NEW YORK – A team led by researchers at the University of Utah has developed a mass spec-based approach to high-throughput discovery of protein-metabolite interactions.
Detailed in a paper published this month in Science, the approach, called MIDAS (for mass spectrometry integrated with equilibrium dialysis for the discovery of allostery systematically), combines equilibrium dialysis with mass spec and allows users to rapidly screen hundreds of metabolites against protein targets of interest.
Jared Rutter, professor of biochemistry at the University of Utah and senior author on the study, has cofounded a company, Atavistik Bio, that aims to use the MIDAS approach to identify drug targets and develop small molecule agents against them.
MIDAS is based on equilibrium dialysis, an approach that Rutter noted has long been used to study protein-metabolite interactions. In an equilibrium dialysis experiment, a protein of interest is put in a chamber separated by a semipermeable membrane from another chamber containing metabolites of interest. When the system reaches equilibrium, the concentration of unbound metabolites will be the same on both sides of the chamber. However, in the case of metabolites that interact with the target protein, the concentrations will be different in the two chambers, which means researchers can identify metabolites interacting with the protein of interest by comparing the concentration of metabolites present in the two chambers after the system has reached equilibrium.
MIDAS uses mass spectrometry to measure the metabolite concentrations following equilibrium dialysis experiments, making the technique more amenable to large-scale screening efforts, Rutter said.
"Instead of taking one potential [metabolite] ligand and asking does it bind [a target protein] and with what affinity, we are now doing it with a library of metabolites and asking the question, which of all these potential ligands really do interact with the putative receptor," he said.
The technique is, in a sense, the inverse of mass spec-based thermal profiling or limited proteolysis methods developed for studying protein-metabolite or protein-drug interactions in that those approaches allow researchers to screen small numbers of metabolites or other molecules against large numbers of proteins.
In the Science study, Rutter and his coauthors screened a library of 401 compounds against 33 enzymes involved in glycolysis, gluconeogenesis, the tricarboxylic acid cycle, and the serine biosynthetic pathway, identifying 830 potential protein-metabolite interactions. He highlighted as particularly interesting the discovery of isoform specific interactions between acyl coenzyme A (acyl-CoA) and two forms of lactate dehydrogenase (LDH) that offered new insights into how different tissues use fat and carbohydrate metabolism.
Rutter said the method could be useful for studying a wide range of biological questions and systems. He and his cofounders established Atavistik Bio to apply the technique to drug development, where they believe it could help identify allosteric regulation sites on protein targets of interest as well as provide structural information that could advance development of small molecules targeting those sites.
Based in Cambridge, Massachusetts, Atavistik launched in 2021 with a $60 million Series A funding round led by The Column Group and joined by Lux Capital and Nextech Invest.
The company's AMPS (Atavistik Metabolite Protein Screening) platform is fundamentally similar to the MIDAS system described in the Science study but with a number of refinements to improve throughput and accuracy, said Marion Dorsch, Atavistik's president and CSO.
Dorsch said the company had miniaturized the equilibrium dialysis portion of the workflow so that it fit a 96-well plate format. Additionally, while for the Science study Rutter and his colleagues used flow injection analysis mass spectrometry, in which samples are injected directly into the instrument, Atavistik has moved to using liquid chromatography to separate samples prior to mass spec analysis.
"If you use flow injection, it's not always possible to tell each [metabolite] apart," Dorsch said. "What [adding LC] helped us with is the unambiguous detection of each metabolite."
Atavistik needed higher performance mass spec analysis in part because it wanted to expand the panel of metabolites it screens against target proteins, she said, noting that the AMPS platform uses a metabolite library roughly twice as large as the 401 compounds screened in the Science study. She said the company plans to continue adding to this library to capture as much biology as possible but added, "We're very confident right now that with the current makeup of our library, we are covering a very good percentage of the existing metabolome."
One challenge to broad mass spec analysis of the metabolome is that different metabolites require different conditions for optimal mass spec detection. In the Science study, the researchers analyzed the ideal conditions for each metabolite separately and ultimately divided them into four pools, each of which they analyzed separately.
Atavistik's researchers have similarly worked to refine their experimental conditions, Dorsch said. "We have a very experienced team that is really expert in mass spec, and they have really optimized our detection methods."
She said the company could screen its full metabolite library against a protein of interest in several days.
From there Atavistik applies a variety of approaches including additional biochemical screens and bioinformatic analyses to decide which interactions are worth pursuing in more depth. The company's ultimate goal is to identify allosteric interactions that could be used to target proteins of interest and develop small molecule agents against those proteins.
"There's huge interest [within drug development] in finding those allosteric sites, and we feel that metabolites are a perfect bait for that because they leverage existing physiological regulation in cells, so they are very good at pointing you toward this allosteric regulatory site," Dorsch said.
Rutter suggested that identification of metabolites that allosterically regulate target proteins could also help with designing therapeutic small molecule agents.
"Nature is probably doing it better than humans will ever be able to do it, so can we use that information?" he said. "If we find a way that nature allosterically modulates some very important disease-relevant protein, where millions of years of evolution have selected for this activity, that might be a better place to start."
Internally, Atavistik is focused on drug development for oncology and inborn errors of metabolism. Dorsch said, though, that the company expects that with the AMPS platform it will generate more leads than it has the capacity to pursue internally, and that it hopes to partner with outside pharma firms on their development. She said Atavistik also plans to make the platform available to pharma firms interested in screening proteins of interest for metabolite interactions.
The company is in discussions with several pharma firms at the moment but has not announced any collaborations, Dorsch said.