The Frasch Lab          


Vanadyl as a Probe of the Function of the F1-ATPase-Mg2+ Cofactor

The F1Fo ATP synthase uses the nonequilibrium transmembrane gradient derived from photosynthetic light energy or the oxidation of metabolites to drive the reaction

ADP + Pi « ATP + H2O beyond the point of equilibrium, and thereby maintain high cellular concentrations of ATP.  Many enzymes use ATP hydrolysis to return the ATP/ADP.Pi chemical gradient toward equilibrium as an energy source for catalysis.  Under some conditions, the enzyme can catalyze ATP hydrolysis that pumps protons in the reverse direction across the membrane.  However, the F1Fo from mitochondria and chloroplasts employ specific mechanisms to minimize this reverse reaction.

The intrinsic membrane Fo protein complex mediates proton translocation.  The extrinsic membrane F1 protein complex can be solubilized from the membrane where, in the absence of Fo, it can catalyze ATP hydrolysis.  Partial structures of soluble F1 have been determined from bovine mitochondrial F1 shown in Figure 1 (Abrahams et al. 1994), rat liver mitochondria (Bianchet et al., 1998), the thermophilic Bacillus PS3 (Shirakira et al., 1997), E. coli (Hausrath et al., 1999), and the F1Fo from yeast (Stock et al., 1999).  The three a and three b subunits, which fold in a similar manner, are arranged alternately like segments of an orange around a large portion of the g subunit.  The binding sites for the nucleotides are at the interfaces between a and b subunits.  The catalytic sites are predominantly in the b subunits with some contributions from groups on the a subunits, and conversely with the noncatalytic sites.

Figure 1. Cross section of the 2.8 Å crystal structure of F1 from bovine heart mitochondria

(Abrahams et al., 1994) that shows the asymmetry between the empty (bE) and Mg2+-ADP containing

(bDP) catalytic sites.







In the structure of F1 from bovine mitochondria, the three catalytic sites are asymmetric in that one contains bound Mg2+-ADP (bDP), one contains bound Mg2+-AMPPNP (an analog of ATP) (bTP), and one is empty (bE) (Abrahams et al., 1994).  Such asymmetry was predicted from experiments that served as the basis of the binding-change hypothesis (O’Neal & Boyer, 1984).  In this hypothesis, the enzyme adopts a conformation at one of the three catalytic sites in which ADP and phosphate are tightly bound.  In this high affinity conformation, the equilibrium of the ADP + Pi « ATP + H2O reaction is close to unity and, therefore, the synthesis of ATP is not the energy-requiring step.  Instead, input of energy from the proton gradient is used to drive a conformational change that promotes the release of newly synthesized ATP.  Because this conformational change can only occur when an adjacent empty catalytic site fills with substrate, the conformation of catalytic sites was thought to be staggered and work in a cooperative manner.

The observation that ATPase activity of soluble F1 drives the rotation of the g subunit provided insight to questions concerning how the catalytic subunits act in a cooperative manner (Duncan et al., 1995; Sabbert et al., 1996; Noji et al., 1997).  The enzyme is, in fact, a molecular motor.  In the most compelling demonstration of g rotation, an actin filament decorated with fluorescent groups was attached to the g subunit of F1.  Counterclockwise rotation of the actin driven by ATP hydrolysis was observed with a fluorescence microscope (Noji et al., 1997). Based on the direction of g rotation, the sequence of conformations of each catalytic site is bE®bTP®bDP®bE during ATP hydrolysis.

The asymmetry of the catalytic sites that is necessary for the binding-change mechanism depends on the g subunit and Mg2+.  In the absence of the g subunit and Mg2+, the crystal structure of the a3b3 complex from the thermophilic Bacillus PS3 has three fold symmetry (Shirakira et al., 1997).  In the rat liver F1 structure, which was crystallized in the absence of Mg2+, the a3b3 portion of the structure shows three fold symmetry despite the presence of the g subunit (Bianchet et al., 1998).

The Mg2+-induced asymmetry of the catalytic sites is also responsible for the differences in nucleotide affinity between these sites.  Using the bY331W mutant in E. coli F1, Weber et al. (1993) monitored the catalytic site occupancy from the quenching of the tryptophan fluorescence that occurs when nucleotide is bound.  In the absence of Mg2+, the three catalytic sites bind ATP with the same affinity (Weber et al., 1996).  However, when ATP binds as a complex with Mg2+, the affinity for nucleotides can differ by as much as five orders of magnitude.  This suggests that changes in the ligand environment of Mg2+ may be responsible for the observed differences in nucleotide affinity among the catalytic sites. 

Magnesium is difficult to study due to the lack of spectroscopic probes.  It is also difficult to identify in a protein crystal structure because it has a similar size and electron density to water. We have found vanadyl (VIV=O)2+ to be a valuable tool to characterize the environment of the Mg2+-binding sites in the F1-ATPase.  Vanadyl has four equatorial ligands and one axial ligand that is trans to the oxo group.  As a result, the VO2+ ligands adopt an octahedral configuration like that of Mg2+.  Due to the nuclear spin I = 7/2 and the anisotropy of the molecule, the single unpaired electron of VO2+ gives rise to an EPR signal that consists of two sets of 8 transitions from the fractions of molecules where the V=O bond is aligned parallel and perpendicular to the magnetic field (Figure 2).  The center of each group of 8 transitions and the spacing between them is determined by the values of g and A, respectively.  The magnitude of these values depends on the strength of the hyperfine coupling between the unpaired electron and the 51V nucleus.

The 51V-hyperfine coupling of VO2+ is sensitive to the types of groups coordinated at the equatorial positions (Chasteen, 1981).  Figure 2 shows the EPR spectra and hyperfine coupling of VO2+ when the four equatorial ligands are either water or hydroxyl groups.  The dependence of the position of the -5/2|| transition of the EPR signal on the type of groups ligated is also shown in the inset of Figure 2.  Each ligand contributes independently to the 51V-hyperfine coupling, enabling the complement of the four different ligands to be discerned (Chasteen,1981; Markham, 1984).  Based on the measured coupling constants of A|| from studies of model complexes, the hyperfine coupling for a given group of equatorial ligands can be calculated from the equation

                  A||calc = SniA||i/4      (Eq. 1)

where i counts the different types of equatorial ligand donor groups, ni (ni = 1-4) is the number of ligands of type i, and A|| is the measured coupling constant for the equatorial ligand donor group of type i (Chasteen, 1981).  Similar equations can be written for g|| and for Aiso, although the changes in A|| are the largest and most easily discerned.  As shown in Table 1, there are 210 possible sets of equatorial ligands to protein-bound VO2+ not including the smaller differences due to oxygen coordination by carboxyl groups like aspartate, carbonyl from groups like asparagine, and phosphate, and when rarely occurring groups that result from post-translational modification are excluded.

The intensity of the EPR signal from VO2+ free in frozen solution is suppressed by aggregation in HEPES buffer relative to that bound to the protein (Chasteen, 1981).  Furthermore, the signal intensity that remains from the unbound VO2+ in solution can be distinguished from that which arises from enzyme-bound VO2+ when examined at room temperature.  At this temperature the enzyme-bound VO2+ species will still resolve the parallel and perpendicular features usually observed at liquid nitrogen temperatures.  However, at room temperature, the VO2+ in solution will only show the rotationally averaged EPR signal.

Figure 2. Equatorial ligands determine 51V-hyperfine Parameters of VO2+. EPR spectra of VO2+

with waters (top) or hydroxyl groups (bottom) as equatorial ligands. The parallel transitions (–7/2||, -

5/2||, +3/2||, +5/2||, and +7/2||) that are not superimposed with perpendicular transitions are identified.

Inset: dependence of the -5/2|| transition on the type of equatorial ligands.













Purified spinach chloroplast F1-ATPase can use VO2+ as a functional cofactor for ATP hydrolysis (Houseman et al., 1994a).  At metal concentrations equal to or less than that of ATP, higher rates of ATPase activity were observed with VO2+ than with either Mg2+ or Ca2+.  In a manner similar to the free-metal inhibition caused by Mg2+, lower rates of ATPase activity are observed when the VO2+ concentration exceeds that of ATP.  The rates of VO2+-dependent photophosphorylation catalyzed by CF1Fo in thylakoids are also comparable to those observed with Mg2+ as a cofactor which indicates that VO2+ is a functional surrogate for Mg2+ in the intact enzyme.

Some studies suggested that some of the six total metal-binding sites known to exist on F1 were regulatory sites that did not complex with nucleotide (Schobert, 1993).  Although the crystal structure confirmed the presence of six metal binding sites, each was found to be coordinated to the bound nucleotides, and no additional regulatory metal-binding sites were observed (Abrahams et al., 1994).  Specific catalytic or noncatalytic binding sites of CF1 can be depleted and selectively filled with metal-nucleotide complexes (Bruist & Hammes, 1981; Shapiro et al., 1991).  Three nucleotide-binding sites were initially characterized that were designated Sites 1-3 in order of decreasing affinity for nucleotide.  The dissociation constants for ATP or ADP binding to catalytic Site 3 are in the micromolar range and can be removed from CF1 via gel filtration chromatography.  Site 2 binds Mg2+-ATP specifically and with very high affinity.  This noncatalytic site remains occupied even after extensive turnover of the enzyme.  Depletion of Mg2+-ATP from Site 2 requires partial unfolding of CF1 by precipitation in ammonium sulfate in the presence of EDTA.  Gel filtration chromatography of the resuspended protein with buffer that contains EDTA removes the metal nucleotide from this site.  Site 1 is catalytic and contains Mg2+-ADP that remains bound to the enzyme after extensive dialysis or gel filtration.  However, ADP or ATP in the medium will exchange with the nucleotide bound at this site.  Subsequently, an additional catalytic site was identified as Site 4.  This site binds nucleotide extremely tightly like Site 1, but can be distinguished from Site 1 in that only Site 1 can bind TNP-nucleotides (Soteropoulos et al., 1994).  Depletion of Sites 1 and 4 of metal-nucleotide complex is facilitated by removal of the e subunit of CF1 (Xiao & McCarty, 1989).

The catalytic activity of chloroplast F1 purified from Fo and the thylakoid membrane is latent, although ATPase activity can be activated via reduction of a disulfide bond on the g-subunit (cfFrasch, 1994).  The oxidation state of this disulfide provides one of multiple levels of interrelated regulatory mechanisms to minimize ATPase activity of CF1Fo in thylakoids in the absence of a proton gradient.  The dark decay of ATPase activity in thylakoids is accelerated by the addition of ADP that becomes tightly bound in the latent state (Smith & Boyer, 1976).  Inhibition of ATPase activity and the increase in affinity of the site for ADP only occur upon the addition of Mg2+ (Feldman & Boyer, 1985; Zhou et al., 1988).  Under these conditions, the bound ADP apparently forms a complex with Mg2+ that prevents further catalysis.  This entrapment of Mg2+-ADP serves a regulatory function in F1-ATPases from other organisms as well (Jault & Allison, 1993; Jault et al., 1994).

Figure 3a shows the -5/2|| transition of VO2+ bound as the VO2+-ATP complex to noncatalytic Site 2 of latent CF1 (Houseman et al., 1994b; Houseman et al., 1995).  This EPR signal designated species A was identified as the VO2+-ATP complex at Site 2 because this species was not depleted by extensive gel filtration, but required precipitation of the protein in ammonium sulfate and EDTA to remove the metal-nucleotide from this site.  The integrated intensity of Species A calibrated by atomic absorption spectroscopy of vanadium showed that this species saturated upon binding of about one VO2+ per CF1.  Species A exists as a 5.2 G doublet that results from superhyperfine coupling of a 31P nucleus from a single phosphate coordinated at an equatorial position.

Figure 3. The –5/2|| transition(s) of VO2+ bound to CF1 as: (a) VO2+-ATP at latent Site 2, (b) VO2+-ADP bound to latent Site 3, (c) VO2+-ADP bound to activated Site 3, (d) VO2+-ATP bound to activated Site 3; and (e) VO2+-ATP bound to Site 1 using CF1-ε.






Figure 3b shows the –5/2|| line of the EPR spectrum after the addition of one equivalent of VO2+-ADP to latent CF1 depleted of metal-nucleotide from catalytic Site 3 (Houseman et al., 1994b; Houseman et al., 1995).  The binding of VO2+-ADP to Site 3 gave rise to two EPR species designated B and C, with the Species B form predominant.  Activation of ATPase activity of this sample with dithiothreitol caused the signal intensity of species B to convert to species C (Figure 3c).  Though 31P splittings are not well defined in species C, line-shape analyses suggest that this species is comprised of a 1:2:1 triplet characteristic of a bidentate VO2+-ADP complex.  The conversion of species B to C upon activation indicates that EPR spectra of bound VO2+ can reveal structural changes in the metal-binding site of the enzyme in solution.  It is noteworthy that these experiments were carried out at pH 8 which is the pH of the stroma when transmembrane proton gradient maintained by light-driven electron transfer reactions has reached steady state and the F1Fo is activated by thioredoxin-mediated reduction of the g subunit disulfide.  In the dark, the stroma reverts to about pH 7 upon the collapse of the proton gradient when the enzyme reverts to the latent form.  At this pH, the VO2+-nucleotide bound to site 3 of latent CF1 is almost exclusively in the form of EPR species B.

When VO2+-ATP is bound to Site 3 of latent CF1, EPR species B and C are also observed in the same proportion as observed in Figure 3b (Houseman et al., 1995).  Upon activation, the signal intensity from species B is converted into species C and an additional species designated D (Figure 3d).  The transitions in Species D are significantly broader than observed in the other EPR species.  This may be the result of coordination by as many as all three phosphates, although the splitting due to 31P-superhyperfine coupling is not resolved in the spectrum.  The metal-nucleotides bound to high affinity Site 1 is more easily depleted from CF1 after the e-subunit has been removed (Xiao & McCarty, 1989).  Addition of an equivalent of VO2+-nucleotide to CF1-e results in the formation of EPR species designated E and F (Figure 3e).

Site-directed mutations of the P-loop threonine, a known metal ligand in bovine MF1 (Abrahams et al.,1994), were made to the b subunit of Chamydomonas CF1 to determine if perceptible changes in the EPR spectrum of VO2+ bound to catalytic site 3 could be observed (Chen et al., 1999).  Each of the three mutations examined was found to alter the 51V-hyperfine tensors of the bound VO2+ (Figure 4).  These changes were specific for the VO2+-nucleotide bound in the active conformation of Site 3(EPR species C).  In the latent, catalytically inactive conformation that results in EPR species B, the P-loop threonine mutants did not alter the EPR spectrum of VO2+-nucleotide complex bound to Site 3.  These data indicated that this technique could be used to identify specific groups as metal ligands.  It also demonstrated that the latent form of CF1 did not have the P-loop threonine as a ligand but that the activation process caused the insertion of this hydroxyl group into the coordination sphere to make the functional form of the enzyme. 

Figure 4. Changes in the EPR spectrum from VO2+-ATP bound to Site 3 of latent Chlamydomonas CF1 as the result of site-directed mutations of CF1bT168 (Chen et al., 1999). The parallel transitions of the EPR spectrum that do not overlap with perpendicular transitions are shown for wild type (a), T68D (b), T168C (c), and T168L (d). The transitions that result from EPR species B and C are indicated.




Site-directed mutations of the catch loop tyrosine (MF1bY311, CF1bY317 in Chlamydomonas) also changed the values of A|| and g|| from EPR species C but had little effect on those of species B when VO2+-nucleotide was bound to Site 3.  These data indicate that this residue was also a metal ligand in the activated but not the latent conformation, of catalytic Site 3.  Thus, the activated form of Site 3, the catalytic site with lowest affinity for the metal-nucleotide complex, contains two hydroxyl groups as ligands to the metal cofactor.

Three carboxyl groups, MF1bE188, MF1bE192, and MF1bD256, are within about 5 Å of the Mg2+ bound at the catalytic sites in bovine MF1 such that they can be considered candidates as metal ligands (Abrahams et al., 1994).  In this structure, the former carboxyl is hydrogen bonded to a water molecule that is 4.4 Å from the g-phosphate in bTP of bovine MF1 (Abrahams et al., 1994) such that this residue could alternatively activate the water to promote nucleophilic attack of the terminal phosphate.  The latter carboxyl is in the Walker homology B (WHB) motif that is conserved among several Mg2+-nucleotide binding proteins (58).  The WHB-aspartate has been suggested to hydrogen bond to a water molecule that is coordinated to the metal, or to coordinate to Mg2+ directly (Al-Shawi et al., 1992; Yohda et al., 1988; Weber et al., 1998).

None of the mutants to the carboxyl residues analogous to MF1bE188, MF1bE192, and MF1bD256 in the Chlamydomonas CF1 significantly changed the 51V-hyperfine coupling of EPR species C, the activated form of VO2+-nucleotide bound to Site 3 (Hu et al., 1996; 1999; Chen et al., 2000).  The 51V-hyperfine parameters of EPR species B, the latent form of Site 3, were affected only by mutations made to the WHB-aspartate (Hu et al., 1999).  Thus, the WHB aspartate is a ligand in the latent form of Site 3, but is displaced by the hydroxyl groups as ligands upon activation.  The other carboxyl residues are not equatorial ligands to VO2+ in Site 3 in either the latent or activated forms.

The same mutants to the Chlamydomonas CF1 residues analogous to MF1bE192, and MF1bD256 were examined for their effects on VO2+-ADP bound to Site 1 of CF1-e (Chen et al., 2000).  Depletion of the e subunit activates the enzyme and facilitates higher occupancy of Site 1 with metal-nucleotide complex.  The changes in EPR species E and F from catalytic Site 1 that result from these mutants indicated that both carboxyl residues are metal ligands to the metal-nucleotide bound at this site.

In the chloroplast enzyme, the catalytic site with highest affinity (CF1Site 4) binds the metal-nucleotide so tightly that it is difficult to study.  The analogous site in E. coli F1 (EF1Site 1) is much more easily loaded with metal-nucleotide (Weber et al., 1993).  The EPR spectrum of VO2+ bound as VO2+-ADP to EF1 is closely similar to EPR species D observed with CF1 (Figure 3d).  The types of equatorial ligands derived from the best fit of the 51V-hyperfine parameters to Equation 1 suggest the presence of 3 oxygens from either phosphate or carboxyl (asp or glu) and 1 oxygen from a hydroxyl (ser or thr).

The analysis of each mutant by EPR spectroscopy of the VO2+ bound to F1 not only indicates whether the 51V-hyperfine parameters differ from the wild type, but also provides an estimate of the types of groups that the protein used as ligands in its attempt to compensate for the mutation based on Equation 1.  These analyses provide insight into the evolutionary pressures that resulted in the metal ligands used by the F1.  For example, mutation of either a hydroxyl or a carboxyl ligand to a sulfhydryl typically results in the sulfhydryl group as a metal ligand, but will often cause a second ligand to be substituted for a coordinated water. 

Mutations of hydroxyl groups for carboxyl groups or the converse tend to cause the greatest disruption to the coordination sphere of the metal bound to the catalytic site.  The most likely reason for this is that the enzyme uses the hydroxyl ligands and carboxyl ligands in the conformations with low and high affinity for nucleotide, respectively.  The conversion between low and high affinity conformations appears to be critical to the function of the enzyme.

Figure 5 summarizes the metal ligands identified in the sequential progression of conformational states of each catalytic site during a catalytic cycle of ATPase hydrolysis.  The amino acids that serve as ligands in one or more conformations are shown in three-dimensional space approximately as they exist in the catalytic site.  These groups roughly define the corners of a cube in the catalytic site.  The equatorial ligands to VO2+ in each conformation are those residues shown at the corners of the blue square.  Sidechains indicated with question marks are the best estimate of the ligand in that conformation at this time.  In some cases, more work is also required to identify the number of coordinated phosphates.

When the metal-ATP complex initially binds in the low affinity, Site 3 conformation, hydroxyl groups from the P-loop threonine and the catch loop tyrosine form equatorial ligands to VO2+ (Figure 5A).  These groups define the bottom plane of the binding pocket.  Conversion to the high affinity (EF1Site 1, CF1Site 4) conformation (Figure 5B) causes the insertion of a carboxyl group that is likely the Walker B-aspartate (MF1bD256), and causes the displacement of the catch loop tyrosine from the coordination sphere.  Upon formation of CF1Site 1 (Figure 5C), the metal-nucleotide moves to the top of the binding pocket where the carboxyls analogous to MF1bD256 (WHB-aspartate) and MF1b193 are equatorial metal ligands.  The hydroxyl group (probably the P-loop threonine) is not an equatorial ligand to VO2+ in this conformation but may be an axial metal ligand.  In Chlamydomonas CF1Site 1 the VO2+-ADP can exist in two conformations that differ only by one ligand that is either a tyrosine hydroxyl or a water molecule (Figure 5C and D).  If the tyrosine ligand proves to be the catch loop tyrosine, this form may be an intermediate state between the high affinity site 1 and the low affinity site 3 in the final conformational change that completes the catalytic cycle as shown in Figure 5.  Alternatively, one of the conformations in Chlamydomonas CF1Site 1 may represent the nonfunctional (entrapped) metal-ADP complex thought to regulate the enzyme (Drobinshaya et al., 1985).  Recent evidence suggests that this is the site at which the entrapped metal-ADP complex binds (Digel et al., 1998).  More experiments are necessary to resolve these possibilities.

Figure 5. Changes in metal ligands during the catalytic cycle of F1-ATPase identified by EPR spectroscopy of VO2+. The equatorial ligands in each conformation are indicated as the residues at the corners of the blue square. The box shown provides an indication of the approximate three dimensional relationship of the residues. Positive identification of the residues that give rise to EPR species D awaits confirmation with mutagenesis studies.








In the crystal structure of bovine MF1 (Abrahams et al., 1994), the helical domain that contains the DELSEED sequence is significantly closer to the P-loop in the catalytic sites that contain Mg2+-nucleotide than is observed in the empty catalytic site.  This implies that the helical bundle oscillates up and down during catalysis.  Since the Mg2+-nucleotide complex is tethered to the helical bundle via p-bonding of the adenine ring with MF1bY345, the upward movement of the helical bundle during the transition of the catalytic site from low to high affinity should also move the metal-nucleotide in the same direction.  This movement is consistent with the change in metal ligands from the bottom to the top of the metal-binding pocket shown in Figure 5.  Carboxyl groups coordinate hard metals such as Mg2+ with higher affinity than do hydroxyl groups.  It is noteworthy that the carboxyl and hydroxyl ligands are found in the high and low affinity metal-nucleotide conformations, respectively. 

The presence of the catch loop tyrosine as a metal ligand in the low affinity site may also contribute significantly to the mechanism of ATP synthesis due to its location.  In the bovine MF1 structure (Abrahams et al., 1994), hydrogen bonds between g subunit residues MF1gR254 and gQ255 and b subunit residues MF1bD316, MF1bT318, and MF1bD319 form the “catch” of the catch loop.  Under conditions of ATP synthesis, the transmembrane proton gradient will provide a constant torque on the g subunit.  Rotation of this subunit when the catalytic sites are empty would dissipate the proton gradient in a nonproductive manner.  However, the energy in the hydrogen bonds between the b and g subunits including those at the catch loop when one of the three catalytic sites is empty is approximately sufficient to prevent this rotation.  When Mg2+-ADP binds to the empty catalytic site, metal coordination by the bulky catch loop tyrosine may deform the catch loop sufficiently to break the hydrogen bonds at this location, thereby triggering g subunit rotation and conformational changes of all three catalytic sites.  Because all three sites change in response to the rotation, one catalytic site will necessarily form the conformation that leads to product dissociation.  Hydrogen bonds between the g subunit and the catch loop of this newly formed empty site would limit the rotation of the g subunit to 120º per catalytic event and provide an escapement mechanism to insure that proton flux through Fo only occurred when the catalytic sites synthesized ATP.

Vanadyl has been a useful tool to characterize the changes in the ligands of the metal cofactor through the catalytic cycle.  As these changes are more completely discerned, the molecular basis for the Mg2+ requirement for the catalytic site asymmetry and the associated differences in affinity for nucleotides that enables the enzyme to maintain the chemical gradient of ATP that provides cellular energy are becoming more clear. 


This work was supported by NIH grant GM-50202.


Abrahams, J. P., Leslie, A. G. W., Lutter, R., and Walker, J. E. (1994) Nature 370, 621-628.

Al-Shawi, M. K., and Senior, A. E. (1992). Biochemistry 31, 878-885

Bianchet, M. A., Hullihen, J., Pedersen, P. L., and Amzel, L. M. (1998) Proc. Natl. Acad. Sci. (USA) 95, 11065-11070.

Bruist, M. F. and Hammes, G. G. (1981) Biochemistry 20, 6298-6305

Chasteen, N. D. (1981) in, Biological Magnetic Resonance, (Berliner, L. and Reuben, J., eds.), Plenum Press, pp 53-119

Chen, W., Hu, C.-Y., Crampton, D. J., Frasch, W. D. (2000) Biochemistry 39, 9393-9400

Chen, W., LoBrutto, R., and Frasch, W. D. (1999) J. Biol. Chem. 274, 7089-7094

Digel, J. G., Moore, N. D., and McCarty, R. E. (1998) Biochemistry 37, 17209-17215.

Drobinshaya, I. Y., Kozlov, I. A., Murataliev, M. B., and Vulfson, E. N. (1985) FEBS Lett. 182, 419-424.

Duncan, T. M., Bulygin, V. V, Zhou, Y., Hutcheon, M. L., and Cross, R. L. (1995) Proc. Natl. Acad. Sci. (USA) 92, 10964-10968.

Feldman, R. I., and Boyer, P. D. (1985) J. Biol. Chem. 260, 13088-13094

Frasch, W. D. (1994) in, The Molecular Biology of Cyanobacteria, (Bryant, D., ed.), Kluwer, The Netherlands, pp. 361-380

Hausrath, A. C., Gruber, G., Matthews, B. W. and Capaldi, R. A. (1999) Proc. Natl. Acad. Sci. USA 96, 13697

Houseman, A. L. P, LoBrutto, R., Bell, M., and Frasch, W. D. (1995) in, Photosynthesis: from Light to Biosphere, (P. Mathis, ed.), Kluwer Acad. Publ., Netherlands, vol. III, 127-130

Houseman, A. L. P., Morgan, L., LoBrutto, R., and Frasch, W. D. (1994) Biochemistry 33, 4910-4917.

Houseman, A. L., LoBrutto, R., and Frasch, W. D. (1994). Biochemistry 33, 10000-10006

Houseman, A. L., LoBrutto, R., and Frasch, W. D. (1995). Biochemistry 34, 3277-3285

Hu, C.-Y., Chen, W. and Frasch, W. D. (1999), J. Biol. Chem. 274, 30481-30486

Hu, C.-Y., Houseman, A. L. P., Morgan, L., Webber, A. N., and Frasch, W. D. (1996) Biochemistry 35, 12201-12211

Jault, J.-M. and Allison, W. S. (1993) J. Biol. Chem. 268, 1558-1566

Jault, J.-M. and Allison, W. S. (1994) J. Biol. Chem. 269, 319-325

Markham, G. D. (1984) Biochemistry 23, 470-478

Noji, H., Yasuda, R., Yoshida, M., Kinosita, K. (1997) Nature 386, 299-302.

O’Neal, C. & Boyer, P. D. (1984) J. Biol. Chem. 259, 5761-5767.

Sabbert, D., Engelbrecht, S., and Junge, W. (1996) Nature 381, 623-625.

Schobert, B. (1993) Biochemistry 32, 13204-13211.

Shapiro, A. B., Huber, A. H., and McCarty, R. E. (1991) J. Biol. Chem. 266, 4194-4200

Shirakihara, Y., Kambara, M., Saika, K., Kagawa, Y. and Yoshida, M. (1997) Structure 5, 825-836.

Smith, J. P. and Boyer, P. D. (1976) Proc. Natl. Acad. Sci. USA 73, 4314-4318

Soteropoulos, P., Ong, A. M., and McCarty, R. E. (1994) J. Biol. Chem. 269, 19810-19816

Stock, D., Leslie, A. G. W. and Walker, J. E. (1999) Science 286, 1700.

Walker, J. E., Saraste, M., Runswick, M., J., and Gay, N. J. (1982). EMBO J. 1, 945-951

Weber, J. & Senior, A. E. (1997) Biochim. Biophys. Acta 1319, 19-58.

Weber, J., Hammond, S. T., Wilke-Mounts, S. and Senior, A. E. (1998) Biochemistry 37, 608-614

Weber, J., Wilke-Mounts, S., Lee, R.S.-F., Grell, E., & Senior, A. E. (1993) J. Biol. Chem. 268, 20126-20133.

Xiao, J., and McCarty, R. E. (1989) Biochim. Biophys. Acta 976, 203-209

Yohda, M., Ohta, S., Hisabori, T., and Kagawa, Y. (1988) Biochim Biophys Acta 933, 156-164

Zhou, J.M., Xue, Z., Du, Z., Melese, T., and Boyer, P.D. (1988) Biochemistry 27, 5129-5135

Figure 1.  Cross section of the 2.8 Å crystal structure of F1 from bovine heart mitochondria (Abrahams et al., 1994) that shows the asymmetry between the empty (bE) and Mg2+-ADP containing (bDP) catalytic sites.

Figure 2.  Equatorial ligands determine 51V-hyperfine Parameters of VO2+.  EPR spectra of VO2+ with waters (top) or hydroxyl groups (bottom) as equatorial ligands.  The parallel transitions (–7/2||, -5/2||, +3/2||, +5/2||, and +7/2||) that are not superimposed with perpendicular transitions are identified.  Inset: dependence of the -5/2|| transition on the type of equatorial ligands.

Figure 3.  The –5/2|| transition(s) of VO2+ bound to CF1 as: (a) VO2+-ATP at latent Site 2, (b) VO2+-ADP bound to latent Site 3, (c) VO2+-ADP bound to activated Site 3, (d) VO2+-ATP bound to activated Site 3; and (e) VO2+-ATP bound to Site 1 using CF1-e.

Figure 4. Changes in the EPR spectrum from VO2+-ATP bound to Site 3 of latent Chlamydomonas CF1 as the result of site-directed mutations of CF1bT168 (Chen et al., 1999).  The parallel transitions of the EPR spectrum that do not overlap with perpendicular transitions are shown for wild type (a), T68D (b), T168C (c), and T168L (d).  The transitions that result from EPR species B and C are indicated. 

Figure 5. Changes in metal ligands during the catalytic cycle of F1-ATPase identified by EPR spectroscopy of VO2+.  The equatorial ligands in each conformation are indicated as the residues at the corners of the blue square.  The box shown provides an indication of the approximate three-dimensional relationship of the residues.  Positive identification of the residues that give rise to EPR species D awaits confirmation with mutagenesis studies.


Last updated 2/20/2006

Frasch Lab

Arizona State University