Around 1,600 people die every day in the U.S. as a result of cancer. While surgery can frequently be an effective cure at early stages, once cancer cells start to disseminate away from the primary tumor and colonize other tissues, effective therapy becomes a major challenge.
But why do cancer cells move? Why doesn't the primary tumor continue to grow but remain where it is? The answer to this question requires an in-depth understanding of how tumors develop, combined with insights from evolutionary biology.
Cancer initiates with the acquisition of genetic or epigenetic events that provide pro-proliferative signals together with bypass of senescence. Subsequent proliferation is accompanied by the generation of intra-tumor heterogeneity. Genetic heterogeneity is well-known and can give rise to mutations that bypass blockade by targeted therapies. Since genetic heterogeneity is proportional to the number of cancer cells in the patient, a targeted therapy will be more effective at a stage when the numbers of such cells are low, such that the probability of genetic resistance is reduced. This may be on first diagnosis, or after the first round of therapy has substantially diminished the number of viable cancer cells.
Superimposed on genetic heterogeneity is so-called phenotypic heterogeneity, in which cells within the tumor adopt specific states that are triggered by the intra-tumor microenvironment. Within a tumor, cancer cells co-exist with infiltrating immune cells, fibroblasts, and other stromal cells. Signals from these cells together with changes in oxygen and nutrient availability can have a major impact on cancer cell behavior. The effect of the microenvironment may account for why, within a tumor, perhaps only as few as 20 percent of cancer cells may be actively proliferating, even though all may have an activated oncogene.
In addition to those cells that are actively cycling, tumors also contain a proportion of dormant cells that may cause relapse, even many years after an apparently successful therapy, and invasive cells that can enter the blood or lymphatic vessels to seed new tumors. It is now increasingly recognized that a major trigger for invasive behavior is the influence of the microenvironment, though genetic lesions may sensitize cells to microenvironmental factors. Surprisingly, for an increasing number of cancers, it seems proliferation and invasion may be inversely correlated.
Over recent years, studies have identified many of the molecular triggers for cells to move, including hypoxia and exposure to an inflammatory environment, and it is known that cell migration requires the engagement of specific mechanisms implicated in remodeling the cytoskeleton. However, a key question remains as to whether each trigger for cell movement engages a distinct molecular mechanism to facilitate movement or whether many triggers for invasiveness converge on a common mechanism. If so, what is the mechanism and how does it relate to the therapeutic resistance often associated with cells that have undergone the proliferative-to-invasive phenotype switch? Moreover, would understanding why cells move, as opposed to how, provide new therapeutic opportunities?
In single-celled organisms, invasion is triggered by starvation. Cells that have depleted resources move or become invasive in an attempt to find new sources of nutrient. This response is conserved in higher organisms, including humans. Could cancer cells also become invasive in response to starvation?
We recently explored this possibility and found that starvation of glutamine could trigger melanoma cells to become invasive (Genes & Dev 2017;31:18-33). More importantly, pre-starving melanoma cells before injection into mice led to massively enhanced tumor formation capacity, induced hallmarks of BRAF-inhibitor resistance, and imposed a gene expression program related to one that had recently been identified in melanoma patients that were resistant to anti-PD-1 immune checkpoint inhibitor therapy.
We also dissected the molecular mechanism underpinning these biological consequences downstream from glutamine deprivation. The key trigger for invasion was shown to be translation reprogramming mediated by phosphorylation of the protein translation initiation factor eIF2α. In fact, pharmacological activation of eIF2α phosphorylation, even in nutrient-rich conditions, both increased invasion and the tumor-forming potential of the treated cells.
Why should translation reprogramming be important? Starving cells need to respond in two ways. They increase nutrient supply, which they can do by becoming invasive and seeking new resources. They also boost expression of nutrient importers on their cell surface and decrease their nutrient demand, which can be achieved by shutting down the majority of translation. Both responses are coordinated by eIF2α phosphorylation.
However, cancer cells are found in a microenvironment in which nutrient limitation is just one of many stresses. Could other stresses also trigger invasion by the same mechanism? We examined whether inflammation driven by TNFα would also impose an invasive phenotype by the same mechanism. Again, we found it could. We also found the same translation reprogramming mechanism was important for neural crest migration in development, as well as invasion by yeast that form invasive hyphae in response to nutrient limitation.
A simple analogy to explain these observations is that primordial cells long ago evolved a bicycle to get pizza and in doing so were able to more efficiently get the nutrients they required to survive and divide; it is much more efficient to go to the pizza store than wait for the pizza delivery van to pass by chance. But as organisms became more complex and the stresses that cells encountered became more diverse than simply nutrient limitation, they realized the same bicycle originally evolved to get pizza could be used as a means to escape other stresses as well. The bicycle, in this case, is translation reprogramming and the many stresses that trigger eIF2α phosphorylation include nutrient deprivation, inflammation, chemotherapies, and hypoxia.
There remain a number of outstanding questions, such as what proteins translated under stress are necessary for invasion or tumor-initiation and whether a similar “starvation” response underpins invasion, metastatic spread, and therapy resistance in other types of cancer.
That said, the results provide a number of insights important for cancer therapy. The first is that at sub-lethal doses, drugs that increase stresses that trigger eIF2α phosphorylation may promote invasion and metastatic spread. The second is that, if reducing nutrient demand were critical for the survival of invasive cells, then perhaps strategies designed to increase demand in this key phenotypic subpopulation would trigger cell death. The search is now on to identify specific molecular targets for drugs that can alter the supply-demand balance within cancer cells. Such drugs might be more effective against cells in which an activated oncogene predisposes the cancer to a high-demand state.
Have Stories Delivered Directly to You
Sign up for our free e-newsletter and receive email notifications when each issue becomes available and when other new content is posted online. New to Oncology Times? Visit the publication's website at www.oncology-times.com where you can get free access to archived issues, subscribe to the print magazine, and create your own personal online article collections.
COLIN GODING, PHD, is a member of the Ludwig Institute for Cancer Research at the University of Oxford and also a Professor of Oncology at the University of Oxford and Director of the Oxford Stem Cell Institute.