Sequence-Specific Protein Secondary-Structure Assignment with Isotope Reverse-Labeled Amide I Spectroscopy

The dramatic advances made recently by protein-structure prediction tools like AlphaFold offer an opportunity for the development of new experimental methods designed to test sequence-based predictions rather than build full atomistic structures from scratch. While atomistic methods such as X-ray crystallography, nuclear magnetic resonance (NMR) spectroscopy, and cryogenic electron microscopy (CryoEM) will always play a critical role in structure determination, their cost and stringent sample preparation requirements can be prohibitive for routine characterization. Spectroscopic methods such as circular dichroism (CD) and Fourier-transform infrared (FTIR) spectroscopy are often cheaper and faster but offer insight only into bulk secondary-structure content such as the total fraction of α-helix or β-sheet present in a protein. In the present contribution, we suggest that isotope-labeled FTIR spectroscopy has the potential to fill this gap, and we demonstrate a new experimental approach to producing isotope-enriched spectra at a low sample cost. We illustrate these concepts experimentally using the protein Top7 V48 V (whose crystal structure is known) as an example. By selectively ¹²C-labeling individual amino acid residues in only 5 mL of ¹³C-enriched protein-expression culture, we collect FTIR spectra of each reverse-labeled protein construct at a cost of only a few dollars of isotope-enriched material per sample. Analysis of the resulting isotope reverse-labeled FTIR spectra provides a path to secondary-structure assignments for each amino acid individually rather than only the total secondary structure estimates that are available from traditional FTIR or CD measurements. Combining these residue-specific secondary-structure assignments with knowledge of the protein's primary sequence, we obtain assignments for the secondary structure of each stretch of amino acids in the protein that agree well with the published crystal structure of the protein. Based on these results, we suggest that this approach offers a fast and efficient means of extracting sequence-specific structural information that can be applied to proteins in a wide variety of contexts, from live cells to solubilized membranes.