INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Measurement of the kinetics of an enzyme-catalyzed reaction at the level of a single molecule in an aqueous environment is a landmark in biophysics (Funatsu et al., 1995; Lu et al., 1998; Nie and Zare, 1997; Weiss, 1999; Xie and Trautman, 1998). The approach, based on observing the fluorescence of a substrate or cofactor when bound to immobilized enzyme, lends itself in principle to the analysis of other biological mechanisms, such as ligand-receptor and protein-DNA interactions. The reaction chosen by Funatsu et al. (1995) was the hydrolysis of a fluorescent ATP analog by the motor protein myosin. Over the past few years analysis of elementary steps of individual myosin molecules has provided convincing evidence about the nature of force development, motility, and work output in muscle and nonmuscle cells (Finer et al., 1994; Harada et al., 1987, 1990; Ishijima et al., 1994, 1998; Kron and Spudich, 1986; Mehta et al., 1999; Molloy et al., 1995; Toyoshima et al., 1987, 1990; Uyeda et al., 1990; Veigel et al., 1999; Warshaw et al., 1998). An important reason for analyzing nucleotide hydrolysis at the single-molecule level is that it opens up new approaches to the study of mechanochemical energy transduction in the myosin superfamily.

The advance made by Funatsu et al. (1995) raises a number of points, some of which are fundamental to the development of single-molecule studies for probing biological mechanisms. We need to know the extent to which individual stochastic versus ensemble (i.e., macroscopic) observations lead to identical results. Is there unique information about enzyme mechanisms that can be obtained by studying single enzyme molecules? Are there factors that derive from the state of the system under investigation (e.g., filamentous versus solubilized myosin)? In the single-molecule studies of Funatsu et al. (1995) myosin was adsorbed on a surface, while kinetic analysis of the myosin ATPase has generally been carried out with water-solubilized protein; the influence of the state of myosin therefore needs to be assessed. We address these issues by comparing kinetics of reactions involving myosin and fluorescent analogs of ATP in different systems. It is important to use analogs that closely mimic properties of ATP in its interactions with myosin and muscle, and therefore we have chosen to introduce the fluorophore on the 2'- or 3'- hydroxyl group of the ribose moiety of ATP (Alessi et al., 1992; Conibear et al., 1996; Cremo et al., 1990; Crowder and Cooke, 1987; Ferenczi et al., 1989; Hiratsuka, 1983, 1984; Sowerby et al., 1993; Tonomura, 1973; Woodward et al., 1991). The fluorescent ATP analog used by Funatsu et al. (1995), in contrast, was modified on the purine moiety of ATP, though in more recent work this group has used ribose-modified ATP analogs (Ishijima et al., 1998). Replacing the 2'- or 3'-hydroxyl group with probes has generally resulted in a mixture of compounds. If the link between the hydroxyl group and the probe is via a carboxylate ester there is relatively rapid isomerization at neutral pH between the resulting 2'- and 3'-O-esters of nucleotides (reviewed in Jameson and Eccleston, 1997). We have resolved this by synthesizing carbamoyl esters and showing that their isomerization is sufficiently slow for individual isomers to be separated. It has thus been possible to assess the effect of working with separated as opposed to mixed isomers of ATP analogs.

It is frequently advantageous in single-molecule kinetic studies to monitor at least two fluorophores. For example, besides detecting the myosin-bound substrate, the position of the myosin molecule itself had to be identified by Funatsu et al. (1995). Accordingly, we have used two fluorophores and compared their properties in ATP analogs. Part of the assessment involved measurement of their photobleaching properties. This is a crucial parameter, as monitoring the disappearance of a fluorophore from an enzyme bound to a surface is at the heart of many single-molecule experiments, and the rate of fluorescence loss from the surface due to photobleaching must be known. Like Funatsu et al. (1995) and Kuhlman and Bagshaw (1998), we have used the sulfoindocyanine dyes Cy3 and Cy5, developed by Mujumdar et al. (1993), which have large extinction coefficients and are fluorescent in the long-wavelength region of the visible spectrum.

The ability of Cy3 and Cy5 fluorescent ATP analogs to mimic ATP had to be tested in a variety of ways to assess their potential for energy transduction studies in muscle. This has been done through the delineation of the kinetics of elementary steps in the myosin- and actomyosin-catalyzed hydrolysis of the analogs in solution and the ability of the analogs to support muscle contraction and relaxation in permeabilized fibers, and by comparing their properties with those of ATP in in vitro motility assays.

Single-molecule kinetics have been monitored here, using the total internal reflection fluorescence (TIRF) technique (Axelrod et al., 1984; Funatsu et al., 1995). Conibear and Bagshaw (1996) have applied this approach, using myosin filaments adsorbed on a surface. We extend TIRF studies of the myosin triphosphatase and analyze the interaction between myosin and ADP analogs. The latter serves as a model for ligand-receptor interactions. We have compared ensemble and single-molecule triphosphatase studies of myosin and actomyosin directly with the TIRF technique and assessed the influence of steric interactions arising from binding of myosin filaments to a glass surface.

For these studies we have used the mixed isomers designated Cy3-EDA-ATP 1 and Cy5-EDA-ATP 2 and the corresponding diphosphates. The isomers of the Cy3-EDA-nucleotides have been separated and their structures assigned by NMR spectroscopy. This has enabled us to determine how the properties of 2'- and 3'-substituted nucleotides differ. Preliminary results have been reported in abstract form (Anson and Oiwa, 1997, 1998, 1999; Eccleston et al., 1996; Oiwa et al., 1995, 1996, 1997, 1998, 1999).

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Chemicals

The monosubstituted cyanine carboxylic acids Cy3.29.OH and Cy5.29.OH and their succinimidyl derivatives Cy3.29.OSu and Cy5.29.OSu were purchased from Biological Detection Systems (Pittsburgh, PA). 2'(3')-O-(N-Methylanthraniloyl)ATP (mant-ATP) was synthesized and purified as described by Hiratsuka (1983) and Jameson and Eccleston (1997). Disuccinimidyl carbonate, dimethylformamide, ethylenediamine, high-performance liquid chromatography (HPLC)-grade acetonitrile, and pyridine were purchased from Aldrich Chemical Co. (Milwaukee, WI). Triethylammonium bicarbonate (TEAB) buffer was prepared by bubbling CO₂ through an ice-cold aqueous solution of redistilled triethylamine until the pH decreased to 7.6. All other reagents were of the highest available purity and were used without further treatment.

Synthesis and characterization of 2'-O-Ac-EDA-ADP and 3'-O-Ac-EDA-ADP

One hundred micromoles of the triethylammonium salt of EDA-ATP (2'(3')-O-[N-(2-aminoethyl)carbamoyl]ATP) (Jameson and Eccleston, 1997) was stirred in 20 ml aqueous 40 mM NaHCO₃ and N-acetylated by the addition of 70 µl acetic anhydride. The reaction was maintained at pH 8.0 with solid NaHCO₃ at 20°C for 5 min. The mixed isomers were purified by DEAE-cellulose chromatography using a TEAB salt gradient and rotary evaporation of solvent followed by methanol evaporations to yield 26 µmol 2'(3')-O-Ac-EDA-ATP as the triethylammonium salt. A sample was converted to its Na⁺ salt for ¹H-NMR analysis (see below).

View larger version (13K): [in this window] [in a new window]

Kinetics of isomerization between 2'-O-Ac-EDA-ADP and 3'-O-Ac-EDA-ADP

Pure 2'-O-Ac-EDA-ADP or 3'-O-Ac-EDA-ADP obtained from the analytical reverse-phase HPLC column in 100 mM KH₂PO₄/K₂HPO₄ at pH 5.5 was treated with 1 M K₂HPO₄, Tris, or Na₂CO₃ to bring the pH to 7.1, 8.4, or 9.4, respectively. These solutions were incubated at 22°C, and the isomerization rate was followed by quenching reaction samples (acidification to pH 5.5) and HPLC analysis using the same reverse-phase column. The isomerization was allowed to proceed to equilibrium, and the equilibrium constant was calculated from the ratio of the peak areas in the HPLC elution profile. When starting with 3'-O-Ac-EDA-ADP, k_obs ( = k₂₃+ k₃₂) (k₂₃ and k₃₂ are defined in Fig. 1) was calculated from the slope of the linear plot of ln(1  [2'-O-Ac-EDA-ADP]_t/[2'-O-Ac-EDA-ADP]_t) against time, t. Because k₂₃/k₃₂ = ([3'-O-Ac-EDA-ADP]_t /[2'-O-Ac-EDA-ADP]_t) (i.e., the ratio of isomer concentrations at equilibrium), k₂₃ and k₃₂ could be calculated.

Synthesis of Cy3-EDA-ATP and Cy5-EDA-ATP

Cy3-EDA-ATP and Cy5-EDA-ATP were prepared by identical protocols as described by Jameson and Eccleston (1997). Typically a succinimidyl ester was formed from 4 µmol Cy3.29.OH or Cy5.29.OH and then condensed with 20 µmol EDA-ATP. The reaction was followed by thin-layer chromatography on silica plates (silica gel 60_F254; Merck, Germany) in a solvent of propan-2-ol:H₂O:NH₄OH (7:2:1, v/v). Cy3.29.OH (or Cy5.29.OH) had an R_f of 0.2 (0.3), with a minor impurity at R_f 0.8 in both cases. The reaction mixture containing crude Cy3-EDA-ATP or Cy5-EDA-ATP showed an intense pink (Cy3) or blue (Cy5) spot remaining at the origin and the presence of some Cy3.29.OH (or Cy5.29.OH) and other impurities. The presence of a new spot remaining at the origin indicates ribose modification of EDA-ATP by the fluorophore (Hazlett et al., 1993). The reaction mixture was purified on a DEAE-cellulose column (12 × 2 cm). The column was washed with 10 mM TEAB (pH 7.6) at a flow rate of 0.6 ml min¹ until no more pink (or blue) material was eluted. Then a linear gradient of 10 mM to 0.8 M TEAB was applied (total volume 600 ml). EDA-ATP eluted at 0.2 M TEAB, followed by a major peak with absorbance maxima at 260 nm and either 550 nm (for Cy3-EDA-ATP) or 650 nm (for Cy5-EDA-ATP) that eluted at ~0.4 M TEAB. The appropriate fractions were pooled and evaporated almost to dryness in vacuo. Residual triethylamine was removed by three additions and evaporations of methanol. The final material was redissolved in water and stored at 20°C. The ATP analogs were each obtained in 25% yield based on the initial weight of Cy3.29.OH or Cy5.29.OH.

Characterization of Cy3-EDA-ATP and Cy5-EDA-ATP

The absorbance spectra of Cy3.29.OH and Cy3-EDA-ATP in 50 mM Tris-HCl (pH 7.5) were recorded between 220 nm and 850 nm. Taking  for the Cy3 chromophore to be 150,000 M¹ cm¹ at 552 nm (Mujumdar et al., 1993), a value of 6950 (± 1400 limit of error range) M¹ cm¹ at 260 nm was measured for Cy3.29.OH. Assuming that  of the Cy3 chromophore in Cy3-EDA-ATP remains unchanged, and for adenosine,  is 15,200 M¹ cm¹ at 260 nm, the molar ratio of Cy3 to adenosine in Cy3-EDA-ATP was calculated to be 1.04 (± 0.09) and is good evidence that Cy3-EDA-ATP contains equimolar amounts of Cy3 and adenosine. Corresponding values for Cy5-EDA-ATP were as follows:  = 250,000 M¹ cm¹ (Mujumdar et al., 1993) and 8680 (± 2170) M¹ cm¹ for Cy5.29.OH at 650 nm and 260 nm, respectively, and the molar ratio of Cy5 to adenosine in Cy5-EDA-ATP was calculated from the absorbance spectrum to be 1.15 (± 0.19).

These molar ratios depend on reliable  values of Cy3 and Cy5. Furthermore, the reported value of  for Cy5.29.OH is exceptionally large, possibly exceeding the theoretical maximum for the intensity of the electronic absorption transition (assuming no intermolecular interactions such as chromophore stacking; see Discussion). Accordingly,  for Cy5.29.OH was remeasured and found to be 218,000 M¹ cm¹ at _max 650 nm in aqueous solution at pH 7, using a dried and weighed sample of Cy5.29.OH. The value of 218,000 M¹ cm¹ is a lower limit because any colorless impurities in our sample would make the measured  less than the true value. Except where noted otherwise,  is taken to be 250,000 M¹ cm¹ at 650 nm (Mujumdar et al., 1993).

The fluorescence emission spectra of Cy3.29.OH and Cy3-EDA-ATP were identical, as were those of Cy5.29.OH and Cy5-EDA-ATP. Negative-ion fast-atom bombardment mass spectroscopy for Cy3-EDA-ATP and Cy5-EDA-ATP gave molecular ion peaks at m/z = 620.5 and 633.2, corresponding to doubly charged molecular masses (1241 and 1267) for (Cy3-EDA-ATP + K⁺)² and (Cy5-EDA-ATP + K⁺)², respectively.

Cy3-EDA-ATP was characterized by HPLC with a diode array (220-600 nm) or by 550-nm absorption for detection, using a reverse-phase column (Nova-Pak C₁₈, 150 × 3.9 mm; Waters). The flow was 1.5 ml min¹, first isocratically for 2 min in 10 mM KH₂PO₄/K₂HPO₄ at pH 6.8 and then with a linear gradient of 1% (by volume) acetonitrile min¹ for 30 min. Cy3-EDA-ATP eluted after 15.6 min. Neither Cy3.29.OH nor other nucleotides (<1% level) were detected as contaminants. For Cy5-EDA-ATP conditions were identical, except that detection was by 260- and 650-nm absorption. Cy5-EDA-ATP eluted after 20.9 min. Cy5.29.OH (22.1 min) was a contaminant (<3%) and was the major impurity; no other nucleotides were detected (<1% level). Cy3 and Cy5 nucleotides were also routinely monitored by fluorescence, using HPLC, which had at least 100-fold greater sensitivity than absorption.

Synthesis of the ATP analogs was confirmed by showing that they were substrates for myosin subfragment 1 and yielded products, subsequently characterized as ADP analogs (see below), that were easily resolved by analytical HPLC from their parent ATP analogs.

Cy3-EDA-ATP was also characterized by ¹H-NMR spectroscopy (500 MHz); its structure was fully consistent with the ¹H spectrum. Ribose protons in the 2' and 3'-O isomers were identified based on ribose proton assignments in 2'(3')-O-Ac-EDA-ATP (see above). Selective TOCSY experiments (Xu and Evans, 1996) were used to confirm assignments in the isomer mixture by establishing the connectivity of ribose protons in each isomer. The H-2' and H-4' peaks of 2'-O-Cy3-EDA-ATP and the H-1', H-2', H-3', and H-4' peaks of 3'-O-Cy3-EDA-ATP were well resolved. The H-1' and H-3' peaks of the 2' isomer overlapped with peaks from the Cy3 group and water, respectively. For 2'-O-Cy3-EDA-ATP the ¹H-NMR spectrum had  5.48 (H-2') and 4.37 (H-4'), and 3'-O-Cy3-EDA-ATP had  6.10 (H-1'), 5.35 (H-3'), 5.00 (H-2'), and 4.46 (H-4'). The average intensity of the two resolvable ribose protons of 2'-O-Cy3-EDA-ATP relative to that of the four resolvable ribose protons of 3'-O-Cy3-EDA-ATP was 0.70. Thus the ratio of 3'-O-Cy3-EDA-ATP to 2'-O-Cy3-EDA-ATP was 1.43 in the mixture, which was then shown to be at equilibrium by HPLC analysis (see below).

Preparation of Cy3-EDA-ADP and Cy5-EDA-ADP

Cy3-EDA-ADP and Cy5-EDA-ADP were obtained by hydrolysis of the corresponding triphosphate (8.5 µM Cy3-EDA-ATP or 10 µM Cy5-EDA-ATP) catalyzed by rabbit skeletal muscle myosin (1 mg ml¹) for 2 h in a stirred solution containing 1 mM MgCl₂, 0.2 mM dithiothreitol (DTT), and 10 mM Tris-HCl at pH 7.0 and 4°C. Myosin was removed by centrifugation, and the supernatant was stored at 20°C.

Cy3-EDA-ADP (100% yield) was analyzed by reverse-phase HPLC (Novapak C₁₈; 150 × 3.9 mm, 550 nm absorption) and eluted isocratically with acetonitrile:aqueous 10 mM K₂HPO₄/KH₂PO₄ at pH 6.8 (13:87 v/v) at 0.5 ml min¹. Cy3-EDA-ADP eluted as two peaks (the putative 2'- and 3'-isomers) with almost baseline resolution, with retention times of 21.5 and 25.2 min. (In the same system Cy3-EDA-ATP eluted as a poorly resolved doublet with retention times of 17.3 and 18.7 min.) Cy5-EDA-ADP (100% yield) was analyzed using the same HPLC conditions but monitored by 650-nm absorption. The isomers of Cy5-EDA-ADP eluted as a well-resolved doublet with retention times of 20.2 and 21.5 min. (Under these conditions Cy5-EDA-ATP eluted as a doublet with retention times of 17.5 and 18.2 min.)

Separation and isomerization of 2'-O-Cy3-EDA-ATP and 3'-O-Cy3-EDA-ATP

The isomers of Cy3-EDA-ATP were separated on a 10-nmol scale by HPLC using the Nova-Pak C₁₈ column with absorbance detection at 550 nm. Samples were eluted isocratically at 1.5 ml min¹ in acetonitrile:aqueous 100 mM TEAB at pH 7.4 and 22°C (1:10 by volume). The two isomers eluted as a doublet (area ratio of 1:1.4) with peak retention times of 31 and 35 min. The solution from each peak was evaporated to dryness in vacuo, and the solid was redissolved in water and stored at 80°C. Samples from each peak were analyzed on the same HPLC column, but using fluorescence detection (550 nm excitation, 565 nm emission). The first peak, designated isomer I, was 100% pure, and the second peak, isomer II, was 97% pure (3% contamination by isomer I).

The intensity ratio of the fluorescence of the leading peak (isomer I) to that of isomer II was 0.65. Noting that the fluorescence of isomer II is 92% that of isomer I (see below), the molar ratio of isomer I to isomer II is 0.65/0.92 = 0.71, showing that isomer I is 2'-O-Cy3-EDA-ATP and isomer II is 3'-O-Cy3-EDA-ATP. That the sample of 2'-O- and 3'-O-Cy3-EDA-ATP used in the NMR experiment had reached equilibrium was checked by incubating the mixture at pH 10 and 22°C for 3 h (estimated equilibration rate of 2.5 h¹; see Results) and analyzing by HPLC. The fluorescence ratio of the 2'-O-Cy3-EDA-ATP peak to that of 3'-O-Cy3-EDA-ATP was identical (0.65) before and after the incubation. This establishes that the sample analyzed by NMR was at equilibrium and that, from the H' peak intensities, the equilibrium constant (analogous to k₂₃/k₃₂) for the Cy3 nucleotides is 1.43.

The equilibration rate of 2'-O-Cy3-EDA-ATP and 3'-O-Cy3-EDA-ATP isomers was measured at 20°C by the addition of 0.2 M 3-(4-morpholino)- propanesulfonic acid (MOPS) adjusted to pH 7.2 with aqueous KOH or by the addition of 0.5 M NaHCO₃ adjusted to pH 9.6 with NaOH, and monitoring the reaction by HPLC. Whether isomerization occurs during hydrolysis catalyzed by myosin subfragment 1 was tested by incubating 5 µM 2'-O-Cy3-EDA-ATP with 100 nM subfragment 1 in 100 mM KCl, 5 mM MgCl₂, 1 mM DTT, and 50 mM Tris adjusted to pH 7.5 with HCl at 20°C. After reaction times of up to 8 min, aliquots of the reaction mixture were quenched in acid and analyzed under the same HPLC conditions that also separated the two isomers of Cy3-EDA-ADP.

Preparation and separation of 2'-O-Cy3-EDA-ADP and 3'-O-Cy3-EDA-ADP

Cy3-EDA-ATP (31 nmol) was hydrolyzed to Cy3-EDA-ADP as described above. The hydrolysis was monitored by reverse-phase HPLC (as in Preparation of Cy3-EDA-ADP and Cy5-EDA-ADP). When all of the Cy3-EDA-ATP had been hydrolyzed to Cy3-EDA-ADP, the solution was applied to the HPLC column. The two peaks, the first of 2'-O-Cy3-EDA-ADP and the second of 3'-O-Cy3-EDA-ADP, were collected. A sample from each peak was analyzed by the same HPLC system; the 2' isomer was 100% pure, whereas the 3' isomer was 98% pure (and contained 2% 2' isomer). The remainder of each peak was evaporated to dryness in a Speedvac SVC100 (Savent) apparatus and then taken up in the original volume of water. The two solutions of Cy3-EDA-ADP isomers therefore each contained 10 mM potassium phosphate at pH 6.8.

Protein preparations

Rabbit skeletal muscle myosin and heavy meromyosin were prepared essentially as described by Margossian and Lowey (1982), and myosin subfragment 1 was prepared according to the method of Weeds and Taylor (1975). F-actin was prepared from acetone powder by the method of Pardee and Spudich (1982). For measurements in the microscope, F-actin was stabilized against depolymerization at low concentrations with phalloidin or, where visualization by fluorescence was required, with rhodamine-labeled phalloidin (Faulstich et al., 1988). Concentrations of myosin, heavy meromyosin, and subfragment 1 were determined from A_{280 nm} values for 1 mg ml¹ of 0.53, 0.65, and 0.79 cm¹ respectively, and the concentration of actin was determined from A_{290 nm}  A_{310 nm} of 0.62 cm¹.

Synthetic bipolar filaments of rabbit skeletal muscle myosin for TIRF studies were prepared by slowly diluting a myosin solution, initially 1 ml at a concentration of 50 µg ml¹ in 0.5 M KCl, 2 mM MgCl₂, 1 mM DTT, and 10 mM piperazine-N,N'-bis(2-ethanesulfonic acid) (PIPES) adjusted with KOH to pH 7.0 at 4°C, into 7.5 ml of the same buffer but containing 0.08 M KCl (Nagashima, 1986). The myosin filaments were further diluted in the same buffer to a final concentration of 2 µg ml¹ for infusion into the microscope observation chamber. Electron micrographs with negative staining showed that the length of the filaments ranged from 1.6 to 3.2 µm with an average of 2.6 ± 0.5 µm (mean ± SD, n = 83).

Labeling of myosin with Cy3.29.OSu and Cy5.29.OSu

To measure the photobleaching rates of Cy3 and Cy5 fluorophores under aqueous conditions it was necessary to immobilize fluorophores on myosin filaments to prevent exchange with free molecules in solution. Model myosin filaments were made using the succinimidyl esters of the dyes to couple covalently but nonspecifically to the constituent monomeric myosin. One milliliter of myosin solution at 5 mg ml¹ in 0.5 M KCl, 20 mM KH₂PO₄/K₂HPO₄ at pH 6.8 was added to 0.1 ml of 1 M sodium carbonate (pH 9.3) followed by 0.2 mg of either the Cy3.29.OSu or Cy5.29.OSu reagent. The reaction mixture was incubated at 4°C for 1 h with gentle stirring, then diluted 20-fold into 2 mM MgCl₂, 1 mM DTT, and 20 mM PIPES (adjusted to pH 7.0 with KOH) to form myosin filaments. The labeled myosin filaments were separated from unconjugated dye by low-speed centrifugation followed by resuspension, washing at pH 7, and further centrifugation. Molar concentrations of dye and myosin were calculated from the absorbance of the labeled myosin solution ( 150,000 M¹ cm¹ at 550 nm for Cy3.29.OH, 250,000 M¹ cm¹ at 650 nm for Cy5.29.OH and 275,000 M¹ cm¹ at 280 nm for myosin; from A_280 nm; see previous section). The molecular ratios of dye to myosin were ~10:1. Filaments suitable for analysis by TIRF were formed from the labeled myosin as described for unlabeled myosin.

Interaction of Cy3-EDA-ATP and Cy5-EDA-ATP with myosin subfragment 1

Stopped-flow measurements were carried out using a Hi-Tech Scientific SF61-MX instrument (Salisbury, UK) in the single-push fluorescence mode, and records were analyzed, generally as single exponential decays, using Hi-Tech software or Origin 5. The relatively small Stokes shift in Cy3 and Cy5 fluorescence necessitated careful choice of excitation and emission wavelengths. A quartz-halogen lamp was used with monochromatic excitation at 520 and 600 nm for Cy3 and Cy5, respectively. These wavelengths correspond to shoulders on the absorption spectra 40-50 nm to the blue of the maxima (Mujumdar et al., 1993). Emitted light was transmitted through a Wratten 21 cutoff filter (0.1 and 50% transmission at 535 and 555 nm, respectively) to record Cy3 fluorescence and through a Wratten 70 cutoff filter (0.1 and 50% transmission at 640 and 675 nm, respectively) to record Cy5 fluorescence. Mant-ATP fluorescence was detected by energy transfer from tryptophan in subfragment 1 (Woodward et al., 1991). Excitation was at 290 nm from a 75-W mercury-xenon lamp, and emitted light was transmitted through a Wratten 47B bandpass filter (430 nm, bandwidth 50 nm). Actosubfragment 1 dissociation by ATP analogs was monitored at 350 nm, using the light-scattering signal orthogonal to the incident beam. Experiments were carried out at 20°C in solvent A (Table 2) to allow direct comparison with rate constants obtained with ATP and mantATP (Woodward et al., 1991) or in solvent B (Table 2) to allow direct comparison with TIRF measurements. Except where otherwise noted, concentrations quoted are those after mixing in the flow cell.

Myosin subfragment 1 and its actin-activated Cy3-EDA-ATPase and Cy5-EDA-ATPase activities were measured at 20°C in a reaction solution of 1 mM MgCl₂, 0.2 mM DTT, typically 40 µM Cy3-EDA-ATP or Cy5-EDA-ATP, 10 mM Tris adjusted to pH 7.5 with HCl, with between 0 and 72 or 36 µM F-actin, respectively. The reaction was started by the addition of 0.1-0.5 µM subfragment 1 in the absence of actin and 0.02 µM subfragment 1 in its presence. Aliquots of the reaction solution were then taken and quenched with acid at intervals of 30 s and then centrifuged, and aliquots from these supernatants analyzed by reverse-phase HPLC (see above), using isocratic rather than gradient elution with a mobile phase of acetonitrile:aqueous 100 mM KH₂PO₄/K₂HPO₄ at pH 6.8 (13:87 v/v). From the ratios of the diphosphate to triphosphate peak areas the Cy3- and Cy5-EDA-ATPase rates were determined. Actin-activated subfragment 1 ATPase was measured as described above, using 1.36 mM ATP, except that anion exchange HPLC (Whatman SAX) was used to resolve ATP and ADP with absorbance monitored at 260 nm.

Muscle fiber preparation and physiological measurements

Single muscle fibers from the rabbit psoas muscle were dissected, permeabilized, and mounted between a force transducer and a motor as described by Thirlwell et al. (1994). Cy3(Cy5)-EDA-ATP and ATP were compared as substrates in fibers by measuring the force that they induced in activated isometric fibers at 2.4-2.6 µm sarcomere length, following protocols essentially as in Alessi et al. (1992). The kinetics of fiber relaxation and activation were likewise compared. The rates of relaxation from the activated isometric state were measured as in Alessi et al. (1992), as were the rates of activation from the relaxed state, as described in the next paragraph.

The efficacy of Cy3(Cy5)-EDA-ATP was also determined by measurements of maximum shortening velocity under zero load, using the slack-test method of Edman (1979) as described in Thirlwell et al. (1995), except that insufficient Cy3(Cy5)-nucleotides were available to obtain enough data for statistical analysis. The muscle fiber was initially incubated in relaxing solution at 20°C that contained Cy3(Cy5)-EDA-ATP or ATP, 15 mM EGTA, 10 mM MgCl₂, 1 mM glutathione, 60 mM N-tris-(hydroxymethyl)methyl-2-aminoethanesulfonic acid adjusted to pH 7.1 with KOH, and K⁺ propionate to bring ionic strength to 0.15 M. The fiber was then transferred to a similar solution that also contained sufficient Ca²⁺ to give 32 µM free Ca²⁺, and tension developed. When an isometric plateau was reached, the length of the muscle fiber was decreased to drop the force to zero. The muscle fiber shortened under zero-load conditions to take up the slack and then force redeveloped. The muscle length was reduced sufficiently to allow for a measurable shortening time at zero load. The muscle length was returned to the initial value 0.5 s later. The process was repeated twice with different extents of shortening. The maximum shortening velocity was calculated from the time taken for the fiber to take up the slack and the extent of shortening.

In vitro motility assay

The in vitro motility assay was carried out as described by Anson (1992) with the modifications described below. A nitrocellulose-coated coverslip formed the bottom of a 40-µl flow cell, and 100 µl of heavy meromyosin at 50 µg ml¹ was used to coat the surface. After washing, 10-20 nM F-actin, labeled and stabilized with rhodamine-phalloidin, was introduced and allowed to bind to the heavy meromyosin in rigor. Unbound F-actin was washed out, and the cell was flushed with motility buffer, 25 mM KCl, 4 mM MgCl₂, 1 mM EGTA, 0.5 mg ml¹ bovine serum albumin, 5 mM DTT, 25 mM HEPES adjusted to pH 7.5 with KOH, containing 100 µg ml¹ glucose oxidase, 20 µg ml¹ catalase, and 3-5 mg ml¹ glucose as an oxygen-depleting antifade mixture (Harada et al., 1990). The flow cell was transferred to an inverted epifluorescence microscope (Axiovert 35; Carl Zeiss Jena, Jena, Germany). Rhodamine fluorescence was excited either by the 546-nm line of a 100-W mercury arc lamp (HBO-100 W2) selected by a 546-nm interference filter (bandpass 8 nm) and a BG18 red-absorbing filter, or a 1-mW, 543-nm He-Ne laser (1674P; Uniphase, San Jose, CA). Imaging was via a 560-nm dichroic reflector and a 580-nm interference filter (bandwidth 30 nm; 560DRLP02 and 580DRF30; Omega Optical, Brattleboro, VT) by a Plan-Neofluar 40×, 1.3 NA oil-immersion objective (Carl Zeiss) through a 4× magnifying lens onto an intensified CCD TV camera (Darkstar 800; Photonic Science, Robertsbridge, UK). TV-rate (CCIR, 25 frames s¹) images were rolling averaged over four frames by an image processor (Argus 10; Hamamatsu Photonics, Hamamatsu, Japan). The averaged image was stored with time-marking, using a time-date generator (VTG33; ForA, Tokyo, Japan) and an S-VHS (PAL) videotape recorder (BRS-800E; JVC, Yokohama, Japan). After focusing and recording actin filaments in rigor as a reference, we started motility on the microscope by perfusing into the flow cell 100 µl of motility buffer containing Cy5-EDA-ATP or, for reference, mant-ATP. All solutions were stored on ice, and the microscope stage and flow cell were stabilized at 25°C for the assays.

Measurement of F-actin velocity

Computer analysis of the video recordings was carried out in a manner similar to that used by Anson et al. (1995), but employing different video equipment. Videotapes of the F-actin movement were played back off-line by the BRS-800E VCR, and the images were converted in real time by a CVR22 digital standards converter (Snell and Wilcox, Petersfield, UK) from PAL (25 frames s¹, 625 lines) to NTSC (30 frames s¹, 525 lines). This allowed digitization of the images with a VP110 video processor (Motion Analysis, Santa Rosa, CA) and their automatic tracking by computer, using the Expert Vision system (Motion Analysis) running on a 486DX2 66 MHz PC. In view of the low actin velocities (<1 µm s¹) produced by the low concentrations of ATP analogs used, video frames were digitized at 1 frame s¹ (Homsher et al., 1992).

TIRF microscopy

The apparatus

8

2

9

2

1

8

Manipulations of myosin filaments and solutions in the flow cells

1

1

1

1

Kinetics of solution exchange in the flow cell

1

1

1

Data analysis and curve fitting

Statistical data analysis, nonlinear curve fitting, and graphical presentation were carried out with either Origin 5 (Microcal Software, Northampton, MA) or Sigmaplot 4.0 (SPSS, Chicago, IL) software running on IBM Pentium PC platforms. Critical analyses were checked using both programs. Standard deviations (SD) are quoted as mean ± SD. However, in several of the stopped-flow experiments there was insufficient cyanine nucleotide to repeat experiments with different subfragment 1 preparations. In these cases we estimate that the maximum range of the rate constant k was from 0.7 k to 1.4 k.

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Interconversion kinetics of 2'- and 3'-O-Ac-EDA-ADP

Fig. 1 illustrates an HPLC analysis of an experiment at pH 9.4 and 22°C recording the time course of conversion from 3'-O-Ac-EDA-ADP to 2'-O-Ac-EDA-ADP. Measurements (not shown) were also made at 22°C, starting from the 3'-O-ester at pH 7.1 and 8.4 and from the 2'-O-ester at pH 9.4. Based on the isomerization scheme in Fig. 1, values of k₂₃ are 0.010, 0.14, and 1.40 h¹, and values of k₃₂ are 0.007, 0.098, and 1.00 h¹ at pH 7.1, 8.4, and 9.4, respectively. The mean value of the equilibrium constant for the isomerization (k₂₃/k₃₂) equals 1.4 when it is measured from the equilibrium proportions of [3'-O-Ac-EDA-ADP] to [2'-O-Ac-EDA-ADP]. The isomerization rate, expressed as (k₂₃ + k₃₂), is proportional to [OH] over the pH range 7.1 to 9.4 and equals 30[OH] s¹.

View larger version (16K): [in this window] [in a new window]   FIGURE 1   Equilibration of 3'-O-Ac-EDA-ADP to a mixture of the 2'- and 3'-O-esters. HPLC absorbance records at 254 nm after the incubation of 3'-O-Ac-EDA-ADP for times as indicated at pH 9.4 and 22°C.

The chemistry of Cy3-EDA- and Cy5-EDA-nucleotides

Separation of 2'- and 3'-isomers was achieved on a 10-nmol scale for Cy3-EDA-ATP and a 30-nmol scale for Cy3-EDA-ADP. The Cy5-EDA-ADP isomers were similarly separated but, unlike the Cy3 isomers, were not subjected to further experiments. For both Cy3- and Cy5-EDA-ATP (as for the 2'(3')-O-Ac-EDA-ADP) the earlier eluting isomer peak had the smaller absorption (range 1:1.4 to 1:1.5), and for Cy3-EDA-ATP it was identified by ¹H-NMR as the 2'-O-isomer. Isomerization rates were measured for 2'-O- and 3'-O-Cy3-EDA-ATP. At pH 7.2 and pH 9.6 the rates were <0.01 h¹ and 1.0 h¹, respectively, at 20°C. No isomerization was detected during the hydrolysis of 5 µM 2'-O-Cy3-EDA-ATP to 2'-O-Cy3-EDA-ADP by 100 nM subfragment 1 at 20°C and pH 7.5.

Fluorescence of 2'-O- and 3'-O-Cy3-EDA-ATP(ADP)

After isomer separation it was possible to compare the ratio of the fluorescence of 2'-O-Cy3-EDA-ATP and 3'-O-Cy3-EDA-ATP (assuming identical  values). It was important to know this, as otherwise it would not be clear to what extent a fluorescence difference when the 2'- or 3'-isomer bound to myosin subfragment 1 in solution was due to a difference when the isomer was not bound. This in turn influences the analysis in TIRF experiments. At the emission peak 3'-O-Cy3-EDA-ATP fluorescence was 89% of that of 2'-O-Cy3-EDA-ATP in solvent A. In a further study the two isomers of Cy3-EDA-ADP were separated from a (nonequilibrium) mixture by HPLC and monitored using absorption and fluorescence HPLC detectors. The ratio of the absorption peaks was 1:1.72, and that of the fluorescent peaks was 1:1.59, indicating that the fluorescence of 3'-O-Cy3-EDA-ADP was 92% of that of 2'-O-Cy3-EDA-ADP.

Elementary processes of Cy3-EDA-ATPase and Cy5-EDA-ATPase activities

Data in Table 1 summarize the results obtained by analysis of the interaction of free myosin subfragment 1 in solution with Cy3- and Cy5-EDA-nucleotides in the presence and absence of actin. Most of the data in Table 1 are derived from mixed isomers; those from separated isomers are shown in brackets. Except for ss k_cat, results relating to subfragment 1 alone were obtained by transient kinetic methods as follows. The second-order association rate constant, k₊₁, was calculated by measuring the dependence of the observed rate of Cy3 fluorescence increase (Fig. 2 A) or Cy5 fluorescence decrease (Fig. 2 D) on subfragment 1 concentration (0.5-5 µM) when the ATP analogs were mixed with subfragment 1 in excess. This yielded values for k₊₁ of 1.2 and 1.4 µM¹ s¹ for Cy3- and Cy5-EDA-ATP, respectively.

View larger version (24K): [in this window] [in a new window]   FIGURE 2   Stopped-flow fluorescence records of the interaction of Cy3-EDA-ATP and Cy5-EDA-ATP with subfragment 1. (A and B) 2.5 µM subfragment 1 mixed with 0.25 µM Cy3-EDA-ATP. (C) 0.5 µM subfragment 1 preincubated for <5 s with 1.0 µM Cy3-EDA-ATP and mixed with 50 µM ATP. (D and E) 1 µM subfragment 1 mixed with 0.4 µM Cy5-EDA-ATP. (F) 1 µM subfragment 1 preincubated for <5 s with 2 µM Cy5-EDA-ATP and mixed with 20 µM mant-ATP. Nonlinear least-squares fit of single (double in A) exponentials is drawn through the data with rate constants of (A) 2.9 s1 (second phase 0.018 s1), (B) 0.0125 s1, (C) 0.017 s1, (D) 0.71 s1, (E) 0.034 s1, and (F) 0.038 s1. Note that the fluorescence decay in B has an unexplained phase of small amplitude that precedes the main decay phase. Reaction chamber concentrations are listed (hence preincubations in C and F were carried out at double the indicated concentrations). The fluorescence of Cy3 and Cy5 is recorded in A-E. In F mant fluorescence is recorded as described in Materials and Methods. Reactions were carried out at 20°C in solvent A (Table 2). Zero time is the time at which flow stopped after solution mixing.

Under single turnover conditions (i.e., [subfragment 1] > [ATP analogue]) there is a characteristic formation and decay of an intermediate. This is the same as the steady-state intermediate of the triphosphatase reaction when the ATP analog is in excess (Bagshaw and Trentham, 1973). This decay was evident as a Cy3 fluorescence decrease (Fig. 2 B) and a Cy5 fluorescence increase (Fig. 2 E) in the cases of Cy3-EDA-ATP and Cy5-EDA-ATP, respectively. The exponential decay rates were independent of subfragment 1 concentration and were equated with k_cat (or k₊₃; more correctly, k₊₂k₊₃/(k₊₂ + k₂)) (Table 1), because at least in the case of ATP, the ATP cleavage step is reversible (Bagshaw and Trentham, 1973).

Several checks were made on the above results and their interpretation. The rapid phase associated with Cy5-EDA-ATP binding to subfragment 1 was also observed as a 10% protein fluorescence decrease (data not shown) that contrasts with the 20% protein fluorescence increase on ATP binding. As a second test each nucleotide analog was added to subfragment 1 and within 5 s mixed with a 100-fold excess of ATP (Cy3 experiment) or mant-ATP (Cy5 experiment). (The use of mant-ATP rather than ATP and monitoring mant fluorescence for the Cy5-EDA-ATP experiments gave an improved signal-to-noise ratio.) Under these conditions the Cy3- or Cy5-EDA-ATP bound to subfragment 1 will be displaced from its steady-state intermediate. As predicted, the exponential rates in Fig. 2, C and F, matched those in Fig. 2, B and E, respectively. Finally Mg²⁺-dependent subfragment 1 Cy3-EDA-ATPase and Cy5-EDA-ATPase activities (called ss k_cat in Table 1) were measured using a steady-state assay used for the actin-activated analysis (see Materials and Methods); rates were 0.030 and 0.018 µmol min¹ (mg subfragment 1)¹ corresponding to turnover rates of 0.056 and 0.034 s¹, respectively.

The interaction of Cy3-EDA-ADP with myosin is a key element in our single-molecule analysis (see below), so the reaction kinetics of Cy3-EDA-ADP (mixed isomers) with subfragment 1 were measured in some detail. Displacement of, typically, 1 µM Cy3-EDA-ADP from 2.5 µM subfragment 1 by 100 µM ATP under the same conditions as in Fig. 2 was accompanied by a decrease in Cy3 fluorescence; the exponential decay was independent of ATP concentration and had a rate constant, k₊₄, equal to 1.20 (± 0.09) s¹. Conversely, the association occurred with an increase in Cy3 fluorescence, and under the same conditions the observed rate constant, k_obs, increased linearly with Cy3-EDA-ADP concentration (2.5 to 7.5 µM and 0.5 µM subfragment 1). k_obs can be equated with k₄[Cy3-EDA-ADP] + k₊₄. From these data k₄ was 0.53 µM¹ s¹, k₊₄ was 1.3 s¹, and the dissociation constant of Cy3-EDA-ADP from subfragment 1, K₄, was 2.6 µM. Further studies were carried out to measure the Cy3-EDA-ADP dissociation rate constant from subfragment 1 over the same temperature range as in the single-molecule studies. At 5 and 10°C, values of k₊₄ were 0.07 (± <0.01) and 0.22 (± 0.01) s¹, respectively.

The separated isomers 2'-O-Cy3-EDA-ATP (and ADP) and 3'-O-Cy3-EDA-ATP (and ADP) were isolated in sufficient quantities for their interaction with subfragment 1 to be analyzed. Surprisingly, the fluorescence of 2'-O-Cy3-EDA-ATP decreased on binding to subfragment 1, while that of 3'-O-Cy3-EDA-ATP increased (Fig. 3, A and B). From such data, binding of 0.25 µM 3'-O-Cy3-EDA-ATP to subfragment 1 was measured over a 1-8-µM range of subfragment 1, and k₊₁ equaled 1.1 × 10⁶ M¹ s¹ (Table 1). Records as shown in the slow phases in Fig. 3, C and D, were used to calculate k₊₃ values for 2'-O-Cy3-EDA-ATP and 3'-O-Cy3-EDA-ATP, respectively, which are recorded in Table 1. The dissociation rate constants of 2'-O- and 3'-O-Cy3-EDA-ADP were 2.6 and 1.7 s¹, respectively (Fig. 3, E and F).

View larger version (22K): [in this window] [in a new window]   FIGURE 3   Interaction of 2'-O- and 3'-O-Cy3-EDA-ATP (ADP) with subfragment 1. (A) 0.5 µM 2'-O-Cy3-EDA-ATP was mixed with 5 µM subfragment 1, and the Cy3 fluorescence change was recorded. (B) As in A but with 0.5 µM 3'-O-Cy3-EDA-ATP. (C) As in A, with an extended time scale. (D) As in B, with an extended time scale. (E) 50 µM ATP was mixed with 0.77 µM 2'-O-Cy3-EDA-ADP and 2.5 µM subfragment 1 (1.53 µM and 5.0 µM, respectively, before mixing). (F) As in E but with 0.40 µM 3'-O-Cy3-EDA-ADP (0.80 µM before mixing) instead of the 2' isomer. Nonlinear least-squares fits of single exponentials are drawn through the data with rate constants of (A) 1.26 s1, (B) 1.151, (C) 0.085 s1, (D) 0.0185 s1, (E) 2.6 s1, and (F) 1.7 s1. Reaction conditions were as in Fig. 2.

Taken together with the K_d ( = K₄) value of 2.6 µM (Table 1) for the Cy3-EDA-ADP interaction with subfragment 1, the amplitudes of the records (Fig. 3) show that the fluorescence decrease when 2'-O-Cy3-EDA-ADP binds to subfragment 1 is 5% compared to 12% for 2'-O-Cy3-EDA-ATP, while the fluorescence increase for 3'-O-Cy3-EDA-ADP is 26% compared to 70% for 3'-O-Cy3-EDA-ATP. Cutoff filters were used to obtain these data, and it is useful to be able to relate relative fluorescence intensities to a specific emission wavelength. The 70% increase observed when 3'-O-Cy3-EDA-ATP binds to subfragment 1 (Fig. 3 B) may be compared to the 80% increase for the formation of 3'-O-Cy3-EDA-ATP subfragment 1 steady-state complex recorded in an SLM fluorimeter at the maximum emission of 562 nm.

Because Cy5 nucleotides gave a relatively small fluorescence change on interaction with subfragment 1, an indirect fluorescence approach was used to measure the transient kinetics. Mant-ATP was mixed with a solution containing Cy5-EDA-ADP and subfragment 1 at equilibrium (Fig. 4 A), and mant-ATP fluorescence was excited by energy transfer from excited tryptophans in subfragment 1. A biphasic increase in fluorescence was recorded. In the absence of Cy5-EDA-ADP, a monophasic fluorescence increase occurred (Fig. 4 B), the rate constant of which equaled that of the rapid phase in Fig. 4 A and the amplitude of which corresponded to the overall amplitude in Fig. 4 A. On the other hand, if the Cy5-EDA-ADP concentration was increased, the ratio of slow to fast phases also increased. The slow phase in Fig. 4 A may be equated to the dissociation rate constant, k₊₄, of 1.6 s¹ of Cy5-EDA-ADP from subfragment 1, and the fast phase may be equated to the association of mant-ATP with free subfragment 1 (the process observed in Fig. 4 B). The 1:2 ratio of amplitudes of the slow to fast phases gives the ratio between subfragment 1 with Cy5-EDA-ADP bound and free subfragment 1 before mixing in the stopped-flow apparatus. From this the dissociation constant, K₄, can be calculated to be 2.4 µM.

View larger version (9K): [in this window] [in a new window]   FIGURE 4   Displacement of Cy5-EDA-ADP from subfragment 1. (A) 1 µM subfragment 1 and 2 µM Cy5-EDA-ADP (obtained by incubating 2 µM Cy5-EDA-ATP with subfragment 1 for 5 min) mixed with 10 µM mant-ATP. (B) As in A without Cy5-EDA-ADP. Reaction conditions were as in Fig. 2. Mant-ATP fluorescence was recorded at 430 nm (bandpass filter). In record B the fluorescence before time zero arises from mant-ATP during flow. The nonlinear least-squares fit in A was to two exponentials with rate constants of 24.9 s1 and 1.6 s1 (relative amplitude 45:55), and the fit in B was to a single exponential with a rate constant of 23.9 s1.

Steady-state kinetic methods were used to measure Michaelis-Menten constants, K'_m for actin and V_max of the actin-activated subfragment 1 Cy3-EDA-ATPase and Cy5-EDA-ATPase. However, Cy3-EDA-ATP and Cy5-EDA-ATP were not available at saturating concentrations. Nevertheless, V_max values are comparable to those obtained with saturating ATP, although K'_m for actin is two- to threefold greater for the Cy3- and Cy5-nucleotides (Table 1). HPLC records showed that the ratio of isomers of Cy3-EDA-ADP and Cy5-EDA-ADP was unchanged through the time course of the actin-activated triphosphatase assay, indicating that pairs of isomers hydrolyzed at the same rates (within a factor of 1.5; this assay of relative triphosphatase rates has low sensitivity). Transient kinetic methods were used to measure the rates of dissociation of actosubfragment 1 induced by Cy3-EDA-ATP and Cy5-EDA-ATP (Fig. 5 and Table 1). Second-order rate constants were threefold less than for dissociation induced by mant-ATP or ATP.

View larger version (15K): [in this window] [in a new window]   FIGURE 5   Dissociation of actosubfragment 1 induced by Cy3-EDA-ATP and Cy5-EDA-ATP. (A) Stopped-flow record of the light scattering change at 350 nm when 0.25 µM subfragment 1 and 0.37 µM actin (reaction chamber concentrations) were mixed with 5 µM Cy5-EDA-ATP. Solution conditions were as in Fig. 2. The solid line is a nonlinear least-squares single exponential fit to the data and has a rate constant, kobs, of 4.76 s1. The light scattering before zero time is that of the end of the reaction of the previous experimental trial. (B and C) Concentration dependence of kobs on Cy3-EDA-ATP and Cy5-EDA-ATP. The slope of the least-squares fit to the data (where the solid lines are constrained to pass through the origin) is 0.95 µM1 s1 in B and 0.93 µM1 s1 in C. SD are shown, except where the deviation was less than the radius of the symbol.

To compare the kinetics of the Cy3(Cy5)-EDA-nucleotides interacting with subfragment 1 and the kinetics of their interactions with myosin filaments, we needed to measure k_cat and k₊₄ in the same solutions. Experiments were carried out in the appropriate solvent (solvent B, Table 2), but using 1 mM rather than 10 mM DTT. The results are recorded in Table 2. k_cat values in solvent B were typically twofold greater than those in solvent A (cf. columns 1 and 2 in Table 2). Measurements in other buffer systems showed that this was due primarily to the lower ionic strength of solvent B compared to that of solvent A. 

Cy3- and Cy5-EDA-ATP as physiological substrates for muscle fiber contraction

Cy3-EDA-ATP and Cy5-EDA-ATP were compared with ATP as substrates in single permeabilized muscle fibers. Each analog generated the same isometric force in muscle as ATP. Cy3(Cy5)-EDA-ATP relaxed muscle fibers from the activated state but at a rate fivefold slower than that obtained with the same concentration of ATP. The data from the two assays were very similar to those of a ribose-modified spin-labeled ATP analog (figure 2 of Alessi et al., 1992). The half-times of activation from the relaxed state by Ca²⁺ addition were indistinguishable: between 1.3 and 1.8 s for 0.5 mM analogs and ATP. However, here rates were limited by Ca²⁺ diffusion (Moisescu, 1976), as illustrated by the half-time for activation being 20 ms after photolysis of caged ATP under otherwise comparable conditions (Ellis-Davies and Kaplan, 1994).

The maximum unloaded shortening velocities were 0.63 and 1.31 muscle lengths s¹ at 100 µM Cy5-EDA-ATP and ATP, respectively, and 1.73 muscle lengths s¹ in 6.25 mM Cy3-EDA-ATP compared to 2.93 muscle lengths s¹ in 5 mM ATP (Fig. 6). Thus Cy3- and Cy5-EDA-ATP induce unloaded shortening velocities about twofold more slowly than those achieved with ATP, and this, together with the results of force measurements, shows that they have physiological properties similar to those of ATP.

View larger version (16K): [in this window] [in a new window]   FIGURE 6   Shortening velocity of a rabbit psoas muscle fiber with Cy3-EDA-ATP and ATP. The records are from a slack test performed in 6.25 mM Cy3-EDA-ATP or 5 mM ATP at 20°C and pH 7.1. The sarcomere length was set to 2.6 µm before fiber activation. The lower two records show the force response to length changes shown in the upper record. After the first fiber length decrease from 2.8 to 2 mm, there are delays of 72 ms in the case of Cy3-EDA-ATP and 48 ms in the case of ATP before force begins to increase. With the smaller length decreases the delays are barely discernible and are in the range of 5-20 ms. The decrease in isometric force from 215 kN m2 at the start of the Cy3-EDA-ATP reco