“In early drug discovery, typical chemistry workflows synthesize about 20 000 times more chemical material than researchers need for initial profiling,” says Jonathan Grob, a research investigator at the Novartis Institutes for BioMedical Research (NIBR). “That’s one example of where the scales of biology and chemistry can be better aligned – we also wanted to further accelerate drug discovery.”

The chemistry workflows used today date back to the 1950s – a time when investigational molecules were often directly tested in animals. In that setting, researchers needed grams of a drug candidate early in the process. But today, as a consequence of the molecular biology revolution of the 1980s and 1990s, researchers initially test new molecules in cells or in biochemical screens before moving to in vivo testing in animals. These tests need only nanograms of each drug candidate – but due to technical limitations, drug batch sizes have not shrunk accordingly.

Today’s standard drug discovery process starts with testing millions of compounds in plastic plates to find chemical starting points to engage the biological target, generating tons of plastic waste and hundreds of liters of hazardous liquids. The chemical starting points are then optimized in flasks and vials producing a surplus of drug compounds beyond what is required for initial biological testing.

Beyond the inefficiency of producing far more than is needed, making and testing new molecules takes weeks. New drug candidates must be synthesized, purified and assessed for their chemical and biological properties – and each step is completed in separate specialized labs. As a result, for each new idea, sometimes waiting times for teams can amount to weeks before they have compiled all the data and know whether they are on the right track. Because of the time and expense invested in each round of optimization, scientists are also limited in the number of hypotheses they can test.

But what if this could all be done faster, with fewer resources and less waste? “The opportunity to run smaller-scale chemistry gave us the chance to ask all of the questions a team would want to know – and to have a huge impact on hazardous-waste generation and energy use too,” says Grob, who in 2017, together with his colleague, Alexander Marziale, and others founded the MicroCycle team. It became part of the Novartis Genesis Labs initiative, which gives in-house researchers start-up conditions to pursue unconventional ideas during a period of 18 months.

“The key motivation was to push down the cycle time from idea to data and to allow that iterative process to run faster and more efficiently,” explains Grob. “We wanted to develop a platform that allows teams to have more iterations within the regular drug discovery timeline.”

One of the team’s central strategies for reducing the cycle time was to bring all of the early discovery steps together into a single multidisciplinary lab space. In the MicroCycle platform, every step – synthesis, purification, analytics, biological and physicochemical profiling, virtual screening and data science – is no more than a few meters away from the next. This eliminates the handover of compounds between groups and increases the efficiency of the process.

But beyond the necessary technological innovations, the team also had to recruit experts for each specialization, although volunteers weren’t hard to find. “Genesis Labs allows this unique approach to team building that’s a really good way for improving the possibilities of what we could achieve together,” says Grob. “Whenever we asked if people wanted to get involved, the answer was almost always yes!” “It’s been an unprecedented level of engagement and enthusiasm,” adds Marziale.

Learn more about environmental action at Novartis here!

In order to fit everything into a lab space, miniaturization and lab automation have also been critical components of the MicroCycle platform. While biochemical and cellular screens can already be run in miniature, scaling down chemistry was a key innovation. Specifically, the team needed to acquire, adapt and repurpose several technologies for miniaturized synthesis, purification and quantification. “We’ve brought several emerging technologies into chemistry labs and we’re contributing to an orientation towards miniaturization, lab automation and digitalization of drug discovery,” says Alexander Marziale. “We’re now doing drug discovery on a scale that is about 100-fold smaller than researchers have previously been able to do in this early stage of the process.” Since its launch, MicroCycle has already contributed to over 10 NIBR portfolio projects.

Another potential revolution in the realm of drug discovery is Mic-Drop, which “enables new projects by finding the needle in the haystack – for example, promising new chemical starting points,” says Ken Yamada, a medicinal chemist at NIBR.

Since the 1990s, so-called high throughput screening (HTS), the industry standard, has enabled the rapid biochemical and cellular testing of millions of different molecules. These screens use plastic plates with tiny wells that each contain an experimental sample of only a few microliters in volume. Even so, after tens of thousands of plates, this adds up to hundreds of liters of hazardous waste and tons of plastic, which cannot be recycled due to the possibility of toxic contamination.

Mic-Drop aims to miniaturize screening to the size of a single chip that fits in the palm of your hand. By miniaturizing compound loading and biological assays to the size of a single cell, the platform will make it possible to perform screens with patient-derived cells that are scarce and have been out of reach in traditional screening paradigm.

Mic-Drop began as a conversation between Ken Yamada and Piro Siuti, a synthetic biologist at NIBR who had done his Ph.D. in microfluidics. Yamada was looking for a rapid way to perform multiplexed screens against large compound libraries and droplet microfluidics – the precise control of microscopic droplets – offered the possibility of miniaturizing standard biological tests by up to 100 000-fold.

“In academia, microfluidics applications were never more than a Proof-of-Concept,” says Siuti. “I was really interested to see if microfluidics could be applied in pharma.” Microfluidics have not yet been successfully applied to drug discovery, mostly because of the need for a diverse pool of talent and expertise. But this technology has the potential to miniaturize biochemical and cellular experiments into droplets just one-tenth of a millimeter in diameter.

“The Mic-Drop platform we’re developing makes it possible to reduce reagent use 100 000-fold to allow screens with more biologically relevant cells that are rare or expensive, which also results in one million-fold less liquid waste,” says Ken Yamada. “Whereas a typical biological screen against a million compounds uses approximately 1000 plastic screening plates, we can do the same screen in just one chip that fits in the palm of your hand.”

In theory, Yamada and Siuti knew that microfluidics could potentially make it possible to store and measure billions of microscopic droplets – each containing a unique experiment – in a total volume of just a few milliliters. The droplets could be created and kept separate from each other by using a solution of oil. This ability to mix the droplets together would also be the key to screening faster.

As they developed their proposal, Yamada and Siuti were joined by over a dozen colleagues in Basel, Cambridge and at the Genomics Institute of the Novartis Research Foundation, which had related side projects. They had all seen potential synergies with microfluidic technologies, but none of them had ever been able to dedicate more than a few spare moments to their ideas.

“It was interesting to find out that there were people around Novartis who were interested in these approaches and who had been working on them on the side,” says Siuti. “But nothing was organized until the Genesis Labs project – that was what ignited it and put it all together.” The cross-functional team thus formed taps into the vast diversity of expertise available at Novartis, from engineering to chemistry, biology, informatics and analytical sciences.

In addition to microfluidics, a central component of the Mic-Drop concept is an effective method for identifying droplets of interest after millions of droplets have been mixed together. During the biological measurement, only a few molecules – and therefore a few droplets – show activity, for example by lighting up with fluorescence. Droplets that light up can thus be sorted from those that do not. The team then uses a unique barcoding strategy to identify the molecules in the illuminated droplets.

“When we started out, we were wondering if we could actually do this, but with the team fully equipped with the right talents and resources to realize this dream platform, we’ve been able to prove that it’s definitely feasible,” says Yamada. “We’ve validated all of the individual pieces and now we just need to put them all together.”

Moving towards efficiency and sustainability

As of early 2020, the Mic-Drop prototype has reduced waste and resource use just as much as anticipated. The platform is on track to perform previously week-long screens in only a few hours. Like Micro-Cycle, the Mic-Drop team hopes that their platform will become an official ongoing Novartis project as well.

“We want to make the Mic-Drop platform as broadly applicable as possible,” says Piro Siuti. “The Genesis Labs program has broadened our audience and we have many people contacting us now to collaborate.” Ken Yamada adds that interest is now so big, “we have to prioritize which projects to start with.”

Meanwhile, the MicroCycle team has since created two platforms – one for each of the Basel and Cambridge campuses. Each lab contains all of the instrumentation needed for the complete workflow, as well as experts for each step of the process. “MicroCycle isn’t a replacement for existing approaches, but it offers a new tool in addition to the big toolkit that’s available for drug hunters in this organization,” says Alexander Marziale. “We’re also continuously improving the platform, adding new components and deepening our collaborations with project teams.”

The MicroCycle team sees potential for improving the sustainability of the platform too. “So far, we’ve only gone down 100-fold in scale,” says Jonathan Grob. “There’s still a 200-fold opportunity to go further.”

While highly impactful, both Mic-Drop and MicroCycle are active in niche areas so far. But further innovations are already underway to expand the chemical and assay space for both MicroCycle and Mic-Drop to deepen their impact across the research portfolio of Novartis.