High-Throughput DNA Nanoswitches Accelerate Drug Discovery by Tracking Protein Interactions

The fundamental constraint in developing new therapeutics is not merely finding a molecule that sticks to a target, but finding one that alters how that target behaves. Proteins govern human biology by interacting with one another, and diseases often arise when these complex interactions fail or misfire.

For decades, the pharmaceutical industry has relied on binding-based screening to identify potential drug candidates. This approach tests whether a small molecule attaches to a single isolated protein, leaving the crucial question of its functional impact for later, more expensive stages of clinical testing.

"Determining which molecules interact, and measuring the strength of these interactions is fundamental for many areas of research, from drug discovery to understanding the mechanisms underlying disease," said Wesley Wong, a biophysicist at Boston Children's Hospital and the Wyss Institute.[4]

Conventional high-throughput screening requires robotics to test compounds individually across massive multi-well plates. The process demands expensive optical instruments and significant time, creating a bottleneck that slows the pipeline from initial discovery to clinical trials.

Now, a multi-institutional team has engineered a way to bypass this limitation entirely. By leveraging the predictable geometry of nucleic acids, they have moved the screening process out of the plastic well and onto the molecule itself.

Engineering the FIND Platform

In a study published on November 5, 2026, in the journal Science, researchers unveiled the FIND platform. Developed by scientists at the Wyss Institute at Harvard University, Boston Children's Hospital, and Harvard Medical School, the system directly tracks protein-protein interactions.[1][2][3]

FIND stands for Function-based Identification of Nanoscale Drug candidates. Instead of isolating single proteins, the platform uses programmable DNA nanoswitches to observe how two distinct proteins behave in the presence of a massive library of potential drug compounds.[1][2]

The nanoswitches begin as long, linear strands of single-stranded DNA, constructed using a viral scaffold of exactly 7,249 nucleotides. Researchers attach the two target proteins to specific locations along this flexible DNA backbone using complementary staple strands.[4]

When a candidate drug successfully promotes an interaction between the two target proteins, they bind together. This binding forces the flexible DNA scaffold to bend, reconfiguring the entire complex from a linear strand into a closed loop.[1][4]

"Our idea was to build each experiment onto a single DNA molecule, so that many experiments could run in parallel in one tube," Wong explained, describing the shift from macroscopic wells to nanoscale reactions.[3]

Reading the Shape of DNA

The genius of the FIND platform lies in how easily this nanoscale shape change can be detected. The researchers rely on gel electrophoresis, a foundational laboratory technique that has been used in molecular biology for over 50 years.[1][4]

In gel electrophoresis, an electric field pulls molecules through a porous agarose matrix. Because the linear, unactivated nanoswitches are streamlined, they move rapidly through the gel and travel a significant distance in a short time.[4]

AI

Conversely, the triggered nanoswitches that have folded into loops are physically bulkier. This clunkier shape causes them to drag and move much slower through the porous matrix, separating them entirely from the inactive strands.[4]

Simply by looking at the distinct bands on the resulting gel, researchers can instantly identify which compounds successfully modulated the protein interaction. The distance traveled provides a definitive, binary readout of the drug's functional efficacy.[4]

This physical separation allows scientists to screen large, pooled libraries of molecules simultaneously. Rather than testing 100,000 compounds in 100,000 separate wells, the entire library can be evaluated in a single test tube for less than five cents per sample.[4]

From Small Molecules to Nanobodies

To prove the platform's versatility, the Harvard-led team deployed FIND against several disease-relevant targets. They successfully screened libraries containing both traditional small molecules and nanobodies, which are small, antibody-like proteins gaining traction in modern medicine.[1][2]

The researchers used the DNA nanoswitches to discover compounds that actively promoted specific protein-protein interactions. These interactions are notoriously difficult to target with conventional screening, yet they represent a massive untapped reservoir for therapeutic intervention.[1]

Many of the most effective drugs on the market already operate by modulating protein interactions. The diabetes medication Ozempic mimics the natural GLP-1 hormone to activate its receptor, while inhaled albuterol triggers beta-2 adrenergic receptors to relax airway muscles.[1][2][3]

By identifying molecules that facilitate these complex interactions early in the discovery pipeline, FIND dramatically increases the odds of finding viable clinical candidates. The platform filters out compounds that merely bind without providing a functional benefit.[1][2]

The technology also builds on more than a decade of foundational work in DNA nanotechnology. Since 2014, Wong's laboratory has been refining DNA nanoswitches for various applications, steadily improving their stability, programmability, and ease of use.[4]

Democratizing Drug Discovery

Beyond accelerating the search for new therapeutics, the FIND platform fundamentally alters the economics of drug discovery. By eliminating the need for million-dollar robotic screening facilities, it lowers the barrier to entry for smaller research institutions.[4]

Academic laboratories and startup biotechnology companies can now perform high-throughput functional screens using standard benchtop equipment. The primary requirement is a basic gel electrophoresis setup, which is already present in nearly every molecular biology laboratory worldwide.[4]

The Wyss Institute has actively supported the commercialization of this technology through its Lab-on-a-Molecule project. In 2023, the initiative received backing from Northpond Labs to help de-risk the platform and drive its development toward widespread industry adoption.[1]

Researchers are also exploring how the underlying nanoswitch architecture can be adapted for other uses. Recent studies have demonstrated their potential in point-of-care diagnostics, where the same looping mechanism can detect viral biomarkers or single-gene mutations in patient samples.[4]

The ability to measure multiple distances on natively folded single-molecule complexes represents a paradigm shift for structural biology. By combining reconfigurable nanoscale devices with force-based barcoding, the team has achieved spatial resolutions of 1.5 to 2.5 angstroms, matching X-ray crystallography.[6]

The next critical checkpoint for the technology is its integration into the automated pipelines of major pharmaceutical companies. If the nanoswitches can maintain their fidelity when screening libraries of millions of compounds, the timeline from initial target identification to a viable clinical lead could shrink from years to a matter of months.