The new sensor, made by Vincent Rotello and colleagues, consists of a gold nanoparticle (AuNP) bound to three fluorescent proteins – one red, one green and one blue. When bound to the nanoparticle, the proteins’ fluorescence quenches. When the nanoparticle-protein complex is then added to a solution of human cells that have been treated with drugs the gold nanoparticle binds to the cell surface and displaces the proteins – which subsequently light up again.

“By looking at the fluorescence pattern generated, we can observe changes in the cell surface that indicate specific drug action mechanisms,” Rotello tells “Our nanosensor can be used for all types of cells and the technique does not require any other processing steps before analysis, which greatly speeds up the time to readout a drug’s mode of action compared to conventional assays.”

The Amherst team began by first non-covalently conjugating a benzyl headgroup-terminated AuNP (called BenzNP) with the three fluorescent proteins EBFP2 (blue), EGFP (green) and tdTomato (red). The researchers were already familiar with BenzNP, having studied it in previous experiments to profile cell surface phenotypes.

Fluorescence "turns" on and off

In their new BenzNP-fluorescent protein supramolecular complexes (which they called BenzNP-FP), the cationic AuNP strongly binds to the anionic fluorescent proteins. It is this mechanism that quenches the proteins’ fluorescence. When the complex is then exposed to biological cells, it binds to the cells and displaces the FPs from the AuNP surface. And fluorescence is “switched on” again.

“This florescence ‘turn on’ varies considerably depending on the drugs that have been used to treat the cell surfaces,” explains Rotello.

“A key issue in designing our sensor was to select the appropriate FPs from the broad range available,” he adds. “Thanks to comprehensive tests on different FP variants, we finally chose a three-colour FP, optimized to bear a net negative charge and which has minimum spectral ‘crosstalk’ (that is, with well separated excitation and emission spectra).”

Another important point was to choose FPs that could reversibly bind to the AuNP. The researchers say that they found the best FPs for the job by undertaking detailed fluorescence quenching studies that allowed them to determine how stable the complexes were.

Testing out the technique

Rotello’s team tested out its technique using a straightforward protocol on several routinely employed anti-cancer drugs, whose mode of action is well known. For example, Paclitaxel, which works by disrupting cell mitosis, Thio-TEPA, which alkylates DNA, and MG-132, which degrades cell proteins.

“On interacting with the drug-treated cells, our sensor generated characteristic fluorescence fingerprints for the three FPs we studied, says Rotello.

“Application-wise, this platform might be used to rapidly screen therapeutics and drug ‘cocktails’ – especially in the emerging field of personalized medicine. The method could also be a way to quickly characterize potential environmental hazards – for example, the 70,000 or so industrial chemicals that we yet do not have any available toxicology data for.”

The researchers say that they are now actively looking to translate the technology into a commercially useful platform, for both pharmaceutical and environmental applications.

The paper describing the research is published in Nature Nanotechnology doi:10.1038/nnano.2014.285.