Targeting tumor hypoxia with the epigenetic anticancer agent, RRx-001: a superagonist of nitric oxide generation
Abstract This study reveals a novel interaction between deoxyhemoglobin, nitrite and the non-toxic compound, RRx-001, to generate supraphysiologic levels of nitric oxide (NO) in blood. We characterize the nitrite reductase activity of deoxyhemoglobin, which in the presence of bound RRx-001 reduces nitrite at a much faster rate, leading to markedly increased NO generation. These data expand on the paradigm that hemoglobin generates NO via nitrite reduction during hypoxia and ischemia when nitric oxide synthase (NOS) function is limited. Here, we demonstrate that RRx-001 greatly enhances NO generation from nitrite reduction. RRx-001 is thus the first example of a functional superagonist for nitrite reductase. We hypothesize that physiologically this reaction releases the potentially cytotoxic effector NO selectively in hypoxic tumor regions. It may be that a binary NO–H2O2 trigger is indirectly responsible for the observed tumoricidal activity of RRx-001 since NO is known to inhibit mitochondrial respiration.
Introduction
With apologies to Cole Porter and Ella Fitzgerald, birds, bees and fleas are not the only ones who do it. Bacteria, arthropods, molluscs, annelids and humans, to name a few other species, also do it—that is, they reduce the anion nitrite (NO2-) to nitric oxide (NO) via the nitrite reductase activity of deoxyhemoglobin (deoxyHb) [1]. Unlike nitric oxide synthases (NOSs), which are dependent on the presence of oxygen, the non-enzymatic hemoglobin nitrite reductase reduction to NO is hypoxia mediated [2]. Although oxyhemoglobin (oxyHb) rapidly and irreversibly reacts with NO, thus limiting NO bioavailability, hypoxic hemoglobin nitrite reductase provides a mechanism by which NO is generated when NOS becomes oxygen limited [3].Nitric oxide is a universal endogenous free radical that has been anthropomorphically described as a ‘‘two-faced’’ molecule with dual characteristics of ‘‘friend’’ and ‘‘foe,’’ suppressor and promoter, in cancer [4]. This biphasic, Dr. Jekyll and Mr. Hyde character is associated with local concentrations of NO in the tumor microenvironment and the overall redox state in the cell [5]. At low levels of exposure, NO is a ‘‘foe,’’ a fertilizer of the ‘‘seed and soil,’’ promoting tumor growth, angiogenesis and metastasis through second messenger-induced activations on effectors such as cyclic guanosine monophosphate (cGMP) and vascular endothelial growth factor (VEGF) to transiently alter their activity [6]. At higher levels of exposure, a threshold is reached and nitrosative stress predominates over nitrosative signaling, potentially resulting in a pro- foundly cytotoxic and pro-apoptotic effect on tumor cells that would limit angiogenesis and malignant progression [5]. Among the multiple other contradictory functions of NO are both radiosensitization and radioprotection of tumor cells [7, 8].
The term nitrosative stress refers to the formation of a toxic cocktail of highly reactive nitrogen species (RNS), such as peroxynitrite (ONOO-), nitrogen dioxide (NO2) and dinitrogen trioxide (N2O3) through reaction of NO with oxygen and superoxide anion, respectively [9]. Therefore, the double-edged sword indeed cuts both ways: NO is a tumor promoter and also potent therapeutic weapon, the activity of which is very context specific, depending on the local concentration of NO, the envi- ronment under which it is released and the tissue that it comes into contact with [6].RRx-001 [10] is a multifunctional C-geminal dinitro compound under investigation in several Phase II anti- cancer clinical trials as both a nitro-oxidative stress and epigenetic-mediated [11, 12] sensitizer of drug resistant tumors to standard chemotherapy, immunotherapy and radiation. In a Phase 1 clinical trial, RRx-001 was very well tolerated without clinically significant toxic effects at the doses investigated. In addition, the reversal of acquired chemoresistance after exposure to RRx-001, that is, sen- sitization to previously effective but now failed therapies, was observed [13, 14]. RRx-001-induced oxidative and nitrosative stress occurs in the absence of DNA and RNA toxicity through selective thiol alkylation of free glu- tathione and hemoglobin [15], resulting in enhanced NO production under hypoxic conditions. As reported previ- ously [15, 16], both in vitro and in vivo studies on the metabolism and disposition of RRx-001 showed that, on infusion, RRx-001 binds rapidly, selectively and irreversibly to the beta-Cys 93 residue of hemoglobin and to reduced glutathione (GSH). Binding is so rapid that the glutathione adduct of RRx-001 was chosen as the bioana- lyte for preclinical and clinical pharmacokinetic studies [14, 17].
The binding target of RRx-001 on hemoglobin, beta- Cys-93, is highly conserved in mammalian species; this binding is responsible, in part, for controlling the oxygen affinity of hemoglobin and nitric oxide transport [18, 19]. Thus, RRx-001 allosterically modifies and activates the catalytic domain of the nitrite reductase activity of deox- yHb [16]. Given the rapidity of its reaction with whole blood, only a subpopulation of red blood cells is modified. The size of this population and grade of modification will depend on the concentration of the compound as well as the hematocrit. In this subpopulation, the modification dra- matically enhances the efficiency of the enzyme and con- fers superagonist properties in the presence of the endogenous inorganic anion, nitrite. In addition, because oxyHb is also stabilized by this selective covalent adduc- tion [16, 17], the oxygen saturation curve is left shifted, which reduces tissue oxygen delivery, theoretically resulting in increased tumor hypoxia compared with nor- mal tissue and therefore enhanced nitrite reductase-medi- ated NO release inside the tumor [8].RRx-001-induced NO generation is hypothesized to occur in two ways: (1) through RRx-001-mediated dis- placement of endogenous NO bound to the b-cysteine residue on the red blood cell (RBC) immediately on infu- sion and (2) through catalysis of deoxyHb nitrite reductase activity.
The principle of combining a hypoxically activated NO release with a mechanism to increase tumor hypoxia rep- resents a promising therapeutic strategy for overcoming tumor chemoresistance and radioresistance, often associ- ated with upregulation of the glutathione redox system, through the in situ generation of NO and associated reac- tive nitrogen species (RNS). Indeed, RRx-001 has also exhibited dual radiosensitizing and radioprotection prop- erties in tumor cells and normal tissue, respectively [10]. The hypoxia-dependent induction of nitrosative stress is preceded by a burst of ROS due to the depletion of plasma glutathione (GSH) from the reactivity of RRx-001 with free thiol nucleophiles [10]; in addition, NO is known to bind to cytochrome C oxidase, the terminal electron acceptor in the mitochondrial electron transport chain, and inhibit mitochondrial respiration, which leads to increased oxygen availability and ROS production [20]. The vicinal increase in ROS and RNS together can lead to a witches’brew of reactive intermediates capable of damaging the tumor, increasing its susceptibility to radiotherapy and electrophilic chemotherapeutic agents [21] (Fig. 1) while also inducing an immune response directed at tumor tissue. In this paper, we study the effects of RRx-001 on whole blood and focus on RBC integrity in particular. Further- more, we characterize the nitrite–deoxyhemoglobin reac- tion and demonstrate that RRx-001 is a nitrite reductase- stimulating superagonist which binds to the b-Cys 93 residue on deoxyHb and also shifts the oxygen– hemoglobin affinity curve to the left.
RRx-001 used in these studies was obtained from Epi- centRx, Mountain View, CA. RRx-001 is a cyclic nitro compound with a chemical structure of C5H6BrN3O5 and a molecular weight of 268.02 (Fig. 6). The synthesis and characterization of RRx-001 are reported in detail else- where [22].Complete blood count (CBC) for whole-blood analysis, samples were characterized by complete blood count (Advia 120 hematology analyzer, Siemens, Munich, Germany).Oxygen equilibrium curves (OEC) for RBCs were mea- sured as described elsewhere by re-oxygenation of N2- equilibrated samples in the Hemox buffer at 37 °C, using a Hemox Analyzer (TCS Scientific Corporation, New Hope, PA) [23]. In brief, to determine the oxygen equilibrium curve of RBC treated with different concentrations of RRx- 001, RBC was washed and re-suspended in phosphate- buffered saline to a hematocrit value of approximately 40 % (hemoglobin concentration *12 g/dL). The suspen- sions were incubated with 0 (control), 1 % DMSO, 0.1, 0.3, 1.0 or 3.0 mM RRx-001 (all containing 1 % DMSO final concentration), at 37 °C for 1 h. Of each RBC sample, 50 lL was added to 4 mL of Hemox buffer, containing 0.1 % bovine serum albumin (BSA, Sigma-Aldrich, St. Louis, MO, USA) and 0.05 % antifoaming agent (pH 7.4). Finally, all runs were subjected to OEC software analysis.
RBC deformability was monitored by ektacytometry using a Technicon ektacytometer (Bayer Diagnostics, Tarrytown, NY, USA) as a function of osmolality at a constant applied shear stress of 170 dynes/cm2, as described previously [24]. Osmotic fragility RBC samples were diluted 250 times using a range of NaCl concentrations (0.1–0.9 %) and incubated for 30 min at room temperature. After incuba- tion, cells were centrifuged for 5 min at 10009g. To cal- culate the amount of free hemoglobin resulting from lysis of the cells, the supernatant was measured at a wavelength of 540 nm on a Spectramax 340PC microplate reader (Molecular Devices, Sunnyvale, CA, USA).Whole blood of normal volunteers (n = 4) was collected in EDTA/K3 tubes and incubated at 37 °C for 1 h on a rocker untreated and with 1 % DMSO, 0.3 mM and 3.0 mM RRx- 001 (both in 1 % DMSO final concentration). Directly after incubation, samples were spun down, plasma and buffy coat were removed, and RBC pellet was washed 29 in HEPES-buffered saline (HBS) followed by snap freezing in liquid nitrogen. Samples were stored at -80 °C till further processed for detection of a panel of amino acids and other cellular components on high-performance liquid chro- matography (HPLC)-linked tandem mass spectrometry using a method described previously by Morris et al. [25].
Sample RBC preparation For RBC samples, fresh whole blood from two different healthy subjects (coauthors San- dra Larkin and Marcel Fens) was collected in anticoagu- lated EDTA/K3 tubes. Whole-blood samples were characterized by a complete blood count (Advia 120 hematology analyzer, Siemens, Munich, Germany), and RBC was collected by centrifugation, and plasma and buffy coat were removed by twice washing with PBS at pH 7.4. Packed RBC was diluted with HBS containing 2 % BSA, to 10 g/dL hemoglobin.A 300 mM in DMSO stock solution of RRx-001 (Epi- centRx, Inc. Mountain View CA) was prepared freshly. To obtain the indicated final concentration of RRx-001, the stock was further diluted in HBS or directly in blood samples.The method to study NO2- to NO reduction by RBC samples was described in detail previously [26]. In brief, a single gas or a mixture of gases with defined composition is led through a tonometer (Instrumentation Laboratory, Bedford, MA) at a set flow rate of 180 mL/min. Inside the tonometer, a glass cuvette, containing 500 lL (RRx-001 treated) blood sample, is heated to 37 °C and set for a continuous set of spinning cycles to promote rapid equili- bration at the liquid–gas interphase. Nitrite incubations were initiated after 5-min initial incubation of the RBC mixture in the tonometer, to equilibrate the samples to temperature and gas conditions. Before addition of nitrite (5 mM), gas was collected for background NO values. Next, 100 lL of a sodium nitrite solution (Sigma-Aldrich) was added to the blood sample by pipetting directly into the cuvette using an elongated pipet tip. This allowed addition without changing the gas state in the tonometer. Gas leaves the tonometer solely through an opening in the lid and is collected in mylar balloons at predetermined time points (generally after 0, 10, 20, 40, 60, 80 and 100 min) for 2 consecutive minutes. NO concentration present in the mylar balloons is measured in a Nitric Oxide Analyzer (NOA 280, GE Water & Process Technologies, Trevose, PA, USA) by detecting chemiluminescence resulting from the reaction of ozone with NO. The NO in the mylar bal- loons was determined similar way to the analysis of NO in exhaled air by the Sievers NOA. All NO peaks were recorded by the NOAnalysisTM software, allowing for peak quantification. Finally, NO values collected were used to calculate the NO release rate and to calculate the rate and sum of total moles NO released per mole hemoglobin during the run.
For the nitrite influx studies, 8 lL was collected from the mixture immediately after adding nitrite. The same volume was sampled at 5, 10, 20, 40 and 100 min during the run.Sample collection was performed without changing the gas state in the tonometer. The samples were spun down immediately upon collection, and the supernatant and RBC pellet, after being washed with HBS, were stored at -20 °C until nitrite was measured. The nitrite concentra- tion was measured, at 550 nm on a Spectramax 340, by the colorimetric Thomson-Griess assay (Sigma-Aldrich) [27]. Curve fit and regression were performed using Kalei- daGraph (Synergy Software, Reading, PA, USA). Data were statistically analyzed by nonparametric two-way ANOVA with Bonferroni’s posttest using GraphPad Prism 6 for Mac OS X (GraphPad Software, San Diego, CA, USA). The institutional review board (IRB) at Children’s Hospital & Research Center Oakland (CHRCO) approved all experimental protocols. All procedures followed were in accordance with the ethical standards of the responsible committee on human experimentation (institutional and national) and with the Helsinki Declaration of 1975, as revised in 2008. Informed consent was obtained from patients included in the study.
Results
Hematological parameters (CBC) were determined after incubation at 0.1, 0.3, 1 and 3 mM RRx-001 (in 1 % DMSO final concentration) for 1 h at 37 °C. No differ- ences were seen for RBC, platelet or white blood cell (WBC) parameters for any of the treated samples, indi- cating that change in detected cell features and cell loss was minimal (see Table 1).In in vitro radiolabelled metabolism studies, RRx-001 was discovered to bind irreversibly to hemoglobin [16]. The objective of this experiment was to determine whether in binding to hemoglobin, RRx-001 acted as an allosteric effector, shifting the hemoglobin–oxygen dissociation curve using a Hemox analyzer. Fresh whole blood with K3/ EDTA as an anticoagulant was centrifuged to sediment the RBC. Plasma components and buffy coat were removed by washing with PBS. Hematocrit was adjusted to 40 % to mimic the hematocrit of whole blood. RRx-001 solutions at concentrations of 0.1, 0.3, 1.0 and 3.0 mM were investi- gated. At pharmacologically and toxicologically relevant doses of 0.1 mM and 0.3 mM, RRx-001 had no effect on the hemoglobin–oxygen dissociation curve of normal blood (Fig. 2a). At a suprapharmacologic concentration of 3 mM, RRx-001 significantly shifted the hemoglobin–oxygen dissociation curve toward a lower value of oxygen partial pressure, as represented by p50, a conventional measure of hemoglobin affinity for oxygen. The measured p50 values are 28.6, 29.6, 28.6, 28.2, 27.9 and 23.1 for untreated control RBC, 1 % DMSO-treated RBC and 0.1, 0.3, 1.0 and 3.0 mM RRx-001-treated RBC, respectively.Deformability of RBCs after treatment with RRx-001 was measured using ektacytometry as described previously [24]. Deformability of RBCs treated with any of the RRx- 001 concentrations was found to be unchanged. Similarly, the osmotic resistance or fragility of treated RBC was also found unchanged (Fig. 2b, c). Loss of phospholipid asymmetry and the exposure of phosphatidylserine on the surface of the red cell lead to a process referred to as eryptosis [28], similar to apoptosis in nucleated cells. Both induction of phospholipid scrambling and inhibition of the aminophospholipid translocase [29] may lead to the exposure of PS. Treatment of RBC with up to 3 mM RRx- 001 did not lead to a significant increase in the PS exposing subpopulation as measured using di-annexin [30] (data not shown).
Changes in a panel of amino acids and other cellular components involved in the cells’ metabolism were detected using a HPLC-linked tandem mass spectrometry method. Alanine, arginine, asparagine, aspartate, beta-ala- nine, cysteine, cysteinylglycine, glutamate, glutamine, glutathione, glycine, histidine, leucine, lysine, methionine, methylhistidine, phenylalanine, proline, serine, spermidine, threonine, tryptophan, tyrosine and valine were detected, and the majority did not show marked changes upon RRx- 001 (0.3 and 3.0 mM) incubation. The molecules that did show marked differences (only at the highest RRx-001 concentration) are depicted in Fig. 3: glutathione, cys- teinylglycine and glutamine showed a significant reduction in RBCs treated with 3.0 mM RRx-001 (Fig. 3a–c). On the RRx-001, green line; 0.3 mM RRx-001, yellow line; 1.0 mM RRx- 001, purple line; 3.0 mM RRx-001, red line. RBC samples were pre- incubated with RRx-001 for 1 h at 37 °C. A representative graph is shown. c Osmotic fragility is not changed upon incubation with RRx- 001 other hand, phenylalanine, glutamate and leucine showed a marked increase for the 3.0 mM RRx-001-treated RBC (Fig. 3d–f) depicts the actual NO signal detected in part per billion (ppb), the NO release rate and the cumulative amount of NO formed by the reduction of nitrite by RRx- 001 (3.0 mM)-treated and untreated RBCs (Fig. 4a, b) and hemolysate (Fig. 4c, d) samples over a 100-min time per- iod. NO is generated from reduction of NO2 by deoxy- genated RBCs but only under hypoxic conditions [26]. Upon addition of 5.0 mM nitrite, both RBC samples and hemolysates, which were pre-treated with 3.0 mM RRx- 001, showed a significantly increased NO release rate as well as significant increase in total NO release when compared with the untreated and 1 % DMSO-treated blood samples. After an initial lag phase, the NO release rate became steady which leads to a linear increase in total NO formation. Interestingly, the blood samples pre-treated with RRx-001 showed an accelerated NO release compared with the controls. Regarding the nitrite influx kinetics, no dif- ferences between groups were detected (Fig. 4c).
Discussion
In this study, we demonstrate that RRx-001 induces a significant potentiation of hemoglobin nitrite reductase activity leading to the generation of NO under hypoxic conditions. RRx-001 also left shifts the oxygen affinity of hemoglobin and changes RBC metabolism, by decreasing samples treated with the highest RRx-001 concentration (3 mM) showed a significant decrease in glutathione, cysteinylglycine and glutamine (a–c) and a marked increase in phenylalanine, glutamate and leucine (d–f) glutathione, cysteinylglycine and glutamine while increasing phenylalanine, leucine and glutamate. These modifications occur without significant changes in whole- blood toxicity markers (platelet, white blood cell and RBC parameters) or deformability and osmotic fragility and do not lead to appreciable exposure of PS, an early marker for apoptotic–like processes in the red cell.Previous studies [15–17] using radiolabeled disposition analysis, metabolite identification and proteomic analysis have demonstrated that RRx-001 binds rapidly, selectively and irreversibly to deoxyhemoglobin at b-Cys 93. Here, we show that RRx-001 increases hemoglobin–oxygen affinity and potentiates the nitrite reductase of deoxyhemoglobin, and together with the previously reported data, we hypothesize that this effect is due to binding at the beta- Cys-93 residue. The b-Cys 93 residue is a ubiquitously conserved residue that forms mixed disulfides with glu- tathione and plays a pivotal role in the activities of hemoglobin, GSH and NO [31–33]. Binding to this residue has previously been shown to perturb local hemoglobin structure, alter hemoglobin solubility, increase hemoglo- bin–oxygen affinity [34] and enhance nitrite reductase activity [2, 35]. The effect of RRx-001 on hemoglobin– oxygen affinity is concentration dependent.
Because of the rapid binding kinetics of RRx-001 on infusion only a small proportion or ‘‘quorum’’ of circulating RBC are modified while the bulk of the cell population is unaffected. With the relatively long-circulating half-life of red blood cells, this quorum of adducted RBC increases with each weekly administration of RRx-001 and, presumably, is converted into long-circulating bioreactors for the constant delivery of NO in vivo to the hypoxic regions of the tumor. Indeed, the rapid action of RRx-001 hijacks a subpopulation of RBCs to deliver oxidative and nitrosative stress to the tumor. This hypothesis is graphically presented in Fig. 5. The binding of RRx-001 to the b-Cys 93 residue, which could be referred to as the ‘‘NO switch,’’ regulates hemo- globin nitrite reductase activity by affecting two factors, the hemoglobin–oxygen affinity and the hemoglobin redox potential. These effects are augmented by the release of nitrite metabolites from RRx-001 and occur on the indi- vidual cell level at clinical doses of RRx-001. The results reported here are measured as bulk whole blood, rather than at the individual red cell level. Therefore, higher doses of RRx-001, the chemical structure of which is shown below, are required to demonstrate the effects on the entire population (Fig. 6).Based on the results of multiple experiments, we pro- pose a hypothesis for RRx-001 binding and downstream RBC effects: On infusion, RRx-001 rapidly permeates into RBCs and binds to the beta-Cys-93 residue on hemoglobin and to glutathione [16] leading to an immediate depletion of GSH and increased oxidative and nitrosative stress, manifested by higher hydrogen peroxide and superoxide levels in the RBCs. RRx-001 binding to the beta-Cys-93 residue results in a left shift of the hemoglobin–oxygen Nitrite-induced NO release and cellular nitrite influx under hypoxia.
Blood samples containing approximately 12 g/dL hemoglo- bin were pre-incubated with 3 mM RRx-001 or DMSO 1 % for 30 min before sample was transferred to a tonometer in order to create a hypoxic environment, to allow addition of nitrite (5 mM) and to detect released NO. The NO released from the tonometer collected in mylar balloons is detected by a Nitric Oxide Analyzer [recorded as parts per billion (ppb)]. Next, NO concentration is converted to moles per minute and corrected for the hemoglobin concentration present in the tonometer to obtain the NO release rate. Finally, the total amount of NO released from the tonometer, during an entire run, could be calculated per molecule hemoglobin [19]. Both RBC (a, b) and hemolysate samples (c, d) show superior NO release when pre-treated with 3 mM RRx-001 when compared with untreated and 1 % DMSO- treated controls. Nitrite influx into RBC kinetics of RBC samples with or without pre-treatment did not show any differences between groups(e). Experiments were conducted at 0 % oxygen affinity curve, i.e., increased oxygen affinity, as reported here; the potentiation of hemoglobin nitrite reductase activity under hypoxic conditions further augments the oxidative and nitrosative stress on the RRx-001 RBCs and, by inference, their immediate environment. These modifi- cations of hemoglobin and related RBC-mediated effects likely persist for duration of lifetime of the RBC [16], delivering nitrogen species in hypoxic areas. In contrast to normal tissue, tumor cells have highly deregulated antioxidant mechanisms and relatively small increases in oxidative stress can lead to senescence and apoptosis [36]. In addition, having recently reported on the ROS-mediated epigenetic activity of RRx-001 [11], key enzymes that control gene transcription such as DNA demethylases and histone deacetylases with active-site cysteine residues may be oxidized and inhibited [37]. It is currently unknown whether changes in the metabolomics profile observed upon addition of RRx-001 are maintained for the duration of the RBC in the circulation. Glycolysis-fueled metabo- lism in the red cell may lead to a reversal of the observed changes over time.
Due to the persistent pro-oxidative milieu of cancer cells, compounds that specifically and selectively alter the redox status of the tumor have the potential to induce cell damage and death [38]. Accordingly, RRx-001’s antipro- liferative and apoptogenic effects on cancer cells are likely mediated by prooxidant mechanisms; the bromomethyl pharmacophore in RRx-001 imparts a unique chemical reactivity that specifically and selectively involves covalent adduction to GSH and the b-cysteine 93 residue in deox- yHb [39]. Moreover, the surplus NO may inhibit mito- chondrial cytochrome c, diverting oxygen to non- respiratory substrates and increasing ROS generation [20]. In this way, RRx-001 can lead to the colocalized produc- tion of both NO and superoxide anion (O -) resulting in the generation of the highly reactive and tumoricidal, peroxynitrite [40–42]. We hypothesize that ONOO-, a product of the dilution-limited reaction of NO and O2-, results in tumoricidal activity and subsequent oxidation of cysteine-dependent epigenetic enzymes through this syn- ergy of ROS/RNS [43, 44].
Conclusion
Since nitrates were abundant in early earth, while oxygen was not, respiration of nitrogen oxides may have originally preceded oxygen respiration [45] and therefore partial denitrification [46], or the reduction of nitrate (NO3-) or nitrite (NO2-) to NO and dinitrogen oxide (N2O) would have assumed primordial importance. A relic of these ancestral properties, a nitrite reductase functionality, has been conserved in humans to produce nitric oxide under hypoxic or anoxic conditions. Under physiologic condi- tions, the enzyme appears to operate at low levels; how- ever, RRx-001 has unmasked a latent catalytic mechanism based on high-affinity binding to a conserved b-Cys 93 residue in deoxyHb. As a result of this binding, RBCs become not only vehicles that carry O2 to the tissues but also direct effectors of the hypoxic conversion of nitrite into cytotoxic NO in tumors. Activation of nitrite reductase by RRx-001 therefore represents a promising novel thera- peutic target in cancer.