Blinking fluorophores: what do they tell us about protein dynamics?

The ability to detect emission from a single fluorophore presents a powerful tool to probe the dynamic properties of protein molecules during their interactions with ligands. Here, different classes of experiments are reviewed and a distinction is drawn between experiments that monitor signals from a large number of proteins, one molecule at a time, from those that follow a single protein molecule over many individual cycles. The latter approach is potentially capable of resolving dynamic heterogeneity, such as that displayed by enzymes that fluctuate between high and low activity states. Other factors that need to be considered are the origin of the fluctuations in the emission signal and the extent to which this represents the properties of the protein under investigation, as opposed to the probe itself. Most fluorophores show fluctuations in their emission rates, termed flickering, blinking or intermittency, which may occur on a similar time-scale as the event under investigation.

A prism combination for near isotropic fluorescence excitation by total internal reflection.

Total internal reflection fluorescence (TIRF) microscopy is finding increasing application for selectively detecting molecules at or near a glass-water surface. As with all fluorescence methods, the efficiency of excitation of a fluorophore is potentially sensitive to the polarization state of the source. In TIRF, s-polarized excitation produces an evanescent field that is perpendicular to the incident plane (y direction), whereas p-polarized light generates a more complex pattern but one dominated by a field that is vertical to the surface (z direction). Thus, fluorophores whose absorption dipoles are fixed in the x direction are not favourably aligned for excitation. Here we describe a beam-splitting prism arrangement that allows excitation by two orthogonal beams, thus giving isotropic excitation in the x-y plane with s-polarized light. With linearly polarized light at the magic angle, near isotropic excitation in three dimensions should be achieved. This prism design should find application in polarized fluorescence microscopy to investigate the rotational motions of macromolecules or to minimize flickering of fluorescence emission arising from molecular rotations in single molecule studies.

Kinetic resolution of a conformational transition and the ATP hydrolysis step using relaxation methods with a Dictyostelium myosin II mutant containing a single tryptophan residue.

The fluorescence emission intensity from a conserved tryptophan residue (W501) located in the relay loop (F466 to L516) of the Dicytostelium discoideum myosin II motor domain is sensitive to ATP binding and hydrolysis. The initial binding process is accompanied by a small quench in fluorescence, and this is followed by a large enhancement that appears coincident with the hydrolysis step. Using temperature and pressure jump methods, we show that the enhancement process is kinetically distinct from but coupled to the hydrolysis step. The fluorescence enhancement corresponds to the open-closed transition (k(obs) approximately 1000 s(-1) at 20 degrees C). From the overall steady-state fluorescence signal and the presence or absence of a relaxation transient, we conclude that the ADP state is largely in the open state, while the ADP.AlF(4) state is largely closed. At 20 degrees C the open-closed equilibria for the AMP.PNP and ADP.BeF(x) complexes are close to unity and are readily perturbed by temperature and pressure. In the case of ATP, the equilibrium of this step slightly favors the open state, but coupling to the subsequent hydrolysis step gives rise to a predominantly closed state in the steady state. Pressure jump during steady-state ATP turnover reveals the distinct transients for the rapid open-closed transition and the slower hydrolysis step.

The dynamics of the relay loop tryptophan residue in the Dictyostelium myosin motor domain and the origin of spectroscopic signals.

Steady-state and time-resolved fluorescence measurements were performed on a Dictyostelium discoideum myosin II motor domain construct retaining a single tryptophan residue at position 501, located on the relay loop. Other tryptophan residues were mutated to phenylalanine. The Trp-501 residue showed a large enhancement in fluorescence in the presence of ATP and a small quench in the presence of ADP as a result of perturbing both the ground and excited state processes. Fluorescence lifetime and quantum yield measurements indicated that at least three microstates of Trp-501 were present in all nucleotide states examined, and these could not be assigned to a particular gross conformation of the motor domain. Enhancement in emission intensity was associated with a reduction of the contribution from a statically quenched component and an increase in a component with a 5-ns lifetime, with little change in the contribution from a 1-ns lifetime component. Anisotropy measurements indicated that the Trp-501 side chain was relatively immobile in all nucleotide states, and the fluorescence was effectively depolarized by rotation of the whole motor domain with a correlation time on 50-70 ns. Overall these data suggest that the backbone of the relay loop remains structured throughout the myosin ATPase cycle but that the Trp-501 side chain experiences a different weighting in local environments provided by surrounding residues as the adjacent converter domain rolls around the relay loop.

A comparison of optical geometries for combined flash photolysis and total internal reflection fluorescence microscopy.

Total internal reflection fluorescence (TIRF) microscopy, used in conjunction with flash photolysis, provides a useful way of investigating the kinetics of macromolecular interactions. We compare different TIRF optical geometries to establish an optimal combination. Excitation light was introduced via four different arrangements: (1) a prism positioned on the microscope optical axis, (2) an offset prism with propagation through a silica slide trans to the objective lens, (3) an offset prism with propagation through a silica coverslip cis to a water-immersion objective lens and (4) a prismless arrangement using a high NA oil-immersion objective lens. Photolysis was achieved using a Xe flash lamp and a customised silica condenser lens. Single myosin molecules labelled with a Cy3 fluorophore were used as a test sample. Although the offset trans prism gave the best signal-to-background ratio, a customised thin rhombic prism incorporated, on axis, into the flash condenser assembly was almost as good and was more practical for scanning multiple fields. An oil-immersion lens gave the brightest image for sample depths < 30 micrometer but above this limit, a water-immersion lens was better. The prismless arrangement may offer advantages in other situations but it is important to check the actual numerical aperture of the objective lens.

Resolution of conformational states of Dictyostelium myosin II motor domain using tryptophan (W501) mutants: implications for the open-closed transition identified by crystallography.

When myosin interacts with ATP there is a characteristic enhancement in tryptophan fluorescence which has been widely exploited in kinetic studies. Using Dictyostelium motor domain mutants, we show that W501, located at the end of the relay helix close to the converter region, responds to two independent conformational events on nucleotide binding. First, a rapid isomerization gives a small fluorescence quench and then a slower reversible step which controls the hydrolysis rate (and corresponds to the open-closed transition identified by crystallography) gives a large enhancement. A mutant lacking W501 shows no ATP-induced enhancement in the fluorescence, yet quenched-flow measurements demonstrate that ATP is rapidly hydrolyzed to give a products complex as in the wild-type. The nucleotide-free, open and closed states of a single tryptophan-containing construct, W501+, show distinct fluorescence spectra and susceptibilities to acrylamide quenching which indicate that W501 becomes internalized in the closed state. The open-closed transition does not require hydrolysis per se and can be induced by a nonhydrolyzable analogue. At 20 degrees C, the equilibrium may favor the open state, but with ATP as substrate, the subsequent hydrolysis step pulls the equilibrium toward the closed state such that a tryptophan mutant containing only W501 yields an overall 80% enhancement. These studies allow solution-based assays to be rationalized with the crystal structures of the myosin motor domain and show that three different states can be distinguished at the interface of the relay and converter regions.

Myosin monomer density and exchange in synthetic thick filaments investigated using fluorescence microscopy with single molecule sensitivity.

The number of myosin molecules in synthetic thick filaments, prepared by dialysis at 0.12 M NaCl and pH 7.0, was estimated to be between 400 and 800 molecules per micrometre under conditions appropriate for in vitro motility assays. This estimate was based on a number count of Cy3-labelled myosin molecules incorporated into filaments at a nominal ratio of 1:1000. At this dilution, single fluorescent spots were resolved corresponding to individual labelled myosins. The spots usually bleached with a one- or two-step profile but, in around 30% of the cases, fluctuations were observed indicating that additional photophysical or photochemical events had occurred. Myosin molecules were shown to exchange between filaments in suspension on a time-scale of several hours at 4 degrees C, but the reaction was only 75% complete after 48 h, suggesting a non-exchangeable core. However, myosin exchange does not appear to be the predominant source of the fluctuations in fluorescence intensity in the single molecule assays.

Measurement of nucleotide exchange rate constants in single rabbit soleus myofibrils during shortening and lengthening using a fluorescent ATP analog.

The kinetics of displacement of a fluorescent nucleotide, 2'(3')-O-[N[2-[[Cy3]amido]ethyl]carbamoyl]-adenosine 5'-triphosphate (Cy3-EDA-ATP), bound to rabbit soleus muscle myofibrils were studied using flash photolysis of caged ATP. Use of myofibrils from this slow twitch muscle allowed better resolution of the kinetics of nucleotide exchange than previous studies with psoas muscle myofibrils (, Biophys. J. 73:2033-2042). Soleus myofibrils in the presence of Cy3-EDA-nucleotides (Cy3-EDA-ATP or Cy3-EDA-ADP) showed selective fluorescence staining of the A-band. The K(m) for Cy3-EDA-ATP and the K(d) for Cy3-EDA-ADP binding to the myofibril A-band were 1.9 microM and 3.8 microM, respectively, indicating stronger binding of nucleotide to soleus cross-bridges compared to psoas cross-bridges (2.6 microM and 50 microM, respectively). After flash photolysis of caged ATP, the A-band fluorescence of the myofibril in the Cy3-EDA-ATP solution under isometric conditions decayed exponentially with a rate constant of 0.045 +/- 0.007 s(-1) (n = 32) at 10 degrees C, which was about seven times slower than that for psoas myofibrils. When a myofibril was allowed to shorten with a constant velocity, the nucleotide displacement rate constant increased from 0.066 s(-1) (isometric) to 0.14 s(-1) at 20 degrees C with increasing shortening velocity up to 0.1 myofibril length/s (V(max), the shortening velocity under no load was approximately 0. 2 myofibril lengths/s). The rate constant was not significantly affected by an isovelocity stretch of up to 0.1 myofibril lengths/s. These results suggest that the cross-bridge kinetics are not significantly affected at higher strain during lengthening but depend on the lower strain during shortening. These data also indicate that the interaction distance between a cross-bridge and the actin filament is at least 16 nm for a single cycle of the ATPase.

Motors in muscle: the function of conventional myosin II.

Solution measurements indicate that actin and myosin alternately bind and dissociate during one ATP hydrolysis cycle. Crystallographic studies indicate at least two basic conformations of the myosin head exist in which the regulatory domain swings through an angle of about 70 degrees. Actin must further modulate these conformations, but high-resolution information about the actomyosin interface is lacking. One-to-one coupling between the ATPase cycle and a mechanical cycle involving myosin-head bending could account for about a 10 nm stroke size. At high sliding velocities, discrepancies remain which suggest that a myosin head may undergo repetitive interactions with actin for each ATP hydrolysed.

Crystal structure of the vinculin tail suggests a pathway for activation.

Vinculin plays a dynamic role in the assembly of the actin cytoskeleton. A strong interaction between its head and tail domains that regulates binding to other cytoskeletal components is disrupted by acidic phospholipids. Here, we present the crystal structure of the vinculin tail, residues 879-1066. Five amphipathic helices form an antiparallel bundle that resembles exchangeable apolipoproteins. A C-terminal arm wraps across the base of the bundle and emerges as a hydrophobic hairpin surrounded by a collar of basic residues, adjacent to the N terminus. We show that the C-terminal arm is required for binding to acidic phospholipids but not to actin, and that binding either ligand induces conformational changes that may represent the first step in activation.

ATPase kinetics of the Dictyostelium discoideum myosin II motor domain.

Structural characterization of the mode of interaction of nucleotides bound to myosin has relied upon the crystal structure of the Dictyostelium discoideum myosin II motor domain. This fragment, denoted S1dC, lacks the regulatory domain and light chain subunits and may therefore fail to display the normal ATPase activity of the intact myosin molecule. Here we show that the elementary steps of the S1dC ATPase pathway and the effects of actin are similar to those of the complete myosin head fragment. This indicates that truncation at residue E759, with the removal of the light chain binding sites, is not crucial to catalytic activity. In particular, S1dC does not show the anomolous tight binding of ADP displayed by slightly shorter M754 construct reported elsewhere. We also show that the fluorescent analogue Cy3-EDA-ATP is a good substrate for S1dC and demonstrate the use of fluorescence correlation spectroscopy to determine the affinity of Cy3-EDA-ADP using microgram quantities of proteins.

Measurement of ATP turnover during shortening and lengthening of rabbit psoas myofibrils using a fluorescent ATP analog.

In order to study ATP turnover during shortening and lengthening of rabbit psoas myofibrils, we have used fluorescence microscopy in which the displacement of a fluorescent nucleotide analog, 2'(3')-O-[N-[2-[[Cy3] amido] ethyl] carbamoyl]-adenosine 5' triphosphate (Cy3-EDA-ATP) bound to cross-bridge on flash photolysis of caged ATP was measured [Chaen et al. (1997) Biophys. J. 73, 2033-2042]. In the previous paper, we reported that when a myofibril was imposed to shorten with a constant velocity by a piezo-electric actuator, the nucleotide displacement rate constant initially increased to 0.7 s-1 with increasing shortening velocity and then declined with a further increase in shortening velocity. The rate constant during lengthening measured in the present experiment was found to be not significantly affected. These results suggest that the cross-bridge kinetics show a asymmetrical dependence on the mechanical strain in the cross-bridges, namely, the rate constants are not significantly affected at higher strain during lengthening but depend on the lower strain during shortening.

Measurement of ATPase activities of myosin at the level of tracks and single molecules.

In order to determine the degree of mechanochemical coupling in actomyosin in vitro motility assays, it is desirable to measure the sliding velocity and the associated ATP turnover simultaneously at the single filament level. Actin sliding over tracks of immobilised heavy meromyosin (HMM) has been initiated by flash photolysis of caged ATP. Flash photolysis has also been used to displace fluorescent Cy3-EDA-nucleotides from HMM tracks to monitor the ATPase activity. These assays are now being combined using total internal reflectance fluorescence (TIRF) microscopy with a dual-view detection system for Cy3/Cy5 labels on ATP and actin respectively. In other experiments, we are exploring the use of the single molecule kinetic technique developed by Funatsu et al. (Nature 374, 555-559, 1995) to scale down ATPase assays of Dictyostelium myosin fragments and to elucidate the mechanism of regulation of the molluscan (scallop) myosin ATPase. Although fluctuations occur from the binding and release of Cy3-EDA-nucleotides during turnover and might provide a measure of the ATPase activity, other sources of fluctuations also need to be considered.

Measurement of nucleotide release kinetics in single skeletal muscle myofibrils during isometric and isovelocity contractions using fluorescence microscopy.

Rabbit psoas muscle myofibrils, in the presence of the fluorescent nucleotide analog 2'(3')-O-[N-[2-[[Cy3]amido]ethyl]carbamoyl]-adenosine 5' triphosphate (Cy3-EDA-ATP), showed selective fluorescence staining of the A-band with a reduced fluorescence at the M-line. Addition of Cy3-EDA-ATP to a myofibril in the presence of Ca2+ caused auxotonic shortening against a compliant glass microneedle. These results indicate that Cy3-EDA-ATP is a substrate for myosin in the myofibril system. The kinetics of nucleotide release from a single myofibril, held isometrically between two needles, were measured by the displacement of prebound Cy3-EDA-ATP on flash photolysis of caged ATP. The A-band fluorescence of the myofibril decayed exponentially with a rate constant of 0.3 s(-1) at 8 degrees C, an order of magnitude faster than that for isolated thick filaments in the absence of actin. When a myofibril was imposed to shorten with a constant velocity by a piezoelectric actuator, the nucleotide displacement rate constant initially increased to 0.7 s(-1) with increasing shortening velocity and then declined with a further increase in shortening velocity. These results demonstrate that the displacement rates of Cy3-EDA-nucleotides bound to the cross-bridges in the contracting myofibril reflect a process that shows strain dependence.

X-ray crystal structure and solution fluorescence characterization of Mg.2'(3')-O-(N-methylanthraniloyl) nucleotides bound to the Dictyostelium discoideum myosin motor domain.

Mant (2'(3')-O-(N-methylanthraniloyl)) labeled nucleotides have proven to be useful tools in the study of the kinetic mechanism of the myosin ATPase by fluorescence spectroscopy. The sensitivity of the mant fluorophore to its local environment also makes it suitable to investigate the exposure of bound nucleotides to solvent from collisional quenching measurements. Here we present the crystal structure of mant-ADP and beryllium fluoride complexed with Dictyostelium discoideum myosin motor domain (S1dC) at 1.9 A resolution. We complement the structural approach with an investigation of the accessibility of the mant moiety to solvent using acrylamide quenching of fluorescence emission. In contrast to rabbit skeletal myosin subfragment 1, where the mant group is protected from acrylamide (Ksv=0.2 M-1), the fluorophore is relatively exposed when bound to Dictyostelium myosin motor domain (Ksv= 1.4 M-1). Differences between the Dictyostelium structure and that of vertebrate skeletal subfragment 1, in the region of the nucleotide binding pocket, are proposed as an explanation for the differences observed in the solvent accessibility of complexed mant-nucleotides. We conclude that protection of the mant group from acrylamide quenching does not report on overall closure of the nucleotide binding pocket but reflects more local structural changes.

Kinetic and spectroscopic characterization of fluorescent ribose-modified ATP analogs upon interaction with skeletal muscle myosin subfragment 1.

The interaction of fluorescent ATP analog 2'(3')-O-[N-[2-[3-(5-fluoresceinyl)thioureido]-ethyl]carbamoyl]adenosine 5'-triphosphate (FEDA-ATP) with rabbit skeletal myosin subfragment 1 (S1) and acto-S1 was studied. This and related ATP analogs are potentially useful for determination of the ATPase activity of single myosin filaments using fluorescence microscopy [Sowerby et al. (1993) J. Mol. Biol. 234, 114-123]. However, it is necessary that such analogs mimic ATP in their kinetics of turnover. The apparent second-order association rate constants for FEDA-ATP binding to S1 and for FEDA-ATP-induced dissociation of acto-S1 are about 4 times slower than those for ATP. As with ATP, the hydrolysis step is fast, so that the M.FEDA-ADP.P(i) complex is the major steady-state intermediate. The turnover rate is the same for the 2' and 3' FEDA-ATP derivatives and similar to that of ATP itself. The dissociation rate constant for FEDA-ADP from S1 is identical to that for ADP. Actin-activated turnover is comparable for both FEDA-ATP and ATP. The corresponding rhodamine and sulfoindocyanine, Cy3.18 (Cy3), derivatives also appear to be reasonable analogs. FEDA-ATP binding leads to a 25-40% reduction in fluorescein fluorescence. Spectral properties of the bound nucleotide were explored by trapping FEDA-ADP as its aluminum fluoride complex. The fluorescence quenching is a consequence of a reduction in both absorbance and excited-state lifetime, but there is little change in spectral shape.

Measurement of nucleotide exchange kinetics with isolated synthetic myosin filaments using flash photolysis.

The kinetics of nucleotide release have been measured at the level of isolated synthetic myosin filaments. This was achieved by displacing a rhodamine nucleotide analog from a filament by flash photolysis of caged-ATP with selective observation using total internal reflectance fluorescence microscopy. The procedure gave improved time resolution over previously used flow methods. Kinetics were measured both in the presence of the ADP analog and during steady-state hydrolysis of the ATP analog. These studies open the way for kinetic measurements during actin filament sliding on myosin tracks.

Kinetic mechanism of chloramphenicol acetyltransferase: the role of ternary complex interconversion in rate determination.

Chloramphenicol acetyltransferase (CAT) catalyzes the acetyl-CoA-dependent acetylation of chloramphenicol (Cm) by a ternary complex mechanism and with a random order of addition of substrates. A closer examination of the mechanism of the reaction catalyzed by the type III CAT variant (CATIII) has included the measurement of the individual rate constants by stopped-flow fluorimetry at 5 degrees C. Under all conditions employed, product release from the binary complexes in both forward and reverse reactions was found to be too slow to account for the observed overall rate of turnover for the reaction. Additional, faster routes for product release are achieved via the formation of the nonproductive ternary complexes (CAT:3-acetyl-Cm:acetyl-CoA and CAT:CoA:Cm). The release of 3-acetyl-Cm from the binary complex is 5-fold slower than kcat (135 s-1 at 5 degrees C), whereas the dissociation rate constants of 3-acetyl-Cm from the ternary complexes with CoA and acetyl-CoA are 120 and 200 s-1, respectively. Arrhenius plots of dissociation rate constants indicate a slow release of products over a broad temperature range. Computer simulations based on the rate constants of CATIII applied to a ternary complex mechanism, assuming random order of substrate addition and product release, yielded nonlinear initial rates of product formation unless both nonproductive ternary complexes were included in the model. Simulated steady-state kinetic analyses based on the latter assumption yielded kinetic parameters that compared favorably with those determined experimentally.(ABSTRACT TRUNCATED AT 250 WORDS)