Regulation of antioxidants in cancer

Summary

Scientists in this field often joke, “If you don’t have a mechanism, say it’s ROS.” Seemingly connected to every biological process ever described, ROS has numerous pleiotropic roles in physiology and disease. In some contexts, ROS act as secondary messengers, controlling a variety of signaling cascades. In other scenarios, they initiate damage to macromolecules. Finally, in their worst form, ROS is deadly to cells and surrounding tissues. A set of molecules with detoxifying abilities, termed antioxidants, are the direct counterpart to ROS. Notably, antioxidants exist in the public domain, touted as a “cure-all” for diseases. Research has disproved many of these claims and, in some cases, shown the opposite. Of all the diseases, cancer stands out in its paradoxical relationship with antioxidants. While the field has made numerous strides in understanding the roles of antioxidants in cancer, many questions remain.

eTORC blurb

The paradoxical role of antioxidants in cancer has been a matter of intense debate in recent years. Here, we highlight the latest breakthroughs that shed light on the molecular mechanisms by which antioxidants are regulated in cancer to impact their initiation, progression, and metastatic capabilities.

Introduction

Antioxidants can be broadly defined as “substances that delay, prevent, or remove oxidative damage to specific target molecules by reacting with oxidants,” such as reactive oxygen species (ROS)¹. Antioxidants and oxidants in equilibrium maintain redox homeostasis, and an imbalance in favor of oxidants can lead to oxidative stress, which has been linked to the pathogenesis of a myriad of diseases, including neurodegenerative and cardiovascular disorders, diabetes, and cancer². Cancer tissues are believed to produce higher levels of ROS due to their altered metabolism, inflammation, hypoxic environment, and oncogene-driven upregulation of ROS-generating enzymes³, thus, suggesting the need for antioxidants to buffer excessive ROS. Here, we will discuss pertinent developments and knowledge gaps in the regulation of antioxidants in cancer. We will only minimally discuss, however, topics surrounding ferroptosis^4–6 and ROS signaling^7,8, for which details can be found in several excellent reviews.

Rather than prevent, antioxidant supplementation promotes cancer

Unchecked oxidative stress can cause DNA damage, mutations, and oncogenesis. However, an oversimplistic interpretation of these facts has led to the prevalent misconception that ROS are intrinsically deleterious while antioxidants would have beneficial effects on almost every aspect of human health. This premise potentially contributed to the widespread consumption of complementary and alternative medicine products (CAM), vitamins, minerals, and antioxidant dietary supplements, often without the knowledge of their physicians⁹, to prevent cancer, improve prognosis, or alleviate side effects of radio- and chemotherapy. The reported prevalence of supplement use varies considerably depending on the cancer type and patients’ demographics but is frequently reported to be higher than 50%¹⁰, and is positively associated with age, higher education levels, and female sex^9–11. While at first some observational studies hinted at a possible benefit, large randomized controlled clinical trials conducted in the ‘90s and 2000s have not only consistently failed to demonstrate clinical benefits but also revealed that some antioxidants increased all-cause mortality^12–21. Considering new evidence, recommendations against the use of antioxidants were published²². This subject has been revisited recently, with a new recommendation statement published by the US Preventive Services Task Force recommending against the use of β-carotene or vitamin E supplements for the prevention of cancer or cardiovascular disease^23,24. Of note, the absence of positive outcomes in clinical studies could hypothetically be attributed to differences in the impact of the tested antioxidants when consumed as isolated chemicals versus when they are ingested through food, which contain intricate combinations of antioxidants, vitamins, and minerals. Thus, if antioxidants have any anti-tumor activity, we have not yet discovered when or how to dose, combine, and deliver these molecules to achieve clinical benefit. Certain antioxidants act as “pro-oxidants” at elevated levels (i.e., Vitamin C)^25,26. Additionally, elucidation of the key pathways involved in the overabundance of reducing capabilities (i.e., reductive stress)^27–30 will potentially assist with disentangling the contributions of antioxidant supplements. Separately, whether antioxidant supplements have systemic deleterious effects beyond cancer in humans, as seen with lower organism models³¹, remains to be answered.

Oxidative damage is believed to be an important part of the cytotoxicity of several anti-cancer agents as of radiotherapy³², raising the concern that antioxidants could potentially protect tumors by decreasing therapy efficacy, leading to higher recurrence and death rates. That could provide some mechanistic insight into why antioxidants have worsened prognostic when administered during therapy in some studies^33,34. However, it doesn’t help explain why dietary antioxidants seem to increase cancer incidence, as seen in primary prevention trials^14,20. Moving from “bedside-to-bench,” several cell line and animal studies have provided solid evidence that antioxidants increase tumor progression and reduce survival in mouse models of cancer through multiple mechanisms^35–37. Developing a better mechanistic understanding of antioxidant supplements in cancer is a key step toward dispelling the current narrative of antioxidants as a cure-all and potentially finding coherent treatment avenues with them.

Antioxidants promote tumor initiation and progression

Besides dietary antioxidants, endogenously produced antioxidants govern pivotal roles within cancer cell biology. Glutathione reigns as the most abundant non-protein thiol in mammals (Fig. 1)³⁸. This tripeptide is a key contributor to redox homeostasis as it is a substrate for antioxidant enzymes (e.g., glutathione peroxidases, glutathione-S-transferases, glyoxylases)³⁹. These roles for GSH do not include its impact in modulating protein activity through S-glutathionylation⁴⁰ and C-glutathionylation⁴¹. To a significantly lower extent, GSH spontaneously scavenges different ROS (e.g., hydroxyl radical (OH^•), hypochlorous acid (HOCl), alkoxyl radical (RO^•), nitric oxide radical (NO^•), amongst others)⁴². To the best of our understanding, the synthesis of GSH is controlled by a two-step reaction. First, glutamate and cysteine are combined by the catalytic subunit of glutamate-cysteine ligase (GCLC) to produce gammaglutamylcysteine. Mice lacking GCLC undergo early embryonic lethality⁴³, suggesting the need for GSH in developing tissues. This reaction is feedback inhibited by GSH; however, the modifying subunit of glutamate-cysteine ligase (GCLM) can significantly increase the threshold required for this inhibition⁴⁴, allowing for an accumulation of intracellular GSH. GCLM is not required for animal survival; however, its loss prevents tumor initiation and progression⁴⁵. A pharmacological inhibitor of GCLC, buthionine sulfoximine (BSO), can lower levels of GSH in vitro and in vivo⁴⁶, although its anti-tumor effect is typically only found when combined with other therapies^47–49. While BSO was first described nearly half a century ago⁵⁰, improved compounds against GCLC have yet to be developed. The second step in GSH synthesis is the condensation of gamma-glutamylcysteine and glycine by glutathione synthetase (GSS) to produce GSH. Interestingly, targeting GSS genetically or pharmacologically in tumor models has largely remained unexplored, potentially due to its essential requirement for embryogenesis⁵¹. An alternative approach to lower GSH production is limiting the availability of the amino acid precursors glutamate, glycine, and cysteine (discussed in more detail later). And while these approaches have demonstrated anti-tumor effects, the pleiotropic, GSH-independent endpoints of these amino acids can confound some interpretations.

Figure 1. Dynamic regulation of antioxidants in cancer.

NRF2 is constantly degraded by the proteasome through ubiquitination when bound to KEAP1 and CUL3. The KEAP1/CUL3 complex disengages from NRF2 when oxidized by electrophiles or ROS, leading to NRF2 stabilization and increased transcription of gene targets (shown in yellow). These include enzymes in GSH synthesis (GCLM, GCLC, GSS) and regeneration (GSR), GSH-dependent detoxification (GPX), thioredoxin (TXN), and thioredoxin-related enzymes (TXNRD, PRX). Additionally, NRF2 promotes the transcription of genes involved in quinone reduction (NQO1), iron homeostasis (HMOX1), and NADPH production (G6PD, PGD, ME1, IDH1). Other NRF2 target genes (i.e., ALDH3A1) can induce liabilities, such as reductive stress. Further, the compartmentalization of antioxidants and related metabolites across organelles (i.e., nucleus, mitochondria, ER) plays critical roles in maintaining redox homeostasis in the cell.

Nuclear factor erythroid 2-related factor 2 (NRF2), the master transcriptional regulator antioxidant enzymes (including GCLC, GCLM, and many others), can also support tumorigenesis⁵² (Fig. 1). NRF2 level is controlled by the KEAP1-CUL3 E3 ubiquitin ligase complex, which constantly tags NRF2 to proteasomal degradation. Under oxidative stress, reactive cysteines in KEAP1 are oxidized, impeding its interaction with NRF2, thus, allowing for NRF2 to accumulate and engage in an antioxidant response. Somatic mutations in NRF2, KEAP1, or CUL3 that prevent NRF2 degradation have been documented in several types of cancer with varying frequencies, reaching up to 25% of lung squamous cell carcinoma, 15% of lung adenocarcinoma and 23% of esophagogastric squamous cell carcinoma⁵³. In all cancers combined, it averages 1%. Human cancers with elevated levels of NRF2 are associated with poor prognosis^54,55. It has been suggested that, in animal models, Nrf2 stabilization still requires an oncogenic driver (e.g., Kras) or loss of a tumor suppressor (e.g., Pten) to induce tumorigenesis, as Nrf2 or Keap1 mutations alone do not suffice, even in the absence of tumor suppressors like TP53 or LKB1⁵⁶. Interestingly, somatic alterations in the NRF2 pathway frequently co-occur with alterations in the PI3K pathway⁵³, highlighting a synergistic effect between these two alterations that has also been observed in pre-clinical models^57–61.

Nrf2 also rewires glucose metabolism towards the pentose phosphate pathway (PPP)⁵⁷. This is achieved by transcriptionally upregulating several members of the PPP (e.g., G6PD, PGD), boosting NADPH production, and fueling purine synthesis, a major mechanism through which Nrf2 enables cell proliferation⁵⁷ (Fig. 1). Interestingly, G6PD expression was not required to grow certain cancers driven by oncogenes (i.e., K-Ras^G12D)⁶². However, if G6PD expression is disrupted in cancers with the same oncogenes but also activated Nrf2, cells engage in alternative pathways to produce NADPH (e.g., IDH1, ME1) and drain intermediates from the TCA cycle, which can negatively impact tumor growth⁶³. Overexpression of G6PD in immortalized human or murine cells, resulting in an augmented synthesis of NADPH and nucleosides, is sufficient to make them tumorigenic when xenografted into mice⁶⁴. Remarkably, the sole addition of exogenous nucleosides or N-acetylcysteine can trigger the same tumorigenic effect. This relationship, however, is continually evolving. Recently, the inactivation of GADPH by oxidative stress was discovered to de-repress G6PD, facilitating flux through the PPP and NAPDH production⁶⁵. Importantly, several other pathways support NADPH generation and are linked to tumor growth and survival^66–69. Beyond Nrf2, further studies are required to deconvolute the impact of metabolic pathways on antioxidant abundance and tumors.

Nrf2’s antioxidant transcriptional program includes genes involved in GSH and thioredoxin (TXN) production/regeneration/utilization, labile iron handling, quinone reduction, and NADPH production; together, these programs can promote resistance to anti-cancer therapies, such as receptor tyrosine kinase inhibitors^70,71. However, Nrf2’s antioxidant programs can deplete intracellular metabolites, creating targetable metabolic vulnerabilities^72–74. One of these vulnerabilities, a reliance on de novo glutamate production through glutaminolysis, has been recently explored in a clinical trial (KEAPSAKE, NCT04265534) which tested the effect of the GLS inhibitor CB839 in KEAP1/NRF2-mutant non-small cell lung carcinoma (NSCLC). While this trial and others focused on NRF2-specific vulnerabilities have not advanced in the clinic, additional trials (i.e., BeGIN NCT03872427) are ongoing. Recently, more unbiased approaches (i.e., chemical and genetics screens) have been employed to identify new vulnerabilities in NRF2-driven cancers. It was shown that the survival of Keap1-mutated tumors depends on Slc33a1, a gene that encodes an ER-resident acetyl-CoA transporter involved in the unfolded protein response (UPR)⁷⁵ (Fig. 1). Another screening campaign identified that a subset of NSCLC cell lines reduces proliferation upon Keap1 disruption due to Nrf2-mediated upregulation of the enzyme ALDH3A1, which overproduces NADH causing reductive stress⁷⁶. Finally, Keap1 loss was identified to sensitize cancer cells to inhibition of the DNA repair protein ATM⁷⁷. The vulnerabilities unveiled by these studies represent promising targets that warrant further research.

Antioxidants as metastasis-promoting agents

Loss of attachment to the extracellular matrix is a barrier to metastasis faced by epithelial cancer cells⁷⁸. It was shown that metastatic cells experience elevated oxidative stress compared to primary tumors and cope with this stress by relying on NADPH production^79–82. The folate pathway contributes to NADPH production, which potentially helps explain the role of folate in supporting metastasis and may help explain the detrimental effect of folate supplementation on cancer incidence and mortality⁸³. Aside from NADPH production, it was shown that lymph nodes act as a safe haven from oxidative stress for metastatic cancer cells⁸⁴. In contrast, increased oxidative stress in muscle tissue deters metastasis⁸⁵. These findings suggest that local pools of antioxidants and ROS significantly impact the potential of some cancers to metastasize and progress.

Another protein that connects antioxidants and metastasis is BACH1, a pro-metastatic transcriptional factor⁸⁶ that senses intracellular free heme levels. When free heme levels rise (e.g., released from hemoproteins during oxidative stress), BACH1 is marked for proteasomal degradation. Long-term treatment of Kras-TP53-mutated mice with dietary antioxidants (N-acetylcysteine or Vitamin E) reduces free heme levels, stabilizing BACH1 levels and promoting lung metastasis⁸⁷. BACH1 levels are also stabilized in NRF2/Keap1-mutated tumors due to the upregulation of the classic NRF2 target heme oxygenase 1 (HMOX1), the enzyme responsible for the first and rate-limiting step in heme catabolism. Elevated BACH1 levels promote metastasis in animal models and are associated with increased metastatic burden, poor survival, and higher grade in human lung adenocarcinoma⁸⁸. The extent to which NRF2 supports other transcription factors to impact tumor metastasis requires additional investigations.

When do antioxidants prevent or slow cancers?

In the previous sections, we discussed the compelling evidence showing how antioxidants support tumorigenesis and metastasis of different tumor types. However, the literature reveals a more complex scenario, as several studies found that activation of NRF2 can have a minimal effect on tumorigenesis⁸⁹ and, in some cases, slow cancer progression and metastasis^90,91. A key distinction is the context in which Nrf2 plays a role. Regarding tumor initiation driven by a carcinogen, NRF2 can act as a suppressor of initiation but a supporter of progression⁹². More recent studies have demonstrated that in terms of tumor initiation with a genetic driver, NRF2 can support initiation but prevent progression to high-grade tumors⁵⁶. Further, NRF2 and other antioxidant-related proteins (i.e., TIGAR) can aid in preventing the development of metastasis⁶⁹. These results emphasize that the role of antioxidants during carcinogenesis is complex and dynamic, changing over time as cells in different stages of carcinogenesis require different features to thrive. Notably, paradoxical functions depending on the threshold of oncogene activation are not uncommon in cancer biology. While oncogenic activation of cMYC and RAS strongly stimulate proliferation and are key drivers of several malignancies, overactivation of these oncogenes induces apoptosis and senescence, respectively^93,94.

Aging is another context where antioxidants have demonstrated tumor-suppressive activity. Several mouse models of antioxidant depletion have shown increased cancer incidence, particularly in the later stages of an animal’s life^95–97. However, some antioxidants, such as Prdx1, also promote cancer in fast-growing oncogenic-driven models⁹⁸. It could be surmised that balancing antioxidant regulation is a knife-edge between cancer promotion and age-related damage, placing cancer on either side.

The dichotomy in the impact of antioxidants is observed in other scenarios, such as when a tumor suppressor becomes a tumor “supporter”. The best example is the tumor suppressor BRCA, which promotes the stabilization of NRF2^99,100. While BRCA is critical in preventing DNA damage^101,102 and loss of BRCA leads to the development of a range of tumors¹⁰³, BRCA doesn’t necessarily stay lost. The use of DNA damage-based therapies in BRCA1/2-mutated cancers has led to an interesting phenomenon in some cases where cancers become resistant to platinum and PARP inhibitors by undergoing reversion mutations that restore the functionality of BRCA^104–106. Hypothetically, the cancer has already undergone the necessary mutations to become tumorigenic, and restoration of BRCA1/2 would now support tumor progression. While the most logical impact of BRCA re-expression is to support DNA repair in the face of DNA damage-inducing agents, stabilizing NRF2 is potentially an important factor in this therapy’s resistance and tumor progression. Indeed, NRF2 is critical in preventing DNA damage¹⁰⁷. Further research is required to understand whether this patient group is potentially an ideal candidate for therapies that block NRF2 activity or its downstream antioxidant targets.

Cysteine in cancer

At the center of the battle between antioxidants and oxidative stress lies the tower of cysteine. Where cysteine comes from and where it goes is an ever-expanding list. Cells can acquire cysteine from the extracellular environment or synthesize it (from methionine) using the transsulfuration pathway. Uptake of extracellular cysteine (in the oxidized form of cystine) through the system x_c⁻ (xCT and 4F2hc) has been a major focus in cancer research^108–110 (Fig. 2). Indeed, blocking cystine abundance, uptake, and availability genetically or pharmacologically can impair tumor growth^111–114. The extent to which the transsulfuration pathway fuels cysteine production in cancer is less clear. Multiple reports have implicated a dependency of cancers on methionine and enzymes in the transsulfuration pathway^115–117. Recent studies have shown that only select tissues (i.e., liver and pancreas) have this capacity to synthesize cysteine from methionine in vivo, and this capacity is diminished during tumorigenesis¹¹⁸. Nonetheless, limiting cystine and methionine in the diet is potentially a tractable approach through adopting a plant-based diet¹¹⁹; whether this approach is sufficient to impair tumor growth and progression remains to be determined.

Figure 2. Acquisition of cysteine from GSH catabolism conserves intracellular metabolites.

System x_c⁻ (xCT and 4F2hc) imports cystine at the expense of glutamate export and NADPH consumption. Loss of these metabolites places a burden on other enzymes (i.e., GLS, G6PD) to replenish them and maintain homeostasis. However, GGT-mediated catabolism of GSH costs nothing. Breakdown of GSH by GGT yields glutamate and cysteinylglycine, which upon import, can be further broken down by peptidases into cysteine and glycine. Since this is potentially a more economical route, cancer may favor acquiring cysteine through GSH breakdown.

There are ways for cells to acquire cysteine independent of xCT and transsulfuration pathways. The earliest evidence is from Harry Eagle, showing that supplementation with GSH can rescue cells grown without cystine¹²⁰. At the time, it was not clear how this rescue occurred. While GSH can be exported from a cell¹²¹, the import of GSH is more complicated. A cellular transporter of GSH exists, termed HGT1¹²²; however, this protein is not conserved beyond yeast. Interestingly, GCLM is conserved across all species except yeast. Whether a handoff occurred over time where cells lost the ability to import GSH through HGT1 but gained the ability to accumulate GSH through GCLM is unclear. Nonetheless, while a mechanism to import GSH into mammalian cells has yet to emerge, a family of proteins called gamma-glutamyltransferases can initiate the catabolism of extracellular GSH into its individual amino acids. First postulated more than half a century ago¹²³, GSH is believed to be broken down into glutamate and cysteinylglycine, which can then fuel cysteine abundance in the cell, thus, explaining the observation by Eagle that GSH supplementation could rescue cystine deprivation.

GGT-mediated GSH breakdown might even be a more economical path for the cell to acquire cysteine compared to xCT-mediated cystine import. As previously mentioned, the xCT antiporter expels glutamate in exchange for importing cystine (Fig. 2). Further, upon entering the cell, cystine is reduced to cysteine in a NADPH-dependent reaction¹²⁴. Thus, to use xCT to acquire cysteine, the cell loses a molecule of glutamate and NADPH¹²⁵. In contrast, acquiring cysteine from GSH costs nothing. Upon being released from GSH, cysteinylglycine can enter the cell through PEPT1/2 transporters¹²⁶; afterward, it is further broken down into cysteine and glycine by dipeptidases. The differential importance between GGT and xCT is clearly demonstrated in vivo, where the loss of GGT1 (the putative dominant isoform of GGT) in mice causes perinatal lethality and serum depletion of cysteine¹²⁷. In contrast, loss of SLC7A11 in mice is well-tolerated and causes no differences in survival¹²⁸. Further, the importance of GGT1 is found in humans, as patients with inborn errors in GGT1 suffer from glutathionuria and neurological disorders¹²⁹. While these processes are poorly understood, further research is needed to understand the importance of GGT-mediated cysteine acquisition in tumors. If tumors rely on a mechanism that acquires cysteine while conserving glutamate, this could explain the lack of clinical impact glutaminase inhibitors (i.e., CB-839) have shown in the clinic.

Antioxidants in tumor microenvironments

The tumor microenvironment (TME) is a major determinant of the growth of any given tumor cell. Amino acid availability¹³⁰, oxygen content¹³¹, fluid viscosity¹³², and other local features define the suitability of any cancerous growth. The milieu surrounding the individual cells of the tumor is the tumor interstitial fluid (TIF)¹³³, a potentially critical source of metabolites (Fig. 3). In pancreatic tumors, the TIF is, surprisingly, low in antioxidant precursors, namely cystine¹³⁴. While this could reflect a rapid uptake of cystine by tumors, an alternative hypothesis is that other molecules (i.e., GSH) could potentially serve as a source of antioxidant precursors to tumors. Interestingly, cancer cells may also acquire cysteine from surrounding stromal cells. Chronic lymphocytic leukemia (CLL) cells, which exhibit low xCT expression, are supported by marrow stromal cells where the latter consume cystine and subsequently release reduced cysteine¹³⁵. Other precursors to GSH, such as glycine, are also abundant in the TME of certain malignancies and fuel GSH synthesis in cancer cells¹³⁶. Together, the antioxidant content of the TME could provide a supportive and nutritive buffering zone for cancer cells to resist stressors.

Figure 3. The tumor microenvironment supplies resources to support antioxidant pools.

The tumor microenvironment (TME) is a heterogeneous composition, which includes cancer cells, stromal, and extracellular matrix. Tumor interstitial fluid (TIF) is the medium within the TME comprised of antioxidants, amino acids, and metabolites. Cancer cells acquire these extracellular molecules through transporters and scavenging processes (i.e., macropinocytosis). Once inside, these molecules (e.g., cysteine) feed antioxidant-dependent (i.e., GSH synthesis, persulfides) and -independent pathways (i.e., protein translation, CoA synthesis).

Tumor and stromal cells can compete for nutrients in the TME¹³⁷. In this nutrient-scarce environment, extracellular protein scavenging, in the form of micropinocytosis and lysosome-mediated degradation, is necessary to survive^138–143 (Fig. 3). Indeed, both stromal and cancer cells engage in macropinocytosis to break down extracellular molecules, such as albumin, to yield precursors for GSH synthesis, such as cysteine and glutamine^144,145. Taken together, these observations strongly suggest that macropinocytosis may maintain antioxidant pools by contributing to nutrient uptake in cancers.

Compartmentalization of antioxidants in cancer cells

Subcellular compartmentalization of ROS and antioxidants is an important topic that is, to a large extent, poorly understood in tumors. Several innovative tools have informed the distribution of ROS and antioxidants across the cell^146–150, but less is known about how antioxidants transit in and out of organelles or are produced locally within cellular compartments. Only recently was it discovered that mitochondria import GSH through SLC25A39 to maintain the iron-sulfur clusters in the ETC^151,152. NADPH is also suggested to have fluxes within the cytosol and mitochondria that are distinct and independent^153–155. More recently, using a combination of cysteine profiling and functional genomic screening, it was discovered that an accumulation of ROS in the nucleus could initiate a signaling cascade to dampen ROS production in the mitochondria¹⁵⁶. The extent to which similar circuits exist between ROS production and antioxidant synthesis in other organelles warrants further investigation, with potentially exciting findings to be made.

While mitochondria are considered a central location for ROS production, the ER is unique in its elevated redox state, potentially due to the protein folding that occurs there¹⁵⁷. Compared to GSH’s highly reduced state in other organelles, glutathione in the ER is predominantly found either in its oxidized state¹⁵⁸ or bound to proteins¹⁵⁹. Sec61 has been identified as a GSH importer into the ER¹⁶⁰; however, whether this process is conserved in mammalian cells is unclear. Further, the production of NADPH through the pentose phosphate pathway can support GSH levels in the ER and maintain appropriate protein folding¹⁶¹. GSH-independent antioxidant enzymes, such as PRDX4, have also been implicated in maintaining the ER redox homeostasis in tumor cells¹⁶². Together, these findings implicate the ER as a unique location for redox buffering in the cell.

Conclusion

Along with the questions surrounding antioxidants and cancer posed throughout this review, numerous questions remain. Do tumors from distinct cell lineage (i.e., breast vs. intestinal) have a differential reliance on antioxidants? How does the persulfidation of antioxidant molecules impact tumor growth^163–165? How are cofactors (i.e., NADPH) shared between enzymes that support antioxidants and those that drive biosynthetic processes, such as lipogenesis¹⁶⁶? How does cysteine’s role in redox buffering balanced by the demand for it by other key tumor processes (i.e., protein translation, CoA synthesis, and iron-sulfur clusters)^167–172? How is selenocysteine, a key amino acid in antioxidant enzymes, distributed amongst antioxidant proteins (i.e., TXNRD1/2, GPX4) in cancer cells^173,174? Finally, what molecules remain to be discovered as “new” antioxidants? Similar to Sidney Farber’s initial tragic trials with folate¹⁷⁵, our poor understanding of antioxidants has led to disappointing and sometimes dangerous outcomes in the clinic. Importantly, however, key aspects of standard-of-care treatments (i.e., radiation and chemotherapies) are linked to the depletion of antioxidants and the induction of oxidative stress. If we can elucidate the regulation of antioxidants in cancer, we can potentially reveal, like Farber, an entirely new world of therapies for patients.