Inorganic pyrophosphate: one substrate, two mechanisms?
Adrian Goldman, Igor Fabrichniy, Esko Oksanen, Lari Lehtiö
Chemists and enzymologists believe that hydrolysis of phosphate esters (1), such as pyrophosphate and nucleic acids, proceeds in solution through a dissociative “E1-type” mechanism, with formation of a metaphosphate PO3 intermediate. It is thus curious that most crystallographers describe the self-same reactions, when they occur on enzymes, as proceeding by associative “SN2-type” mechanisms, with a trigonal bipyramidal O-PO3-O transition state. Such mechanisms have been postulated for enzymes ranging from inorganic pyrophosphatase to polymerases (eg 2). We have therefore revisited this issue with detailed x-ray crystallographic studies.
Our studies on the two structurally-unrelated soluble inorganic pyrophosphatases (PPase), yeast family I PPase (YPPase) and B. subtilis family II PPase (BsPPase), have led us to the conclusion that they have different folds and different mechanisms, even though they share a superficially structurally-similar active site. Our most recent results demonstrate that BsPPase has a novel 4-metal ion active site that forms only when substrate binds. Three of the metal ions form a triangular arrangement that separates the nucleophilic “water” from the electrophilic phosphorus. Consequently, the water must be preionised as a hydroxide ion in the active site – but in addition, no lone pair is available for attack on the electrophilic phosphorus; the three lone pairs on OH- are each associated with a metal ion. This and other data suggest that BsPPase clearly proceeds through a dissociative metaphosphate intermediate, while our earlier studies of YPPase (2) seem to indicate that catalysis proceeds by an associative mechanism. Our work thus suggests that we should pay more attention to alternative mechanisms. Finally, BsPPase is ten times more active than YPPase; this increase in catalytic speed might be coupled to the change in mechanism, as the energy barrier for the uncatalysed reaction is intrinsically lower for the dissociative, rather than the associative mechanism.
(1) Mildvan, A.S. (1997). Proteins 29, 401-416.
(2) Heikinheimo, P. et al. (2001). Proc. Natl. Acad. Sci. USA 98, 3121-3126.