It Looks Like a Duck, It Talks Like a Duck, but It is Not a Duck. A Case Where Computational Functional Predictions of GNAT Enzymes are Inaccurate
James Chen, Jessie Leung, Michael Simpson, Gabriela Tozlovan, Hazel N. Leiva Martel
Department of Chemistry & Biochemistry
Faculty Supervisor: Misty L. Kuhn
Predicting protein functions is one of the greatest scientific challenges of the 21st century. Misannotation of functions frequently arise when homologous proteins are assigned functions based solely on computational predictions without experimental validation. This is particularly evident for proteins from the Gcn5-related N-acetyltransferase (GNAT) superfamily, which are functionally diverse and acetylate a wide range of substrates. Our laboratory focuses on improving the functional annotation of GNATs, and in particular, spermidine/spermine N-acetyltransferases (SSATs). These enzymes acetylate polyamines, are often annotated as putative diamine acetyltransferases, and in vitro experimental validation is lacking. We selected and kinetically screened a set of previously uncharacterized enzymes that had sequence similarity to SSATs. However, we found that they acetylated L-methionine sulfone, L-methionine sulfoximine, and phosphinothricin; these compounds are not chemically or structurally similar to polyamines. To determine how well these enzymes acetylated the substrates, we performed kinetic assays and compared their catalytic efficiencies. We also generated structural models to investigate which protein residues may be important for substrate recognition, and we compared them to crystal structures of enzymes that acetylate polyamines. Our results highlight the functional diversity within the GNAT superfamily and emphasize the importance of in vitro experiments to validate computational predictions of protein functions.
This study was funded in part by the National Institutes of Health Department of Health and Human Services, National Institute of General Medical Sciences under Award Number R35GM133506 (to MLK).