It has become increasingly apparent within the last several years that
unusual N-formylated sugars are often found on the O-antigens
of such Gram negative pathogenic organisms as Francisella
tularensis, Campylobacter jejuni, and
Providencia alcalifaciens, amongst others. Indeed, in some
species of Brucella, for example, the O-antigen contains
1,2-linked 4-formamido-4,6-dideoxy-α-D-mannosyl groups. These sugars,
often referred to as N-formylperosamine, are synthesized in
pathways initiating with GDP-mannose. One of the enzymes required for the
production of N-formylperosamine, namely WbkC, was first
identified in 2000 and was suggested to function as an
N-formyltransferase. Its biochemical activity was never
experimentally verified, however. Here we describe a combined structural and
functional investigation of WbkC from Brucella melitensis. Four
high resolution X-ray structures of WbkC were determined in various complexes to
address its active site architecture. Unexpectedly, the quaternary structure of
WbkC was shown to be different from that previously observed for other sugar
N-formyltransferases. Additionally, the structures revealed
a second binding site for a GDP molecule distinct from that required for
GDP-perosamine positioning. In keeping with this additional binding site,
kinetic data with the wild type enzyme revealed complex patterns. Removal of GDP
binding by mutating Phe 142 to an alanine residue resulted in an enzyme variant
displaying normal Michaelis-Menten kinetics. These data suggest that this
nucleotide binding pocket plays a role in enzyme regulation. Finally, by using
an alternative substrate, we demonstrate that WbkC can be utilized to produce a
trideoxysugar not found in nature.