Proteome-wide probing of the dual NMT-dependent myristoylation tradeoff unveils potent, mechanism-based suicide inhibitors

Abstract

N-myristoyltransferases (NMTs) catalyze protein myristoylation, a major and ubiquitous lipid modification. Originally thought to modify only N-terminal glycine alpha-amino groups (G-myristoylation), NMTs are now known to modify lysine epsilon-amino groups (K-myristoylation), the significance of which is uncertain. Here we exploited systematic structural proteomics analyses and a novel pipeline involving the Shigella IpaJ protease to discriminate K- and G-myristoylation with unprecedented accuracy and identify the specific features driving each modification. NMT-dependent K-myristoylation occurs post-translationally and only on lysines 1, 2, or 3 following G-myristoylation or caspase cleavage. Direct interactions between the substrate′s reactive amino group and the NMT catalytic base slow K-myristoylation catalysis. IpaJ unmasked novel K-myristoylation sites in a dozen human proteins. The unique properties of NMT-driven K-myristoylation allowed us to design potent, mechanism-based suicide NMT inhibitors. These analyses unravel the respective paths towards K-myristoylation, G-myristoylation, or NMT inhibition, which rely on a very subtle tradeoff embracing the chemical landscape around the reactive group.

Mapping the myristoylome through a complete understanding of protein myristoylation biochemistry

“…Compared to N-terminal myristoylation, the current data indicate that K-myristoylation is scarce, reversible due to de-myristoylation catalysts, and less efficient [44]. The associated catalysts need further characterization.…”

“…Together, these data suggest the presence of distinct catalysts, unless the acylation is non-enzymatic [40], which could occur for serum albumin due to its natural avidity for myristate (see Section 3.2). Most recently, K-myristoylation has been described in vitro involving at least Lys3 of the small G-protein ARF6 in vivo [41][42][43][44]. The mechanism and catalyst are like those of G-myristoylation.…”

“…The mechanism and catalyst are like those of G-myristoylation. However, the catalyst, Nmyristoyltransferase (NMT, see Section 4), is not predicted to myristoylate most internal lysine residues [44]. Nevertheless, a lysine myristoyltransferase activity -still uncharacterizedspecifically recognizes small peptides of 14 aa around target lysine residues as a substrate and uses MyrCoA as the myristate donor [37] (Table 1).…”

“…NMT has strikingly high specificity for N-terminal Gly residues, and replacement with Ala or other amino acids established that a single substitution of Gly2 made N-terminal myristoylation impossible [213]. When a vicinal Lys occurs however, such change may promote Lysmyristoylation [42,44]. The search for NMT targets started with the early observation that single substitutions in known substrates resulted in a dramatic loss of myristoylation; for instance, the single substitution K7N [214] in c-Src precludes myristoylation.…”