Targeting mitochondrial metabolism for metastatic cancer therapy

Abstract

Primary tumors evolve metabolic mechanisms favoring glycolysis for ATP generation and antioxidant defenses. In contrast, metastatic cells frequently depend on mitochondrial respiration and oxidative phosphorylation (OxPhos). This reliance of metastatic cells on OxPhos can be exploited using drugs that target mitochondrial metabolism. Therefore, therapeutic agents that act via diverse mechanisms, including the activation of signaling pathways that promote the production of reactive oxygen species (ROS) and/or a reduction in antioxidant defenses may elevate oxidative stress and inhibit tumor cell survival. In this review, we will provide (1) a mechanistic analysis of function-selective extracellular signal-regulated kinase-1/2 (ERK1/2) inhibitors that inhibit cancer cells through enhanced ROS, (2) a review of the role of mitochondrial ATP synthase in redox regulation and drug resistance, (3) a rationale for inhibiting ERK signaling and mitochondrial OxPhos towards the therapeutic goal of reducing tumor metastasis and treatment resistance. Recent reports from our laboratories using metastatic melanoma and breast cancer models have shown the pre-clinical efficacy of novel and rationally designed therapeutic agents that target ERK1/2 signaling and mitochondrial ATP synthase, which modulate ROS events that may prevent or treat metastatic cancer. These findings and those of others suggest that targeting a tumor’s metabolic requirements and vulnerabilities may inhibit metastatic pathways and tumor growth. Approaches that exploit the ability of therapeutic agents to alter oxidative balance in tumor cells may be selective for cancer cells and may ultimately have an impact on clinical efficacy and safety. Elucidating the translational potential of metabolic targeting could lead to the discovery of new approaches for treatment of metastatic cancer.

INTRODUCTION

Cancer is the leading cause of death for men and women ages 40-79 and in 2022 an estimated 1,918,030 new cases of cancer and 609,360 cancer deaths are projected to occur in the United States¹. Incidence for breast cancer in women has increased by about 0.5% per year from 2014-2018. There are also more than 3 million survivors in the U.S. with about 80% of them being post-menopausal². Since breast cancer is a major public health issue, identifying new non-toxic metastasis-prevention agents would be highly significant. Melanoma of the skin is a highly metastatic cancer that is expected to affect almost 57,100 men and 42,000 women in 2022 accounting for 6% and 5% of the total cancer incidence, respectively. Metabolic changes in tumor cells result from onco/suppressor genetic events that activate specific signaling and transcriptional networks. Discovering the molecular connections between mitochondria and metastasis will determine whether inhibitors of mitochondrial OxPhos could have therapeutic impact for cancer treatment. Understanding how kinase signaling and metabolic pathways cooperate in promoting tumor progression will lead to discovery of novel combination approaches to increase efficacy and lower toxicity. Work in this area is significant and innovative because a focus on mechanisms of mitochondrial redox regulation, which include inhibition of ATP synthase, could reveal more effective therapeutic targets. This could eventually provide a rationale for translational and preclinical development of novel compounds to prevent or treat metastatic and drug-resistant tumors^3-5. The goal of this review is to provide an overview of signaling and cancer metabolism pathways as they relate to metastatic cells and how metabolic regulation of glycolytic and mitochondrial OxPhos pathways play a role in survival and spread of metastatic cancers and their acquisition of treatment resistance. Mitochondrial generation of ATP and oxidative stress are two areas of therapeutic intervention. In particular, the targeting of kinase signaling pathways to increase reactive oxygen species (ROS) or directly targeting the production of ATP with ATP synthase inhibitors to generate ROS are two strategies with potential promise to alter cellular bioenergetics. Specific recent examples will be highlighted to show that inhibiting mitochondrial OxPhos and ERK1/2 signaling may increase ROS levels to compromise metastatic tumor cell survival.

Cancer treatments are changing the lives of people with primary breast cancer^6-8. However, metastatic breast cancer is a continuing public health problem^8,9. The difficulties in managing metastatic cancer and drug resistance are the motivation for designing new and effective treatments. Many hormone-responsive breast cancers are treatable with selective estrogen receptor modulators (SERMs) or aromatase inhibitors, which reduce estrogen levels. But treating metastatic breast cancer, especially triple-negative breast cancers (TNBC) that lack estrogen receptor, progesterone receptor, and HER2 expression, is currently less effective^5,10. Approximately half of all melanomas express mutations in the BRAF gene, which leads to constitutive activation of the downstream MEK1/2 and ERK1/2 proteins that drive cancer cell survival. While targeted inhibition of mutated BRAF showed short term efficacy in treating late-stage metastatic melanoma, current clinical practice includes use of BRAF plus MEK1/2 inhibitors and/or immunotherapy approaches for late-stage metastatic melanoma^11,12. For example, BRAF inhibitors – vemurafenib (Zelboraf), dabrafenib (Tafinlar) or encorafenib (Braftovi) and MAPK kinase (MEK) inhibitors – trametinib (Mekinist), cobimetinib (Cotellic) or binimetinib (Mektovi) are standard of care for treating advanced melanoma that has spread to other parts of the body^11,12. In 2020, the FDA designated the ERK1/2 inhibitor ulixertinib available for the Expanded Access Program to treat cancers with aberrant ERK1/2 activity^11,12. Trametinib is indicated in combination with dabrafenib after surgery as adjuvant therapy for stage III melanoma patients. Combination approaches may have several advantages, including reduced toxicity because of the use of lower individual drug levels and reduced risk of cancer recurrence and treatment resistance, resulting in more durable responses. While BRAF and MEK inhibitors transiently reduce metastatic burden and extend patient survival by a few months, most patients eventually develop acquired resistance to these therapies.

Evidence for reactivation of ERK1/2 in patients with resistance to BRAF and MEK inhibitors has promoted the development of the previously mentioned ulixertinib and other selective ERK1/2 inhibitors to improve therapeutic outcomes in melanoma or other cancers driven by constitutive ERK1/2 activation¹³. In addition, increased dependence on OxPhos is one reason for the development of acquired resistance to BRAF and MEK inhibitors¹⁴.

Over the last several years, a focus on the Warburg effect has diverted attention from the role of mitochondrial respiration in tumor progression. However, the premise and rationale for targeting metastatic cancer metabolism depends on the observations that metastatic tumors exhibit active mitochondrial OxPhos, which plays a central role in generation of ROS and cell death, survival, or metastasis^15-18. Redox balance is regulated by the very high glucose uptake evident in tumor cells and the concomitant generation of TCA cycle metabolites that feed electrons into the mitochondrial electron transport chain (ETC). Primary cancers, including melanoma and breast cancers, are often treatable via a variety of interventions including surgery¹⁹. However, metastatic tumors, such as TNBC or melanoma with BRAF or NRAS mutations, are often resistant to treatment and may lack effective targeted therapies²⁰. Metastatic cells in the circulation and cells homing to the metastatic site may depend on mitochondrial OxPhos for their energy needs^5,10 because the ETC and mitochondrial adenosine triphosphate (ATP) synthase promote OxPhos and play a central role in regulation of cancer cell oxidative damage, growth, and survival^21,22. Metastatic cells under hypoxic conditions exhibit functional OxPhos with variable glycolytic activity to promote a hybrid metabolic state¹⁵ that supports metastatic cell survival²³. Additionally, kinase signaling pathways that include ERK1/2, phosphoinositide 3-kinase/protein kinase B (PI3K/Akt), and the Janus kinases (JAK) play a central role in linear (vertical) and crosstalk (horizontal)-dependent metabolic plasticity⁷. Active mitochondrial function and cell differentiation can reduce cell type evolution (heterogeneity), inhibit drug resistance, and prevent cancer initiation. However, in tumor cells mitochondrial reprogramming^5,15,24-28 or constitutive activation of intracellular kinase signaling or both^7,29-33 may promote metastatic progression. Therefore, inhibition of these pathways may restrict tumor cell proliferation or promote cell death through feedback inhibition of respiratory capacity, increased ROS, and lower glucose utilization. Our recent published data^34-36 support the hypothesis that mitochondrial OxPhos and elevation of ROS are important drivers of oncogenesis. Partial inhibition of MAPK (ERK1/2) signaling or targeting mitochondrial ATP generation via ATP synthase targeting can disrupt the redox balance in tumor cells. Strategies to improve therapeutic efficacy by inhibiting glucose uptake, mitochondrial OxPhos, and anaerobic (glycolytic) metabolism may, therefore, be promising approaches to prevent or treat metastatic progression^37,38.

CANCER METABOLISM IN METASTATIC CELLS

Tumor metabolic heterogeneity reflects the tissue of origin, oncogenic events that support specific survival or proliferation, and changes in the microenvironment of the tumor that promote an epithelial mesenchymal transition (EMT) and tumor progression³⁹. A diversity of cells in the microenvironment of a tumor, including fibroblasts, tumor-associated macrophages, and adipocytes provide metabolic intermediates and nutrients that drive bioenergetics and escape from cell death signaling pathways. Because of the complexities that regulate metastatic spread of a tumor, there is high heterogeneity in drug response among different patients diagnosed with the same cancers. Understanding the metabolic mechanisms that regulate metastasis would, therefore, provide valuable information about precision medicine strategies to prevent or treat metastases. Glycolysis does play an important role in tumorigenesis because separately knocking down several glycolysis-regulating genes (HMMR, KIF20A, PGM2L1, and ANKZF1) resulted in less glucose consumption, less lactic acid production, and reduced cell migration and invasion of prostate cancer cells⁴⁰. A risk score based on glycolysis-related genes may serve as an accurate prognostic marker for prostate cancer patients with biochemical recurrence. For these reasons, most research in recent years has focused on glycolytic pathways driving tumor growth (Warburg effect). However, there is still considerable interest in the “reverse Warburg” effect, which involves the capture of glycolytic lactate by cancer cells from adjacent stromal cells that is converted to pyruvate, which enters the TCA cycle and provides electrons for the mitochondrial ETC and ATP generation (OxPhos)^15,39. Because of the complex tumor microenvironment, lactate can also be derived from fibroblasts, tumor cells or both.

Glycolysis-driven lactate dehydrogenase (LDH) can stimulate ROS production from mitochondria. In fact, the role of lactate in promoting metastasis is well-documented^41-43. Lactate can regulate immunosuppression in the tumor microenvironment⁴⁴ and lactate transporters like monocarboxylate transporter (MCT1-4) can promote MCT-dependent cancer metastasis⁴⁵. One inhibitor of MCT1 (AZD3965) is currently under evaluation in clinical trials⁴⁶. For these reasons, MCT1 inhibitors are currently in development to reduce lactate transport and enhance radiosensitivity⁴⁷. Hexokinases (HKs), which mediate the utilization of glucose, are present in the cytoplasm where they phosphorylate intracellular glucose, the first rate-limiting step of glycolysis^48,49. Of the four HK subtypes (encoded by different genes), HK1 is expressed in normal tissues while HK2 is highly expressed in cancer cells and facilitates chemoresistance and metastasis of hepatocellular, colorectal, lung, gastric, melanoma, and ovarian cancers^50-57. HK2 can translocate from the cytoplasm to the outer mitochondrial membrane and interact with the voltage-dependent anion channel (VDAC) to promote resistance to chemotherapy by inhibiting apoptosis. Mitochondrial-associated HK2 (Mito-HK2) is also in close proximity to mitochondrial ATP production by ATP synthase and may prevent cell death by multiple mechanisms⁵⁸. Mito-HK2 promotes survival by inhibiting mitochondrial permeabilization and the consequent release of pro-apoptotic cytochrome c⁵⁹. The suppressive mechanism involved appears to include the inhibition of ROS accumulation⁵¹. Another anti-apoptotic mechanism includes HK2 inhibition of Bcl2/BAX association, which releases BAX from the mitochondria and promotes Bcl2 inhibition of apoptosis⁶⁰. Mito-HK2 is also involved in regulating prostate cancer metabolic activities^61,62. HK2, together with these interacting proteins, plays an important role in maintaining cancer cell malignancy and metastasis although no HK2 inhibitors have as yet shown efficacy in clinical trials.

Current translational medicine approaches targeting metabolism⁶³, including mitochondrial metabolism, are beginning to have an impact on clinical therapeutics^64-66. Recent findings indicate that glycolysis and mitochondrial OxPhos cooperate under hypoxic and nutrient-deprived selective pressures to promote drug resistance and fuel the metastatic potential of tumor cells and circulating metastatic cells that rely on mitochondrial OxPhos^5,10. Therefore, mitochondria are an important therapeutic target in cancer²⁴ because uncoupling mitochondrial membrane potential (MMP) from ATP production has been shown to inhibit glucose uptake, OxPhos, and also anaerobic (glycolytic) metabolism that support a metastatic phenotype^24,37,38. A major regulator of mitochondrial respiration, or oxygen consumption rate (OCR), is the ATP synthase, which drives mitochondrial OxPhos in the presence or absence of glycolysis^15,67. ATP synthase is also important in ATP production in tumor cells characterized by high glucose uptake under hypoxic conditions (a classic Warburg effect), suggesting cooperative tumorigenic effects of glycolysis and mitochondrial OxPhos^15,23. In fact, a subgroup of breast cancer patients with significantly worse prognosis was identified on the basis of higher ATP synthase expression⁶⁸. Drug-resistant metastatic tumors, such as triple negative breast cancers (TNBC), are often dependent on mitochondrial respiration (OxPhos) to generate energy and promote survival. Since there are no targeted therapies for TNBC and since most mitochondrial-targeting drugs exhibit substantial toxicity, there is a need to find new and safer therapeutic agents.

Mitochondrial dysfunction may be associated with increased cancer incidence and poor outcomes in select populations⁶⁹. African American patients have higher cancer mortality rates and shorter survival times relative to European American patients⁷⁰. Mitochondrial OxPhos and its dysfunction have been linked to racial disparity for cancer risk and survival of African American patients with prostate cancer⁷¹. A set of genes upregulated in several tumors was associated with enhanced OxPhos, transcription factors that promote mitochondrial biogenesis, and increased mitochondrial number in these patients⁷². By measuring mitochondrial content in patient tissues using a multi-cancer tissue microarray approach (TMA), OxPhos genes were found to be enriched in tumors from African American patients⁷². These tumors also exhibited enrichment for ERR1-PGC1α–mediated transcriptional programs that are associated with mitochondrial biogenesis. Changes in mitochondria may, therefore, be a distinguishing feature among tumors from African American patients and provide a rationale for repurposing mitochondrial inhibitors to treat these cancers⁷³.

Mitochondrial ATP production has been a therapeutic target for cancer treatment for many years²⁵. Cancer cells with high ATP levels are aggressive and exhibit multi-drug resistance, invasiveness, and spontaneous metastasis. Therefore, cancer cells exhibiting enhanced bioenergetics may resist environmental or chemotherapeutic pressure while simultaneously maintaining survival because of their increased ATP content, which allows for tumor recurrence and metastatic progression. Drug treatments that target mitochondrial ATP synthase and/or induce mitochondrial uncoupling (such as Bedaquiline or Niclosamide) have shown some promise in depleting cellular ATP and inhibiting metastasis^25,74. New approaches to control cancer metabolism are also beginning to take into consideration the role of the tumor immune microenvironment in regulating metastasis⁷⁵. As these cell-intrinsic metabolic pathways are defined, the role of the immune system in regulating metastatic metabolism⁷⁶ and exploiting the metabolic vulnerabilities of normal and cancer cells to control metastasis may also be elucidated⁶³.

MITOCHONDRIAL OXPHOS AND ATP SYNTHASE – IMPLICATIONS for METASTASIS

Production of ATP can promote tumor cell drug resistance and metastasis²⁵, which has implications for therapeutic approaches that lower mitochondrial ATP levels. In fact, high levels of ATP in the extracellular milieu activates hypoxia-inducible factors, which are associated with cancer progression, metastasis, poor prognosis, and resistance to chemotherapy⁷⁷. Inhibiting ATP synthase to reduce ATP levels can generate increased oxidative stress resulting from reduced antioxidant levels and lower rates of respiration (or oxygen consumption rates)^25,35 and may have benefits for treating tumor growth and preventing metastasis. However, for mitochondrial OxPhos analysis in response to therapeutic drugs several caveats need to be considered^78,79, including cellular energy availability, ATP production, glycolysis, extracellular acidification (ECAR) levels, energy partitioning, and mitochondrial dysfunction, which are essential in the design of anti-metastatic approaches for cancer therapy⁸⁰. These considerations may be relevant since there are considerable differences in the response of primary and metastatic tumors to therapeutic drugs⁸¹. Inhibiting ATP synthesis and OxPhos may affect the survival of respiration-competent metastatic cancer cells²⁵. We recently showed³⁵ that targeting breast cancer metabolism with a novel inhibitor of mitochondrial ATP synthase may have therapeutic potential for breast cancer treatment by targeting mitochondrial OxPhos. Using a direct drug discovery approach and computer-assisted drug design (CADD), we identified novel small molecules that interfere with transcription factor Runx2 protein-DNA binding and transcriptional activity⁸². While normal epithelial cells were relatively resistant to these molecules, a lead compound (CADD522) inhibited breast cancer cell proliferation and tumorsphere formation and inhibited mitochondrial ATP synthase and respiration (oxygen consumption) while increasing ROS³⁵. CADD522, was shown to not only reduce primary tumor growth in vivo but also to reduce experimental metastasis of tumor cells in the lungs of immune-compromised mice after tail-vein injection⁸². This compound lowered mitochondrial oxygen consumption rates (OCR) and ATP levels in human breast cancer cells. The inhibition of mitochondrial OxPhos resulted from the direct targeting of ATP synthase enzymatic activity. A key observation providing a clue to the mechanisms essential for these activities was the demonstration that the CADD522 compound inhibits mitochondrial respiration OCR³⁵ and maximum respiratory capacity (MRC)⁸³. Mitochondrial respiratory parameters in the future may be a more accurate depiction of metastatic capacity of tumor cells^26,78,84,85. To measure ATP levels, MCF7 or MDA468 breast cancer cells were treated with CADD522 under conditions (presence of pyruvate or galactose) where cellular metabolism favors mitochondrial OxPhos-driven ATP production³⁵. CADD522 lowered the levels of ATP under these conditions. Associated with this clear suppression of ATP production was the related increase in intracellular ROS levels. Results indicate that some of the ROS stimulated by CADD522 may be derived from mitochondria, which are implicated in tumor metastasis⁸⁵.

CADD522 also inhibited ATP synthase enzymatic in a dose-dependent manner and interacted with specific ATP synthase subunits³⁵, suggesting that lower ATP levels in response to CADD522 could be due to direct inhibition of mitochondrial ATP synthase activity, the central regulator of cell bioenergetic activity of the cell^21,86. These results suggest that the effects of CADD522 involve its interaction with subunits of the F1-ATP synthase complex (Figure 1)⁸⁶. Many metabolic poisons have been found to interact with mitochondrial ATP synthase and the enzyme complex has been the target of numerous approaches⁸⁶ to inhibit cancer metabolism⁸⁷. However, many of the approaches targeting the Fo-subunits of the ATP synthase (such as oligomycin) are too toxic for clinical use^88,89. Whereas drugs that interact with and inhibit the F1-ATP synthase subunits (such as CADD522, metformin, resveratrol)^86,90 have more favorable toxicity profiles and may be of particular interest in treating metastatic cancers. With regard to toxicity, the CADD522 has so far been shown to reduce tumor cell growth but not induce apoptosis^35,82. In fact, normal epithelial cells and non-transformed breast epithelial (MCF10A) cells grown in 3D cultures were largely unaffected by CADD522³⁵.

Figure 1. (A) ATP synthase, respiration (OCR), and ROS production.

ATP synthase plays a key role in mitochondrial ATP production, glucose oxidation, OCR, and ROS production. This regulation has implications for stem cell maintenance, drug resistance, and metastasis. We found a small molecule inhibitor of the Runx2 transcription factor (CADD522) that also inhibits ATP synthase leading to lower OCR and elevated ROS, inhibited PDH and TCA cycle glucose flux, and inhibited electron transport. (B) Lead Compound = fusion of groups found in nature: plant-derived core structures norbornene and dichlorophenyl identified by computer-assisted drug design (CADD). ETC, electron (e−) transport chain; OCR, oxygen consumption rate; TCA, tricarboxylic acid cycle; NADH, succinate, major electron donors; ETC Complex IV, electron acceptor (O₂); PDH, pyruvate dehydrogenase. (C) Inhibitory sites of ATP synthase and putative (*) CADD522 interaction site. ATP synthase inhibitors act as scaffolds for new drug development because of their conserved function in multiple species (modified from Hong 2008)⁸⁶. Some drugs (Oligomycin) inhibit ATP synthase activity (ATP production coupled to proton flux) because they block proton (H+) translocation across the mitochondrial inner membrane (Fo subunits). Rotation of the central stalk against the surrounding alpha(3)beta(3) F1-subunits leads to synthesis of ATP at three separate catalytic domains on the beta subunits. Many F1-complex-targeted drugs interact with alpha/beta sites. Cell-free mitochondrial lysates do not exhibit proton-coupled or Fo-dependent enzymatic activity, which resides on the F1 complex of subunits (the alpha/beta interaction sites) inhibited by CADD522.

Previous reports have shown that the general ROS scavenger N-acetyl cysteine (NAC) may provide protection against cytotoxic oxidants^91,92 and inhibit apoptosis⁹³ induced by ROS. Inhibition of ATP synthase by CADD522 enhanced ROS levels under conditions of oxidative stress and may be one explanation for its anti-tumor effects^35,94. NAC was found to neutralize some of the CADD522-stimulated ROS in breast cancer cells and NAC attenuated the inhibitory effect of CADD522 on breast cancer cell growth. Expression of the catalytic subunit ß-F1-ATP synthase is tightly regulated by post-transcriptional signaling mechanisms that affect mRNA localization, stability, and translation^95-99. CADD522 treated tumor cells may exhibit gene expression changes in response to this pro-oxidant molecule. Consistent with its effect on ATP synthase, the mRNA level of MT-ATP5B (mitochondrial ATP synthase F1 subunit beta) was markedly decreased by CADD522 treatment in breast cancer cells³⁵ and it significantly inhibited expression of several mitochondrial biogenesis-related genes examined (PGC-1α, NRF-1, HRF2b, TFAM, TFB2M). Notably, the mRNA expression level of the key mitochondrial biogenesis-promoting transcription factor PGC-1α was almost completely suppressed by CADD522³⁵. These changes in gene expression are commonly observed after treatment with drugs that target mitochondria in metastatic tumor cells^100,101. This is also observed for mitochondrial targeted therapies to treat tubular cell acute kidney injury¹⁰², multi-targeted metabolic therapy for cancer that activates multi-gene pathways¹⁰³, targeting mitochondrial malignancy and ROS generation for cancer therapy¹⁰⁴, and targeting the redox imbalance in mitochondria to increase ROS to selectively inhibit cancer cells through altered gene expression promoting apoptosis, autophagy, and necrosis thus inhibiting their metastatic potential¹⁰⁵. In summary, our data suggest that inhibition of mitochondrial ATP synthase and ROS generation are contributors to the effectiveness of CADD522 to delay tumor growth, metastasis, and cancer progression.

MAPK SIGNALING, REDOX REGULATION, AND THERAPEUTIC STRATEGIES

Mitogen-activated protein kinase (MAPK) family proteins, including ERK1/2, regulate and link intra- and extracellular signaling events to mediate changes in cellular gene expression, proliferation, and survival. Aberrant ERK1/2 activation in cancer cells is due to upstream mutations or overexpression of receptor tyrosine kinases, RAS G-proteins, and RAF kinases^106-110. Mutations in MEK1, a direct activator of the ERK1/2 proteins, may also deregulate ERK1/2 activity¹¹¹. In addition, an ERK2 mutation that changes a conserved glutamate to lysine (E322K) within a protein docking domain causes constitutive ERK2 activation and promotes cutaneous T-cell lymphoma progression^112-114. Typically, anticancer drugs that target ERK1/2 signaling block the ATP binding or catalytic sites on receptor tyrosine kinases or serine/threonine kinases like BRAF and MEK1/2. These compounds are effective in blocking ERK1/2 signaling^115-118. However, because ATP binding sites are conserved in many protein kinases, this strategy often results in off-target toxicity and the invariable development of acquired drug resistance^119,120. Cellular ROS and oxidative stress play a confounding role in both promoting or inhibiting cancer cell proliferation, depending on the cellular context. There are several examples that demonstrate increased or decreased oxidative stress will sensitize cancer cells to growth inhibition and cell death^85,121. Much research has supported the use of antioxidant strategies to protect against some cancers¹²¹. However, clinical trials have yet to show conclusive evidence for the beneficial effects of antioxidant supplements to reduce cancer risk¹²². In the context of the ERK1/2 pathway, antioxidants like NAC and vitamin E may even promote metastasis in lung cancer with KRAS mutations¹²³. This mechanism is thought to occur through the stabilization of the transcription factor BACH1, which regulates the expression of genes involved in oxidative phosphorylation, EMT, and metastasis. In contrast, many anticancer drugs, including ERK1/2 pathway inhibitors, promote tumor cell death by increasing oxidative stress¹²¹. Melanocytes and melanoma cells are particularly sensitive to ROS and subtle changes in ROS levels can influence proliferation and survival of these cells¹²⁴. Moreover, there is evidence to suggest that increased oxidative stress may inhibit melanoma cell metastasis⁹². Even if ROS elevation by individual therapeutic agents may not be sufficient to inhibit all cancer cells, enhanced oxidative stress may improve anticancer efficacy of combination therapies¹²⁵. Recent data show that therapeutic approaches that increase ROS may benefit patients who have developed resistance to BRAF and MEK1/2 inhibitors¹²⁶.

As an alternative to ATP/catalytic site inhibitors of protein kinases, we have used computer-aided drug design (CADD) to identify function-selective ERK1/2 inhibitors that disrupt ERK1/2 regulation of select substrate proteins^127-130. Subsequent studies identified a novel compound, SF-3-030, that forms a covalent interaction with a cysteine residue near the F-recruitment site (FRS) on ERK1/2 that is distinct from the D-domain recruitment site (DRS) involved in interactions with MEK1/2 (Figure 2). The FRS on ERK2 facilitates interactions with substrates containing an F-site (FXF motif) that include the FOS family of transcription factors and c-Myc. Consistent with ATP competitive ERK1/2 inhibitors, SF-3-030 was more effective at inhibiting the proliferation of cancer cell lines containing activating mutations in the ERK1/2 pathway and it inhibited mutant BRAF expressing melanoma cell proliferation by increasing ROS production³⁴.

Figure 2.

(A) Ribbon structure of ERK2 (PDB:4GT3) highlights substrate docking sites. The D-domain recruitment site (DRS) or F-recruitment site (FRS) are indicated by blue/green or brown spheres, respectively. The conserved lysine found in the ATP binding site and the activation site are the yellow and red spheres, respectively. Unlike ATP binding/catalytic site inhibitors that block all kinase activity, function-selective ERK1/2 inhibitors target a select group of substrates that interact with the DRS or FRS. (B) SF-3-030 forms a cysteine adduct near the FRS. Model of ERK2 shows the position of all cysteine residues (green). High-resolution LC-MS identified C252 (red font) to be the primary (90%) residue modified by SF-3-030. The adjacent FRS (red) and activating TXY motif (magenta) are shown for reference.

In support of its mechanism in disrupting ERK1/2 interactions at the FRS, SF-3-030 inhibited epidermal growth factor (EGF) or phorbol ester (PMA) induction of immediate early genes of the FOS family (c-Fos, FosB, and Fra-1) and c-Myc in HeLa cells¹³⁰. In addition, SF-3-030 inhibited Activator Protein-1 (AP-1)-mediated transcription in HeLa cells treated with EGF or PMA and in A375 melanoma cells expressing a constitutively active BRAF (V600E) mutation^34,130. AP-1 has been implicated in regulating the expression of genes that promote cell invasion and metastasis¹³¹. Comparing the effects SF-3-030 with ATP-competitive/catalytic site inhibitors of MEK1/2 and ERK1/2 in A375 cells containing activating BRAF and Ras mutations provided additional clues to the molecular mechanisms mediating the biological effects of SF-3-030. Transcriptome analysis by RNAseq comparing A375 cells treated with either SF-3-030 or the MEK inhibitor AZD6244 (positive control for ERK1/2 pathway inhibition) revealed overlap between SF-3-030 and AZD6244-regulated genes. A key finding was that both SF-3-030 and AZD6244 inhibited c-Myc levels at both the RNA and protein levels suggesting common effects of these compounds on ERK1/2-mediated signaling. In non-transformed airway smooth muscle cells, we have shown that SF-3-030 inhibits transforming growth factor-beta (TGFß) signaling¹³², which promotes cell invasion and metastasis¹³³. Although SF-3-030 inhibits key regulators of cell invasion and metastasis, such as c-Myc, AP-1, and TGFß signaling, experimental evidence demonstrating this compound inhibits cancer cell metastasis in vivo has yet to be determined.

Since Myc proteins are key regulators of metabolic changes in cancer cells that support a metastatic phenotype¹³⁴, we also determined whether SF-3-030 affected proteins that regulate cell metabolism. High-resolution liquid chromatography-tandem mass spectrometry was used to analyze protein changes in A375 cells in the first 12 hours after exposure to SF-3-030. Unique to SF-3-030 treated A375 cells was enhanced expression of markers of mitochondrial dysfunction and oxidative stress. In particular, SF-3-030 significantly increased the oxidative stress-induced growth inhibitor-1 (OSGIN1) protein, which is regulated by oxidized lipids and the transcription factor NRF2 to promote cytochrome c release and induce apoptosis¹³⁵. These data indicate that SF-3-030 induced signaling events that reduce mitochondrial function and increase ROS. Another ROS responsive protein upregulated by SF-3-030 was the zinc finger transcription factor ZNF744, which has been reported to repress hepatocellular carcinoma cell invasion and metastasis¹³⁶.

While ROS activation by SF-3-030 was consistent with elevated ROS in melanoma cells treated with ATP binding site and catalytic inhibitors of BRAF and MEK1/2^14,137,138, the rapid induction of NRF2 by SF-3-030 raised questions about the implications for cancer progression and metastasis. NRF2 competes with the previously mentioned BACH1 transcription factor to regulate genes involved in oxidative stress such as heme oxygenase-1 (HO-1)¹³⁹. Oxidative stress displaces BACH1 repression to allow NRF2 to activate the HO-1 promoter. Consistent with the activation of NRF2, SF-3-030 is a potent inducer of HO-1. While HO-1 has been implicated in promoting cancer progression and metastasis, HO-1 may also have anti-metastatic functions¹⁴⁰. The effects of SF-3-030 on HO-1 are consistent with the upregulation of HO-1 observed in melanoma cells that have developed acquired resistance to BRAF and MEK1/2 inhibitors³⁶. SF-3-03 induction of NRF2 and HO-1 could suggest a protective response against oxidative stress that has been observed with other anticancer drugs and contributes to drug resistance¹⁴¹. However, the use of NRF2 inhibitors to block induction of HO-1 and other NRF2-regulated genes did not affect SF-3-030 inhibition of A375 cell proliferation. This finding suggests that SF-3-030–mediated inhibition of A375 melanoma cell proliferation and a possible metastatic phenotype is through a mechanism that is dependent on elevated ROS but not NRF2.

MAP kinases are essential intracellular signaling enzymes that mediate extracellular signals and receptor activation with gene expression changes that promote tumor growth and stress responses that lead to metastatic progression⁷. Inhibitors of kinases within this network have been extensively used in clinical settings¹³ and employ the use of maximum tolerated dose approaches to achieve maximum clinical benefits. Ironically, these approaches often do not yield durable responses due to the development of acquired drug resistance, unacceptable toxicities, and activation of alternative pathways, which may be difficult to target. One solution to overcome drug resistance may be to identify targets and combinations of drugs to inhibit multiple oncogenic pathways with lower drug concentrations that avoid the toxicity often associated with maximum tolerated dosing. Increased OxPhos is a common metabolic adaptation that supports drug resistance^17,36,142. We have observed a number of protein changes in melanoma cells with acquired resistance to BRAF and MEK1/2 inhibitors, including increased protein levels of pyruvate kinase, mitochondrial pyruvate carrier 1, and mitochondrial pyruvate dehydrogenase subunits all of which support increased glycolytic ATP production and mitochondrial oxidative phosphorylation³⁶. Other proteins that promote a metastatic phenotype and are upregulated in BRAF/MEK1/2 inhibitor resistant melanoma cells include matrix metalloproteinases (MMP1/3), tetraspanins (TSPAN3,6, 8,31), and caveolin1/2 isoforms³⁶. Similarly, MMP inhibitors (TIMP1/3) were reduced in BRAF/MEK1/2 inhibitor resistant cells, which supports an environment conducive to ECM degradation, cell invasion, and metastasis. Moreover, the metastatic suppressor NDRG1 (N-MYC downstream regulated gene-1), which inhibits EMT and cell migration¹⁴³ was downregulated in BRAF/MEK1/2 inhibitor-resistant melanoma cells. These findings support the link between inhibitors of the ERK1/2 pathway and MYC-mediated metabolic reprogramming involved in metastasis¹⁴⁴ providing a rationale for further development of MYC selective inhibitors.

Several new anticancer targeted strategies that focus on prevention of drug resistance have been proposed⁷. The main hypothesis behind these approaches is that the use of specific drugs targeting cancer promoting genes may benefit from novel low-dose multi-drug combinations that target key kinase signaling networks and branch points, which are often the rate-determining steps in tumor progression. There are several aspects to these strategies that need to be evaluated including re-examining single agent approaches and choice of targets, vertical inhibition strategies (linear), horizontal inhibition strategies (crosstalk), low dose multi-drug strategies, and combination low-dose approaches. These studies should aim to use bioinformatics and mathematical modeling to identify novel combination approaches⁷. These considerations highlight several novel strategies that could be used for targeting interconnected oncogenic networks in metastatic cells with single or multiple targeting agents. Compensatory networks are activated when there is redundancy in oncogenic signaling and drug exposure is prolonged. For example, targeting only a single node with low-dose drug may avoid drug toxicity but may lead to reduced efficacy if oncogenic signaling is not completely blocked. On the other hand, targeting an individual node with high-dose drug could inhibit pathway activity with some efficacy but alternative pathways could be activated causing drug resistance^7,25,145. Therefore, it would be of therapeutic benefit if the strategy of drug delivery is designed for the specific activating events present in a particular tumor and with the design of avoiding drug resistance⁷.

CONCLUSIONS

Chemotherapy improves overall survival of cancer patients and can be employed for the treatment of metastatic tumors but chemoresistance is an ongoing problem in achieving durable responses to therapy. Mitochondrial metabolism is an essential driver of tumor metastasis and chemoresistance in cancer. Many metabolic mechanisms of glycolysis-induced chemoresistance have been studied, including MAPK signaling pathways within the tumor microenvironment^146-148. However, mitochondrial contributions to chemoresistance are also observed¹⁴⁹. It is now becoming clearer that tumor cells respond to changes in ROS levels^150,151. Early studies suggested that elevated ROS promoted tumorigenesis but there is no evidence that the use of antioxidants improves patient survival and, in some cases, antioxidant use may worsen outcomes⁹². In contrast, the use of prooxidants to further elevate ROS may improve therapeutic outcomes by inhibiting cancer cell metastasis¹⁵². One approach to increase ROS is through function-selective ERK1/2 inhibitors (such as SF-3-030)³⁴ that target substrate binding sites and block some but not all enzyme activity. This is in contrast to current ERK1/2 (and most other kinase) inhibitors that target the ATP binding or catalytic sites and block all enzyme activity. Targeting ERK1/2 functions that selectively disrupt MYC and AP1-mediated transcription and cancer cell growth has the potential to reduce selective pressure that invariably drives acquired resistance to the current kinase inhibitors. For the ATP synthase inhibitor, CADD522, ROS is likely from mitochondrial sources since the drug directly inhibits ATP synthase, inhibits OCR, and increases ROS ³⁵. Therefore, both SF-3-030 and CADD522 appear to disrupt mitochondrial function. Targeting these pathways with novel compounds that increase ROS to inhibit cancer cell proliferation and metastasis has the potential to mitigate acquired drug resistance. These findings are consistent with previous studies that implicate the ERK1/2 pathway in regulating OxPhos and the induction of ROS after treatment with inhibitors of BRAF and MEK1/2^14,137,138. In summary, we present evidence that selective targeting of unique ERK1/2 and ATP synthase mitochondrial functions increases ROS in tumor cells and may be a valid therapeutic approach to reduce cancer cell metastasis. Consideration of mitochondrial bioenergetics⁷⁹ may, in future, lead to development of combination targeting approaches to overcome therapeutic limitations⁷. Such combination approaches should focus on the use of low dose, multidrug combination therapy to target key kinase signaling networks rather than individual kinases or signaling pathways, as a way of realizing the full potential of combination treatments.

DATA AVAILABILITY STATEMENT

Data sharing is not applicable to this article as no new data were created or analyzed in this study.