Posted on Leave a comment

With lessons learned from computers, a new platform could help boost production of lifesaving biological therapies

The power of the Station B platform lies in pulling all those pieces of the puzzle together in one integrated system, Phillips said. Both initial deployments will occur in labs that are overseen by health, safety, ethical and medical regulators.

“It marries Microsoft’s deep expertise in programming languages, modeling capabilities and machine learning with lab automation and the power of the cloud and intelligent edge — that combination of tools doesn’t exist anywhere in this industry today,” Phillips said.

To solve one key challenge, the platform uses Synthace’s lab automation system to allow users to run experiments from the cloud and precisely replicate each step in complicated scientific protocols.

Synthace’s Antha software allows the user to replace subjective instructions like “shake a test tube vigorously” with digital language that isn’t open to misinterpretation and that lab robots can execute. Building on top of Azure IoT, Antha is a high-level language for describing biological experiments that allows an array of lab machines made by different manufacturers to run them, much like printer drivers allow any make or model of printer to print PDF documents.

That ability to run experiments exactly the same way each time gives users confidence that the results they’re seeing are meaningful, and not just a fluke in the way the experiment happened to be set up that day.

Synthace’s system — which can handle experiments that simultaneously test dozens of different parameters or genetic constructs rather than one or two at a time — speeds up the research process exponentially. Combined with machine learning capabilities, it also gives customers the ability to pose and learn from much more sophisticated lines of inquiry.

“The near infinite power of biology can only be unlocked by bringing software abstraction and automation to biological R&D and manufacturing, and by enabling biologists to build atop their collective work. That is what the Antha platform does successfully,” said Tim Fell, Synthace chief executive officer.

Sarah-Jane Dunn stands in front a mural with her arms crossed
Sarah-Jane Dunn, scientist for Microsoft Research Cambridge, UK. Photo by Jonathan Banks.

’This could have huge reach’

The Station B platform will be tested first in the lab of Bonnie Bassler, chair of Princeton’s Department of Molecular Biology, a Howard Hughes Medical Institute Investigator and recipient of a MacArthur genius grant, who studies how bacteria wield outsized power by acting as collectives. The Princeton team includes Bassler’s longtime collaborator Ned Wingreen, a physicist and professor in Princeton’s Lewis-Sigler Institute for Integrative Genomics.

“Historically we’ve thought of bacteria as only having harmful behaviors, like infecting us and causing disease, but more recently scientists have discovered the microbiome, a rather magical bacterial community that lives in and on us and that keeps us alive,” Bassler said. “What my lab has always wondered about is how do bacteria manage to either kill us or keep us alive? They’re so tiny.”

Bassler discovered the widespread use of a phenomenon called quorum sensing in the bacterial world. It’s a form of molecular communication that bacteria use to determine when their numbers have reached a critical mass. When they reach the “quorum,” together they trigger behaviors that are only successful when bacteria act as a coordinated group — such as unleashing virulent diseases.

In a proof-of-concept pilot, the team will deploy the Station B platform to investigate how cholera bacteria use quorum sensing to form biofilms, thin layers of bacteria that grow on almost all surfaces. Bacteria living in biofilm communities can be 1000 times more resistant to antibiotics than non-biofilm bacteria.

Princeton researchers will use the Station B platform and Synthace’s lab automation tools to construct and test different versions of two proteins that are key to biofilm formation — which are also genetically programmed to light up. The light allows the scientists to see and measure how much of each protein is produced under many different conditions and in different regions of the biofilm.

Bassler compares the working microbiologists in her lab to master craftspeople, creating elegant and complicated genetic constructs to produce a desired result. But that artisanal process yields only a few prospects at a time and doesn’t allow the team to massively attack the problem.

The Station B platform will be able to build and test dozens of engineered proteins at once — in whatever combinations a researcher can dream up and type into the system for a liquid handling robot to produce. The platform will then help the scientists learn which of the protein constructs behave most like the natural proteins and yield an accurate picture of how biofilm cells organize, Bassler said.

The goal is to build on that basic understanding and find an Achilles heel that might weaken virulent biofilms or increase their sensitivity to antibiotics.

“The platform will allow us to ask more questions, get more results and do more experiments than a graduate student or postdoc, no matter how clever, can do today. So, it gets us to the winning genetic constructs faster,” Bassler said.

Equally important, the platform will also collect and help analyze data from every single lab experiment — including ones that fail, Bassler said. By necessity, scientists have to pursue their most fruitful lines of inquiry, but that can leave an untapped trove of information about why something didn’t succeed.

“If this extra information can help us discover the underlying patterns in what works and what doesn’t work and why, that would be a transformative leap for us,” she said.

The value of deploying the Station B platform in Bassler’s lab is that those researchers have already built an extensive inventory of genetic components, chemical mixtures and models in the years that they’ve been studying bacteria like cholera.

If the team can begin to uncover the rules and principles that govern those systems, Wingreen said, they may be able to program them in transferrable ways. That could potentially enable a doctor who studies cancer or an engineer working on low-carbon fuels to imagine a genetic construct that they’d love to test and get an exact blueprint for assembling it — without spending years at a lab bench.

“From my perspective, this could have a huge reach,” Wingreen said. “Just as the tech sector was democratized by software that lets you ask for what you want in a microchip design and have someone make it, we need that same revolution in biology.”

Top image: Breech Odu works in an Oxford Biomedica lab, where the Station B platform will be deployed to accelerate discovery and manufacturing of gene and cell therapies. Photo by Jonathan Banks.


Jennifer Langston writes about Microsoft research and innovation. Follow her on Twitter.

Leave a Reply