Human Cells Can Pass Damaged DNA Directly to Neighbors Through Nanotube Bridges
A landmark Cell study reveals cells transfer cytoplasmic DNA fragments to neighbors via nanotubes, stably reshaping recipient genomes.
Summary
Researchers at UT Southwestern discovered that when human cells suffer genomic damage — from radiation, drug treatment, or CRISPR cutting — fragments of chromosomal DNA that escape into the cytoplasm can travel through nanotube-like connections directly into neighboring cells. These transferred DNA fragments are not degraded; they persist and function in recipient cells, even conferring drug resistance. The process works across multiple cell types, including normal and cancerous cells, and requires direct physical contact. This finding introduces a horizontal gene transfer-like mechanism in mammals, suggesting that genome instability can spread non-cell-autonomously between cells sharing a tissue environment.
Detailed Summary
For decades, the mammalian genome was considered strictly cell-autonomous — confined within the nucleus of a single cell and isolated from its neighbors. This landmark study published in Cell overturns that assumption by demonstrating that genomic instability can trigger the physical transfer of nuclear DNA fragments from one human cell into the genome of an adjacent cell, via microscopic nanotube connections. The implications are profound: genome damage is not just a private cellular crisis — it can propagate laterally across a tissue.
The research team at UT Southwestern used hTERT-immortalized human retinal pigment epithelial (RPE-1) cells and renal proximal tubule epithelial cells (RPTECs), along with HeLa cancer cells, to study this phenomenon. Genomic instability was induced through multiple orthogonal methods: treatment with CENP-E and Mps1 inhibitors to cause mitotic segregation errors, prolonged nocodazole arrest followed by release, Mad2 depletion to accelerate mitotic exit, CRISPR-Cas9-induced chromosomal double-strand breaks on chromosome 3p, and 2 Gy ionizing radiation. In all cases, cytoplasmic DNA — visualized with SiR-DNA dye and fluorescently tagged histone H2B — was observed transiting through nanotube structures connecting adjacent cells.
To definitively prove intercellular transfer (rather than cytoplasmic contamination or bridges), the team co-cultured cells expressing H2B-GFP with cells expressing H2B-mCherry. Successful DNA transfer produced recipient cells with mismatched H2B labeling between nucleus and cytoplasm — a signature that was quantified across 2,867 cells from three independent experiments. Between 1.1% and 3.9% of micronuclei showed mismatched H2B signals, a figure the authors note is an underestimate since same-color transfers are invisible to this assay. Transfer frequency was density-dependent and was completely abolished by a 4-μm pore transwell filter that prevented physical cell-cell contact, confirming the contact-dependent nanotube mechanism.
The nanotubes themselves were characterized as microtubule-rich structures (positive for α-tubulin), distinct from chromatin bridges arising from dicentric chromosomes. Average DNA transport speed within nanotubes was measured at ~390 nm/min. Plasma membrane labeling with CAAX-Halo confirmed that cargo traversed genuine membrane-enclosed nanotube connections. Approximately 26.7% of transferred micronuclei retained Lamin B1 coating, while the remainder were uncoated — indicating both lamina-enclosed and bare cytoplasmic DNAs can be transferred.
Most strikingly, transferred DNA was not merely a passive passenger — it was functionally integrated. Transferred fragments were stably inherited across multiple cell generations as extrachromosomal genetic elements. In a proof-of-concept experiment, the team demonstrated that DNA encoding drug resistance genes was transferred from donor to recipient cells, which subsequently acquired heritable resistance to the corresponding drug — a de novo phenotypic change conferred entirely by horizontal-like DNA transfer. This positions nanotube-mediated DNA transfer as a previously unrecognized mechanism of non-cell-autonomous genome reshaping in mammals, with potential relevance to cancer evolution, tissue aging, and disease progression.
Key Findings
- 1.1%–3.9% of micronuclei in treated co-cultures showed mismatched H2B labeling between nucleus and cytoplasm, indicating successful intercellular DNA transfer (assessed across 2,867 cells from 3 independent experiments)
- DNA transfer frequency was abolished entirely when cell-cell contact was prevented using a 4-μm pore transwell filter, confirming contact-dependent nanotube mediation
- Average speed of DNA transport through nanotubes spanning 10–60 μm was ~390 nm/min (360 nm/min without mitotic inhibitors)
- DNA transfer was triggered by all tested genomic insults: CENP-E/Mps1 inhibition, nocodazole release, Mad2 depletion, CRISPR-Cas9 chromosome 3p breaks, and 2 Gy ionizing radiation
- ~26.7% of transferred micronuclei retained Lamin B1 nuclear envelope coating, indicating both lamina-coated and uncoated cytoplasmic DNAs undergo transfer
- Transferred DNA fragments persisted as functional extrachromosomal elements in recipient cells, conferring de novo drug resistance as a heritable phenotypic trait
- Transfer occurred across multiple human cell lines — RPE-1, RPTEC, and HeLa — demonstrating the mechanism is not cell-type specific and extends to cancer cells
Methodology
This is a mechanistic cell biology study using live-cell time-lapse imaging, immunofluorescence, and fixed co-culture assays in hTERT-immortalized non-transformed human cell lines (RPE-1, RPTEC) and HeLa cancer cells. Genomic instability was induced via pharmacological (CENP-E/Mps1 inhibitors, nocodazole, cytochalasin D), genetic (Mad2 depletion, dominant-negative TRF2, CRISPR-Cas9), and radiation (2 Gy IR) approaches. Intercellular transfer was quantified in 2,867 cells across 3 independent experiments using differentially labeled H2B-GFP/mCherry co-cultures with mismatched cytoplasmic labeling as the readout. Contact-dependency was confirmed with transwell separation controls; nanotube identity confirmed by α-tubulin/β-actin immunostaining.
Study Limitations
The study was conducted entirely in cell culture models and does not yet demonstrate nanotube-mediated DNA transfer in intact tissues or in vivo, limiting direct clinical translation. Transfer frequency measurements are acknowledged as underestimates because same-color H2B transfers are invisible to the dual-reporter assay. The authors do not fully characterize the molecular machinery governing nanotube formation or cargo selectivity, and the rate of functional integration versus degradation of transferred DNA fragments remains to be quantified systematically.
Enjoyed this summary?
Get the latest longevity research delivered to your inbox every week.
