Article In Brief
New research suggests that synapses exist between neurons and glioma cells involving AMPA receptors, that the neuron-glioma interaction is bidirectional, and that neuronal activity can induce glioma growth, while gliomas can increase neuronal activity.
Scientists have shown that glioma tumor cells are communicating with local neurons by making synapses that accelerate the progression of the tumor growth. The finding, published September 18 in Nature, offers the possibility that scientists could target these synapses or their signaling consequences to stop or slow these virulent cancers.
“Our earlier work made us consider this wild idea that neurons might be forming synapses with cancer cells,” said Michelle Monje, MD, PhD, associate professor of neurology and neurological sciences at Stanford University. “We were startled when our experiments demonstrated that they do.”
The story behind the latest finding begins in 2015, when the pediatric neuro-oncologist and her colleagues made xenografts (using glioma tumor cells from patients and transplanting them into mouse brain) and used optogenetics to show that excitatory neuronal actively promotes the growth of high-grade glioma tumors.
They took healthy brain slices and collected the fluid surrounding these slices in the lab dish and studied it. They wanted to know if healthy brain cells were secreting growth factors in an activity-regulated way that may cause glioma cells to proliferate. Indeed, they did find substances in the fluid: Brain-derived growth factor (BDNF) and a secreted form of neuroligin-3. This second substance was surprising since neuroligin-3 is a post-synaptic adhesion molecule, and not known to be a mitogen that promotes the growth of cells. They found that neuroligin-3 triggered the PI3k-mTOR pathway, a classical oncogenic signaling pathway, in glioma cells.
In another series of experiments, they created xenografts in either wild-type or neuroligin-3 knockout mice and showed that high-grade human gliomas did not grow in the mice that were not producing neuroligin-3. Still, it made no sense that the loss of neuroligin-3 from the brain microenvironment had such a strong effect on glioma growth, because in general cancer cells receive many different growth factors and also have an intrinsic propensity to proliferate. Why would glioma depend so much on one factor in the microenvironment? Was there something more fundamental that neuroligin-3 contributes to?
The researchers went on to expose glioma tumor cells to neuroligin-3 and measured gene expression changes. And this is when they saw an increase in synapse-related gene expression. This finding, reported in Nature in 2017, set the stage for this new work. No one had ever considered that neurons might be making synapses with high-grade glioma tumor cells.
Study Methods, Findings
In the current study, Dr. Monje and her colleagues designed a series of experiments using primary tumor cells—pediatric and high-grade gliomas—removed during biopsy or autopsy to test their idea that there is electrical and synaptic integration of glioma into neural circuits. They examined gene expression activity in the primary human glioma biopsy samples and found expression of glutamate receptor genes and post-synaptic structural genes in the glioma cells.
Dr. Monje said that this gene enrichment was found in glioma cells that resembled oligodendrocyte precursor cells, a glial cell type that normally forms synapses with neurons. They also replicated this finding in the xenograft model.
They performed immune-electron microscopy using green fluorescent protein labeled glioma cells and found 10 percent of the malignant glioma cells were engaged in post-synaptic activity. They co-cultured the glioma cells expressing the fluorescent markers with wild type or neuroligin-3 knockout neurons and again found less synaptic activity in the absence of neuroligin-3 compared to wild-type neurons.
Using classical patch clamp electrophysiology, Dr. Monje's team found that there is a strong electrochemical communication through AMPA-receptor dependent neuron-glioma synapses. They also found a second form of neuronal activity-evoked current in the glioma cells that represents (non-synaptic) potassium currents. These neuronal activity-evoked currents are amplified through gap junction coupling between malignant glioma cells.
In other words, there is a complete network that involves neuronal-glioma signaling. When they blocked that signaling (genetically or with drugs) it inhibited the growth of the glioma xenografts. The mice had smaller tumors and lived longer. When they turned up the signal, the glioma cells multiplied and mouse survival was decreased.
“We know that the microenvironment of the brain is very important in understanding high-grade gliomas,” explained Dr. Monje. “Neurons are a critical component of the glioma microenvironment.”
The scientists went on to target these neuron-glioma currents with AMPAR-blocking antiepileptic drugs. The xenografted mice (with human high-grade pediatric glioma tumors) were administered the antiepilepsy drug, perampanel, and showed a 50 percent decrease in proliferation compared to mice that were not treated with the medication. They also used other medications to block the potassium currents. That also led to a decrease in glioma tumor growth.
“Neuron-glioma interactions are bi-directional,” the team wrote in the Nature paper. “Neuronal activity increases glioma growth, and gliomas are thought to increase neuronal activity.”
They then used intraoperative electrocorticography to measure electrical activity in three glioblastoma patients during a resting state before resection and found increased cortical excitability right around the tumor compared to nearby healthy tissue.
“In glioma, we have demonstrated bona fide neuron-to-glioma synapses, reminiscent of the synapses found on normal oligodendrocyte progenitor cells,” the scientists wrote. “In addition, we have shown neuronal activity-evoked potassium currents in glioma cells, reminiscent of activity-dependent currents in normal astrocytes.”
The hope, said Dr. Monje, is to interrupt this crosstalk between neurons and glioma cells to slow or stop tumor progression.
In another paper in the same issue of Nature, a team of German scientists published similar findings. “We were surprised to learn that neurons form synapses with cancer cells, which we thought only healthy cells do,” said Hrvoje Miletic, MD, PhD, professor of biomedicine at the University of Bergen, a co-author of the study. “The new insight has allowed us to open a completely new field to understand malignant brain cancer and how to attack it. Controlled testing of antiepileptic medicine is perhaps a possible new strategy,” Dr. Miletic said.
“This is potentially a very important paper that builds on a growing body of evidence suggesting that glioma growth may be regulated by neuronal activity,” said Patrick Y. Wen, MD, MD, FAAN, director of the Center of Neuro-Oncology at the Dana Farber/Brigham and Women's Cancer Center and professor of neurology at Harvard Medical School. “This paper, together with another paper in the same issue of Nature, suggest that synapses exist between neurons and glioma cells involving AMPA receptors. This neuron-glioma interaction is bidirectional; neuronal activity can induce glioma growth, while gliomas can increase neuronal activity.
“This discovery has important potential therapeutic implications. Inhibition of AMPA receptors with drugs such as the FDA approved antiepileptic, perampanel, appear to inhibit tumor growth and may have therapeutic potential. In addition, seizures may potentially exacerbate tumor growth, suggesting that more rigid control of seizures may be important in glioma patients.”
“The important thing is the field in general hasn't thought enough about the tissue context in shaping glioma pathologies, particularly the brain's unique electrophysiological environment,” said Paul Mischel, MD, distinguished professor of neurology at University of California, San Diego (UCSD) and a member of the Ludwig Institute for Cancer Research at UCSD. “This study sheds new light on how the electrophysiology of the brain through synapses might be involved. Future studies will shed much more needed light on how to leverage these findings for the benefit of patients.”
Tracy T. Batchelor, MD, MPH, chair of the department of neurology at Brigham and Women's Hospital and Miriam Sydney Joseph professor at Harvard Medical School, added: “This is elegant science that has identified yet another essential element of the tumor microenvironment—the neuronal milieu in which gliomas occur and grow. It is pathbreaking work that continues to get stronger with each study published. While these are early days in understanding these connections, it offers us unique angles to attack gliomas.”
Dr. Monje serves on the scientific advisory board of Cygnal Therapeutics. Dr. Wen has disclosed grants, research, and clinical trial support from Agios, AstraZeneca, Beigene, Eli Lilly, Immunocellular Therapeutics, Kadmon, Karyopharm, Kazia, Lilly USA, Merck, Novartis, Oncoceutic, and Vascular Biogenics. He has also disclosed participation in a speaker's bureau for Merck. He has receipt of consulting fees from Agios Pharmaceuticals Inc, Novartis, Roche, and Taiho Oncology, and participation on advisory boards for Abbvie, AstraZeneca, Eli Lilly, Genentech/Roche, GW Pharmaceuticals, Immunomic Therapeutics, Kadmon, Merck, Puma, Taiho Oncology, Vascular Biogenics, and Ziopharm. He also disclosed participation on data safety monitoring boards for Monteris and Tocagen. Dr. Mischel is a co-founder of Pretzel Therapeutics, has equity in the company, and serves as a consultant. Dr. Batchelor has received honoraria from Champions Biotechnology, GenomiCare, Imedex, Merck, NX Development Corp, and UpToDate; has served in a consulting or advisory role for Amgen, GenomiCare, Merck, N NX Development Corp XDC; and has been reimbursed for travel, accommodations, or other expenses by GenomiCare and Merck.