Over 400 P450s have been identified to date in prokaryotes and eukaryo
tes, plants and animals, mitochondria and endoplasmic reticulum. These
enzymes function in areas such as xenobiotic metabolism and steroidog
enesis. The eukaryotic members of this gene superfamily of proteins ha
ve proved difficult to study because of the hydrophobic nature of thei
r substrates, their various redox partners, and membrane association.
To better understand the structure/function relationship of P450s-what
determines substrate specificity and selectivity, what determines red
ox-partner binding, and which regions are involved in membrane binding
-we have compared the three crystallized, soluble bacterial P450s (two
class I and one class II) and a model of a steroidogenic, eukaryotic
P450 (P450arom), to define which structural elements form a conserved
structural fold for P450s, what determines specificity of substrate bi
nding and redox-partner binding, and which regions are potentially inv
olved in membrane association. We believe that there is a conserved st
ructural fold for all p450s that can be used to model those P450s that
prove intransigent to structural determination. However, although the
re appears to be a conserved structural core among P450s, there is suf
ficient sequence variability that no two P450 catalytic cycle. This en
zyme has usually been thought of as a simple globular protein; however
, sequence analysis has shown that NADPH-P450 reductase is related to
two separate flavoprotein families, ferredoxin nucleotide reductase (F
NR) and flavodoxin. Recent studies by Wolff and his colleagues have sh
own that the FAD-binding FNR domain and FMN-binding flavodoxin domain
of human NADPH-p450 reductase can be independently expressed in Escher
ichia coli. The subdomains can be used to reconstitute, however poorly
, the monooxygenase activity of the P450 system. We have been utilizin
g the reductase domain of P450BM-3 to study the mechanism of electron
transfer from NADPH to P450 in this complex multidomain protein. We ha
ve overexpressed both the FNR subdomain and the flavodoxin subdomain i
n E. coli and fully reconstituted the cytochrome c reductase activity
of this enzyme. Our studies have shown that electron transfer from NAD
PH through the reductase domain to the P450 requires shuttling of the
FMN subdomain between the reductase subdomain and the P450. Studies of
the factors that control the molecular recognition and interaction am
ong these three proteins are complicated by the weakness of the associ
ation and changes in the strength of the interaction depending on the
redox state of each of the components. How these structural and mechan
istic studies of a soluble bacterial P450 can be extended to gain a be
tter understanding of the control of membrane-bound eukaryotic P450-de
pendent redox systems is discussed. (C) 1997 by Elsevier Science Inc.