"This flips our understanding of cancer cell death on its head," said senior author Matthew J. Hangauer, Ph.D., assistant professor of dermatology at UC San Diego School of Medicine and Moores Cancer Center member. "Cancer cells which survive initial drug treatment experience sublethal cell death signaling which, instead of killing the cell, actually helps the cancer regrow. If we block this death signaling within these surviving cells, we can potentially stop tumors from relapsing during therapy."

Global cancer burden and early resistance

Cancer is responsible for about one in six deaths worldwide. Many of these deaths occur because tumors initially respond to treatment, then later acquire resistance and come back. Typically, resistance develops over months to years through new mutations, much like bacteria gradually evolving resistance to antibiotics. These mutation-driven changes are hard to manage with the limited number of drug combinations available.

The newly identified mechanism, however, operates at the very beginning of resistance and does not depend on genetic mutations. Because it appears so early and is not tied to permanent changes in DNA, it represents a promising new point of attack for future therapies.

"Most research on resistance focuses on genetic mutations," said first author August F. Williams, Ph.D., a postdoctoral fellow in the Hangauer lab at UC San Diego. "Our work shows that non-genetic regrowth mechanisms can come into play much earlier, and they may be targetable with drugs. This approach could help patients stay in remission longer and reduce the risk of recurrence."

Persister cells, death enzymes and tumor relapse

In the new study, the researchers found:

• In models of melanoma, lung and breast cancers, a subset of "persister" cells that survive treatment showed ongoing, low-level activation of a protein involved in normal cell death that breaks down DNA, called DNA fragmentation factor B (DFFB).
• The level of DFFB activation was too low to kill these cells, but high enough to disrupt how they respond to signals that would normally keep their growth in check.
• When this protein was removed, persister cancer cells stayed dormant and did not regrow during drug treatment.
• DFFB is not required in normal cells, but is necessary for regrowth of cancer persister cells, which marks it as a promising target for combination therapies designed to prolong responses to targeted treatments.

Study publication and research support

The findings were reported in Nature Cell Biology and were supported in part by grants from the Department of Defense, the National Institutes of Health and the American Cancer Society. Hangauer is a cofounder, consultant and research funding recipient of BridgeBio subsidiary Ferro Therapeutics.

"Our ability to link certain cues or stimuli with positive or rewarding experiences is a basic brain process, and it is disrupted in many conditions such as addiction, depression, and schizophrenia," says Alexey Ostroumov, PhD, assistant professor in the Department of Pharmacology & Physiology at Georgetown University School of Medicine and senior author of the study. "For example, drug abuse can cause changes in the KCC2 protein that is crucial for normal learning. By interfering with this mechanism, addictive substances can hijack the learning process."

The study, supported by the National Institutes of Health (NIH), was published December 9 in Nature Communications.

How KCC2 Shapes Dopamine Activity and Reward Learning

The team found that changes in learning can occur when levels of the KCC2 protein shift. When KCC2 levels are reduced, dopamine neurons fire more rapidly, which encourages the formation of new reward associations. These dopamine neurons produce and release dopamine, a neurotransmitter essential for motivation, reward processing, and motor control.

To better understand this relationship, researchers studied rodent brain tissue and monitored the behavior of rats during Pavlovian cue-reward tests. In these classic experiments, a brief sound alerts the rats that a sugar cube is on the way. Beyond analyzing how KCC2 affects the pace of neuron firing, the investigators discovered that neurons firing in a coordinated pattern can amplify dopamine activity in a surprising way. Short bursts of dopamine appear to serve as potent learning signals that help the brain assign meaning and value to shared experiences.

Why Everyday Cues Can Trigger Cravings

"Our findings help explain why powerful and unwanted associations form so easily, like when a smoker who always pairs morning coffee with a cigarette later finds that just drinking coffee triggers a strong craving to smoke," notes Ostroumov. "Preventing even relatively benign drug-induced associations with situations or places, or restoring healthy learning mechanisms, can help develop better treatments for addiction and related disorders."

How Diazepam and Other Drugs Influence Neuron Coordination

The researchers also examined whether drugs that act on specific cellular receptors, including benzodiazepines such as diazepam, could alter learning processes. Earlier work showed that shifts in KCC2 production, and therefore in neuron activity, can change how diazepam (valium) produces its calming effects in the brain. The current study adds another layer to this understanding by showing that neurons do more than increase or decrease activity. They can coordinate their firing patterns, and when that coordination occurs, they transmit information more effectively. The team found that diazepam can support this coordinated activity in their experiments.

Methods and the Importance of Using Rats for Behavioral Tests

"To reach our conclusions, we combined many experimental approaches, including electrophysiology, pharmacology, fiber photometry, behavior, computational modeling, and molecular analyses," says the study's first author Joyce Woo, a PhD candidate in Ostroumov's lab.

She explained that rats were chosen for the behavioral portion of the research because they typically perform more consistently than mice on longer and more complex tasks. Their reliability in reward-learning experiments allowed the research team to gather more stable and informative data.

Broader Implications for Brain Disorders and Treatment Strategies

"We believe these discoveries extend beyond basic learning research," says Ostroumov. "They reveal new ways the brain regulates communication between neurons. And because this communication can go wrong in different brain disorders, our hope is that by preempting these disruptions, or by fixing normal communication when it's impaired, we can help develop better treatments for a wide range of brain disorders."

Additional Georgetown contributors include Ajay Uprety, Daniel Reid, Irene Chang, Aelon Ketema Samuel, Helena de Carvalho Schuch and Caroline C Swain.

Ostroumov and his co-authors report having no personal financial interests related to the study.

This work was supported by NIH grants MH125996, DA048134, NS139517, DA061493, as well as grants from the Brain & Behavior Research Foundation, the Whitehall Foundation and the Brain Research Foundation.