How growth-coupled selection could rebuild metabolism

Within the whirl of metabolic reactions lies the chemistry that binds all living things together. Over the years, many scientists have tweaked metabolism in organisms such as bacteria and yeast toward the production of useful molecules. But some scientists want to go further, breaking apart metabolic pathways and networks into modular components that use the tools of metabolism to unlock a wider range of chemical transformations.

That vision was the ambition of synthetic biologist Arren Bar-Even, who led a research group at the Max Planck Institute of Molecular Plant Physiology until he died in 2020.

According to Ron Milo, Bar-Even’s PhD supervisor at the Weizmann Institute of Science, Bar-Even “had this very open, honest, no-bulls— approach that was rough. . . . But at the same time, it led to complete integrity, and not needing to take anything for granted.”

As part of Milo’s group, Bar-Even pursued synthetic metabolism, not to improve on nature but to replace it altogether. “He didn’t care that that’s what nature did for the last 4 billion years. We could do something else. And that’s just a state of mind that not everybody has,” Milo says.

The conventional attitude toward metabolism is that it is fixed—pressed and refined by the weight of selection and time into rigid networks of reactions. In this view, metabolic engineering is confined to mixing and matching pieces of the pathways that nature has already produced.

But as Tobias Erb, a biochemist at the Max Planck Institute for Terrestrial Microbiology, says, “Nature rarely does a jump innovation. So only a couple of paths have been realized, and there is a lot of transformation which nature has never touched upon.”

Before his death, Bar-Even pursued the theory and methods for using metabolic reactions to do things they hadn’t evolved to do. He also demonstrated how that ability could be used to tackle societal problems like how we transform carbon. As the world grapples with issues regarding climate change and access to fossil fuels, the case for circular biobased economies is growing. Six years after his death, Bar-Even’s colleagues and students believe his approach to metabolism has the potential to accelerate and bolster that case for bioeconomies.

A thermodynamic vision of metabolism

In a laboratory or a cell, chemical reactions transform one molecule into another. Metabolism contains all the tools nature evolved to make these transformations, and Bar-Even excelled at understanding how they work. How do macromolecules such as enzymes behave? How and why did they evolve? And importantly, what rules do they follow?

According to Pablo Ivan Nikel, a biotechnologist at the Technical University of Denmark, Bar-Even’s insight was to think “of metabolism as a whole, with a way more integrating type of vision, where he could see the connections and factors that were not usually part of traditional metabolic engineering.” Bar-Even, Nikel says, realized that the way to explore the possible was with thermodynamics.

“Knowing what reaction mechanisms are feasible on a thermodynamic principle, what molecules could be in theory turned into other molecules, is the first discipline you need to master,” says Jan Krüsemann, a former student of Bar-Even’s, “and that is something he could do extremely well.”

Bar-Even’s favorite demonstration of this idea was the Haldane relationship, which describes how the ratio of an enzyme’s forward and reverse catalytic efficiency is fixed to the thermodynamic equilibrium constant of the reaction. “I remember in a meeting he would just ask the room, ‘What’s the Haldane relationship?’ ” recalls Krüsemann, “and then he just gets out a whiteboard and out of his head draws this connection of thermodynamics and kinetics.”

To Bar-Even, the apparent constraint that improving an enzyme’s forward efficiency meant that reverse efficiency must also improve by the same factor proved liberating. It shows that metabolism is not linear or fixed. An enzyme could reverse direction based on factors such as the conditions of the reaction environment or the concentrations of substrates and products.

Thermodynamics is the theoretical basis of what can be built using the metabolic toolkit. Armed with this insight and a deep understanding of the rules of biochemistry, Bar-Even designed wholly new and elegant metabolic pathways, often using what he called plausible chemistry. He sought out theoretical enzymes that had not been discovered or designed but would work in theory, based on the principles of thermodynamics and organic chemistry. He also used his deep understanding of metabolism to make the theoretical real by breaking larger networks into modules that perform distinct transformations. This approach allows researchers to play with the resulting pieces, recombining them or adding new modules to find the most promising path from molecule A all the way to molecule Z.

An early example from Bar-Even’s time in the Milo lab was a strategic severing of the Escherichia coli gluconeogenesis pathway into two modules, one that harvests energy and the other which builds biomass. Once separated, each module could be engineered for a new function.

This approach allowed the researchers to create a nonnative Calvin-Benson-Bassham cycle that converts CO₂ into sugars and is powered by energy harvested from pyruvate. They called this new mode of growth “hemiautotrophic” because it is a blend of heterotrophy and autotrophy. This work demonstrated that metabolism could be reprogrammed to run on new carbon sources. If optimized, sources like the atmosphere’s overabundance of CO₂ could be tapped for production.

For those in the field, there is an elegance to this type of metabolic design.

“Many of these pathways are also aesthetically beautiful,” Erb says. “It’s like Bauhaus style. Form follows function.”

Nature, however, didn’t share this appreciation.

Using growth coupling to build new metabolic pathways

Life wants to grow and survive, not work for humans. Nature is protective of its pathways and cannot afford changes to fundamental processes such as metabolism.

Bypasses, new connections, and promiscuous enzymes all thwart designers at every step—and isolating a step that has failed and why is near impossible.

To successfully install designs in a cell, Bar-Even once again saw potential in something others hadn’t. The strategy is called growth coupling, and while Bar-Even didn’t invent it, “I think he put it to an extreme, and I think he smartly used it,” says Erb.

Growth coupling is the inextricable linking of a metabolic module’s success to the survival of the cell.

Using clever deletions and mutations, biologists can create new strains of common laboratory microbes like E. coli that lack a product from or are somehow metabolically dependent on a specific module. The growth and survival of these selection strains depend on the pathway, and natural selection goes to work to ensure the microbe’s survival. Once growth and survival are linked, the incentive for natural selection to resist the module is reversed.

Growth coupling can be applied directly to strains already used in bioproduction to boost efficiency, but Bar-Even saw it as a tool to bring even his most complicated new designs to life. A new design could be broken into modules, making it easier for a selection strain to optimize and accept. Each module is optimized in parallel with the others, and then all are connected piece by piece in new selection strains until the entire pathway is running in cell. It’s a clever and powerful method, but it requires a deep familiarity with metabolism.

By the time Bar-Even was leader of his own lab group at the Max Planck Institute in Potsdam, growth-coupled modular design was his signature technique, and one that many researchers from his lab still use.

“It’s a lot of fun,” says Steffen Lindner-Mehlich, who was a postdoctoral researcher with Bar-Even in Potsdam and is now head of his own group at the Charité University Medicine Berlin. “I always compared it to crafting, like playing with Lego. You can break something down, reconstruct it, build your strain, then test it.”

With the technique in place, the lab in Potsdam went back to work on the carbon fixation challenge Bar-Even had started at the Weizmann Institute: a bioeconomy based on carbon produced using renewable power and CO₂ . In 2020 the researchers created the first E. coli bacteria that survives off of formate and methanol alone using a synthetic reductive glycine pathway.

Simultaneously, the Bar-Even group also went after the inefficient photorespiration cycle in plants. Together with Erb and his team, the researchers produced the tartronyl–coenzyme A (TaCo) pathway using a mix of synthetic and known enzymes to replace photorespiration. TaCo achieves the same function as photorespiration but without the lost carbon.

Bar-Even died unexpectedly soon after the reductive glycine pathway was published and within weeks of receiving news that he had beat Mother Nature to publication. In October 2020, a group from Wageningen University and Research reported that they had found the novel pathway in the wild.

Since then, his former collaborators have performed trials that show that the TaCo pathway improves carbon uptake, reduces water and fertilizer usage, and unexpectedly increases resistance to drought.

A legacy of influence

For his students and collaborators, these projects, while impressive, are not the entire legacy of Bar-Even.

For Lindner-Mehlich and Krüsemann, the use of growth coupling holds great potential since selection strains can be used for a variety of tasks. “Is it going to be bioproduction? Is it going to be enzyme engineering? Is it going to be data collection for [artificial intelligence] models?” Krüsemann asks. “Designing catalysis is such a dynamic process, and that’s what [growth coupling] is really good for.”

Like the loops of chemistry in metabolic networks, a back-and-forth exists between designer and nature. Lindner-Mehlich believes that growth-coupled modular design has the power to accelerate biotechnologies. His own work has produced E. coli with entirely new metabolic capabilities that continue the pursuit of usable carbon sources aside from fossil fuels.

Lindner-Mehlich also says that making the techniques more accessible is crucial, and he is focused on sharing and increasing the resources needed for the creation of selection strains. This means building databases of known selection strains, genomic resources for E. coli and other model organisms, and reducing red tape when sharing strains and new models for metabolism.

New technologies can also help as advances in AI enzyme design, automation, laboratory-guided evolution, combined with growth coupling, could further accelerate scientists’ ability to design new pathways using rapidly created synthetic enzymes. “The design space will dramatically increase once we really learn how to develop new to nature enzymes,” Lindner says.

Bar-Even’s approach and style were not typical, but it was fruitful. “The idea to use a different carbon source in the way he was using it, I think this was one sort of game changer,” says Laura Martinelli, CEO of INsociety, a scientific consultancy, and a former collaborator with Bar-Even. “The way he looked at metabolic engineering, his logic was entirely new. To the extent that, to me, he can be named among the most influential scientists in the field.”

Bradley van Paridon is a freelance writer, speaker, and podcast producer covering chemistry, science, and technology. He is from and based in Canada.