MODELING OF NUCLEOTIDE-BINDING DOMAINS OF ABC TRANSPORTER PROTEINS BASED ON A F-1-ATPASE RECA TOPOLOGY - STRUCTURAL MODEL OF THE NUCLEOTIDE-BINDING DOMAINS OF THE CYSTIC-FIBROSIS TRANSMEMBRANE CONDUCTANCE REGULATOR (CFTR)/
Ma. Bianchet et al., MODELING OF NUCLEOTIDE-BINDING DOMAINS OF ABC TRANSPORTER PROTEINS BASED ON A F-1-ATPASE RECA TOPOLOGY - STRUCTURAL MODEL OF THE NUCLEOTIDE-BINDING DOMAINS OF THE CYSTIC-FIBROSIS TRANSMEMBRANE CONDUCTANCE REGULATOR (CFTR)/, Journal of bioenergetics and biomembranes, 29(5), 1997, pp. 503-524
Members of the ABC transporter superfamily contain two nucleotide bind
ing domains. To date, the three dimensional structure of no member of
this super-family has been elucidated. To gain structural insight, the
known structures of several other nucleotides binding proteins can be
used as a framework for modeling these domains. We have modeled both
nucleotide binding domains of the protein CFTR (Cystic Fibrosis Transm
embrane Conductance Regulator) using the two similar domains of mitoch
ondrial F-1-ATPase. The models obtained, provide useful insights into
the putative functions of these domains and their possible interaction
as well as a rationale for the basis of Cystic Fibrosis causing mutat
ions. First, the two nucleotide binding domains (folds) of CFTR are ea
ch predicted to span a 240-250 amino acid sequence rather than the 150
-160 amino acid sequence originally proposed. Second, the first nucleo
tide binding fold, is predicted to catalyze significant rates of ATP h
ydrolysis as a catalytic base (E504) resides near the gamma phosphate
of ATP. This prediction has been verified experimentally [Ko, Y.H., an
d Pedersen, P.L. (1995) J, Biol. Chem. 268, 24330-24338], providing su
pport for the model. In contrast, the second nucleotide binding fold i
s predicted at best to be a weak ATPase as the glutamic acid residue i
s replaced with a glutamine. Third, F508, which when deleted causes si
milar to 70% of all cases of cystic fibrosis, is predicted to lie in a
cleft near the nucleotide binding pocket. All other disease causing m
utations within the two nucleotide binding domains of CFTR either resi
de near the Walker A and Walker B consensus motifs in the heart of the
nucleotide binding pocket, or in the C motif which lies outside but n
ear the nucleotide binding pocket. Finally, the two nucleotide binding
domains of CFTR are predicted to interact, and in one of the two pred
icted orientations, F508 resides near the interface. This is the first
report where both nucleotide binding domains of an ABC transporter an
d their putative domain-domain interactions have been modeled in three
dimensions. The methods and the template used in this work can be use
d to analyze the structures and function of the nucleotide binding dom
ains of all other members of the ABC transporter super-family.