Cancer cells live in a constantly changing environment, where access to oxygen can determine whether they grow, migrate, or survive. As tumors expand, their rapidly growing cells often outpace the development of functional blood vessels, creating oxygen-deprived regions known as hypoxic microenvironments. Under these conditions, cancer cells frequently shift their behavior, slowing proliferation while becoming more migratory and invasive. Interestingly, restoring oxygen does not always reverse these changes immediately. Some cancer cells continue behaving as though they are oxygen-starved, even after returning to oxygen-rich surroundings. This phenomenon, known as “hypoxic memory,” is rather like a person who has learned to survive a prolonged drought and continues conserving resources even after the rain returns. However, an important question remains: does the duration of oxygen deprivation determine how strongly cancer cells remember it, and could this memory influence how aggressively tumors invade surrounding tissues?
This question is particularly important because oxygen availability within tumors is neither uniform nor constant. Abnormal blood vessel architecture, unstable perfusion, and fluctuating oxygen delivery create regions that repeatedly experience hypoxia and reoxygenation. These environmental changes can activate hypoxia-inducible factors, alter cellular metabolism, and trigger transcriptional and epigenetic adaptations that persist beyond the initial exposure. Consequently, two cancer cells experiencing the same oxygen concentration may behave very differently depending on their previous exposure histories. A cell exposed to hypoxia for several hours may rapidly recover its original phenotype, whereas one exposed for several days may retain an invasive state for much longer. This suggests that tumor invasion depends not only on the oxygen concentration that cells experience today, but also on how long they experienced oxygen deprivation in the past. Yet, most existing computational approaches represent hypoxic memory by using a fixed persistence time, overlooking how the duration of exposure might progressively reshape cellular behavior.
Gopinath Sadhu and colleagues address this challenge by combining transcriptomic analysis of breast cancer samples with a computational framework that connects oxygen availability, phenotypic switching, and tumor invasion. Their analysis identifies an enrichment of invasion-associated genes that are both transcriptionally upregulated and epigenetically hypomethylated, supporting a possible link between persistent molecular changes and invasive behavior. The researchers then develop a mathematical model describing two interacting cancer cell states: oxygen-rich normoxic cells, which primarily proliferate, and oxygen-deprived hypoxic cells, which exhibit enhanced migration. Rather than assuming that hypoxic cells return to their original state after a fixed period of reoxygenation, the model incorporates duration-dependent hypoxic memory, allowing prolonged oxygen deprivation to progressively slow the transition back to the normoxic phenotype. By coupling these phenotypic transitions with spatial oxygen gradients, extracellular matrix dynamics, and tumor growth, the framework investigates how past environmental exposure influences future invasive behavior.
Their simulations reveal that the duration of hypoxic exposure can fundamentally reshape tumor invasion. Compared with models incorporating fixed hypoxic memory, dynamic hypoxic memory produces faster invasion by allowing migratory hypoxic cells to persist within oxygen-rich regions and accumulate at the advancing tumor front. These cells effectively carry their previous hypoxic phenotype into new environments, increasing their collective spreading capacity. The model also predicts an enhanced effective diffusion of hypoxic cells, reflecting their greater spatial dispersal, and shows that invasion driven by dynamic memory becomes less sensitive to fluctuations in oxygen supply. In other words, once cancer cells develop a sufficiently persistent hypoxic phenotype, temporary improvements in oxygen availability may no longer be enough to reverse their invasive behavior. This creates a striking biological possibility: oxygen deprivation may continue influencing tumor progression long after the original hypoxic conditions have disappeared. Importantly, these findings arise primarily from computational simulations, providing mechanistic predictions that require further experimental validation.
The broader significance of this study lies in recognizing cellular memory as an important link between the tumor microenvironment and cancer progression. Hypoxia is often evaluated through oxygen measurements or molecular markers at a particular moment, but these measurements may overlook the cumulative effects of previous oxygen deprivation. Incorporating exposure history could therefore improve our understanding of tumor heterogeneity, invasive progression, and responses to therapies that alter tumor oxygenation or vascular function. The findings also highlight opportunities for developing microfluidic tumor models and vascularized organ-on-chip platforms capable of reproducing controlled cycles of hypoxia and reoxygenation while monitoring cancer cell migration. Future investigations could combine such experimental systems with live-cell imaging, oxygen sensing, transcriptomics, and epigenetic profiling to determine how hypoxic memory is established, maintained, and eventually erased. By connecting environmental history with cellular behavior, Sadhu and colleagues offer a compelling perspective on tumor invasion: cancer cells may not simply respond to the environment they occupy—they may also be guided by the environments they remember.
—Prasanna Padmanaban