A myosin head consists of a globular head and α-helical lever arm. The lever arm is believed to undergo a large “swing” resulting from actin-activated hydrolysis of ATP by the myosin head. The swing is responsible for a relative sliding of myosin and actin containing filaments and force development during contraction. The essential Light Chain (LC) of myosin resides on the lever arm and therefore is a convenient site to monitor the swing (orientation of the lever arm). Steady State Fluorescence Anisotropy (SSFA) is a well-known method to monitor the orientation. Therefore, the values of SSFA in different physiological states of muscle are an indication of orientation changes during XB cycle.

To monitor the orientation of LC during ventricle contraction, LC was expressed in E. coli, labeled with fluorescent dye and exchanged with native ventricular LC. Figure 1 illustrates how the anisotropy (measure of the orientation) of labeled LC reflects the kinetics of a myosin cross-bridge (XB).

To observe a small number of myosin molecules In-situ it is necessary to carry out experiments on the A-band of the isolated myocytes from human ventricles. The A-band is the volume where actin and myosin interact to produce contractile force. Myocyte A-band contains ~10⁴ myosin molecules, still too many to obtain kinetic information. To reduce this number, myosin was intentionally labeled inefficiently with fluorescent probe (i.e., only 1 in a 1,000 myosins were labeled). Myocytes were cross-linked with zero-length cross-linker to eliminate any movement during contraction while preserving ATPase activity. Figure 2 explains how the experiments were carried out.

All chemicals were from Sigma-Aldrich (St Louis, MO). The fluorescent dye SeTau-647-mono-maleimide was from SETA BioMedicals (Urbana, IL). The glycerinating solution contained: 50% glycerol, 150 mM KCl, 10 mM Tris-HCl pH 7.5, 5 mM MgCl₂, 5 mM EGTA, 5 mM ATP, 1 mM DTT, 2 mM PMSF, and 0.1% β-mercaptoethanol. The rigor solution: 50 mM KCl, 10 mM Tris-HCl pH 7.5, 2 mM MgCl₂. The EDTA-rigor solution: 50 mM KCl, 10 mM Tris-HCl pH 7.5, 5 mM EDTA. The Ca-rigor solution: 50 mM KCl, 10 mM Tris-HCl pH 7.5, 2 mM MgCl₂, 0.1 mM CaCl₂. The contracting solution: 50 mM KCl, 10 mM Tris-HCl pH 7.5, 5 mM MgCl₂, 0.1 mM CaCl₂, 5 mM ATP, 20 mM creatine phosphate and 10 units/ml of 1 mg/ml creatine kinase. The relaxing solution: 50 mM KCl, 10 mM Tris-HCl pH 7.5, 5 mM MgCl₂, 5 mM ATP, 2 mM EGTA.

LC was genetically modified to contain two thiol groups (rather than 1). This makes it possible to attach two molecules of the dye to the light chain and double the extinction coefficient of the probe. To modify ELC, the human fast skeletal muscle ELC was subcloned into pQE60 vector to produce pQE-ELC expression vector. The LC cDNA insert in pQE60-LC construct contains one Cys at amino acid position 181. We have introduced Cys in place of Gly at amino acid position 174 by site directed mutagenesis to generate pQE60-LCG174C expression vector. The resulting pQE60-LCG174C expression vector contained two Cys residues at amino acid positions 174 and 181. The pQE60-LCG174C construct was generated by PCR-based site directed mutagenesis using the QuickChange kit from Strata Gene (La Jolla, CA), pQE60-LC template plasmid, and two complimentary primers:

F-hLCG174C: 5′-gaagtggaagccctgatggcatgtcaagaagactccaatggctgc-3′ and

R-hLCG174C: 5′-gcagccattggagtcttcttgacatgccatcagggcttccacttcttcc-3′. The sequence of the plasmid pQE60-LCG174C was confirmed by DNA-sequencing of both strands of the entire plasmid. Afterwards, the plasmid pQE60-LCG174C was introduced into Escherichia coli M15 cells (Qiagen). The expressed recombinant proteins were purified on Ni-NTA-Agarose column and confirmed by immunoblotting with human LC monoclonal antibody. (We also tried to cross-link two thiols with bifunctional rhodamine in order to immobilize the dye, a procedure originally developed by Irving's group (Corrie et al., 1998, 1999; Hopkins et al., 2002; Lewis et al., 2012), but in our hands this strategy was unsuccessful).

Samples of human myocardium were collected at the University of Kentucky using procedures that were approved by the local Institutional Review Board. The human cardiac samples used in this study were acquired using the collection protocol described by Blair et al. (2016) hearts procured from patients undergoing cardiac transplants at the University of Kentucky and from organ donors who did not have heart failure. Some samples were cut into strips of ventricular walls and analyzed for tension and ATPase by Dr Campbell. Whole ventricles were then given to a researcher immediately after being removed from the patient. Ventricular tissue was dissected into ~500 mg samples and snap-frozen in liquid nitrogen within 30 min of being removed from the patient. These samples were then stored in the vapor phase of liquid nitrogen at −150°C until use. The University of Kentucky Institutional Review Board approved all procedures, and subjects gave informed consent (IRB 08-0338-F2L).

All samples were passed to a researcher as soon as they were removed from the patient and snap-frozen in liquid nitrogen within a few minutes. Samples were shipped to UNTHSC on dry ice. Immediately upon arrival in Fort Worth, they were tied to wooden sticks (in a cold room) and placed for 24 h in glycerinating solution at 0°C. After 24 h, the glycerinating solution was replaced with a fresh solution and placed at −20°C. Myocytes were made from glycerinated hearts after a minimum of 2 weeks at −20°C.

Human ventricular myocytes were used in all experiments. Use of myocytes eliminates problems associated with the non-uniformity of labeling of sarcomeres and assures better access of ligands to the contractile proteins. Myocytes were prepared from ventricles as follows: ATP present in the glycerinating solution was first removed without causing contraction by washing ventricles three times for 30 min with ice-cold EDTA-rigor solution (50 mM KCl, 10 mM Tris-HCl pH 7.5, 5 mM EDTA) followed by extensive washing with the Ca-rigor solution (50 mM KCl, 10 mM Tris-HCl pH 7.5, 2 mM MgCl₂, 0.1 mM CaCl₂).They were homogenized in the Ca-rigor solution (to avoid forming foam upon turbulent stirring when homogenized in EDTA-solution.) in the Cole-Palmer LabGen 125 homogenizer for 10 s followed by further 10 s homogenization after a cool down period of 30 s.

The measurements were carried out using Anaspec (Fremont, Ca) Sensolyte MG Phosphate Colorimetric Assay at 30°C in a 96-well plate and read on a microplate reader at 650 nm, with and without cross-linker EDC. The myocytes (1 mg/ml) were incubated in EDC as described above. The reaction mixture of 50 mM KCl, 5 mM MgCl₂, 20 mM EDC, 5 mM ATP in Ca-rigor along with 1 mg/ml myocytes was made to 80 μl in volume to which 20 μl Malachite Green (MG) reagent was added. Samples were ran in triplicate. Readings were taken for 30 min at 2 min. intervals. A graph was plotted to calculate the Pi released per s.

One milligram per milliliter myocytes were incubated for 20 min at room temperature with 20 mM water soluble cross-linker 1-ethyl-3-[3-(dimethylamino)-propyl]-carbodiimide (EDC) in the Ca-rigor solution. It is known that cross-linking of skeletal myocytes does not affect tension or ATPase (Herrmann et al., 1994; Bershitsky et al., 1997; Barman et al., 1998; Tsaturyan et al., 1999). The reaction was stopped by adding 20 mM DTT. To check whether cross-linking does not affect ATPase of cardiac MFs, ATPase was measured independently by two researchers. ATPase varied between 0.028 ± 0.001 and 0.032 ± 0.007 mols Pi/s/mol myosin, in agreement with (Alpert and Gordon, 1962). 1 In every case R² was 0.98 or 0.99. The ATPase was slightly higher for the RV than the LV. The pH of the solution (7.5) remained unchanged throughout the 20 min reaction.

Recombinant LC (G174C-hLC) was labeled at positions 174 and 181 with three-fold excess of SeTau 647 Maleimide. SeTau647-maleimide was chosen to label SH groups on LC because it is excited in the red and hence bypasses most contributions of autofluorescence (Lakowicz, 2006). SeTau is also very resistant to photobleaching (the initial rate of photobleaching was 2.4 s⁻¹) because of nano-encapsulation of the squaraine moiety of the dye chromophore system in a mixed aliphatic-aromatic macrocycle. SeTau has a large Stokes shift (44 nm), very high extinction coefficient (200,000 M⁻¹ cm⁻¹) and quantum yield (0.65). Its overall fluorescence intensity was 4.2 times larger than the fluorescence intensity of Alexa647. The reaction was allowed to occur overnight. Excess dye was removed by centrifugal filtration followed by overnight dialysis. Using two SeTau molecules per molecule of LC increases overall extinction coefficient to an extraordinary 400,000 M⁻¹ cm⁻¹. It is possible that SeTau resides on two different myosin molecules rather than single myosin, but this is of no consequence because ultimately there are ~6 molecules of SeTau in the OV (see below). The dyes were well-immobilized by myocytes in rigor and in the presence of MgADP (see below), and their combined transition dipoles have a well-defined orientation.

Following labeling of recombinant peptide with SeTau-647 Maleimide, 5 nM of this protein was exchanged with the native LC of ventricular myosin in the exchange solution containing 15 mM KCl, 5 mM EDTA, 5 mM DTT, 10 mM KH₂PO₄, 5 mM ATP, 1 mM TFP, and 10 mM imidazole, pH 7 (Duggal et al., 2015; Huang et al., 2015). The reaction was allowed to occur at 30°C for 10 min. This ensured that the exchange was inefficient (i.e., that only a small fraction of myosin carried fluorescent label). Two milligram per milliliters of freshly prepared myocytes were used.

The fluorescence was measured by a PicoQuant MT 200 time-resolved fluorescence instrument coupled to an Olympus IX 71 microscope. Before each experiment, the instrument was calibrated by measuring fluorescence of an isotropic solution of a dye with a long fluorescence lifetime (rhodamine 700, excitation at 635 nm). Such fluorescence is completely polarized (steady-state anisotropy r = 0.400, I. Gryczynski, unpublished) because the dye with a long lifetime has no time to reorient itself during a ns short laser pulse. This procedure makes sure that 100% of emitted light was detected by a parallel (∥) channel. If it was not, the prism position was adjusted until it was. Rotating the birefringent prism by 90° made sure that the 100% of emitted light was detected by a perpendicular (⊥) channel. The beam was focused by an Olympus x60, NA = 1.2 water immersion objective on the A-band of a myocyte. The fluorescence intensity was collected into 10 μs bins for 20 s (i.e., there were 2,000,000 bins). To smooth the data, 1,000 bins were averaged into 2,000 bins yielding overall time resolution of 10 ms. Polarization of fluorescence (PF) and steady-state anisotropy (r) were computed from the orthogonal components of fluorescent light. The autocorrelation function (ACF) and the plot of PF vs. frequency with which it occurs over 20 s (histogram), was constructed from these 2,000 measurements. Approximately 20–30 myocytes were examined in each ventricle.

To determine the number of observed myosin molecules, it was necessary to measure fluorescence intensity associated with one molecule of SeTau. This number was determined by Fluorescence Correlation Spectroscopy (FCS). The autocorrelation function at delay time 0 of fluctuations of freely diffusing SeTau molecules entering and leaving the OV is equal to the inverse of the number of molecules contributing to the fluctuations N = 1/ACF(0) (Magde et al., 1974; Elson, 1985, 2007). Fifteen nanomolars SeTau was illuminated with 1–70 μW of laser power at 635 nm. A calibration curve was constructed by plotting the power of the laser vs. the rate of photon arrival per molecule of the dye. From this curve it was determined that the number of photons/s collected from a single SeTau molecule was ~900.

A typical fluorescent signal obtained from ventricular myocyte is shown in Figure 3.

The average number of observed photons per second in a parallel channel was 2,731 ± 645 (Figure 3, red). The average number of photons per second in perpendicular channel (black) was 1,161 ± 364. The average total fluorescence intensity was I_∥ + 2^*I_⊥ = 5,053 ± 1,373 counts/s. Therefore, the average number of observed myosin molecules was 5,053/900 = 5.6. However, it should be noted that as long as the number of molecules is mesoscopic, the exact number does not matter. The same results would have been obtained observing 6 or 60 molecules.

FT300 fluorescence lifetime spectrometer (PicoQuant GmbH, Germany) was used for time-resolved measurements. The muscle was excited by Fianium supercontinuum white light (Fianium Ltd, Whitelase SC400-4). The white light was passed through a monochromator set at 635 nm at 10 MHz repetition rate. The emission was passed through long pass 650 nm filter and detected by a fast microchannel photomultiplier tube. The resolution was kept at 4 ps per channel and the pulse width was less than 100 ps. Fluorescence intensity decays were collected while orienting the emission polarizer in vertical and horizontal position respective to the vertically oriented excitation polarizer for measuring anisotropy. The vertical (parallel) and horizontal (perpendicular) intensity decays were used to calculate the time dependent anisotropy using the equation (t) = (I_Parallel (t)−I_Perpandicular (t))/(I_Parallel (t)+2I_Perpandicular (t)).

The obtained anisotropy decay was analyzed using Fluofit 4.0 program provided by PicoQuant and was fitted using formula r(t) = ∑ R_ie^∧(−t/Ø_i) where r is the total anisotropy, R_i is the fractional anisotropy amplitude associated with ith component, Ø_i is the i-th rotation correlation time and t is the time in nanoseconds. SSFA is the mean of r at all times until the time after which r remains constant (12 ns).

Comparisons between groups were performed using the unpaired Student's t-test by Origin v.8.6 (Northampton, MA). Origin was also used to fit the data by a non-linear routine, calculate histograms, and autocorrelation functions. The significance was defined as P < 0.05.

The chief difficulty in obtaining single molecule data from a ventricle In-situ lies in assuring adequate Signal/Noise ratio. Adequate S/N is possible when observing a single myosin molecule in-vitro (Forkey et al., 2003, 2005). However, in-vitro data cannot be viewed as originating from independent motors (Walcott et al., 2012) (Pate and Cooke, 1991; Baker et al., 2002). In-situ data from a ventricle contains significant contributions from the background. In addition to the very strong autofluorescence due to an extremely dense environment (Bagshaw, 1982), the background consists of a constant fluorescence coming from myosin that is always present in the OV. In contrast to the regular FCS, the myosin molecules do not translate and fluorescent signal does not fluctuate between zero and maximum. In our experiments only the PF fluctuates around the mean. Moreover, to avoid photobleaching the laser beam cannot be focused on the same spot for much more than 20 s.

We emphasize again that the data presented here was obtained from the ventricles from HF hearts. When cross-linked, they do not shorten, but retain full ATPase activity 0.040 ± 0.004 s⁻¹/mol; 0.034 ± 0.004 s⁻¹/mol for control and cross-linked respectively for the LV, and 0.033 ± 0.003 s⁻¹/mol; 0.037 ± 0.004 s⁻¹/mol for control and cross-linked respectively for the RV.

The experiment begins by placing isolated myocyte on an ethanol cleaned cover slip. Because we observe ~6 myosin molecules (and not a single one), the changes of anisotropy occur gradually (Figure 4, and not in steps as illustrated in Figure 1 which shows changes for a single molecule). The time course of SSFA can be simplified to a scheme shown as a Figure 4, red line. The cycle begins when XB is dissociated from a thin filament where SSFA is low (A). Binding to a thin filament causes a small increase in SSFA (B) transition (A → B). Dissociation of Pi and assumption of actin-myosin-ADP complex (transition from apoenzyme to holoenzyme (B → C) causes further increase in SSFA (C) in accordance with (Coureux et al., 2004). Finally, dissociation of myosin from actin (C → A) occurs with the rate k_DISS.

To extract the rate constants from fluorescence data we used the method first introduced by Magde and Elson (Magde et al., 1974; Elson, 1985, 2007). In this method Autocorrelation Function (ACF) of fluctuations of anisotropy is calculated. ACF is an average of the sum of products of the instantaneous values of anisotropy and the values of anisotropy delayed by delay time τ ranging from 0 to 1 s. The process of calculating ACF is shown in Figure S3. The analytical form of ACF of the three-state process is very complex and is fully described in Mettikolla et al. (2011).

A typical example of the ACF data from a HF ventricle is shown in Figure 5. The non-linear fit to the analytical form of ACF is shown as the red line. The experimental data clearly indicates a two-state process. XB binding to actin (rate k_B) is too fast to be reflected in the experimental points. This is due to the limited time resolution of the instrument. It collects photons every 10 μs, but in order to decrease the noise, the 2 M data points are packed in 2,000 bins, 1,000 points per bin, Figure S3. Thus, the instrumental response time is 10 ms. We cannot detect processes faster than 100 Hz (s⁻¹).

We analyzed data from 6 HF ventricles. One ventricle was from a patient suffering from non-compaction. Data from this ventricle was eliminated. Table 1 shows averages from five remaining ventricles. There was no statistical difference between LV and RV for either rate constant. The data from individual ventricles is shown in Figure S4.

A characteristic feature of a distribution of orientations is Full Width at Half Maximum (FWHM) [it is related to Standard Deviation, SD; for Gaussian distributions FWHM = 2Sqrt(2ln2) = 2.355 SD]. The distribution of polarizations was Gaussian. The differences between FWHM's of polarization of fluorescence of different populations of ventricles are statistically interpretable only when the signal from each half-sarcomere is similar. This is because the relative value of FWHM of a strong random signal is small relative to a weak signal (Midde et al., 2013). To comply with this requirement, the power of a laser was adjusted within 1–2 μW of the mean. Examples of measurements of 27 experiments from LVs and 26 experiments from RVs are shown in Figure 6. The distributions are displayed here as histograms (plots of frequency of occurrence of a given orientation.

Polarization of Fluorescence (PF) rather than steady-state anisotropy is used here to describe orientation of the lever arm because it has been used routinely to measure conformation of XBs (Dos Remedios et al., 1972; Nihei et al., 1974; Tregear and Mendelson, 1975; Morales, 1984; Hopkins et al., 1998, 2002; Sabido-David et al., 1998). PF is closely related to the SSFA by SSFA = 2PF/(3-PF). PF is the difference between ∥ and ⊥ components of the fluorescent light emitted by the dye, normalized by their sum.

All data is summarized in Table 2. The distributions show no differences in the value of FWHM. AR² values indicate how well the fitted curve (red) matches a perfect Gaussian (a perfect fit has an AR² value of 1).

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.