathione (GSH) has been implicated as the major intracellular nonprotein thiol involved in protecting hypoxic cells against radiation injury (1–10). In contrast, we have demonstrated increased radiosensitivity for GSH-depleted aerobic and hypoxic A549 human lung carcinoma cells (2–5). This parallel increase in radiosensitivity conflicts with the popular interpretation of existing data regarding the effect of GSH depletion on hypoxic sensitivity alone (5). Subsequently, ratios of (hypoxic slope) / (aerobic slope) may be misleading if GSH depletion alters both responses, but by different mechanisms.

The role of DNA damage modification in the protection of mammalian cells from the lethal effects of radiation by aminothiols has been the subject of much research over the last 20 years since the early demonstrations that cysteamine reduced the level of DNA single-strand breaks (SSBs) in irradiated cells (1, 2). Despite this considerable effort, the actual mechanisms of radioprotection at both the cellular and DNA level remain poorly defined. Perhaps one barrier to our understanding of these effects has been the failure to recognize that many chemical modifiers of radiosensitivity, including sensitizers such as oxygen and protectors such as cysteamine, may not simply dose-modify numbers of DNA lesions but may also change the “spectrum” of lesions induced by low-LET radiations. For example, lesions such as DNA-protein cross-links (DPCs) (3) and 8,5’-cycloadenosine (4) are actually preferentially induced under hypoxia despite the fact that this condition offers considerable radioprotection to the cell. Even these presumably minor lesions could contribute significantly to lethality under appropriate conditions, e.g. in cells exhibiting a particular repair defect (3). Evidence that aminothiols also cause a shift in the spectrum of radiation-induced lesions comes from the recent report by Radfotd (5) that cysteamine altered the levels of 4 types of roiA lesion—SSBs, double-strand breaks (DSBs), DPCs and base damage—in different proportions.

The doses of radiotherapy that normal central nervous system (CNS) parenchyma and stroma can tolerate without injury are too low to provide cure for malignant astrocytic gliomas. One potential solution to this problem is to develop compounds that protect the normal CNS tissues more effectively than tumor cells against irradiation. The phosphorothioate radioprotectors such as WR2721, WR77913 and WR3689 are highly water-soluble and do not readily cross the blood-brain barrier. Injection of these drugs via routes that bypass the blood-brain barrier has allowed us to assess radioprotection in the rat cervical spinal cord.

Glutathione (GSH), a multifunctional tripeptide thiol found in virtually all mammalian cells, has a prominent role in the maintenance and execution of many homeostatic and regulatory processes of the cell (1). Of the many functions of GSH, perhaps the most studied, is its ability to protect the cell against the detrimental effects of ionizing radiation, reactive oxygen compounds, free radicals, and toxic xenobiotics (2). The GSH oxidation-reduction cycle is of primary importance when dealing with reactive oxygen intermediates as well as the detoxification of xenobiotics. Although the elimination of each is catalyzed by different enzymes, GSH peroxidase in the former and GSH S-transferases in the latter, the maintenance of a high ratio of GSH:GSSG by the cycle is of vital importance to both pathways.

The search for chemical substances that induce radiation protection in tissues has been in progress for many years. Interest has tended to concentrate on small molecules such as simple alcohols, dimethylsulphoxide and in particular, numerous thiols and thiophosphates. Major objectives have been two-fold. Firstly, the development of techniques for protecting individuals from the effects of exposure to whole-body radiation, accidental or otherwise and secondly, clinical strategies for reducing normal tissue morbidity in the radiation therapy of cancer. Several of the papers in these proceedings are concerned with various aspects of these topics. This paper, however, addresses another approach based on manipulation of oxygenation of normal tissues and tumours not only for inducing radiation resistance but also for obtaining increased therapeutic efficacy of some anti-tumour strategies.

The curative treatment of cancer by radiotherapy is limited by the sensitivity of critical normal tissues surrounding the tumour that are inevitably included in the irradiation volume. In the last decade two new developments have progressed from the experimental stage in animals to clinical testing: Radiosensitizers that appear specific for tumour cells and radioprotectors that appear specific for normal tissues have both been developed. Of these the prototype compounds are nitroimidazole, misonidazole (miso), as a radiosensitizer (l) and the phosphorylated aminothiol WR-2721 (S-2-(3-amino-propylamino) ethyl phosphorothioic acid) as a radioprotector (2). The specificity of action of these agents does not relate to phenotypic alterations connected with the malignant transformation of cells, but rather to differences in regional oxygen tensions that result from the poor vascularity of tumours. The inadequate vascular array leads to regions of hypoxia which confer radloresistance (by a factor of 2.5 to 3.0). Miso can act as an oxygen-mimetic agent in such hypoxic regions and can repair damage from-radiation-induced radicals, a process which already occurs efficiently if oxygen is present. Miso is ineffective in oxygen. Since hypoxic cells rarely exist in normal tissues the effect is tumour-specific. Lesion repair by hydrogen donation by endogenous thiols is believed to compete with oxygen fixation of lesions, so it follows that the effect of thiols as radioprotectors is also influenced by local oxygen concentrations. Thiols are ineffective in anoxia. They are also relatively ineffective at extremely high oxygen concentrations, because the balance of the competition is loaded towards fixation. However, at oxygen tensions that are close to the transition region for radiosensitivity to radio-resistance (i.e., the K value), added thiols are maximally effective and the balance can most readily be tipped one way or the other. Since oxic cells predominate in normal tissues, the radioprotection by thiols seems to be specific for normal tissue.

Many carcinogens and antineoplastic agents act primarily by damaging DNA. A very large portion of this damage is repaired by normal cells. Defects in these repair processes have serious consequences, as shown by xeroderma pigmentosum, ataxia telangiectasia and other cancer-prone genetic diseases. Drugs can also interfere with DNA repair pathways. They, like the genetic defects, increase the lethality of DNA damaging agents. Unlike these diseases, however, some drugs have been shown to decrease carcinogenicity. The basis for this paradoxical effect may be the ability of DNA repair inhibitors to convert normally sublethal misrepaired lesions into lethal ones. Drugs that interfere with DNA repair thus have promise not only for increasing the efficacy of chemotherapy, but also for decreasing carcinogenicity.

The bleomycins (BLMs) are a family of antitumor antibiotics used clinically for the treatment of squamous cell carcinomas and malignant lymphomas (1–4). Bleomycin A₂ is the major constituent of the clinically used mixture of bleomycins. Although biochemical studies have indicated several loci that may contribute to the overall therapeutic effects obtained with bleomycin, the accumulated evidence suggests that the primary effect is at the level of oxidative DNA degradation (5). DNA degradation mediated by bleomycin is metal ion dependent; with the exception of Co·BLM (6), DNA degradation by metallobleomycins also requires oxygen (5,7). It is believed that the degradative event is actually mediated by a ternary complex consisting of bleomycin, a metal ion and oxygen. There are several mechanistically interesting facets of BLM-mediated DNA degradation, including the observed sequence selectivity for 5’-GT-3’ sequences (7–9) and the fact that double-strand cleavage occurs at a frequency greater than that expected from the random accumulation of single-strand breaks (10,11).

Liposomes are vesicles composed of concentric phospholipid bilayers that are formed spontaneously when an aqueous solution is added to a dried lipid film (Figure 1). Several types of liposomes—multilamellar vesicles (MLV), small unilamellar vesicles (SUV), and reverse evaporation vesicles (REV)—have been described. MLV are used extensively as drug carriers for antineoplastic and antimicrobial drugs: lipophilic drugs can be incorporated in the MLV’s large lipid compartment and hydrophilic drugs can be encapsulated between the vesicles’ lipid bilayers. The liposomes’ size. manbrane charge, fluidity, and other characteristics may modify their in vivo behavior. The changes in drug bioavailability and biodistribution associated with liposome incorporation have been exploited to alter drug toxicity and enhance drug targeting. Design of liposomal drug carriers should be based on the rational selection of the drug or drug analogue; drug carriers; and the target disease, be it focused on a cell, an organ, or a tissue. We review here general pharmacologic concepts of the development of liposamal carriers for antimicrobial and antineoplastic agents.

Hamster osteolytic virus H-l is an autonomous parvovirus. Its genome, consisting of a linear single-stranded DNA molecule of 5176 nucleotides, carries 3 structural genes of capsid and non-capsid protein; the virion replication depends therefore almost completely on the enzyme system of the host cells (1). H-l can inhibit the formation of tumours in experimental animals. For example, infection of Syrian hamster with H-l reduced the incidence of “spontaneous” tumours from 0.05% to 0.0023%, the incidence of tumours induced by the chemical carcinogen DMBA from 95% to 38% (3), and the incidence of tumours induced by adenovirus from 67% to 25% (4), other parvoviruses possess similar functions. The reports from human epidemiological studies indicated that infection by parvovirus lowered significantly the incidence of several cancers (5, 6).

Biological response modifiers (BRMs) are those agents or approaches that modify the relationship between the tumor and host by modifying the host’s biological response to the tumor cells with resultant therapeutic effects (1). BRMs may modify the host responses in several ways: 1.increase the host’s antitumor responses through augmentation and/or restoration of effector mechanisms or mediators of the host’s reaction which may be deleterious;2.increase the host’s defenses by the administration of natural biologicals (or the synthetic-recombinant derivatives thereof) as effectors or mediators of an antitumor response;3.augment the sensitivity of the host’s tumor cells to endogenous mechanisms for the control of tumor growth;4.alter the transformed phenotype by increasing the differentiation/maturation of tumor cells;5.increase the ability of the host to tolerate damage by cytotoxic modalities of cancer treatment.

The major challenge facing the oncologist is how to treat cancer that has disseminated to and is growing in multiple organs throughout the body. Despite major advances in general patient care, surgical techniques, and adjuvant therapies, most deaths from cancer are caused by the growth of metastases that are resistant to chemotherapy or radiotherapy. A major factor which prevents treatment of metastases is the fact that cancer cells populating both primary and secondary neoplasms are biologically heterogeneous (1, 2). By the time of diagnosis, and certainly in clinically advanced lesions, malignant neoplasms contain multiple cell populations exhibiting a wide range of biological characteristics such as cell surface structures, growth rate, sensitivity to various cytotoxic drugs, and the ability to further invade and metastasize. The implication of the heterogeneous responses of tumor cells to conventional anticancer drugs is that the successful therapy of disseminated metastases must circumvent the problems of biologic heterogeneity and be a treatment modality to which resistance is unlikely to develop.

Biological response modifiers (BRMs) are those agents or approaches that influence the relationship between the tumor and host by modifying the hosts response to tumor cells, with resultant therapeutic activity (9). A series of studies with recombinant cytokines designed to better understand their immunomodulatory and therapeutic properties have been undertaken. The present report discusses recombinant murine interferon-gamma (rM IFN-g), recombinant human tumor necrosis factor (rH TNF), rH interleukin-2 (rH IL-2), and rM colony stimulating factor-gm (rM CSF-gm). We discuss the development of Optimal Therapeutic Protocols (OTP) for preclinical studies and initial clinical trials that have been initiated to test the resultant clinical hypotheses.

Early attempts at passive immunotherapy by administering natural antisera derived from recovered cancer patients to patients with active disease failed to result in significant antitumor responses. Many of the limitations of this serotherapy were remedied by the revolutionary innovation described by Kohler and Milstein (1) of monoclonal antibody production by means of hybridoma technology. Large quantities of homogeneous, highly specific antibody to a defined antigenic determinant became available for clinical trials. Most monoclonal antibodies (mAbs) used to date to treat cancer patients have been murine mAbs. Recently, clinical grade human mAbs have been generated in sufficient quantity to permit investigation.

Human malignancies arise commonly after exposure to environmental carcinogens. Bladder cancer is a well-studied and excellent preclinical model for this process. Transitional cell carcinoma (TCC) of the lower urinary tract accounts for approximately 4% of human cancers. Most of the TCC arise in the urinary bladder, with a male:female ratio in the U.S. of about 3:1. Chemical carcinogen-induced mouse and rat tumors in situ and transplantable bladder cancer models, which mimic the pathogenesis of the human disease, have been developed (1–4). Neoplasms can be histologically graded and staged in a manner similar to that for human bladder carcinoma.