INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS AND MATERIALS RESULTS DISCUSSION REFERENCES

THE FRANK-STARLING RELATIONSHIP describes an enhancement of contractile function of the heart in response to an increase in end-diastolic ventricular volume. Increased volume implies an increase in muscle length. It has been known for some time (1, 2, 7, 13, 14, 16) that this phenomenon has, at its basis, an increase in the activation of myofibrils by calcium as the muscle length increases. This behavior is exhibited by both skeletal and cardiac muscle, but in cardiac muscle the phenomenon is much more pronounced. The underlying molecular mechanisms for this increase in Ca²⁺ sensitivity are, however, as yet unclear.

It has been suggested that the enhanced contractile response of cardiac muscle when it is stretched is at least partly because the myofilaments move closer together, thereby enhancing the probability of favorable acto-myosin interaction (7, 13, 22). This, in turn, results in increased sensitivity to calcium by cooperative activation of the thin filament. This is an attractive hypothesis because it proposes a simple physical basis for a very important feature of cardiac muscle physiology.

Although a decrease in lattice spacing with increases in muscle length in both skinned and intact skeletal muscle has been established for a long time (6, 19), there are relatively few such measurements in cardiac muscle (18, 20) that would allow one to substantiate the interfilament spacing hypothesis and to make quantitative predictions. Whereas intact cardiac muscle has been shown to maintain a constant lattice volume (21), there have been no direct measurements to date of lattice spacing in skinned cardiac muscle as a function of sarcomere length over the working range of the heart.

Here, we describe an X-ray diffraction method and apparatus for handling isolated cardiac trabeculas that allowed for accurate simultaneous measurement of sarcomere length and interfilament lattice spacing from good-quality X-ray diffraction patterns. The measurements show that the interfilament lattice spacing in the A bands of relaxed, skinned cardiac muscle from rats is indeed a sensitive function of sarcomere length. In addition, we also show that lattice spacing in intact rat cardiac muscle is significantly smaller than that in skinned muscle over the working sarcomere length range of cardiac muscle, but that the relative change in lattice spacing over this range is similar in both preparations. These results are consistent with the notion that lattice spacing changes may underlie the Frank-Starling relationship.

	METHODS AND MATERIALS

TOP ABSTRACT INTRODUCTION METHODS AND MATERIALS RESULTS DISCUSSION REFERENCES

Preparation of isolated rat trabeculas. Rats, 250-350 g in weight and of either sex, were anesthetized with ethyl ether, and the hearts were rapidly excised. All procedures that were used in the current study were in accordance with institutional guidelines regarding the care and use of laboratory animals. After the heart was excised, it was immediately perfused with a modified Krebs-Henseleit solution containing 20 mmol/l 2,3-butanedione monoxime (BDM) and placed in a dissection dish beneath a binocular microscope equipped with an ocular micrometer (~10 µm resolution) (5). Treatment with BDM has been shown to protect myocardium from cell contracture and muscle damage during dissection (25). Thin, unbranched, and uniform trabeculas were carefully dissected from the free wall of the right ventricle. Typical dimensions of these preparations were 3-5 mm long, 80-150 µm wide, and 50-80 µm thick. After dissection was completed, the trabeculas were transferred to a dish containing cold standard relaxing solution to which 1% (vol/vol) Triton X-100 was added to chemically permeabilize the preparation. The composition of this standard relaxing solution was (in mmol/l) 5.97 Na₂ATP, 6.31 MgCl₂, 20 EGTA, 10 phosphocreatine, and 100 N,N-bis(2 hydroxyethyl)-2-aminoethanesulfonic acid (BES); pH was 7.0 (adjusted with KOH), and ionic strength was 200 mmol/l (adjusted with KCl). The preparations were left in this solution for at least 2 h (typically overnight) at 4°C to ensure complete solubilization of all membranous structures. Trabeculas were attached to aluminum T clips before they were mounted in the experimental setup (4). Intact trabeculas were dissected similarly but were maintained in modified Krebs-Henseleit solution to which 20 mM BDM was added to assure complete relaxation of the muscle preparation. The solution was in equilibrium with 100% O₂ while pH was maintained at 7.35 by 25 mM NaHCO₃. All experiments were performed at room temperature (23 ± 1°C).

X-ray diffraction measurements. The overall experimental arrangement is shown in Figure 1. Experiments were performed using the BioCAT undulator-based beamline at the Advanced Photon Source, Argonne National Labs (Argonne, IL) (15). The high X-ray flux density and low beam divergence delivered by this instrument are highly advantageous for small-angle X-ray studies of small specimens, such as those described here. Experiments were done using a 3-m distance between the sample and the detector and with the X-rays beam energy set to 12 keV (1.03 Å wavelength). All flight paths were evacuated except for a small gap around the sample chamber itself (~1 mm downstream and 2 mm upstream of the sample). The beam size at the sample position was collimated to ~0.4 × 0.8 and ~0.040 × 1 mm (vertical × horizontal) at the detector and contained a maximum incident flux of ~3 × 10¹² photons/s.

View larger version (19K): [in this window] [in a new window]   Fig. 1.   Block diagram of the experimental arrangement in the experimental enclosure 18-ID-D of the BioCAT undulator beam line. A highly collimated beam from the undulator (~0.25 × 1.0 mm, horizontal × vertical, at the sample) containing 1012-1013 photons/s enters from the right, coming to a focus of ~0.03 × 1 mm (horizontal × vertical) at the front face of the charge-coupled device (CCD) detector. The muscle sample is held horizontally between the motor and force transducer. Striations are imaged by the ×40 objective of the inverted microscope and digitized by a frame grabber in the personal computer. Length, force, and X-ray exposure signal are also recorded via the analog-to-digital (A/D) subsystem, and muscle length is controlled via the digital-to-analog (D/A) converter (see METHODS AND MATERIALS).

Data acquisition and sarcomere length measurement. Sarcomere length was measured both immediately before and after the X-ray exposure by a multiple, pseudo-two-dimensional fast Fourier transform analysis technique of digitized striation images, as described previously (8). Briefly, the bright-field microscopic image of the striation pattern (collected with a scientific grade CCD-based video camera, model 2122, Cohu) was digitized using the built-in video capture hardware in a Macintosh 7600/G3 personal computer. A region of the image that encompassed most of the attached fiber was selected, and each horizontal pixel line was transformed by fast Fourier transformation into the spatial frequency domain. The amplitude spectra were averaged, and the peak power of the first-order harmonic in the spatial frequency domain was detected. This value was then converted into a median sarcomere length across the region. The system was calibrated with glass gratings of known spacing. Force, muscle length, and X-ray beam intensity were digitized by an analog-to-digital converter (model PCI-1200, National Instruments) installed in the same computer used for digitizing the striation images. The software to collect, process on-line, and store data for off-line processing has been custom designed (see Fan et al., Ref. 8) using the LabView graphical data acquisition language (National Instruments).

Protocols. Individual trabeculas were stretched to arbitrary lengths between 1.9 and 2.4 µm at the beginning of the experiment. Sarcomere lengths were checked both before and immediately after the X-ray exposure. At this time, solution flow was briefly interrupted to obtain a clear striation pattern so as to allow for accurate video sarcomere length determination. Fiber length was then systematically lengthened or shortened to collect at least 10-12 consecutive data points describing the relationship between sarcomere length and interfilament lattice spacing. Next, an additional two to four data points were collected at various sarcomere lengths back along the curve to check for reproducibility or hysteresis. The maximum fiber length that was examined was determined by the onset of a rapid increase in the level of passive tension. The length at which this occurred varied from fiber to fiber. We still collected data at this sarcomere length, but, in this case, X-ray exposure was delayed (10-30 s) in order for passive tension to decay to a steady-state value, as judged by the tension traces. The level of passive tension in the lattice was recorded when the X-ray exposure was taken.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS AND MATERIALS RESULTS DISCUSSION REFERENCES

For most experiments, we attenuated the beam fourfold using aluminum filters and an exposure time of 800 ms. Our data showed that it is possible to obtain at least 16 exposures (and as many as 36 exposures from a single trabecular muscle preparation with this level of absorbed X-ray dose, 3-7 × 10² Gy or 3-7 rads total) before irreversible changes due to radiation damage are noted by deterioration of the striation pattern in the microscopic image or of the X-ray diffraction patterns.

Figure 2A shows a typical X-ray pattern taken from a skinned preparation in relaxing solution that shows clearly resolved 1,0 and 1,1 equatorial reflections. The separation of these reflections allowed lattice spacings to be determined to about ±0.5 pixels out of 200-600 pixels (i.e., better than 0.25% accuracy; equivalent to a interfilament spacing resolution of 0.1 nm). Figure 2B shows a typical bright-field image of the striation pattern obtained with our apparatus, and Fig. 2C shows the associated power spectrum. The sharpness of the peak allowed calculation of sarcomere length to ~10 nm resolution with an acquisition speed of ~0.5 s.

View larger version (53K): [in this window] [in a new window]   Fig. 2.   A: typical X-ray pattern taken from a skinned preparation in relaxing solution. The 1,0 and 1,1 equatorial reflections are as indicated. B: typical bright-field image of the striation pattern obtained with our apparatus. C: associated power spectrum used for the on-line determination of sarcomere length.

Figure 3 shows lattice spacing (d₁₀) of an individual skinned trabecula as a function of sarcomere length for a stretch-release protocol. Note the reproducibility of the curve as the fiber is stretched and subsequently released. Figure 4 shows the averaged interfilament lattice spacing as a function of sarcomere length over the full working range of sarcomere length for nine trabeculas. The average data were obtained by placing the sarcomere length over 0.05-µm-wide intervals into a bin and next averaging the mean value per fiber per bin. This relationship can be described by biphasic linear behavior with a steeper slope from 1.9 to 2.20 µm (8.12 ± 0.33 nm/µm) than from 2.2 to 2.4 µm (3.85 ± 0.49 nm/µm) but could be equally well described by a curvilinear relationship (e.g., a second-order polynomial).

View larger version (13K): [in this window] [in a new window]   Fig. 3.   Individual measurements () of lattice spacing (d10) of 1 sample of rat cardiac trabecula as a function of sarcomere length for a given stretch-release protocol. The dashed line indicates the order in which the measurements were made.

View larger version (12K): [in this window] [in a new window]   Fig. 4.   Averaged lattice spacing () for 9 relaxed, skinned rat trabeculas as a function of sarcomere length obtained by placing the sarcomere length in 0.1-µm intervals into bins and averaging the average value per fiber per bin. Error bars are means ± SE. The solid line indicates the behavior expected for a constant volume system based on the average lattice volume (see DISCUSSION).

Figure 5 shows the averaged data from intact muscle samples over a similar range of sarcomere lengths. A line representing the best-fit second-order polynomial for the skinned muscle data is shown for comparison. The intact lattice spacing data were roughly parallel to the data from skinned muscle but displaced ~7.5 nm downward. There is more scatter in the data, which was due to a larger variability in the sarcomere length determination, owing to the fact that the striation images in these preparations were generally of poorer quality than those in the skinned samples because of the presence of membrane-bound organelles in the intact samples. It should be noted that the X-ray diffraction images obtained from intact and fully relaxed trabeculas were of equal or better quality than those obtained from skinned trabeculas (cf. Fig. 2).

View larger version (9K): [in this window] [in a new window]   Fig. 5.   Average d10 measurements () from 5-11 intact trabeculas (except for the point near 1.85 µm where n = 3) when placed in bins over a suitable range of sarcomere lengths (~ 0.1 µm) from intact muscle samples. The solid line is the behavior expected for constant lattice volume based on the average lattice volume (see DISCUSSION). The dashed line representing the best-fit second-order polynomial for the skinned muscle data is shown for comparison.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS AND MATERIALS RESULTS DISCUSSION REFERENCES

In the intact, resting muscle, lattice spacing changes with sarcomere length in a way that observes constant lattice volume [/2 × (d₁₀)² × sarcomere length] in all systems studied (skeletal and cardiac muscle), showing that the overriding constraint on lattice spacing is the volume constraint provided by the cell membrane (see Ref. 23 for review). Lattice spacings in intact, resting heart papillary muscle from cats and goats has been shown to behave isovolumetrically, with changes in sarcomere length over the range of 2.0 to 2.5 µm (21), and these spacings are similar to those found in amphibian skeletal muscle. In the data presented here, the lattice volume (averaged over all measurements) was calculated to be 2.91 × 10⁶ nm³ with a standard deviation of 2.1 × 10⁵ nm³. The constant volume line is also plotted as a function of sarcomere length in Fig. 5, showing that our intact data are also consistent with the constant volume hypothesis.

Skinned muscle lacks the volume constraint imposed by the plasma membrane. The observed shrinkage behavior (Fig. 4) of the myofilament lattice with increasing length is not consistent with constant lattice volume because the degree of curvature is too large (c.f., solid line vs. data points). It is also not consistent with a linear decrease in lattice spacing associated with a linear increase in sarcomere length. It is possible that the decreased shrinkage with incremental increases in sarcomere length observed at lengths above 2.2 µm could be due to a large increase in electrostatic repulsive forces as the filaments come closer together (23, 24). Calculations show that although electrostatic repulsive forces are very weak at the largest spacings studied, they increase by several orders of magnitude at the smallest. (Repulsive pressure increases approximately exponentially as the interfilament spacing decreases.) One also needs to address the origin of attractive forces causing shrinkage of the lattice with increasing length in relaxed muscle when the plasma membrane is absent. The most likely origins of these compressive forces are radial components of passive tension. It should be stressed, however, that the present data do not allow us to determine whether these originate in extracellular (e.g., collagen-based) or intracellular (e.g., titin-based) structures (see Ref. 11). Clearly, resolution of this question will require further experiments. Interestingly, recent reports (3, 9) suggest that indeed the relevant structures are titin based. In those studies, however, interfilament lattice spacing was estimated by overall cellular dimensions rather than by X-ray diffraction techniques, a method that may not provide an unambiguous estimate of true lattice spacing.

As observed in skeletal muscle (see Ref. 24 and references therein), lattice spacings in skinned muscle are substantially larger than in intact muscle due to the loss of the osmotic constraint to swelling imposed by the plasma membrane. To compensate for this effect, various concentrations of high-molecular-weight polymers (dextran and polyvinyl pyrrolidone) are commonly added to the bathing solution to attempt to restore the in vivo lattice spacing. We performed preliminary experiments using dextran T500 (Sigma), which showed that whereas the spacing at 2.1 µm sarcomere length was 41.8 ± 0.19 (n = 9) in skinned muscle with no added dextran, it was 38.8 ± 0.33 (n = 4) at 2% added dextran and 36.3 ± 0.29 (n = 3) at 4% dextran. If one assumes a linear relationship, this would indicate that ~5.0% dextran would restore the in vivo spacing of 34.84 ± 0.64 (n = 5).

Numerous studies (1, 2, 7, 13, 14, 16, 22) have shown an increase in the calcium responsiveness of force development of the cardiac sarcomere upon an increase in sarcomere length. A recent study (17) has confirmed that a similar phenomenon occurs in intact, tetanized isolated myocardium. It has been shown that calcium responsiveness of force development is enhanced in intact preparations relative to chemically permeabilized preparations. This observation has been attributed to the loss of moieties from the cytosol upon skinning that confer enhanced myofilament calcium sensitivity (10). Alternatively, technical considerations related to both calibration of calcium transient data as well as uncertainties of solution composition in the skinned fiber experiments may also explain this phenomenon. The results of the present study, however, are consistent with the notion that a general increase in interfilament lattice spacing may cause the overall reduced calcium sensitivity that is observed in skinned preparations. Because alterations in sarcomere length in the intact muscles resulted in relatively similar alterations in interfilament lattice spacing, this hypothesis is also consistent with the presence of length-dependent myofilament activation in intact myocardium (17).

The results of the present study represent an important first step in determining whether the myofilament lattice spacing is in fact the molecular length sensor responsible for the enhanced sensitivity to calcium at longer sarcomere lengths, in that they show that the lattice spacing is indeed a sensitive and predictable function of sarcomere length both in skinned and intact cardiac muscle preparations. Thus our results are consistent with the hypothesis that lattice spacing changes may underlie the Frank-Starling relationship. The data do not in themselves, however, constitute proof of this concept. Further experiments are required. For example, it is not yet clear whether the lattice spacings in contracting muscle differ from those present in relaxed muscle and, if so, which is the relevant spacing for conferring length sensitivity. Compression of the lattice with high-molecular-weight compounds such as dextran or polyvinyl pyrrolidone (see, e.g., Refs. 22 and 26) coupled with X-ray diffraction measurements may help clarify matters by decoupling lattice spacing changes from sarcomere length changes and may thus provide further critical tests of the myofilament lattice spacing hypothesis of length-dependent activation of striated muscle.