DNA Nanomachine Inside Living Cells Measures How Aggressive a Cancer Is

Why It Matters

By delivering real‑time, quantitative measurement of oncogenic microRNAs inside intact cells, the TW‑harvester could enable early assessment of tumor aggressiveness and guide personalized therapy. Its durability and high specificity overcome long‑standing barriers to intracellular diagnostics, paving the way for clinical translation.

Key Takeaways

• TW‑harvester uses three‑wheel DNA nanomachine on gold nanoparticle
• Detects miR‑21 with 5.4 pM limit of detection
• Remains functional for >20 hours inside tumor cells
• Shows four‑fold stronger signal than FISH in MCF‑7 cells
• Can multiplex miR‑21 and miR‑125b without cross‑talk

Pulse Analysis

DNA walkers have long promised nanoscale diagnostics, yet early designs suffered from rapid detachment and enzymatic degradation, limiting their clinical relevance. Recent advances in DNA origami and nanoparticle scaffolds improved stability, but multi‑legged walkers still traded speed for attachment. The TW‑harvester resolves this trade‑off by integrating three sequentially locked wheels onto a rigid tetrahedral framework anchored to a gold nanoparticle, ensuring continuous motion while preserving catalytic activity inside the cytoplasm.

The platform’s performance metrics set a new benchmark for intracellular sensing. A limit of detection of 5.4 pM for miR‑21—nearly 500‑fold lower than prior walker assays—combined with a half‑life exceeding 20 hours under nuclease stress enables prolonged monitoring of tumor biomarkers. Fluorescence intensity surpasses traditional fluorescence in situ hybridization by fourfold, and the linear response spans four orders of magnitude, delivering quantitative data that correlate with qPCR (R² = 0.995). Moreover, the system’s modular design allows rapid reconfiguration for additional microRNAs, as demonstrated with simultaneous miR‑125b detection, without cross‑interference.

Clinically, the ability to quantify microRNA levels directly within living cells offers a powerful tool for assessing malignancy, predicting metastasis risk, and tailoring treatment strategies. By coupling tumor‑specific aptamer targeting with intracellular activation, the TW‑harvester minimizes off‑target signals, a critical requirement for patient safety. Future work will focus on in vivo validation, scaling manufacturing, and integrating the nanomachine with point‑of‑care imaging platforms, potentially transforming precision oncology diagnostics.

DNA nanomachine inside living cells measures how aggressive a cancer is

Feb 14 2026 · By Michael Berger

Detecting a cancer‑linked microRNA inside a living cell, without destroying the cell or losing the signal in biochemical noise, remains one of diagnostic medicine’s most persistent unresolved challenges. Synthetic DNA walkers were supposed to help.

Modeled after the protein‑based motors that shuttle cargo through cells along filament highways, these nanoscale devices travel predefined molecular tracks, snipping fluorescent tags off substrates as they go and accumulating a glow proportional to how much of a target biomarker is present. The concept, first demonstrated in the mid‑2000s, generated enormous excitement. A programmable, self‑powered diagnostic machine small enough to operate inside a single cell seemed within reach.

The results disappointed. The devices kept falling off their tracks. A one‑legged walker, the simplest design, would detach after just a few steps and drift uselessly into the surrounding solution, taking its diagnostic potential with it. Engineers responded by adding two, three, or more legs to raise the energy barrier against detachment. But each extra limb slowed the machine down, because coordinating multiple legs on a nanoscale track requires thermodynamic compromises that sap speed.

Two‑dimensional DNA origami platforms and one‑dimensional DNA footpaths offered limited real‑estate and degraded quickly inside cells, where nuclease enzymes dismantle unprotected DNA in minutes. Gold nanoparticles, with their large surface area and natural resistance to enzymatic attack, provided better three‑dimensional scaffolds.

Still, no design had managed to keep a multi‑legged walker both firmly attached and fast enough to generate a strong, clinically useful signal over the hours‑long timescales that intracellular diagnostics demand.

A team at Wenzhou Medical University and Fuzhou University now reports a device that addresses this problem directly. Their study, published in Advanced Functional Materials (“Non‑Derailed DNA Harvester for Persistently Harvesting Information on Malignant Degree of Tumor Cells via High‑Confidence Quantification of Biomarkers”), describes a three‑wheel driving DNA harvester, or TW‑harvester, that stays locked to its spherical gold‑nanoparticle track while moving quickly enough to produce robust fluorescence inside tumor cells.

Figure 1. The molecular mechanism for the operation of the TH‑AT system and its application in accurate cellular imaging.

(a) Schematic illustration of miRNA‑initiated movement of the TH‑AT system along the track consisting of surface‑confined substrates. The capital letter “L” indicates that the Wheel strand is locked by the locking strand, and DNA TW nano‑harvester is kept in a mechanically‑locked state (highlighted in grey dotted circle), while “S” denotes that the Wheel strand is unlocked via the competitive hybridization of miRNA for the locking strand. When three wheels are unlocked, the DNA harvester is in a started state (highlighted in red dotted circle) and can freely move due to substrate cleavage. “1”, “2”, and “3” mark the serial number of each wheel. (b) Schematic illustration of the membrane‑receptor/aptamer interaction‑mediated internalization of the TH‑AT system and subsequent intracellular miRNA‑initiated movement of TW‑harvester, achieving specific imaging of living cells.

The design centers on a DNA tetrahedron, a rigid cage‑like structure self‑assembled from four DNA strands. Three of its four vertices carry “wheel” strands that grip the nanoparticle track. The fourth vertex holds an AS1411 aptamer, a short DNA sequence that recognizes nucleolin, a protein overexpressed on the surface of many cancer cells but largely absent from healthy ones.

This selectivity allows the aptamer to steer the entire assembly toward tumor cells while ignoring normal tissue, promoting uptake through clathrin‑mediated endocytosis, a process in which the cell membrane folds inward to engulf the nanoparticle. Once inside, the system escapes from lysosomes to reach the cytoplasm.

Each wheel begins locked. A complementary strand holds it in place and keeps its associated catalytic element, a split DNAzyme, inactive. Activation requires a specific intracellular trigger: miR‑21, a microRNA whose elevated levels correlate with tumor proliferation and aggressiveness. When miR‑21 binds to a short exposed “toehold” region on the locking strand and peels it away, the freed wheel pulls together the two halves of the split DNAzyme into a functional catalytic core.

That core cleaves a fluorescently labeled substrate on the nanoparticle surface, releasing a green fluorescent tag. The wheel advances to the next substrate, and the cycle repeats.

Derailment is prevented by sequencing. The three wheels take turns rather than moving simultaneously, so at least two remain anchored at any moment. The tetrahedron’s rigidity keeps them far enough apart that they never compete for the same substrate or collide during movement. The result is continuous, stable travel without the speed penalty of earlier multi‑legged designs.

Durability is also critical for intracellular work, because nuclease enzymes inside cells constantly attack exposed DNA. The combined shielding of the gold nanoparticle surface and the DNA tetrahedron framework gives the TW‑harvester a half‑life of 20.38 h, roughly 100 times longer than unprotected DNA wheel strands exposed to DNase I at 1 600 U/L. This exceptional resistance to degradation allows the device to keep operating for extended periods rather than being dismantled shortly after entering a cell. Fluorescence imaging confirmed a strong signal persisting for at least 10 h after cellular uptake.

In controlled laboratory tests, the device achieved a limit of detection of 5.4 pM for miR‑21, approximately 490‑fold lower than previously reported DNA‑walker assays. Its linear quantitative range covered over four orders of magnitude, from 5.4 pM to 50 nM. Non‑target microRNAs produced near‑zero signal, and even single‑base mismatches were clearly distinguished from the true target. Cell‑imaging sensitivity significantly surpassed the well‑established fluorescence in situ hybridization (FISH) technique, producing roughly four‑fold stronger signal in MCF‑7 breast‑cancer cells.

Biological validation spanned three cell types. MCF‑7 cells, rich in both nucleolin and miR‑21, produced the strongest fluorescence. HeLa cervical‑cancer cells, with moderate expression, gave an intermediate signal. Normal L02 liver cells, which express very little nucleolin, generated almost no fluorescence, confirming that the aptamer‑based targeting effectively excludes healthy cells from the diagnostic readout. These results matched quantitative PCR measurements, confirming that the harvester’s output faithfully mirrors actual intracellular microRNA levels.

To test whether that output could predict tumor behavior, the researchers manipulated miR‑21 levels in HeLa cells using miR‑21 mimics and inhibitors. Cells with boosted miR‑21 showed reduced PDCD4, a tumor‑suppressor protein, alongside faster proliferation and quicker wound closure in standard migration assays. The TW‑harvester’s fluorescence tracked these changes with a correlation coefficient (R²) of 0.9953 against qPCR data, indicating the device can quantitatively estimate malignancy from a single microRNA measurement.

Adaptability broadens the platform’s scope. By swapping locking strands and substrates, the team reconfigured the harvester to detect miR‑125b, a biomarker linked to liver cancer, and showed that both microRNAs could be measured simultaneously in one reaction without cross‑interference.

The significance of this work lies in the convergence of several performance metrics within a single platform: picomolar sensitivity, high specificity, prolonged intracellular operation, sequential two‑biomarker recognition, and quantitative agreement with conventional reference assays. The ability to measure not merely the presence of a cancer marker but the degree of tumor aggressiveness positions the TW‑harvester as a candidate tool for predicting proliferation, metastasis risk, and treatment response.

The authors note broad application prospects in cancer management. Translating these cell‑culture results into clinical practice will require systematic testing in animal models and patient‑derived samples, the standard path that separates a promising laboratory demonstration from a working diagnostic technology.