202 GROUP I AND II METALS IN BIOLOGICAL SYSTEMS
using Na + /K + - ATPase from rat (PDB: 1MO7, without ATP, 1MO8, with ATP).^14
(Note that the PDB accession numbers contain the letter “ O ” and not the
number zero “ 0. ” ) Comparison of the solid - state versus solution state struc-
tures reveals similarities and differences. The NMR structure features a six -
stranded β - sheet rather than the seven – stranded one found for the X - ray
structure, and the NMR structure α - helices are shorter. Disorder found in the
same regions of both structures (residues glu397 – ala402 and two residues in
the phe475 – lys480 region) indicates that the methods used to study the system
do not produce these structural artifacts. However, ATP was found to bind to
the Na + /K + - ATPase N - domain, albeit weakly, in the solution structure (PDB:
1MO8). Nuclear Overhauser enhancements (NOEs) showed interactions
between the ATP molecule and residues phe475, lys480, ala503, and leu546,
indicating that these residues fi rmly anchor the ATP ribose moiety in the N -
domain. In addition, phe475 provides good stacking interaction with ATP ’ s
adenine ring. Only one hydrogen bond between ATP and the N - domain could
be unequivocally assigned, namely between amino protons of the adenine base
and carboxy oxygen O ε 1 of gln482 (3.6 Å ). The largest N - domain conforma-
tional changes occur in the beta sheetsβ 1 and β 6 (the so - called hinge region)
that are bent in the native protein (PDB: 1MO7) and straight in the ATP -
bound form (PDB: 1MO8). Since the β 1 and β 6 regions connect the N - domain
to the P - domain, change in this region may affect the conformations of both
the N - and P - domains in the intact Na + /K + - ATPase enzyme. Figure 2(d) of
reference 14 illustrates the change. Unfortunately, the numbering system found
in theNature Structural Biology article (reference 14) does not agree with that
found in the downloaded PDB: 1MO8 data. However, if one adds the number
7 to each residue number given in theNature Structural Biology article, one
fi nds the referred - to residues — that is, phe475 becomes phe482 and gln482
becomes gln489 when comparing the publication ’ s numbering to the PDB data
downloaded for visualization. Hilge and co - workers believe that the nucleo-
tide - free N - domain (1MO7) corresponds to the E 2 state in the intact enzyme
in view of its structural similarity to the E2(TG) state (PDB: 1IWO) for Ca 2+ -
ATPase. (See discussion of this enzyme in Section 6.4.2 and Figure 6.28 .)
Additionally, they believe that ATP binding causes a shift toward the E1 state
(PDB: 1SU4 in Section 6.4.2.1 and Figure 6.28). Finally, the authors speculate
that, in analogy to Ca 2+ binding in Ca 2+ - ATPase, binding of Na + ions causes
the fi nal N - domain conformational change that brings the γ - phosphate into
the shallow cavity where it binds to the invariant aspartate residue.
In a molecular modeling study, Amler and co - workers established eight
amino acid residues that form the ATP binding site in the intact Na + /K + -
ATPase enzyme.^12 These workers expressed point - mutated versions of a part
of the H 4 – H 5 loop (residues leu354 – ile604) of the Na + /K + - ATPase enzyme that
contains the ATP binding residues. The so - called H 4 – H 5 loop contains N - and
C - terminal subdomains of the Na + /K + - ATPase P - domain and the entire N -
domain. This loop has a self - supporting structure, retains the ability to bind
ATP, but lacks interactions with other Na + /K + - ATPase domains, and is not