Neurons Donate Mitochondria to Cancer Cells to Fuel Metastasis
A Nature study reveals neurons physically transfer mitochondria to breast cancer cells via tunneling nanotubes, boosting metastatic potential.
Summary
Researchers at the University of South Alabama discovered that neurons in and around tumors physically transfer their mitochondria to breast cancer cells, dramatically boosting the cancer cells' energy production and ability to spread. Using a novel genetic tracking tool called MitoTRACER, the team traced cancer cells that received neuronal mitochondria and found they were selectively enriched at metastatic sites. Denervating tumors with botulinum toxin reduced cancer cell mitochondrial content and suppressed metabolic activity. The transfer occurred primarily through tunneling nanotubes requiring direct cell contact. These findings reveal a previously unknown mechanism by which the nervous system fuels cancer progression and suggest nerve-cancer metabolic crosstalk as a potential therapeutic target.
Detailed Summary
Cancer cells are metabolically flexible, switching between glycolysis and oxidative phosphorylation (OXPHOS) depending on their environment. This metabolic plasticity is increasingly recognized as essential for metastasis, yet the non-cell-autonomous mechanisms driving it remain poorly understood. This landmark Nature study demonstrates that neurons physically donate mitochondria to breast cancer cells, providing a direct bioenergetic boost that enhances metastatic capacity — a finding with profound implications for cancer biology and treatment.
The researchers used two complementary breast cancer denervation models: the aggressive 4T1 triple-negative breast cancer (TNBC) mouse model and a human ductal carcinoma in situ (DCIS) xenograft. Tumors were chemically denervated using botulinum neurotoxin type A (BoNT/A). Transcriptomic profiling of cancer cells from denervated versus control tumors revealed distinct gene expression signatures, with Gene Ontology analyses showing predominant downregulation of metabolic processes. In the DCIS model, gene set enrichment analysis identified the tricarboxylic acid (TCA) cycle as the most suppressed pathway after denervation. Critically, denervation reduced the incidence of invasive lesions from 55% in controls to just 12% in denervated mice — a striking 78% reduction in invasive progression.
To investigate the nerve-cancer interface mechanistically, the team developed an in vitro coculture system pairing 4T1 cancer cells with subventricular zone neuronal stem cells (SVZ-NSCs). Cancer cells stimulated rapid neuronal differentiation of SVZ-NSCs, and FACS-isolated 4T1 cells from coculture showed significantly increased OXPHOS capacity by Seahorse assay. Time-lapse confocal microscopy and flow cytometry directly visualized mitochondrial transfer from SVZ-NSCs expressing a mitochondria-targeted GFP (CCO-GFP) to mCherry-labeled 4T1 cells. Transfer rates reached 23.04% under direct contact conditions versus only 0.59% in Transwell (no-contact) conditions (p<0.0001, n=6), confirming that transfer requires physical cell-cell contact. Cytochalasin B, an actin polymerization inhibitor, significantly reduced transfer (p=0.001, n=3), implicating tunneling nanotubes as the primary transfer mechanism. Neuronal cells transferred mitochondria at higher rates than mouse embryonic fibroblasts, suggesting cell-type specificity.
To definitively prove functional mitochondrial transfer, the team depleted 4T1 cells of mitochondrial DNA (mtDNA), creating ρ0 cells that cannot perform OXPHOS and require uridine supplementation for survival. When ρ0 4T1 cells were cocultured with SVZ-NSCs, they progressively reacquired mtDNA (confirmed by PCR), restored normal mitochondrial morphology (confirmed by MitoTracker imaging), regained uridine-independent growth, and recovered OXPHOS capacity and proliferative ability — all hallmarks of functional mitochondrial rescue.
The team then developed MitoTRACER, a genetic reporter system that permanently labels cancer cells receiving mitochondria from donor cells, enabling lineage tracing of recipient cells and all their progeny in vivo. Fate mapping of cancer cells that acquired neuronal mitochondria in primary tumors revealed their selective enrichment at metastatic sites following dissemination. In human tissue, multispectral imaging with machine-learning deconvolution showed increased mitochondrial mass in metastatic cancer cells, and perineural invasion was associated with higher mitochondrial content in cancer cells near nerves. BoNT/A denervation in vivo confirmed reduced mitochondrial load in cancer cells. Together, these findings establish nerve-to-cancer mitochondrial transfer as a key mechanism supporting cancer metabolic plasticity and metastatic dissemination.
Key Findings
- Denervation with BoNT/A reduced invasive lesion incidence from 55% to 12% in a human DCIS xenograft model — a 78% reduction in invasive progression
- Direct nerve-cancer contact enabled mitochondrial transfer at a rate of 23.04% vs. only 0.59% in no-contact Transwell conditions (p<0.0001, n=6 independent cocultures)
- Cytochalasin B (actin polymerization inhibitor) significantly reduced mitochondrial transfer (p=0.001, n=3), confirming tunneling nanotubes as the primary transfer mechanism
- ρ0 breast cancer cells (devoid of mtDNA) cocultured with neurons progressively reacquired mtDNA, restored OXPHOS capacity, and regained proliferative ability — confirming functional mitochondrial rescue
- MitoTRACER lineage tracing showed cancer cells that received neuronal mitochondria in primary tumors were selectively enriched at metastatic sites after dissemination
- Neuronal cells transferred mitochondria at higher rates than mouse embryonic fibroblasts, indicating cell-type specificity of the transfer mechanism
- Transcriptomic profiling of denervated tumors identified the TCA cycle as the most suppressed pathway, with broad downregulation of mitochondrial metabolic gene sets
Methodology
The study used 4T1 TNBC mouse models and human DCIS xenografts with BoNT/A-mediated chemical denervation, SVZ-NSC and 50B11-DRG peripheral neuron coculture systems, Seahorse metabolic flux assays, time-lapse confocal microscopy, and FACS-based cell isolation. The team developed MitoTRACER, a novel genetic reporter enabling permanent lineage tracing of mitochondria-recipient cancer cells in vivo. Statistical methods included Student's two-tailed unpaired t-tests and two-way ANOVA; sample sizes ranged from n=3 to n=6 independent cocultures per experiment. Human tissue was analyzed using multispectral imaging with machine-learning deconvolution to quantify mitochondrial mass in cancer cells near nerves.
Study Limitations
The primary in vivo models are mouse-based (4T1 TNBC and DCIS xenografts), and while human tissue data support the findings, direct causal evidence in human patients is lacking. The study focuses on breast cancer, and it is unclear how broadly these mechanisms apply across other cancer types with varying innervation patterns. The paper does not fully characterize which specific neuronal subtypes are most active in mitochondrial transfer, nor the precise molecular machinery governing tunneling nanotube formation at the nerve-cancer interface.
Enjoyed this summary?
Get the latest longevity research delivered to your inbox every week.