Multi-Pole Approach to Structural Science

May 10 - 13, 2015

Structural and functional characterization of a GNAT acetyltransterase - reproducibility issues
Karolina Majorek

The Gcn5-related N-acetyltransferases (GNATs) catalyze the acetylation of a wide range of substrates, including both small molecules and proteins. GNATs are implicated in numerous aspects of eukaryotic and prokaryotic physiology. Bacterial species generally have many GNATs, but very few of them have been functionally characterized. These observations led us to initiate characterization of bacterial GNATs.

One of the enzymes characterized using several screening methods is PA4794 from a nosocomial pathogen Pseudomonas aeruginosa. Our results show that PA4794 is a protein acetyltransferase specific for C-terminal lysine residues. We also discovered that PA4794 is inhibited by cephalosporin antibiotics, which bind in the substrate binding site by mimicking the conformation of the substrates. We determined many high resolution structures of PA4794 in a variety of complexes, including with substrates, products, and inhibitors. Two notable structures include a complex with HEPES and a structure with a His-tag bound in the substrate-binding site. Our analysis allowed us to identify several sensitive elements of the experimental protocols that can affect the biophysical and kinetic analysis, potentially altering the interpretation of results. These observations prove how critical it is to track and adjust experimental conditions and investigate the influence of these factors on protein activity and structure.

The conditions of protein expression and purification can play a significant role in the outcome of protein functional studies. The presence or absence of affinity tags, the choice of buffer, and experimental methods, all can affect the biological activity of investigated proteins, but these effects are rarely discussed in the literature. Even small changes in the experimental protocols can generate unexpected artifacts and influence the results of the experiments and their interpretation. When subsequent studies are based on questionable data, a “ripple effect” is caused, as the problematic data is propagated.