Janus-Faced Tumor Microenvironment and Redox

Abstract

Significance: Tumor microenvironment (TME) is a complex term that includes extracellular matrix, blood vessels, endothelial, stromal, and inflammatory cells, and other supporting structures of the particular organ; and physiological components such as oxygen, pH, nutrients, waste products, signaling molecules, reducing/oxidizing species, growth factors, protumorigenic factors, etc. TME is now widely recognized as a major contributor to cancer aggression and treatment resistance and as a potential target for therapeutic intervention. Recent Advances: Among important physiological parameters of the TME, tissue hypoxia is considered to be a consequence of imbalanced angiogenesis and is associated with changes in metabolic pathways, including a higher dependence on glycolysis resulting in tissue acidosis. Both hypoxia and acidosis affect the tissue redox status and its key intracellular component, glutathione (GSH). Numerous publications support that these local TME conditions select for outgrowth of cells with appropriate phenotypes, which can reflect underlying genetics. Critical Issues: Here, we hypothesize that specific patterns of local TME, namely, tumor oxygenation, extracellular pH, redox, and GSH homeostasis, acting in orchestrated mechanism, can promote cancer cell survival, while at the same time being highly toxic and mutagenic for normal cells, thus contributing to the growth of cancers at the expense of the normal tissues they are invading. This review summarizes the experimental observations that support the hypothesized Janus-faced character of the redox axis. Future Directions: Normalizing the TME redox parameters may decrease the selection pressure for malignant phenotypes, therefore providing a tool for TME-targeted anticancer therapy. Antioxid. Redox Signal. 21, 723–729.

Introduction

In recent years, cancer research has experienced a paradigm shift from a seemingly obvious target, tumor cells, toward a key support system of cancer, tumor microenvironment (TME) (64). Carcinogenesis is a phenomenon that occurs in tissues, not in individual cancer cells. Malignant cells generated by causal genetic insults behave differently depending on their specific microenvironment. This explains why numerous therapeutic effects and treatment strategies elaborated in vitro are not reproduced in further studies in animals and humans. In general, TME is a complex term that includes two interrelated and interacting principal components: the physical TME (extracellular matrix, blood vessels, endothelial, stromal, and inflammatory cells, and other supporting structures of the particular organ); and physiological TME (oxygen, pH, nutrients, waste products, signaling molecules, reducing/oxidizing species, growth factors, protumorigenic factors, etc.). TME is now widely recognized as an integral, essential part of cancer, a major contributor to cancer aggression and treatment resistance and as a potential target for therapeutic intervention. Among important physiological parameters of the TME, tissue hypoxia (66) is considered to be a consequence of imbalanced angiogenesis and is associated with changes in metabolic pathways, including a higher dependence on glycolysis (70) resulting in tissue acidosis (5, 32). Both hypoxia and acidosis affect the tissue redox status (51) and its key intracellular component, glutathione (GSH) (20, 69). In this study, we hypothesize that specific patterns of tumor redox and GSH homeostasis can promote cancer cell survival, while at the same time being highly toxic and mutagenic for nontransformed cells. This review summarizes the experimental observations that support the hypothesized Janus-faced character of the redox axis of the TME.

TME Redox Is Characterized by Low Redox Potential and High Reducing Capacity

The important role of tissue redox microenvironment and redox signaling in physiology and pathophysiology is widely speculated, but not widely accepted. This discrepancy is due to the difficulty in deriving a quantitative description of the redox microenvironment, which requires accounting for the simultaneous status of the numerous redox couples. Among other redox couples, the glutathione disulfide–glutathione couple (GSSG/2GSH) is of particular importance due to the very high millimolar concentration range of the intracellular GSH, which is considered as an intracellular redox buffer (63). Therefore, the redox state of the GSSG/2GSH is accepted as an important indicator of the intracellular redox environment.

Figure 1 illustrates the relationship between the GSH redox potential and cellular state. The most negative GSH redox potential (from −260 to −220 mV) is characteristic for proliferating cells, while apoptotic and necrotic cells characteristically have the highest GSH redox potentials (≥−170 mV) (63). In proliferating cancer cells, low GSH redox potentials have been measured in vitro (34, 63). Furthermore, a high reducing capacity (4, 42) and high intracellular GSH (4, 42, 62) contents have been reported from in vivo measurements in animal cancer models, as illustrated in Figure 2. A negative redox potential reflects TME redox thermodynamics, and indicates a shift of redox balance between oxidizing and reducing capacities in favor of the latter one. A relevant question is whether a highly reducing TME can be interpreted as deriving from decreased oxidative stress or low rates of reactive oxygen species (ROS) production?

FIG. 1.

Simplified scheme of the relationship between the GSH redox potential and biological cellular status. Redox potential, E_hc, for the glutathione disulfide–glutathione couple (GSSG/2GSH), is an important indicator of the intracellular redox environment that reflects cellular status. E_hc (GSSG/2GSH) has the most negative value for proliferating cells, including cancer cells, and is more positive for differentiated cells. A number of oxidative stress-related pathologies, for example, inflammation, are characterized by further increase in GSH redox potential and are often associated with GSH depletion. If E_hc (GSSG/2GSH) becomes too positive, then apoptotic redox signals are initiated, resulting in the removal of cells that have lost an ability to control their redox environment. Very high values of the E_hc leave only necrosis or necroptosis as a path to cell death. Adapted from Schafer and Buettner (63) with permission of Elsevier, Inc.

FIG. 2.

Tumor reducing capacity and GSH content. The reducing capacity of extracellular tissue microenvironment and concentration of intracellular GSH assessed in the normal mammary glands and mammary tumors of female FVB/N mice using in vivo electron paramagnetic resonance technique. (a) The reduction rate of the nitroxide redox probe in extracellular media of normal mammary glands and tumors. (b) Intracellular GSH concentrations measured using paramagnetic GSH-sensitive probes. Adapted from Bobko et al. (4) with permission.

Role of TME in ROS Generation

In cases wherein there is a high level of oxidative stress, for example, due to elevated levels of ROS generation at low pH or in inflammation, the tissue microenvironment is characterized by high redox potential and high oxidizing capacity. Whereas TME is characterized by low redox potential and high reducing capacity, elevated levels of intra- and extracellular ROS generation (2, 46, 49, 55) and products of lipid peroxidation (19, 43) have been detected in almost all cancers. This is not entirely surprising since the higher reducing capacity of TME may promote electron flux from reducing equivalents to the corresponding substrates, such as oxygen or iron, followed by a cascade of free radical reactions and oxidative damage in spite of the very negative redox potential. Superoxide generation is widely considered to represent the initial step in ROS formation followed by its dismutation to hydrogen peroxide and further aggravated by the possible formation of highly reactive hydroxyl radicals in the Fenton-type reactions, catalyzed by reduced metals. The oxidative phosphorylation chain of mitochondria (9) and membrane-associated NADPH-dependent oxidases (NOX) (2, 38) are two major sources of intracellular and extracellular superoxide generation, respectively. We suspect that additional mechanisms of extracellular superoxide generation are facilitated by specific features of TME.

Highly reducing extracellular TME (4) favors electron flux from reductants to dissolved oxygen resulting in superoxide production. Metal ions in their reduced state, for example, Fe²⁺ or Cu⁺, are well-known reducing agents for oxygen that catalyze superoxide production. In respect to this, iron transport proteins, transferrin and ferritin, as well as copper-binding ceruloplasmin, can act as sources of metal ions. Furthermore, tumor-associated extracellular acidosis (15, 48, 72) may potentiate dissociation of protein-bound metals [e.g., Fe³⁺ from transferrin (13, 58)], while highly reducing TME favors metal ion reduction and, therefore, ion redox cycling. Among the mechanisms of metal reduction could be the reduction of ferric iron, Fe³⁺, as well as other metal cations (e.g., Cu²⁺) by reactive thiol, cysteinyl-glycine (12, 17) whose extraordinary low pK_a=6.4 provides a high fraction of metal-binding thiolate anions. The synthesis of cysteinyl-glycine is catalyzed by plasma membrane-bound γ-glutamyltransferase (GGT) from extracellular GSH. The activity of GGT (12, 21, 27, 29, 30) as well as expression of GSH complex export transporters, GS-X pumps (1, 14, 45), are often significantly elevated in cancer cells, including human malignancies, supporting the potential importance of GGT/GSH-dependent mechanism in TME extracellular superoxide production (17).

Many redox reactions are pH dependent with redox potential being increased at lower pH (63). In regard to superoxide (E⁰’=940 mV), a decrease of extracellular pH (pH_e) by about 0.5 units, often observed in solid tumors, results in a threefold increase in its protonated form, hydroperoxyl radical (E⁰’=1060 mV) capable of oxidation of unsaturated lipids. An enhanced lipid peroxidation during tissue acidosis has been previously reported (6, 31, 47). Therefore, extracellular TME acidosis may potentiate enhanced oxidative stress and oxidizing damage (73).

Numerous studies of ischemia/reperfusion phenomena demonstrate that tissue ischemia induces tissue acidosis (41) and accumulation of reducing equivalents (lower redox potential) (74), followed by a burst of ROS production upon onset of reperfusion (26). Interestingly, the spatial distribution of oxygen in TME is also not constant, but varies over time due to periodic microregional blood flow fluctuations with periodicities ranging from 20 to 60 min (waves) (7) to large-scale circadian changes (tides). These temporal oscillations create local cyclic hypoxia events in TME oxygenation resembling an ischemia/reperfusion model, and therefore may facilitate similar bursts in ROS generation.

Janus-Faced Tumor Redox Microenvironment: Oxidative Stress in Cancer Cells Versus Normal Cells

An imbalance between antioxidants and oxidants in favor of oxidants, potentially leading to damage, is termed “oxidative stress” (65). The presence of random oxidative alterations of DNA, proteins, and lipids are well documented in cancers and cancer cells (18, 43, 46) supporting the presence of some level of oxidative stress. As first suggested by Oberley et al. (55), oxidative stimuli may result in signaling pathways leading to proliferation, while an overall more reducing environment might be the result of adaptation in response to an oxidative stress. Cancer cells express increased levels of antioxidant proteins (2, 46) to detoxify ROS suggesting that a delicate balance of intracellular ROS levels is required for cancer cell function.

High levels of GSH, a major intracellular redox buffer and antioxidant (63), have been found in various tumor types, being up to several-fold greater than that in surrounding tissues (4, 20, 33, 67, 69). The overexpression of GSH in tumors has been related to enhanced cell proliferation (39), decreased levels of apoptosis (28), increased resistance to chemotherapeutic drugs (40, 67), and radiation therapy (53). Moreover, high GSH content in tumor cells is typically associated with higher activities of GSH-dependent antioxidant enzymes, including glutathione peroxidase, glutathione transferase (GSH-Tr), and glutathione reductase (GSSG-Rx) (16). Therefore, elevated levels of GSH and antioxidant GSH-dependent enzymes play a central role in protection of cancer cells against intracellular oxidative stress. On the other hand, as discussed in the previous section, overexpression of GGT in cancer cells is related to the GGT/GSH-dependent mechanism of superoxide production in extracellular TME. An extracellular superoxide generation in TME is aggravated by high reducing capacity, low pH_e, and local cycling hypoxia. In the absence of high levels of GSH, the cells in the TME are subjected to large excursions in redox.

In summary, TME exposes both cancer and nontransformed cells to elevated ROS production. The corresponding oxidative stress is tightly controlled in cancer cells by overexpressed antioxidant defense at a level sufficient to provide a signaling pathway to proliferation, while avoiding a fatal oxidative damage. However, the same elevated level of extracellular ROS generation may be highly toxic and mutagenic for normal cells (50) due to significantly lower activities of antioxidant enzymes and GSH content. Adaptation of the cells in response to an oxidative stress may include an increase in intracellular GSH content and GSH redox potential contributing to signaling for proliferation and malignant transition. The proposed paradigm of Janus-faced TME is illustrated in Figure 3.

FIG. 3.

Janus-faced tumor redox microenvironment. The discussed key components of the TME, oxygen, pH, redox, and GSH, are shown in red. Hypoxia-induced acidosis potentiates accumulation of free metal ions such as Fe³⁺ and Cu²⁺. In its turn, a high reducing capacity of TME promotes metal ion reduction to the Fenton-active state, for example, via γ-glutamyltransferase (GGT)/GSH-dependent generation of reducing cysteinyl-glycine dipeptide, GlyCysS⁻. A cycling local hypoxia in TME facilitates further electron transfer to oxygen with formation of O₂^•− radical, triggering the radical reaction cascade of O₂^•− dismutation to H₂O₂ followed by OH-radical formation via the Fenton reaction. Low reactive oxygen species (ROS), H₂O₂, and O₂^•−, penetrate into tumor cells by diffusion or via anion channels, latter contributing to the increase of intracellular pH (pH_i) (8, 37) with a corresponding decrease in oxidizing potential of ROS. In contrary, low acidic extracellular pH (pH_e) in TME enhances ROS oxidizing potential toward surrounding cells that may result in oxidative damage and mutagenesis as well as adaptive response (e.g., increase of GSH) and changing cellular phenotypes. Increase in H₂O₂, O₂^•−, GSH, and pH has been shown contribute into the triggering cells in the proliferation stage (8). Compared with normal cells, cancer cells are well protected against oxidative stress, in part, by elevated GSH content and activities of GSH-dependent antioxidant enzymes, including glutathione peroxidases (GPx1 and GPx4 denote peroxidases that target hydrogen peroxide and lipid peroxide, correspondingly) and glutathione reductase (GSSG-Rx). To see this illustration in color, the reader is referred to the web version of this article at www.liebertpub.com/ars

Selection Pressure of Specific TME Conditions

Genetic heterogeneity of tumors is recognized as a major cause of therapy resistance. This heterogeneity arises from genome instability in combination with a highly selective TME (25). Local TME conditions select for outgrowth of cells with appropriate phenotypes, which can reflect underlying genetics (24). Numerous publications support that among these specific TME conditions, hypoxia and acidosis lead to genomic instability (3, 23, 24, 59). We believe that one of the mechanisms of the TME influence on mutagenesis may be related to the corresponding alterations of TME redox characterized by both high reducing capacity and enhanced extracellular ROS generation, as illustrated in Figure 3. It is evolutionarily reasonable that, since tumors evolved within hypoxic and acidic environments and related oxidizing pressure, they would recapitulate these as they grow, as it provides them with a selective advantage. Cancer cells are well protected against oxidizing pressure, in part, by elevated GSH content and activities of GSH-dependent antioxidant enzymes. This allows successful cancers to avoid severe oxidative damage and to maintain a delicate balance of intracellular ROS levels required for cancer cell functions such as activation of redox signaling pathways leading to proliferation. On the other hand, exposing normal cells to cyclic oxidizing pressure may result in DNA damage, leading to cell death or malignant transition.

Conclusion

A new hypothesis of Janus-faced TME is proposed. High reducing capacity, hypoxia, and acidosis of extracellular TME expose both cancer and nontransformed cells to elevated ROS production, while providing cancer cells with a selective advantage. Normalizing TME may decrease the selection pressure for malignant phenotypes, therefore providing a tool for TME-targeted anticancer therapy. Indeed, there are data supporting the beneficial effects of targeting selected TME parameters, including tumor hypoxia (22, 52, 54, 71), acidosis (4, 35, 36, 60, 61, 68) (see Fig. 4), redox (10, 57), and GSH content (11, 44, 56). A combination therapy aimed to normalize several of these interrelated TME parameters may provide further therapeutic advantages.

FIG. 4.

Effect of bicarbonate on tumor microenvironment pH_e and survival of tumor-bearing animals. Microscopic pH_e gradients in tumors inoculated into the window chamber were measured by fluorescence ratio imaging of SNARF-1 pH probe. Representative pH_e images are shown for untreated (A) and bicarbonate-treated (B) mice. Red lines, region of interest of tumor. (C) Distributions of pH_e along radial lines for control and bicarbonate-treated tumors. “0” is centroid of tumor, and vertical line indicates tumor edge. (D) Effect of bicarbonate treatment on survival of female SCID mice bearing MDA-MB-231 breast tumors. Mice were started on drinking water (ad libitum) supplemented with 200 mM NaHCO₃ at day 6 postinjection of MDA-MB-231 cells. Data are plotted as a Kaplan–Meier survival curve, the difference in the survival curve for the bicarbonate versus control animals was tested using the log-rank test (p=0.027). Adapted from Robey et al. (60) with permission. To see this illustration in color, the reader is referred to the web version of this article at www.liebertpub.com/ars

Abbreviations Used

GGT

γ-glutamyltransferase

GlyCysSH

cysteinyl-glycine dipeptide

GPx1

glutathione peroxidase 1

GPx4

glutathione peroxidase 4

GSH

glutathione

GSH-Tr

glutathione transferase

GSSG

glutathione disulfide

GSSG-Rx

glutathione reductase

NOX

NADPH-dependent oxidase

pH_e

extracellular pH

pH_i

intracellular pH

ROS

reactive oxygen species

TME

tumor microenvironment

Acknowledgments

This work was partly supported by NIH grants EB014542, CA077575-10, and U54 CA 143970-01.