Just Say NO to Cancer?

In this issue of the Journal, Xie and co-workers ( 1 ) demonstrate that introducing nitric oxide (NO) synthase type II (NOSII, iNOS) into tumor cells produces cytotoxicity not only in the transfected cells but also in bystander tumor cells. Here, we discuss the potential implications of these experiments in the light of present knowledge on the complex role of NO in tumor biology. The growth of solid tumors is regulated by interactions betweenen do the lial cells of the tumor vasculature, tumor infiltrating immune cells (such as T lymphocytes and macrophages), and the tumor cells themselves. In these cellular interactions, the unusual biologic messenger molecule and cytotoxin, NO, may play important pathobiologic roles in addition to its many physiologic functions ( 2 ). Endogenous NO production from l -arginine has been directly or indirectly demonstrated in all of these cell types. In most cases, the inducible NO synthasegene (Nos2)—one of three known human Nos genes ( 3 )—was switched on (presumably by NF-κB-dependent mechanisms). Its gene product, the high-output NOS-II isoform, unlike its constitutively expressed low-output counterparts (NOS-I and-III), is not regulated by the intracellular concentration of free calcium ( 3 ) and is chronically active. Such continuously high exposure of cells to endogenous NO as well as exogenous NO donors will inhibit proliferation and induce cell death ( 4 ). High NO levels inhibit mitochondrial respiration, the citric acid cycle, glycolysis, and DNA replication. Locally high levels of reactive oxygen species (ROS), stemming, for example, from host immune cells or from an insufficient oxygen supply to the tumor tissue, may exacerbate these toxic effects by generating even more reactive compounds, such as peroxynitrite (ONOO − ). The latter compound arises from the diffusion-limited interaction of NO and O 2− and is even more reactive than NO, but it is stable enough to diffuse to and thus harm tumor cells. Apart from these tumoricidal effects, NO has facilitated tumor growth and vascularization in a few experimental models, and, in certain human carcinomas, endogenous NO production was positively associated with tumor grade ( 5 ). Thus, NO has acomplex, at least dual, action on tumor growth that may depend on the local concentrations of NO, additional factors such as the presence of ROS, and the type of tumor and its susceptibility to NO. In those models where NO had a permissive effect on tumor growth, NOS activity was up to two orders of magnitude lower than that associated with NO-dependent tumor toxicity and apoptosis ( 5 ). The dual action of NO ontumor cell proliferation is reminiscent of what has been repeatedly demonstrated for ROS ( 6 , 7 ). Maximal growth promotion by ROS is observed when cells maintain a low but sufficient oxidant signal for the induction of growth-competence genes ( 8 , 9 , as may be the case with moderate NO concentrations. At high ROS concentrations, the equilibrium balance between the cellular antioxidant defense and oxidant levels is shifted, i.e., toward lipid peroxidation and DNA fragmentation. Besides this concentration dependence, the mechanisms of action of NO may also diverge considerably ( Fig. 1 ). Signaling functions at comparatively low NO concentrations involve soluble guanylylcyclases as the principal molecular target and subsequent increases in the intracellular level of the second messenger molecule, cyclic guanosine monophosphate (cGMP) ( 10 ). The tumor-relevant consequences of these increases in cGMP may be the stimulation of (neo) vascularization and angiogenesis. Moreover, endogenous NO may not only promote the growth of existing tumors, but it may also be tumorigenic itself. NO cancause mutations by mediating the deamination of S-methylcytosine to thymine and also by increasing the formation of DNA strand breaks ( 11 ). Increased NO formation has been implicated in several tumor types associated with chronic infections and inflammations, such as gastric, duodenal, esophageal, bladder, and liver cancers ( 12 , 13 ). What are the therapeutic implications of this dual role of NO? One apparent goal may be to enhance NO synthesis to a maximum level only in tumor tissue so that cell proliferation is impaired and cell death is induced. This may be achieved in principle by simply substituting NO in the form of so-called NO donors. Indeed, NO is an effective hypoxic radio sensitizer of tumor tissue ( 14 ), while a nonselective NOS inhibitor increases tumor survival ( 15 ). Alternatively, endogenous NO formation may be induced in the tumor cell by different cytokines (tumor necrosis factor-α and interleukin 2). Tumor necrosis factor-α has been shown to induce both NO and superoxide (O 2− ) production, providing an effective means of local ONOO − formation ( 16 ). However, cytokines induce NOS-II expression systemically in the blood vessel wall, resulting in massive NO − − and cGMP-mediated vasodilatation and hypotension, which has been recognized as the primary limiting factor in high-dose cytokine therapy ( 17 ). A more direct approach is to transfer cytokine or NO synthase genes into tumor cells as a means of effective local NO delivery. However, the fraction of nontransduced bystander cells is a paramount problem when tumor cells are to be eradicated. The introduction of cytokine genes into tumor cells improves their recognition by the host immune system and their subsequent destruction. Transduced tumor cells attract an inflammatory exudate, resulting in the destruction of both the transduced and, most notably, nontransduced bystander cells ( 18 ). This bystander effect has also been observed for the transfer of genes that encode drug-metabolizing enzymes, such as cytosine deaminase or the thymidine kinase of herpes simplex virus, which convert a nontoxic prodrug into an ultimately cytotoxic metabolite. However, in these cases, the killing of bystander cells is limited and observed only when cells are in close contact with one another, indicating diffusion of toxic metabolites solely through gap-junctional intercellular communications as the underlying mechanism of adjacent cell killing ( 19 ). Protumor and antitumor activity of nitric oxide (NO) or modulation of endogenous NO homeostasis. The pathway created by Xie et al. ( 1 ) by NO synthase 2 (Nos2) gene transfer results in the overproduction of NO and tumorcell toxicity, eventually leading to tumor toxicity (shown in bold). NOS-II = product of Nos2 gene; MeArg = NG -methyl- l -arginine, an NOS inhibitor; sGC = soluble guanylyl cyclase, an NO signaling target enzyme; O 2− = superoxide anion radical; ONOO − = peroxynitrite, a toxification product of NO and O 2− ; and cGMP = cyclic guanosine monophosphate. Introducing NOS activity into tumor cells, as demonstrate dearlier by Xie et al. ( 20 ) and others ( 5 ), causes cell death by anautocrine mechanism that is independent of host defenses (as in the case of cytokines) or of additional chemotherapeutic drugs (as in the case of drug-metabolizing enzymes). Although NO has been shown to be involved in the antitumor activity of cytotoxic effector cells, and the technical feasibility of introducing NOS-II activity into tumor cells has been demonstrated, this approach will become relevant to tumor therapy only if NOS-transfected tumor cells affect bystander cells. This conceptual advance to previous work on enzyme gene transfer has now been made by Xie et al. ( 1 ) by their demonstration that the Nos2-dependent autocrine suicide mechanism acts concomitantly on transfected cells and bystander cells, despite the extremely short half-life of NO. Apparently, NO, or more likely ONOO − , reach sufficient levels to exert toxicity on bystander cells also. This important observation is of potential therapeutic relevance. What comes next? One obvious aim now is to reach the 100% effectiveness level in killing both transformed tumor cells and bystander cells, which is necessary for tumor eradication. while this represents a considerable technical challenge, Xie et al. have shown that efforts toward overcoming such obstacles may well be worthwhile.