334 GROUP I AND II METALS IN BIOLOGICAL SYSTEMS: GROUP II
Toyoshima ’ s group has determined the X - ray crystallographic structure of
a Ca 2 E 1 · ATP state^99 (PDB: 1VFP). The structure shows a rotation of the A
domain by about 30 ° around an axis approximately parallel to the membrane
thus making contact with the N domain. The N domain is inclined toward the
P domain facilitated by binding of the ATP analogue, AMPPCP. The AMPPCP
molecule makes contact with the ring of phe487 and theγ - phosphate of the
AMPPCP binds near asp351, the residue to be phosphorylated. Movement of
the A domain moves membrane helices M1 and M2. M2 moves one turn of
theα - helix toward the cytoplasm and M1 is drawn up by two turns of its α -
helix. M1 bends at asp59 and the amphipathic (polar – nonpolar) M1 ′ helix lies
on the surface of the membrane. Movement of M1 also locks glu309 into place,
preventing Site II Ca 2+ ions from escaping back into the cytoplasm — this is
called the occluded state. Thus movement of the A domain and its associated
M1 helix are of critical importance in locking Ca 2+ ions into place within the
membrane. The M3 loop connection to the A domain becomes strained in this
process, in preparation for further movement in later parts of the cycle. In
addition, the calcium coordinating residues glu309 and asp800 become looser,
in preparation for removal of Ca 2+ to the lumen. The AMPPCP molecule is
held in place through (1) stacking interaction of the adenine ring with phe487
(N domain), (2) stabilizing the α - phosphate by hydrogen bonding to arg489
(N domain) and theβ - phosphate to arg560 (N domain), and (3) hydrogen
bonding between residues in the N and P domains. The distance from the γ -
phosphate to the phosphorylation residue, asp351, is 3.4 Å in the PDB: 1VFP
structure. The P domain Rossmann fold, consisting of a central seven - stranded
parallelβ - sheet and associated seven α - helices, moves so that some of its
component residues bond to theγ - phosphate and its associated Mg 2+ ions. The
movements bend the P domain into two almost orthogonal sectors, and this
part of the P domain is brought closer to the N domain.
Two X - ray crystallographic structures, PDB: 1T5T^95 and PDB: 1WPE,^100
describe the product of the next part of the cycle: Ca 2 E 1 · ATP to Ca 2 E 1 P · ADP.
Sorenson ’ s work in reference 95 shows that the Ca 2 E 1 · ATP and Ca 2 E 1 P · ADP
structures are quite similar. The Sorensen Ca 2 E 1 P · ADP model (PDB: 1T5T)
contains the ADP: AlF 4 − molecule, two calcium ions, two magnesium ions, and
a potassium ion. The potassium ion may be important for facilitating the
dephosphorylation of the enzyme in a later part of the cycle. The PDB: 1T5T
structure is intended to mimic the transition state for phosphoryl transfer to
asp351 with the tetrafl uoroaluminate ion acting as a nonreactive γ - phosphate
ion substitute. The differences between the Ca 2 E 1 · ATP (PDB: 1T5T) and
Ca 2 E 1 P · ADP (PDB: 1T5S) structures lie primarily at the phosphorylation site.
The Ca 2 E 1 · ATP model shows that the γ - phosphate and the asp351 residue are
not close enough for phosphate ion transfer. In this structure, an Mg 2+ ion has
coordinated to an asp351 carboxyl ion but not to theγ - phosphate of the
AMPPCP. In the Ca 2 E 1 P · ADP model, a second magnesium ion is found
between theα - and β - phosphates, possibly to stabilize the soon - to - be - leaving
ADP molecule, and the aluminum atom of the AlF 4 − moiety has been pulled