Developing drugs – with tens of thousands of miniscule droplets on a small glass plate

The entire droplet production process is automated by way of a compact device the size of a microscope that can also perform microscopy, cultivate cells and automatically change samples. "It used to take half an hour to set everything up correctly," Breitfeld relates. "Today, thanks to our automation, all it takes is the push of a button and the experiment is up and running," says Maximilian Breitfeld, a scientist in Dittrich's research group. What's more, the combination of precise control and physical oil protection allows the composition of the droplets to be varied in a targeted manner - sometimes with more, sometimes with less active ingredient, resulting in fine concentration gradients. Researchers can make use of this to observe enzymes over a longer period of time, for example, or to investigate the effect of active ingredients on cells in highly parallelised tests. The droplets are then analysed either by fluorescence microscopy or mass spectrometry, allowing enzyme reactions to be tracked precisely.

When hours become minutes

The basic idea stems from earlier work performed by the group in which they already generated droplets for screening applications. "We knew how the technology worked in principle, but it was too slow to be competitive in practice," says Dittrich. The former doctoral students Maximilian Breitfeld and Claudius Dietsche took the decisive steps: their newly developed process not only massively accelerates drop generation but also automates the entire process.

ETH Zurich has patented the process and nominated it as a finalist for this year's Spark Award.

Managing a huge amount of data

The method's tremendous potential also entails a new challenge. "We generate a huge amount of data," says Dittrich. "It's no longer possible to evaluate this manually - we need software solutions that help us analyse the information in a meaningful way."

While the amount of data is rapidly on the rise, resource consumption remains surprisingly low: up to five kilograms of non-recyclable plastic can be saved per experiment, the oil used can be collected, while chemical consumption is drastically reduced compared to conventional methods, given that an entire experimental run requires only microlitres instead of litres of reaction medium.

Nevertheless, the method does have clear limitations. Whereas the tiny droplets are perfect for rapid reactions in small volumes, as is common in microfluidics, the new method is not suitable for larger volumes of liquid or tissue cultures that grow over several weeks.

From research to spin-off

The researchers are now planning a spin-off to launch the method on the market. They plan to sell a complete system consisting of glass plates, equipment and software - and biological tests based on it as an option.

"For me, it is crucial that the system is genuinely reliable and easy to use," says Dietsche. "It can only be used beyond our research laboratory if we can guarantee its user-friendliness." Demand is already evident the researchers report, and the nomination for the Spark Award is lending them additional momentum for the planned spin-off.