Fluorescent epsilon-ATP analogues for probing physicochemical properties of proteins. Synthesis, biochemical evaluation, and sensitivity to properties of the medium.

Despite the significance of the elucidation of proteins' physicochemical parameters to understand various molecular phenomena, direct methods for measuring these parameters are not readily available. Here, we propose the use of 8-[p-amino-Ph]-epsilon-ATP, 3b, as a fluorescent probe for the elucidation of physicochemical parameters of binding sites in certain proteins. We synthesized novel fluorescent nucleotide analogues based on an extension of the epsilon-ATP scaffold. These analogues bear a primary or tertiary p-amino-phenyl moiety on the etheno-bridge. We explored the recognition of the fluorescent analogues by the target proteins: P2Y(1)-receptor (P2Y(1)-R) and NTPDase1. Based on the high affinity to the P2Y(1)-R (EC(50) 100nM), 3b proved a suitable probe for the investigation of this receptor. Next, we elucidated the dependencies of the absorption and emission spectra of 3b on environmental parameters, for establishing correlation equations. These equations will help determine the properties of the ATP-binding site from the spectral data of the protein-bound 3b. For this purpose, the sensitivity of the probe to acidity, dielectricity, H-bonding, viscosity, and to correlation between these parameters was determined. Thus, the pH-dependence of 3b emission intensity is bell shaped. At pH2.8 the quantum yield (phi) is enhanced 150-fold, as compared to neutral pH. The basic nitrogen atoms of 3b were assigned and pK(a) values were determined. A linear relationship was found between log phi and log viscosity, however, emission maxima (lambda(max)) remained constant. A linear relationship was found between both phi and lambda(max) and dielectricity, as measured in protic or aprotic solvents of comparable viscosity. pK(a)-like values were measured in acid-titrated alcohols with varying dielectricity but comparable viscosity, or with varying viscosity but comparable dielectricity. An inverse relationship and a linear relationship were found between the pK(a) values of 3b and the medium dielectricity and viscosity, respectively. These correlations help the calibration of properties of a protein ATP-binding site.

Transient kinetic analysis of N-phenylmaleimide-reacted myosin subfragment-1.

Alkylation of myosin's Cys-707 (SH1) and Cys-697 (SH2) has profound consequences for myosin's ability to interact with actin and hydrolyze MgATP. Pre-steady-state measurements of myosin-S1 alkylated at SH1 and SH2 by N-phenylmaleimide (NPM) in the presence of ATP were taken to identify the steps of the reaction that are altered. It was found that the rate constant most affected by this modification is the apparent rate of the ATP hydrolysis step. This rate constant is reduced 20000-fold, an effect comparable in magnitude to the effect of the same modification on the binding of MgATP to S1 or acto-S1 [Xie, L., and Schoenberg, M. (1998) Biochemistry 37, 8048]. In contrast, the rate constants of phosphate release and dissociation of acto-S1 by ATP were reduced <20-fold. For unmodified S1, the enhancement of fluorescence seen after addition of ATP had the same rate constant as the ATP hydrolysis step (S1.ATP if S1.ADP.Pi) measured by single-turnover experiments in a quench-flow experiment. This is consistent with results previously observed [Johnson, K. A., and Taylor, E. W. (1978) Biochemistry 17, 3432]. However, NPM-modified S1 exhibited virtually no fluorescence enhancement upon ATP binding. This provides further evidence that M.ATP is the predominant intermediate of NPM-S1-catalyzed ATP hydrolysis.

Kinetics of binding and hydrolysis of a series of nucleoside triphosphates by actomyosin-S1. Relationship between solution rate constants and properties of muscle fibers.

We have measured the steady state kinetics of hydrolysis and presteady state kinetics of binding of the nucleoside triphosphate GTP, CTP, aza-ATP (1-N6-etheno-2-aza-ATP), and ATP by rabbit skeletal actomyosin-S1. The maximum rates of steady state hydrolysis at 10 degrees C at low ionic strength are: CTP, 1.9 s-1 > ATP, 1.3 s-1 > aza-ATP, 0.19 s-1 > GTP, 0.03 s-1. A similar dependence of the rate of steady state hydrolysis upon nucleotide structure has been observed in isometrically contracting muscle fibers in the accompanying paper (Pate, E., Franks-Skiba, K., White, H., and Cooke, R. (1993) J. Biol. Chem. 268, 10046-10053) which strongly suggests that the same biochemical step that limits the maximum rate of hydrolysis of nucleoside triphosphates by actomyosin-S1 in solution also limits the rate of hydrolysis by isometrically contracting muscle fibers. The apparent second order rate constants for the dissociation of actomyosin-S1 by nucleoside triphosphates at 10 degrees C are: ATP, 2.7 x 10(6) M-1 s-1 > aza-ATP, 3.4 x 10(5) M-1 s-1 > GTP, 2.5 x 10(5) M-1 s-1 > CTP, 1.4 x 10(5) M-1 s-1. There is an excellent correlation between the second order rate constant for the dissociation of actomyosin-S1 in solution and the dependence of shortening velocity in glycerinated muscle fibers upon the concentration for ATP, aza-ATP, and CTP (as per accompanying article; Pate et al., 1993). We have used the second order rate constants obtained in solution for the dissociation of actomyosin-S1 by these nucleotides and shortening velocity data obtained with the same nucleoside triphosphates in glycerinated psoas fibers in the accompanying article (Pate et al., 1993) to determine the average distance over which cross-bridges remain attached during unloaded shortening to be 5-12 nm.

The use of differing nucleotides to investigate cross-bridge kinetics.

We have investigated the ability of the nucleotides GTP, CTP, and 1-N6-etheno-2-aza-ATP (aza-ATP) to support contraction of chemically skinned rabbit psoas fibers. Working at 10 degrees C, millimolar concentrations of all nucleotides relaxed fibers in the absence of calcium. In active fibers, GTP served as a very poor substrate with isometric tension, isometric GTPase rate, and maximum shortening velocity (Vmax) all less than 10% of those obtained with ATP. Aza-ATP was only a slightly better substrate. CTP, on the other hand, was an effective substrate with mechanical parameters which were 65-100% those obtained with ATP, and with a hydrolysis rate that exceeded that of ATP. For all three ligands, Vmax followed Michaelis-Menten saturation behavior with values for Km which were from 2.5 to 12 times greater than that for ATP, showing that the analogs bound slowly to myosin in the fibers. Increasing concentrations of orthophosphate inhibited tension with CTP, to a lesser extent with aza-ATP, but not all with GTP. A combination of the mechanical data obtained in fibers with the kinetic data obtained in solution (White, H.D., Belknap, B., and Jiang, W. (1993) J. Biol. Chem. 268, 10039-10045) is used to better define the actomyosin interaction in fibers.

Mechanism of the interaction of human platelet profilin with actin.

We have reexamined the interaction of purified platelet profilin with actin and present evidence that simple sequestration of actin monomers in a 1:1 complex with profilin cannot explain many of the effects of profilin on actin assembly. Three different methods to assess binding of profilin to actin show that the complex with platelet actin has a dissociation constant in the range of 1 to 5 microM. The value for muscle actin is similar. When bound to actin, profilin increases the rate constant for dissociation of ATP from actin by 1,000-fold and also increases the rate of dissociation of Ca2+ bound to actin. Kinetic simulation showed that the profilin exchanges between actin monomers on a subsecond time scale that allows it to catalyze nucleotide exchange. On the other hand, polymerization assays give disparate results that are inconsistent with the binding assays and each other: profilin has different effects on elongation at the two ends of actin filaments; profilin inhibits the elongation of platelet actin much more strongly than muscle actin; and simple formation of 1:1 complexes of actin with profilin cannot account for the strong inhibition of spontaneous polymerization. We suggest that the in vitro effects on actin polymerization may be explained by a complex mechanism that includes weak capping of filament ends and catalytic poisoning of nucleation. Although platelets contain only 1 profilin for every 5-10 actin molecules, these complex reactions may allow substoichiometric profilin to have an important influence on actin assembly. We also confirm the observation of I. Lassing and U. Lindberg (1985. Nature [Lond.] 318:472-474) that polyphosphoinositides inhibit the effects of profilin on actin polymerization, so lipid metabolism must also be taken into account when considering the functions of profilin in a cell.

Photoaffinity labeling of rabbit muscle fructose-1,6-bisphosphate aldolase with 8-azido-1,N6-ethenoadenosine 5'-triphosphate.

Steady-state kinetic measurements have shown that 8-azido-1,N6-ethenoadenosine 5'-triphosphate (8-N3-epsilon ATP) can be noncovalently bound to rabbit muscle fructose 1,6-bisphosphate aldolase with Ki = 0.075 mM at pH 8.5. This binding is purely competitive with substrate and occurs at the strong binding site for mononucleotides. Photoaffinity labeling of aldolase in the presence of 8-azido-1,N6-ethenoadenosine 5'-triphosphate results in inactivation of the enzyme. Aldolase is protected against modification in the presence of the inhibitors hexitol 1,6-bisphosphate or ATP. The labeling is saturable, and a good correlation is observed between the loss of enzymatic activity and the incorporation of 8-N3-epsilon ATP into aldolase. In addition, aldolase loses its ability to bind to phosphocellulose following modification. Digestion of labeled protein with trypsin, chymotrypsin, and cyanogen bromide revealed substantial modification of peptide 259-269. Thr-265 was identified as the residue that was covalently modified by 8-N3-epsilon ATP. On the basis of these results and other data we propose a model for the mononucleotide binding site.

Utilization of 1,N6-etheno-2'-deoxyadenosine 5'-triphosphate during DNA synthesis on natural templates, catalyzed by DNA polymerase I of Escherichia coli.

To test whether vinyl chloride-induced mutagenesis might involve ambiguous base pairing of 1,N6-etheno-adenine (epsilon A) during DNA synthesis, we examined the base pairing potential of epsilon dATP during DNA synthesis catalyzed by Escherichia coli DNA polymerase I (Klenow fragment). An electrophoretic assay of chain elongation was used to assess the degree to which epsilon dATP could substitute for each of the normal dNTPs during elongation of a primer annealed to a bacteriophage template. Despite the fact that the etheno bridge completely blocks normal Watson-Crick pairing of epsilon A with T, we observed that epsilon dATP could substitute for dATP during primer elongation (although inefficiently). In addition, detectable substitution of epsilon dATP for dGTP and dCTP occurred, indicating that epsilon A exhibits ambiguous base pairing properties. The relative ease of epsilon dAMP incorporation (opposite template T, C and G) appeared to vary considerably at different positions along the template. The major form of epsilon A incorporation (replacement of A) was confirmed by measurements of epsilon dATP----epsilon dAMP turnover (a commonly used method for detecting misincorporation), and also by the demonstration that epsilon A was present in enzymatic hydrolysates prepared from DNA that was synthesized with epsilon dATP replacing dATP. A model for ambiguous base pairing of epsilon dATP is proposed, in which incorporation occurs via the protonated, syn form of epsilon dATP.

Kinetic mechanism of 1-N6-etheno-2-aza-ATP hydrolysis by bovine ventricular myosin subfragment 1 and actomyosin subfragment 1. The nucleotide binding steps.

The large change in fluorescence emission of 1-N6-etheno-2-aza-ATP (epsilon-aza-ATP) has been used to investigate the kinetic mechanism of etheno-aza nucleotide binding to bovine cardiac myosin subfragment 1 (myosin-S1) and actomyosin subfragment 1 (actomyosin-S1). The time course of nucleotide fluorescence enhancement observed during epsilon-aza-ATP hydrolysis is qualitatively similar to the time course of tryptophan fluorescence enhancement observed during ATP hydrolysis. In single turnover experiments, the nucleotide fluorescence rapidly increases to a maximum level, then decreases with a rate constant of 0.045 s-1 to a final level, which is about 30% of the maximal enhancement; a similar fluorescence enhancement is obtained by adding epsilon-aza-ADP to cardiac myosin-S1 or actomyosin-S1 under the same conditions (100 mM KCl, 10 mM 4-morpholinepropanesulfonic acid, 5 mM MgCl2, 0.1 mM dithiothreitol, pH 7.0, 15 degrees C). The kinetic data are consistent with a mechanism in which there are two sequential (acto)myosin-S1 nucleotide complexes with enhanced nucleotide fluorescence following epsilon-aza-ATP binding. The apparent second order rate constants of epsilon-aza-ATP binding to cardiac myosin subfragment 1 and actomyosin subfragment 1 are 2-12 times slower than those for ATP. Actin increases the rate of epsilon-aza-ADP dissociation from bovine cardiac myosin-S1 from 1.9 to 110 s-1 at 15 degrees C which can be compared to 0.3 and 65 s-1 for ADP dissociation under similar conditions. Although there are quantitative differences between the rate and equilibrium constants of epsilon-aza- and adenosine nucleotides to cardiac actomyosin-S1 and myosin-S1, the basic features of the nucleotide binding steps of the mechanism are unchanged.

Kinetic mechanism of 1-N6-etheno-2-aza-ATP and 1-N6-etheno-2-aza-ADP binding to bovine ventricular actomyosin-S1 and myofibrils.

The fluorescence emission of 1-N6-etheno-2-aza-ATP (epsilon-aza-ATP) at 410-460 nm is enhanced approximately 8-fold upon mixing substoichiometric concentrations of epsilon-aza-ATP with bovine cardiac actomyosin-S1 or myofibrils. The time course of nucleotide fluorescence measured in a front face stopped flow cell upon mixing epsilon-aza-ATP with bovine cardiac myofibrils ([Ca2+] less than 10(-7) M) is essentially the same as that with bovine cardiac actomyosin subfragment-1. In single turnover experiments, the fluorescence rapidly rises to a maximum value, then decreases with a rate constant of 0.04 s-1 at 0 degree C to a final value that is approximately twice the level of the unbound nucleotide. At concentrations of epsilon-aza-ATP greater than 40 microM the kinetics of epsilon-aza-ATP binding is clearly biphasic for both actomyosin-S1 and myofibrils. At 0 degree C, the rate of the more rapid phase is proportional to nucleotide concentration and has a second order rate constant of 1.7 X 10(5) M-1 s-1; the rate of the slower phase extrapolates to a maximum of 4-5 s-1 at high nucleotide concentration. The rate constants for dissociation of epsilon-aza-ADP from bovine cardiac actomyosin-S1 and myofibrils were measured from the decrease in epsilon-aza-ADP fluorescence enhancement observed upon displacement by ATP to be 20 and 18 s-1, respectively, at 0 degree C. These results indicate that most of the cross-bridges in cardiac myofibrils are bound to actin and that the geometric constraints imposed upon the interaction of actin and myosin by the three-dimensional structure of the myofibril do not modify the kinetics of epsilon-aza-ATP binding or epsilon-aza-ADP dissociation.

Angle of active site of myosin heads in contracting muscle during sudden length changes.

The change in orientation of myosin crossbridges in contracting muscle during sudden length changes was examined by fluorescence polarization. This study used a fluorescent ATP analogue, 1,N6-etheno-2-aza-ATP(epsilon-2-aza-ATP) as a probe. Its fluorescence is considerably enhanced upon binding with myosin and is dependent on the chemical state of the myosin-nucleotide complex in muscle. The results showed that nucleotides bound to crossbridges in the intermediate attached state (presumably AM-epsilon-2-aza-ADP-Pi) during isometric contraction are highly oriented at the same angle as that of AM in rigor with bound epsilon-2-aza-ADP. Furthermore the orientation of nucleotides bound to crossbridges in the attached state is not altered during sudden changes in length of isometrically contracting muscle. The results of this time-resolved measurement support the conclusion obtained from a previous steady-state experiment that change in axial orientation of the active site of the myosin head is not involved in force generation.

Predominant attached state of myosin cross-bridges during contraction and relaxation at low ionic strength.

The chemical states of a cross-bridge--nucleotide complex were studied using a fluorescent ATP analogue, 1-N6-etheno-2-aza-ATP(epsilon-2-aza-ATP). The fluorescence of epsilon-2-aza-ATP at specific emission wavelengths was enhanced by 12.5 times upon binding to myosin in a relaxed muscle and the fluorescence from the resultant myosin(M)-epsilon-2-aza-ADP-Pi intermediate was 2.5 times greater than that from a M-epsilon-2-aza-ADP complex. Similar enhancements of the fluorescence of epsilon-2-aza-ATP and epsilon-2-aza-ADP were observed upon binding to heavy meromyosin in solution. Binding of F-actin did not change the fluorescence of epsilon-2-aza-ATP or epsilon-2-aza-ADP bound to heavy meromyosin. When a muscle went from a relaxed state to a state of isometric contraction or contraction with shortening, the fluorescence intensity decreased only slightly or not at all, i.e. the fluorescence of nucleotides bound to most of the myosin heads during contraction is the same as that of the M-epsilon-2-aza-ADP-Pi intermediate. These results suggest that an actomyosin(AM)-epsilon-2-aza-ADP-Pi intermediate is the predominant attached state during contraction. When the ionic strength of the relaxing solution was decreased, cross-bridges formed at 6 degrees C without tension generation. At 20 degrees C, a large tension was produced although the shortening velocity was negligibly small or zero. The fluorescence intensity decreased by 15% at 20 degrees C but only a small decrease of 3% was observed at 6 degrees C, suggesting that the predominant complexes in the attached state were AM-epsilon-2-aza-ATP and/or AM-2-aza-ADP-Pi at 6 degrees C and AM-epsilon-2-aza-ADP at 20 degrees C. Thus, the identification of the actomyosin-nucleotide complexes existing before and after the force-generating step lent further support to the conclusion that the sliding force is generated by conformational changes in actomyosin when the (epsilon-2-aza-)ADP-Pi complex is bound to it.

Detection of the conformational change in the catalytic site of adenosine triphosphatase from beef liver mitochondria by affinity labeling with the dialdehyde derivative of ethenoadenosine triphosphate.

Beef liver mitochondrial F1ATPase was inactivated by the 2',3'-dialdehyde derivative of ethenoATP (epsilon ATP) in a pseudo-first order reaction. The kinetics of protection of the enzyme against inactivation by various nucleoside triphosphates (NTPs) revealed that the dial-epsilon ATP was bound to the catalytic site as an affinity label. Certain anions (sulfate or bicarbonate) were ineffective for protection. In the early phase of the reaction, inactivation was due to the binding of 1 mol dial-epsilon ATP per mol enzyme. In this phase, dial-epsilon ATP bound exclusively to the subunit beta of the enzyme, indicating that the catalytic site is in this subunit. The fluorescence of the ethenoadenosine moiety, bound exclusively to the subunit beta of the enzyme, was measured as a conformational probe of the catalytic site region. Addition of ATP or CTP to the labeled enzyme resulted in a decrease in the fluorescence intensity. GTP and other NTPs were less effective than ATP or CTP. The anions (sulfate of bicarbonate) suppressed the ability of ATP to decrease the fluorescence in a competitive manner. Quantitative analysis of these fluorescence changes suggested that they might originate from the binding of the NTP to the regulatory site of the enzyme. These findings are in good agreement with the two-site model proposed by us (Wakagi, T. & Ohta, T. (1981) J. Biochem. 89, 1205) which was deduced from the steady state kinetics of the NTPase reactions catalyzed by the F1ATPase.

Fluorometric studies of aza-epsilon-adenylylated glutamine synthetase from Escherichia coli.

Glutamine synthetase in Escherichia coli is regulated by adenylation and deadenylation reactions. The adenylation reaction converts the divalent cation requirement of the enzyme from Mg2+ to Mn2+. Previously, the catalytic action of unadenylated glutamine synthetase was elucidated by monitoring the intrinsic tryptophan fluorescence change accompanying substrate binding. However, due to the lack of changes in the tryptophan fluorescence, a similar study could not be done with the adenylated enzyme. In this study, therefore, an extrinsic fluor is introduced into the adenylated glutamine synthetase by adenylating the enzyme with 2-aza-1,N6-ethenoadenosine triphosphate, a fluorescent analog of ATP. The modified enzyme (aza-epsilon-glutamine synthetase) exhibits catalytic and kinetic properties similar to those of the naturally adenylated enzyme. The results of fluorometric studies on this aza-epsilon-glutamine synthetase indicated that L-glutamate and ATP bind to both Mn2+ and Mg2+ forms of the enzyme in a random order, but only the Mn2+ form is capable of forming a highly reactive enzyme-bound intermediate which is a prerequisite for the reaction with NH4+ to form products. The extrinsic fluorescence changes are also used to determine the binding constants of various substrates and inhibitors of both the biosynthetic and gamma-glutamyl transfer reactions.

The interaction of nucleotides with F1-ATPase inactivated with 4-chloro-7-nitrobenzofurazan.

In common with the F1-ATPase from other sources, yeast mitochondrial F1-ATPase was inhibited by 4-chloro-7-nitrobenzofurazan. Total inhibition of the F1-ATPase activity was compatible with the modification of a single tyrosine residue per F1-ATPase molecule. Radioactive labelling experiments localized this modification on a beta-subunit. The inactive modified enzyme retained the capacity to bind the photoaffinity label 8-azido-1,N6-etheno-ATP, which has previously been shown to bind nucleotide sites of low affinity. As well, the inactive modified enzyme bound MgATP with high affinity, yielding a Kd of 14 microM. The results are consistent with the hypothesis of alternating, or cooperative, site catalysis by F1-ATPase.

Fluorescence studies of the interaction between 1,N6-ethenoadenosine monophosphate and nucleotides.

AMP, GMP, TMP and CMP quench the fluorescence of 1,N6-ethenoadenosine monophosphate (epsilon-AMP). The fluorescence spectrum of epsilon-AMP-nucleotide system is identical with that of epsilon-AMP itself, and the fluorescence decay kinetics follow a single-exponential decay law. The dependence of fluorescence yields and lifetimes upon the concentration of nucleotides shows that the fluorescence of epsilon-AMP is principally quenched in a dynamic process by AMP, TMP and CMP, while it is quenched in both dynamic and static processes by GMP. The quenching constants increase in the following order: GMP greater than AMP greater than TMP greater than CMP.

Osmium-labeled polynucleotides: reaction of osmium tetraoxide, with poly-1,N6-ethenoadenylic acid.

Osmium tetraoxide, in the presence of ligands such as pyridine and bipyridine, adds across the etheno bridge of 1,N6-etheno-9-methyladenine and poly-1,N6-ethenoadenylic acid. The Os:P ratio in the labeled polynucleotide was approximately equal to 1 when bipyridine was used as the stabilizing ligand. A similar study with polycytidylic acid, which had been partially modified with chloroacetaldehyde so that some bases were converted to 3,N4-ethenocytosine, gave an OS:P ratio of approximately equal to 1.3. Calf-thymus DNA, in which the adenine and cytosine bases were modified by chloroacetaldehyde, gave an Os:P ratio of approximately equal to 1 after 24 h. These results suggest that 3,N4-ethenocytosine will add two Os labels.