(Z)-and (E)-[3-H-2] phosphoenolpyruvate were prepared chemically by th
e reductive deuteration of (Z)- and (E)-3-bromophosphoenolpyruvate, re
spectively, and were converted into 3-deoxyoctulosonate 8-phosphates d
euterated at the C-3 position by incubation with unlabeled D-arabinose
5-phosphate in the presence of the enzyme, 3-deoxyoctulosonate 8-phos
phate synthase (EC 4.1.2.16) purified from Escherichia coli K-12 conta
ining the plasmid pMW101. Analysis of the stereochemistry of the two 3
-deoxyoctulosonate 8-phosphates deuterated at the C-3 position by H-1
NMR showed-that the (Z)-[3-H-2]phosphoenolpyruvate had produced [3-H-2
]-3-deoxyoctulosonate 8-phosphate of predominantly the 3S configuratio
n and that the E isomer had given predominantly (3R)-[3-H-2]-3-deoxyoc
tulosonate 8-phosphate. The 3-deoxyoctulosonate 8-phosphate synthase r
eaction is therefore stereospecific with respect to the C-3 of phospho
enolpyruvate. The results indicate a si face attack from the C-3 of ph
osphoenolpyruvate, a result identical to that reported for 3-deoxyhept
ulosonate 7-phosphate synthase (EC 4.1.2.15), an enzyme catalyzing an
identical aldol-type condensation, except that it takes place between
phosphoenolpyruvate and D-erythrose 4-phosphate. The stereochemistry w
ith respect to the face of the carbonyl of the attacked aldehyde, in b
oth 3-deoxyoctulosonate 8-phosphate synthase and 3-deoxyheptulosonate
7-phosphate synthase, is re. On the basis of the results of the studie
s reported herein, the presence of a transient methyl group at the C-3
of phosphoenolpyruvate as part of the reaction mechanism seems unlike
ly.