Cellular Stress Responses and Metabolic Reprogramming in Cancer Progression and Dormancy

Abstract

Recurrent disease after prolonged tumor dormancy is a major cause of cancer associated mortality, yet many of the mechanisms that are engaged to initiate dormancy as well as later recurrence remain incompletely understood. It is known that cancer cells initiate adaptation mechanisms to adapt tightly regulated cellular processes to non-optimal growth environments; Recent investigations have begun to elucidate the contribution of these mechanisms to malignant progression, with intriguing studies now defining cellular stress as a key contributor to the development and maintenance of cancer dormancy. This review will discuss our current understanding of stress responses facilitating malignant cell adaptation and metabolic reprogramming to establish tumor dormancy.

Introduction

Our collective understanding of the mechanisms at play in the tumor microenvironment which may drive cancer dormancy, and subsequent relapse, remain incompletely defined. The elucidation of these mechanisms is critical, as long-lived dormant cancer cells are known drivers of disease relapse and distant metastasis; key elements of cancer-related mortality^1–3. Accumulating evidence suggests cellular responses to environmental stress and related metabolic reprogramming are critical in the establishment of cancer dormancy^4–6. However, our understanding of the array of stress response mechanisms and their impact on malignant progression and dormancy is in its infancy. Cellular stress is term that is associated with any disrupted cellular process that results in an imbalance in cellular homeostasis. Currently, cellular stress responses are recognized to be driven by exposure to hypoxic conditions, nutrient or heme deficiencies, viral infections, and disruptions in redox homeostasis^7–9. These stress responses function as a cellular adaptation mechanism to hostile environments which have been demonstrated to promote cancer cell survival, therapeutic resistance, and malignant progression, as well as the paralysis of antitumor immune responses^10–15; these events define the lethality of cancer. Recent studies have delineated mechanisms by which endoplasmic reticulum (ER) stress responses as well as oxidative stress are associated with the development and maintenance of dormant cancer cells through metabolic reprogramming and redox imbalances in a variety of malignancies^4–6,16,17. Therefore, further dissecting the role cellular stress plays in promoting or repressing tumor cell dormancy will likely promote the development of a more refined arsenal of therapeutics to promote optimal outcomes for cancer patients.

The unfolded protein response in cancer progression and dormancy

In the ER, stress occurs due to an accumulation of unfolded and potentially damaged proteins that exceed chaperones capacity, eliciting the so-called unfolded protein response (UPR)⁷. The UPR functions as a cellular rheostat that operates in large part to mitigate cellular ER stress through the induction of chaperones, addressing redox imbalances, and promoting ER-associated protein degradation (ERAD)¹⁸. In the context of cancer, the UPR appears to be detrimental to the host, as recent studies have linked this response to progression of several cancers, including ovarian¹⁹, breast^4,20,21, prostate²², and colon cancer²³, while also promoting T cell paralysis^24,25 and myeloid cell dysfunction^26–28 in the tumor microenvironment as has been recently reviewed elsewhere²⁹. Mechanistically, stress-induced accumulation of misfolded proteins within the ER lumen drives the dissociation of binding immunoglobulin protein (BIP) from three ER-resident molecules that instigate the UPR: the kinases inositol-requiring enzyme 1α (IRE1α) and PKR-like ER kinase (PERK), as well as the transcription factor activating transcription factor 6 (ATF6). These molecules thus promote downstream signaling that facilitates the establishment of a cryoprotective state and promotes cancer cell survival under conditions of cellular stress²⁹.

ER stress and the subsequent UPR have established roles in driving malignant progression of cancer. ER stress-driven IRE1a-XBP1 signaling, for example, has been shown to promote cancer progression and therapeutic resistance through the upregulation of interleukin-6 (IL-6) in a variety of tumors^30–32. Interestingly, recent studies have begun to illuminate the role that ER stress and the subsequent UPR have in regulating a switch between tumor cell acquisition of dormancy or malignant progression. In fact, the ablation of XBP1 in triple negative breast cancer (TNBC) cells was shown to inhibit tumor relapse in association with a reduction in CD44^high/CD24^low cancer stem cell-like tumor cells⁴. The role of XBP1 in TNBC tumorigenesis and relapse potential was shown to be dependent on the formation of a transcriptional complex consisting of XBP1 in association with hypoxia inducing factor (HIF)1α; this complex was found to regulate the expression of HIF1α target genes in an XBP1-dependent manner⁴. A subsequent and related study found that MYC-driven driven breast cancer progression also required the formation of MYC/XBP1 transcriptional complexes²⁰. Similar findings have associated oncogenic Ras-induced proliferation with UPR activation and XBP1 activity³³, while targeting IRE1α attenuated the ability of TNBC to adapt to stress in vivo and resulted in normalization of the tumor vasculature³⁴. Further investigation beyond breast cancer is needed to understand the conservation of these mechanisms, as well as to thoroughly dissect ER stress response activity that favors the establishment of cancer dormancy over that of malignant progression. As stress responses are not binary within a heterogenous tumor and may also drive cellular apoptosis³⁵, these outcomes are likely the result as yet uncharacterized differential responses to the local environment, and may be potentially impacted by differential activity of IL-6³⁶. Nonetheless, these studies have begun to implicate intact ER stress response signaling as a facilitator of tumor progression and maintenance of dormant tumor cells.

Other arms of the ER-driven unfolded protein response are associated with the survival of dormant tumor cells as UPR-responsive ATF6 has been implicated in the survival of tumor cells in a dormant state. Animal survival is prolonged upon knockdown of the α isoform of ATF6 in a murine model of dormant squamous carcinoma⁵. Mechanistically, ATF6α−mediated transcription was shown to drive the upregulation of Reb and the activation of mTOR signaling to enhance survival of dormant tumor cells, while knockdown of ATF6 or Reb rendered dormant tumor cells sensitive to rapamycin. Contrastingly, however, ATF6-mediated EGF signaling was found to promote reactivation of quiescent non-small-cell lung cancer cells through initiation of an angiogenic program³⁷. This suggests a potential differential response on tumor cells establishing a dormant tumor program mediated by ATF6 which may be tumor type specific. Finally, unmitigated PERK kinase activity driven by ER stress has also been implicated in the development of dormant tumor cells in a mouse model of pancreatic ductal adenocarcinoma⁶. Importantly, reestablishment of ER homeostasis through the treatment of tumor-bearing mice with the chemical chaperone 4-phenylbutyrate promoted tumor cell escape from senescence and the subsequent development of macrometastases, suggesting unmitigated ER stress functions as a mechanism for the establishment of pancreatic cancer dormancy while the homeostatic ER stress facilitates outgrowth of pancreatic cancer. Thus, there exists clear evidence that the molecular activity associated with perpetual UPR signaling can promote tumor cell acquisition of a dormant phenotype and survival of the cancer cells while existing in a dormant state. Further investigation of the downstream mediators and chaperones induced by the activity of these molecules in various tumors is clearly needed and will yield critical information regarding the establishment of a dormant tumor phenotype as well as mechanisms which promote later recurrence. Studies targeting these UPR pathways will illuminate their role in the establishment of tumor dormancy and will likely provide the rationale to undertake combinatorial therapeutic approaches utilizing inhibitors of IRE1α, ATF6, and PERK signaling to eradicate rogue dormant cancer cells.

Oxidative stress and metabolic reprogramming as a driver of cancer dormancy

The role of oxidative stress in governing cancer cell fate and dormancy is emerging. Oxidative stress occurs in cells as a result of redox imbalances; that is, a disturbance in the production of reactive oxygen species (ROS) and the a scavenging system which functions to detoxify them, e.g. superoxide dismutase, glutathione reductase, catalases, etc³⁸. The mitochondria is a major producer of ROS as a byproduct of oxidative phosphorylation, while peroxisomes and the endoplasmic reticulum are also known producers. The key transcription factor that responds to this redox imbalance is NRF2, which binds to antioxidant (ARE) elements in the promoter of many ROS scavengers^39,40. The hypoxic tumor microenvironment is a driver of increased ROS in cancer cells^41,42, while ionizing radiation, and chemotherapeutics such as Cisplatin, Doxorubicin, 5-fluorouracil, are known to induce apoptosis via ROS in these decades’ old therapeutic approaches^43,44.

The role of ROS in the establishment of distant metastases from circulating tumor cells remains controversial. Scavenging mitochondrial superoxide has been shown to prevent spontaneous tumor metastasis in mice⁴⁵, while pharmacological inhibition of mitochondrial complex I-mediated oxidative stress decreased metastatic feature of colorectal cells⁴⁶. An investigation by Piskounova et al.¹⁶, however, has illuminated the role of redox imbalances in the formation of melanoma metastases. It was demonstrated that circulating melanoma cells possessed elevated ROS levels and oxidized glutathione compared to the primary tumor, and intriguingly, successful metastases of melanoma presented with metabolic alterations, such as NADPH-generating enzymes derived from the folate pathway, as well as reduced mitochondrial mass and mitochondrial membrane potential, suggesting these mechanisms are activated in order to withstand the oxidative stress that promoted the seeding of metastatic lesions¹⁶. The inability to adapt to this redox imbalance via folate pathway inhibition impaired the formation of distant metastasis, while antioxidants promoted distant metastasis in xenograft studies¹⁶, which has been supported by other studies⁴⁷. Overall, perpetual oxidative stress appears to limit the metastatic potential of at least melanoma cells, however differential levels ROS within tumor cells may drive distinct outcomes^16,45–47. Nonetheless, unresolved oxidative stress has been shown to limit cancer cell survival and progression, as has been discussed elsewhere⁴⁸.

Elevated ROS are now considered hallmarks of the circulating cancer stem-like cells that establish tumor dormancy, and which are associated with a glycolytic signature as well as a ROS-induced metabolic usage of the pentose phosphate pathway, likely allowing adaptation through enhanced buffering against oxidative stress through the generation of NADPH and glutathione⁴⁹, and that may be distinctly associated with the attainment of a dormant tumor phenotype^50,51. Recent investigations of circulating tumor cells (CTCs) in breast cancer identified glycolytic signatures associated with the recurrence and metastasis of residual disease in breast cancer. Elevated expression of glycolytic enzyme phosphoglycerate kinase (PGK) and the pentose phosphate pathway enzyme glucose-6-phosphate-dehydrongenase (G6PD) have been associated with increased risk of disease recurrence and metastasis^52,53, likely through successful maintenance of dormant tumor cells. In fact, it has been demonstrated that high levels of PGK and G6PD in breast CTCs represents a signature associated with higher risk of disease relapse⁵⁴.

Supportive of these concepts, tumor cell exit from dormancy may rely on the ability of the cell to successfully scavenge excessive ROS. Her2 downregulation in breast cancer was found to promote oxidative stress through metabolic modulation, which elicited antioxidant activity through enhanced NRF2 transcription⁵⁵. Interestingly, tumor recurrence in this system was found to be accelerated through the reestablishment of redox homeostasis and nucleotide metabolism by the activity of NRF2. Importantly, tumor cells exiting dormancy and expressing high levels of NRF2 proved to be sensitive to inhibition of glutaminase due to their altered metabolic phenotype, suggestive of a potential therapeutic target for NRF2-addicted tumor cells exiting dormancy⁵⁵. The relationship between oxidative stress and NRF2-driven outcomes in cancer is just beginning to be elucidated, however its activity has clearly been associated with tumorigenesis (often in association with KRAS or PKB/AKT activating mutations) as well as the development of resistance to chemotherapy and ionizing radiation⁵⁶. Moreover, evidence suggests that breast cancer dormancy is dependent on the activity of oxidative stress-driven 5’adenosine monophosphate-activated protein kinase (AMPK) activity, which functions upstream of cellular antioxidant responses⁵⁷. The pharmacological targeting of AMPK promoted the clearance of residual disease⁵⁷, further implicating redox homeostasis as critical in the fate of dormant disease. Further defining the role of elevated ROS in regulating maintenance of tumor dormancy or progression will require additional investigation, but clearly involves NAD⁺ metabolism, as discussed below.

Cellular stress in the maintenance of senescent cisplatin-resistant cancer cells

High-grade serous ovarian cancer remains one of the most lethal malignancies, partly owing to the development of resistance to platinum-based chemotherapies. Platinum-based cisplatin is a robust inducer of oxidative stress which impacts the cellular utilization of central carbon metabolism, which was found to be critical in driving cisplatin-mediated antitumor toxicity⁵⁸. This is in addition to ROS-induced upregulation of the mitochondrial class III histone deacetylase SIRT2, and NAD⁺ dependent histone deacetylase, which prime ovarian cancer cells for sensitivity to cisplatin⁵⁹. Poor responses to cisplatin are associated with elevated activity of NRF2, further suggesting the beneficial role of ROS in impairing malignant progression of ovarian cancer⁵⁹.

Redox imbalance also plays a role in the maintenance of cisplatin-induced senescent cancer cells, as well as their eventual relapse. As such, NAD⁺ biosynthesis has recently been demonstrated to regulate the acquisition of a cisplatin-induced senescent phenotype in ovarian cancer⁶⁰. The investigation of mechanisms associated with cisplatin-induced acquisition of cellular senescence points to the activity of nicotinamide phosphoribosyltransferase (NAMPT), the rate-limiting enzyme of NAD⁺ biosynthesis in the establishment of senescent cancer stem cells¹⁷. Inhibition of NAMPT suppressed cisplatin-resistant cancer outgrowth, suggesting an impaired ability to reduce ROS via NAD⁺ is required for the maintenance of cisplatin-resistant tumor cells in a senescent state. These studies clearly highlight the role of NAD⁺ metabolism in fate of senescent cancer cells, and implicate the need for therapeutic strategies which manipulate their redox balance to maintain these cells in a dormant state or to force their elimination^61,62.

Concluding Remarks

We are now beginning to gain an understanding of myriad mechanisms that elicit programs which promote cancer dormancy. As discussed here, the contribution of cellular stress and related metabolic reprogramming have recently become apparent in these processes. Investigation of the activity of the unfolded protein response, elicited by ER stress, have now clearly defined this response with the acquisition and maintenance of a dormant cancer cell phenotype. However, additional studies will be required to discern this response over that of UPR-induced cancer cell proliferation or cancer cell death. In addition to the potential therapeutic utilization of IRE1α, ATF6, and PERK pathway inhibitors to target dormant cancer, there is now also clear rationale for considering the particular redox biology and related metabolic function of dormant cancer cells and therapeutic-resistant senescent cancer cells established by oxidative stress. Targeting the antioxidant activity of NRF2⁶³ to maintain cancer cells in a dormant state, or conversely, to promote NRF2 activity or to utilize ROS scavengers along with combinatorial chemotherapeutic or immunotherapeutic strategies^64–67 may be a viable strategy to purge dormant cancer cells from the patient in the quest to prevent relapse-related cancer mortality.