Structure Prediction and Synthesis of Pyridine-Based Macrocyclic Peptide Natural Products

The structural and
functional characterization of natural products
is vastly outpaced by the bioinformatic identification of biosynthetic
gene clusters (BGCs) that encode such molecules. Uniting our knowledge
of bioinformatics and enzymology to predict and synthetically access
natural products is an effective platform for investigating cryptic/silent
BGCs. We report the identification, biosynthesis, and total synthesis
of a minimalistic class of ribosomally synthesized and post-translationally
modified peptides (RiPPs) with the responsible BGCs encoding a subset
of enzymes known from thiopeptide biosynthesis. On the basis of the
BGC content, these RiPPs were predicted to undergo enzymatic dehydration
of serine followed by [4+2]-cycloaddition to produce a trisubstituted,
pyridine-based macrocycle. These RiPPs, termed “pyritides”,
thus contain the same six-membered, nitrogenous heterocycle that defines
the thiopeptide RiPP class but lack the ubiquitous thiazole/thiazoline
heterocycles, suggesting that thiopeptides should be reclassified
as a more elaborate subclass of the pyritides. One pyritide product
was obtained using an 11-step synthesis, and the structure verified
by an orthogonal chemoenzymatic route using the precursor peptide
and cognate pyridine synthase. This work exemplifies complementary
bioinformatics, enzymology, and synthesis to characterize a minimalistic
yet structurally intriguing scaffold that, unlike most thiopeptides,
lacks growth-suppressive activity toward Gram-positive bacteria.