Here’s how scientists aim to speed up drug development for treating new viruses

When the COVID-19 pandemic hit, one of the first medicines doctors turned to was remdesivir, part of a family of medicines called nucleoside analogues that are also used to treat HIV, hepatitis and some cancers. These drugs are especially important because there are a surprisingly limited arsenal of drugs that reliably fight viral infections. That’s one of the main reasons why new viral outbreaks cause so much alarm. For nearly a decade, our chemistry research group — The Britton Lab at Simon Fraser University — has been developing faster ways to build nucleoside analogues. In 2020, we published a method that bypassed most of the traditional steps, reducing the time needed to build these kinds of molecules from months to about a week. Now, in a new paper published in the journal Science in collaboration with scientists at biopharmaceutical firm Merck, we took this a step further. We have now developed a modular approach that turns one cheap starting material into a wide range of finished drug candidates within weeks. Using this method, we made dozens of new molecules and, in early laboratory tests, identified three with promising anti-HIV activity. Building nucleosides Nucleosides are the molecular building blocks of DNA and RNA. Nucleoside analogues are carefully designed lab-made lookalikes that trick cells, or the viruses infecting them, into using these fake building blocks instead of the natural nucleosides they rely on. This disrupts how the virus’s genetic material gets copied, stopping the virus from reproducing. More than 30 approved medicines work in this way. The problem is that these molecules are notoriously difficult to make. Traditionally, chemists begin with a naturally occurring sugar molecule, and carry out a long, fixed sequence of chemical reactions — often 15 or more — to produce a single target molecule. If researchers want to test a different version or improve the molecule, it means starting the whole sequence again from the beginning. Early manufacturing of remdesivir was so inefficient that less than two per cent of the starting material ended up as usable drug. The antiviral drug remdesivir was one of the first medications used to treat COVID-19. (Unsplash/Nappy) This one-at-a-time approach clashes with how modern drug discovery works. Researchers typically need to screen thousands of related molecules to find a handful of promising leads, then improve on the best ones. The collection of compounds that led to remdesivir contained only around 1,000 candidates, which is tiny by pharmaceutical industry standards. Even so, assembling that molecular library took many years and the considerable resources of a major pharmaceutical company. Getting ready for the next viral threat means building and testing much larger and more diverse batches of candidate molecules, as quickly as possible. Making molecules in fewer steps Our approach turns the traditional synthetic process on its head. Instead of building every final molecule separately from scratch, we use an amino acid catalyst to build a single versatile core piece, containing a chemical connector. This intermediate can be made in large batches and stored until it’s needed. Then, in a single step, we use a reaction driven by ordinary visible light to join this core piece to many different molecular fragments, creating several brand-new nucleoside analogues in one go. Because this final step is quick, many different molecules can be made in parallel instead of repeating an entire synthesis for each drug candidate. Using this strategy, we quickly built a library of more than 70 nucleoside analogues. It includes versions with sulfur or nitrogen atoms replacing the usual oxygen atom in the sugar ring, small changes that are normally difficult to make but can have a huge effect on biological activity. We also made versions designed to get inside cells more easily — the same strategy used by remdesivir and another drug used to treat hepatitis. While this is not a large number in terms of drug discovery, it demonstrates that producing larger libraries is now feasible using this process. While most of these molecules had never been made before, we also made several previously studied molecules in far fewer steps. One compound that had originally required 16 synthetic steps, we made in just six. We then handed these compounds off to our Simon Fraser University collaborator Ralph Pantophlet, whose laboratory tested them against HIV in a cell-based assay. Three compounds stood out, blocking the virus about as effectively as some of the medications already approved to treat it. While it would still be a long path to clinical trials, being able to identify starting points more quickly is critical to the entire drug discovery process. With this platform, we can generate collections of drug candidates 10 to 100 times larger than was previously practical, in weeks rather than months or years. That’s no guarantee of discovering a new medicine, but it meaningfully improves the odds of finding one quickly, whether for a virus that emerges tomorrow or one we already struggle to treat.