November 3, 2022

Miniaturizing Drug Discovery

November 3, 2022

Miniaturizing Drug Discovery

               “The cleverer I am about miniaturizing the world, the better I am at possessing it.”

                                                            - Gaston Bachelard

At Octant we share the same sentiment– especially when it comes to chemistry! We’ve built a nanoscale chemistry platform that empowers us to conduct tens of thousands of chemical reactions in parallel in a few hours with minimal human effort. We use this platform to rapidly and thoroughly explore local chemical diversity during the hit-to-lead phase of discovery, tens of thousands of analogs at a time. For some perspective, our nanoliter-scale reaction mixtures are so tiny that they are barely visible with the naked eye! 

Our robots make great chemists, which frees us up to get out on Galen’s boat now and then. From left to right: Stacie, Tao, and Galen.

So why thousands of tiny reactions at a time? 

At Octant, we are often working on building small molecule drugs with challenging target profiles– modulating multiple targets, pathways, and/or mutations. And we’re directly measuring cellular and biochemical activities like signaling, trafficking, post-translational modifications, folding, and more, which aren’t as simple as a binding assay. The controlled nature of our multiplexed assay platform allows us to screen chemical libraries against these features in parallel, enabling a scaled “direct to biology” chemical approach. Our method increases the speed and throughput of exploring local chemical space, informing us whether a particular chemical series can satisfy all of our criteria. This provides crucial insights into our understanding of biology and shapes new therapeutic possibilities. 

Our scaled biology platform and multiplexed cell-based assays enable us to do this against tens of thousands of compounds at a time. These cell lines are engineered to precisely read out on the set of activity screens and counter-screens that elucidate how chemicals modulate the mechanisms we hope to hit and avoid. For instance, in one of our programs, we are building small molecules that correct the particular mutations in a target that result in mistrafficked proteins. There, we’ve built reporters that measure the trafficking and functional characteristics of many mutations of the target, as well as possible off target pathways (for example, generally increasing transcription or inhibiting the degradation of the protein). The platform enables us to move beyond just finding a hit, towards finding those molecules that satisfy a demanding set of target criteria (such as rescuing many trafficking mutations at once), and also all the associated criteria of being a good small molecule drug (stability, permeability, exposure, etc).

Moving drops of chemicals with acoustic energy [1]

High Throughput Chemistry at Octant

We interface heavily with automation to enable rapid and high throughput chemistry. Onsite there are several pieces of equipment that we use for liquid transfers. My favorite is the Echo 655, which uses acoustic energy to transfer nanoliters of chemicals from source plates to destination plates. This device, coupled with a robotic arm, can autonomously transfer reagents to and from up to 80 SBS plates (384-1536 wells/plate) at a time. Using this technology we can introduce thousands of new molecules into the universe within the course of one day, all with no pipette tips involved! 

Robots have been doing chemistry at Octant for about 2.5 years now. Not only for synthesis, but also to optimize each reaction condition. Our reactions couple cores (deconstructed hits with a reactive handle) with fragments under many different conditions per flavor of coupling chemistry– altering the catalysts, concentrations, acid/base, and volumes to find conditions that lead to high yields and few byproducts. We then use these optimized conditions to make the whole library, which is typically 103-104 members. Our average time from library synthesis to understanding how the library interacts with our targets is around a week. Minimizing turnaround time allows us to rapidly iterate on our chemistry as we aim to find molecules with new therapeutic properties. For example, in our early mistrafficking programs, we used the platform to assay cell lines harboring mutant forms of rhodopsin that lead to autosomal dominant retinitis pigmentosa, a progressive blindness disease. Since the inception of that program about 1.5 years ago, we have built and assayed ~250,000 custom compounds using our HT-Chem platform. We now have leads that rescue the most prevalent clinically-relevant mutations, are orally bioavailable, and cross the blood-retinal barrier.


So, how does the chemistry actually work? 

First, we start with hits from screening our own discovery libraries (>105 compounds) or by looking at endogenous ligands, tool compounds and scanning the literature. We break these molecules into cores with functional modifications that enable us to add fragments from our in-house fragment libraries through optimized chemistries. We use a number of chemistries at Octant. The simplest ones are amide couplings, where a core with an amine can be reacted with a class of carboxylic acid fragments (or vice-versa!) to yield a library of thousands of unique amides (a quick primer/refresher on amide coupling). Even without cores we could synthesize large libraries by cross-reacting our fragment collections. Our fragment libraries are large enough, and the combinatorial space of chemistry is so vast, that if we performed every pairwise synthesis between our amines and carboxylic acids alone we could synthesize over 100 million compounds, ~10,000 molecules at a time!

Once a library is built we use the Echo to aliquot portions of some representative wells to QC, a process which involves running samples through LC/MS to quantify how well our products formed. This is useful for quality control, optimizing reaction conditions, and other downstream analysis. The crude reaction mixtures are then screened for activity against our engineered cell line assays and an automated pipeline facilitates analysis of our results. 

                          Core-fragment assembly enables rapid exploration of targeted chemical space

Some of our latest challenges…

We are continuously adding new chemistries to our platform. Adding chemistries not only geometrically increases the number of molecules our platform can generate, but also gives us access to additional chemical properties that are important for drug development. A big effort over the last year has been to expand into more challenging chemistries that need to be protected from oxygen and water. For example, palladium coupling can make new sp2 hybridized C-C or C-N bonds, resulting in products with extended pi systems. This can be useful for making drugs that intercalate DNA or bind hydrophobic pockets of proteins.

Onboarding palladium coupling to our platform has proven to be an exciting challenge. Because these reactions require catalysts that are sensitive to oxygen, we’ve had to adapt our workflows to a hermetically-sealed glovebox fitted with non-contact liquid-handling equipment. Since these catalysts are toxic to cells, we also worked on ways of mitigating this toxicity in our screens. A nice thing about multiplexed assays is how easy it is to incorporate controls for general toxicity and non-specific activation of our reporters. This lets our chemistry be dirtier and is at the heart of Octant’s ante-disciplinary approach to drug discovery.

              Jordan (RA) piloting Buchwald amination.

What could you do?

Like any team at Octant, HT chem is highly cross-functional – operating at the intersection of synthetic chemistry, automation, computation and discovery biology. This means there’s always a chance to learn a new in-demand skill and collaborate with others who come from different scientific backgrounds. You’ll also get to watch your compounds progress through the drug discovery pipeline– from primary and secondary screens to pre-clinical trials; every molecule you make has a unique story and you’ll get to see it unfold. Who knows? Maybe one of your compounds will help patients one day. Given you’ve read this far, you might want to consider applying for a role on our chemistry team. Our job listings are here. Also, if you are a fresh undergraduate, check out our Octant Apprenticeship!



Jordan Gewing-Mullins

Research Associate
Back to all Posts