Cyclofluidic Ltd. has published proof-of-concept data showing that its automated microfluidic lead optimization platform can identify new BCR-ABL tyrosine kinase inhibitors.1 The company's next step is showing that automated hit-to-lead optimization will work with other targets and compound classes.

Cyclofluidic, a joint venture between Pfizer Inc. and UCB Group, was formed in 2008 to develop and commercialize drug screening technology centered around a microfluidic robot that integrates the four steps of traditional SAR studies into a single automated cycle (see "Automated SAR studies"). The biotech also received an undisclosed amount of seed funding from the U.K.'s Technology Strategy Board.

"We have integrated the hardware and software to do synthesis, purification, bioassay and analysis," said team leader and CTO Christopher Selway. "We have a very fast cycle time from synthesis to screening. Once you have your biological data, the system is ready to start on the next compound. We can go around this loop every couple of hours, very quickly building up SAR data. What takes a typical CRO six to eight weeks, we can do in a matter of hours."

Although academic and industry researchers have previously developed semiautomated methods for conducting SAR studies, the new platform is the first working demonstration of a fully integrated system for automatic hit-to-lead optimization.

For this proof of concept, Selway's team started with an already well-characterized and biochemically well-behaved compound class-kinase inhibitors.

Selway's team used the Cyclofluidic platform to find extremely potent derivatives of Iclusig ponatinib, a pan-BCR-ABL tyrosine kinase inhibitor that is marketed by Ariad Pharmaceuticals Inc. to treat certain refractory leukemias.

Starting with a library of 10 ponatinib-like templates modified to accept any of 27 substituent groups, the team used computational SAR models to predict 90 new compounds that had the potential to inhibit BCR-ABL at least as potently as the parent compound.

The 90 compounds were synthesized, purified and assayed by the microfluidic apparatus. In line with the computational predictions, four of the compounds had subnanomolar IC50 values for BCR-ABL, on par with ponatinib's previously reported potency.

Like ponatinib itself, the four optimized leads also were effective against a variety of mutant forms of BCR-ABL that are resistant to other inhibitors such as Novartis AG's Gleevec imatinib.

Results were reported in the Journal of Medicinal Chemistry.

"In industry, we crank out whole arrays of compounds, so automating the assaying makes sense. They have the assay in line, which is unusual, then they have an algorithm that helps decide what the next series will be," said Derek Lowe, research fellow at Vertex Pharmaceuticals Inc.

Lowe said it is not clear how the Cyclofluidic approach would work beyond kinase inhibitors or whether the compounds coming out of the optimization are necessarily better drugs than the starting compounds.

Because there already is a wealth of knowledge about how kinase inhibitors bind to their targets and ponatinib is already a potent BCR-ABL inhibitor, Lowe said the study effectively started with a stacked deck.

"The automation worked especially well because all kinase inhibitors have similar binding modes, so you can extrapolate a little more," said Lowe. "With a lot of other target classes, the binding modes are unknown because you don't have the structure or you can have multiple modes, so you can't predict how those compounds might bind."

Thus, Lowe said the next step for the Cyclofluidic team should be to demonstrate similar results with nonkinase targets.

Beyond the kinase kingdom

Selway said the platform is indeed best suited for finding small molecule inhibitors of soluble enzymes such as kinases. He added that the company is now working on broadening the platform's range of SAR prediction algorithms and synthesis technologies to cover other chemical classes and targets. His team also is expanding the platform's repertoire of assays with the eventual goal of screening compounds in cell culture.

Selway said that the BCR-ABL compounds are purely proof of concept for the Cyclofluidic platform and will not be advanced further in development.

Cyclofluidic has a fee-for-service business model, and its platform is "open to any company to use on a commercial basis," said Selway. Since the platform's commercial launch in mid-2012, the company has signed four deals with undisclosed pharmas for pilot studies with undisclosed targets.

Lowe cautioned that even if Cyclofluidic can turn around SAR studies faster than human chemists, beating the price of conventional CROs will be challenging. He said the high upfront development cost of customization for different types of targets and assays means the technology is unlikely to be cost effective for now.

Selway countered that the rapid run time and flexibility of the platform could provide savings. He said the company has an 18-month cash runway and hopes to raise additional venture capital to support ongoing development of the platform.

The leads obtained in this study have not been patented.

Osherovich, L. SciBX 6(13); doi:10.1038/scibx.2013.304 Published online April 4, 2013


1.   Desai, B. et al. J. Med. Chem.; published online Feb. 26, 2013; doi:10.1021/jm400099d Contact: Christopher N. Selway, Cyclofluidic Ltd., Welwyn Garden City, U.K. e-mail:


      Ariad Pharmaceuticals Inc. (NASDAQ:ARIA), Cambridge, Mass.

      Cyclofluidic Ltd., Welwyn Garden City, U.K.

      Novartis AG (NYSE:NVS; SIX:NOVN), Basel, Switzerland

      Pfizer Inc. (NYSE:PFE), New York, N.Y.

      Technology Strategy Board, Swindon, U.K.

      UCB Group (Euronext:UCB), Brussels, Belgium

      Vertex Pharmaceuticals Inc. (NASDAQ:VRTX), Cambridge, Mass.