Collaborative Research: Chemical Speciation of Metals in Cyanobacteria: Testing Artificial Intelligence Predictions with X-ray Absorption Spectroscopy

About This Grant

This project is using artificial intelligence (AI) to better understand how metals are used by cyanobacteria, the most abundant group of organisms to have ever existed in Earth history. Metals are essential for life and are held tightly by proteins inside of cyanobacterial cells. AlphaFold is a powerful AI tool that can predict the shapes of these proteins and how they hold onto metals. So far, these predictions have not yet been tested with laboratory experiments. In this study, college students are growing cyanobacteria in the laboratory and using advanced X-ray techniques to see how metals are bound inside the cells. By comparing the AI predictions with real data, the team hopes to better understand how metals move through environments when these cells die and break apart. This knowledge will provide a better picture of how dissolved metals are recycled in aquatic ecosystems. The project also includes outreach to schools and communities, including science activities for children, story-writing contests, and support for college students to get involved in science. The laboratory studies focus on resolving the chemical speciation of Zn and Fe in cyanobacteria. Experiments are performed by the researchers to assess whether AI-predicted metal-ligand binding environments reflect the actual speciation of Zn and Fe in living cells. Marine and freshwater cyanobacteria are being cultured under metal-controlled conditions, and proteins expressed under different growth phases are being identified using LC-MS/MS proteomics. The three-dimensional structures of the proteins will then be modeled using AlphaFold, and the protein structures will be annotated to identify putative metal-binding sites, coordination numbers, and ligand identities. These predictions will be experimentally tested using High Energy Resolution Fluorescence Detection (HERFD) X-ray absorption spectroscopy at the Zn and Fe K-edges, conducted at the Advanced Photon Source. Spectral data will be analyzed using linear combination fitting and principal component analysis to quantify the distribution of metals among cysteine, histidine, and carboxyl ligands. It is anticipated that AI predictions will correlate with experimental data, particularly in conserved protein families. These findings will provide mechanistic insights into metal-ligand complexation in cyanobacteria and establish a framework for AI-enabled investigations of metal cycling and biogeochemistry in natural aquatic systems. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.