Attenuated sarcomere lengthening of the aged murine left ventricle observed using two-photon fluorescence microscopy

Cardiomyocytes within aged mouse hearts exhibited impaired sarcomere lengthening during left ventricular filling, and cardiomyocytes isolated from aged hearts demonstrated increased stiffness as measured with atomic force microscopy. Impaired sarcomere lengthening with ventricular filling will contribute to cardiac dysfunction with advancing age.

Abstract

The Frank-Starling mechanism, whereby increased diastolic filling leads to increased cardiac output, depends on increasing the sarcomere length (L_s) of cardiomyocytes. Ventricular stiffness increases with advancing age, yet it remains unclear how such changes in compliance impact sarcomere dynamics in the intact heart. We developed an isolated murine heart preparation to monitor L_s as a function of left ventricular pressure and tested the hypothesis that sarcomere lengthening in response to ventricular filling is impaired with advanced age. Mouse hearts isolated from young (3–6 mo) and aged (24–28 mo) C57BL/6 mice were perfused via the aorta under Ca²⁺-free conditions with the left ventricle cannulated to control filling pressure. Two-photon imaging of 4-{2-[6-(dioctylamino)-2-naphthalenyl]ethenyl}1-(3-sulfopropyl)-pyridinium fluorescence was used to monitor t-tubule striations and obtain passive L_s between pressures of 0 and 40 mmHg. L_s values (in μm, aged vs. young, respectively) were 2.02 ± 0.04 versus 2.01 ± 0.02 at 0 mmHg, 2.13 ± 0.04 versus 2.23 ± 0.02 at 5 mmHg, 2.21 ± 0.03 versus 2.27 ± 0.03 at 10 mmHg, and 2.28 ± 0.02 versus 2.36 ± 0.01 at 40 mmHg, indicative of impaired sarcomere lengthening in aged hearts. Atomic force microscopy nanoindentation revealed that intact cardiomyocytes enzymatically isolated from aged hearts had increased stiffness compared with those of young hearts (elastic modulus: aged, 41.9 ± 5.8 kPa vs. young, 18.6 ± 3.3 kPa; P = 0.006). Impaired sarcomere lengthening during left ventricular filling may contribute to cardiac dysfunction with advancing age by attenuating the Frank-Starling mechanism and reducing stroke volume.

NEW & NOTEWORTHY

Cardiomyocytes within aged mouse hearts exhibited impaired sarcomere lengthening during left ventricular filling, and cardiomyocytes isolated from aged hearts demonstrated increased stiffness as measured with atomic force microscopy. Impaired sarcomere lengthening with ventricular filling will contribute to cardiac dysfunction with advancing age.

the frank-starling mechanism describes the heart's intrinsic response to diastolic filling, where an increase in filling associates with an increase in stroke volume [for historical review, see Katz (18)]. At the cellular level, the Frank-Starling mechanism occurs via increases in sarcomere length (L_s) of cardiomyocytes within the chamber wall, with subsequent length-dependent activation of myofilament force production (2, 8, 10, 26). For example, during submaximal Ca²⁺ activation of isolated cardiomyocytes (7, 9, 13) or trabecular/papillary muscle preparations (17, 19, 32), contractile force can increase severalfold through the range in L_s from ∼1.70 to ∼2.20 μm. While force production as a function of L_s has been well defined in isolated cardiac muscle preparations, there is a paucity of information concerned with how left ventricular filling pressure alters L_s of cardiomyocytes in the intact living heart. Fluorescent labeling of the t-tubule network gives accurate indication of z-line spacing and therefore L_s in cardiomyocytes of the rat heart (5, 6). However, these latter studies examined L_s in the absence of ventricular filling pressure. Thus the relationship between ventricular filling and L_s in the living heart remains to be elucidated.

The aged heart exhibits complex chamber remodeling as manifested by cardiomyocyte hypertrophy, cellular disorganization, and stiffening of both cardiomyocytes (23) and extracellular matrix components (20). A decrease in ventricular compliance is evident in the pressure-volume relationship of the aged compared with young heart (3, 27), yet it remains unclear how age-associated ultrastructural changes impact the functional relationship between ventricular filling pressure and cardiomyocyte L_s in the intact heart. For the present study in mice, we developed an isolated heart preparation to monitor the relationship between left ventricular pressure and cardiomyocyte L_s. We tested the hypothesis that sarcomere lengthening in response to ventricular filling is impaired with advanced age.

MATERIALS AND METHODS

Solutions and Chemicals

Chemicals and reagents were obtained from Sigma-Aldrich (St. Louis, MO) unless otherwise noted. Nominally Ca²⁺-free physiological saline solution (Ca²⁺-free PSS) contained (in mM) 135 NaCl, 5 KCl, 1 MgCl₂, 10 d-glucose, and 10 HEPES (pH 7.4) with NaOH. Experiments using isolated cardiomyocytes were performed in nominally Ca²⁺-free PSS or PSS containing 2 mM CaCl₂. Caffeine (10 mM) was prepared daily in 2 mM CaCl₂ PSS. Ca²⁺-free cardioplegia solution consisted of Ca²⁺-free PSS, supplemented with 100 μM EGTA (to chelate contaminating Ca²⁺) and 6% dextran (∼500,000 mol wt, to maintain colloid osmotic pressure and limit tissue edema).

Animals and Cardiac Excision

Procedures were approved by the Animal Care and Use Committee at the University of Missouri and complied with all United States regulations involving animal experimentation. Male C57BL/6 mice were obtained from the colony of the National Institute on Aging (housed at Charles River Laboratories, Wilmington, MA) and housed at the University of Missouri for 1–4 wk before experimentation. Age groups studied were 3–6 mo (defined as young, n = 25) and 24–28 mo (defined as aged, n = 16). Each mouse was anesthetized with intraperitoneal injection of pentobarbital sodium (60 mg/kg). Following confirmation of deep anesthesia, the heart was rapidly (∼30 s) excised and immersed in 4°C Ca²⁺-free PSS, and the mouse was euthanized by exsanguination.

Cardiomyocyte Isolation, Dye Loading, and Confocal Imaging

Fluorescent markers of cardiomyocyte L_s were initially characterized and compared with sarcomere striations evaluated using transmitted light in enzymatically isolated cardiomyocytes. Following cardiac excision, the aorta was cannulated for retrograde perfusion of the myocardium via the coronary arteries. Perfusion began with Ca²⁺-free PSS containing 2 U/ml heparin for 10 min, followed by 15 to 16 min with a MEM-based enzymatic isolation solution, supplemented with (in mM) 10 NaHCO₃, 2 Na-pyruvate, 10 Na-HEPES, 10 HEPES, 8 taurine in addition to 20 μM CaCl₂, 50,000 U/l penicillin-streptomycin (PenStrep, Life Technologies, Grand Island, NY), and 22.5 μg/ml⁻Liberase Blendzyme TH (Roche Applied Science, Indianapolis, IN). Perfusion solutions were maintained at 37°C (pH, 7.35) while oxygenated with 95% O₂-5% CO₂. The left ventricle and septum were then isolated, minced, agitated, and filtered at 22–25°C. Isolated cardiomyocytes were plated on laminin-coated 25 × 25-mm glass coverslips and loaded with 5 μM 4-{2-[6-(dioctylamino)-2-naphthalenyl]ethenyl}1-(3-sulfopropyl)-pyridinium (di-8-ANEPPS; Molecular Probes/Life Technologies, Grand Island, NY) and 5 μM MitoTracker Deep Red (Molecular Probes/Life Technologies) for 10–20 min followed by a 30–60-min wash. Ventricular cardiomyocytes were studied using two-dimensional imaging using the galvano scanhead of the Leica SP5/MP (×40 oil objective; HCX PL APA, numerical aperture, 1.25; 512 × 128 pixels; bidirectional scanning/2× line averaging; pixel size, 0.24 μm). Excitation was at 488 and 633 nm with emission recorded at 500–560 and 650–800 nm for respective dyes. Transmitted light images were simultaneously acquired to monitor L_s. To deplete sarcoplasmic reticulum Ca²⁺ stores and induce cardiomyocyte contracture (producing short L_s), 10 mM caffeine was applied via a gravity-fed pipette. Imaging was performed at 22–25°C.

Isolated-Perfused Heart Preparation with Left Ventricular Cannulation

Imaging platform.

A custom imaging platform was fabricated for use with a Leica SP5/MP inverted confocal microscope. The platform secured an imaging chamber and two cannulae for concurrent aortic and ventricular cannulation. The imaging chamber was sealed to 24 × 50-mm coverslips using vacuum grease and placed within a PH-1 heating element (Warner Instruments, Hamden, CT). The aortic cannula [polyethylene (PE)-90 tubing, coupled to a SH-27B inline solution heater (Warner Instruments)] and ventricular cannula (PE-100 tubing) were positioned at angles to 1) facilitate cannulation procedures beneath the stereomicroscope; 2) permit cannulation of both the left ventricle (via the left atrium) and the ascending aorta (superior to the aortic valve) while avoiding bending or twisting of the aorta; 3) allow the left ventricular free wall to rest gently against the coverglass; 4) prevent physical contact between platform components and the Leica SP5/MP.

Procedures.

Excised hearts were placed in 4°C Ca²⁺-free PSS within the imaging platform, and the aorta was cannulated and then perfused retrogradely with room temperature Ca²⁺-free PSS containing 2 U/ml heparin for ∼3 min at a pressure of 41 cm·H₂O to flush blood from the coronary vasculature, which coincided with loss of spontaneous contractile activity. Note that all pressures refer to a column of PSS above the heart, where 0, 7, 14, 27, 41, and 54 cm of PSS equate to pressures of 0, 5, 10, 20, 30, and 40 mmHg. The aortic cannula was then removed, and the cannula for the ventricle was inserted into the left atrium (via an incision) and guided through the mitral valve into the left ventricle. To secure the tip of cannula within the ventricle, the atrium and pulmonary vessels were ligated with 6-0 silk suture. Ca²⁺-free cardioplegia solution was perfused into the ventricular compartment via the cannula (∼10 mmHg for ∼10 s) to flush the chamber, and the heart was rotated to position the left ventricular free wall against the coverslip (see Fig. 2, A and B). The aorta was then recannulated, and coronary perfusion was reintroduced at 30 mmHg, with a ∼1-mm incision made in the right atrium to drain coronary venous effluent. Coronary flow draining into the left ventricular chamber via the Thebesian circulation was removed by opening the ventricular cannula, set at 0 mmHg. Elevation of the ventricular solution column above the aortic solution column resulted in aortic valve opening and equilibration of the columns and confirmed that the ventricular compartment filled at the pressure set by the ventricular column. The time required for ventricular cannulation (following perfusion of the myocardium with Ca²⁺-free cardioplegia solution) was typically 10–15 min.

Fig. 2.

L_s measurements in an isolated, perfused heart preparation. A and B: photographs of an isolated perfused heart preparation held at left ventricular (LV) pressures of 0 (A) and 40 (B) mmHg. The aortic and LV cannulae are marked aortic and LV, respectively. In this configuration, the LV free wall gently rests against the coverslip for imaging. The superfusion pipette, vacuum suction pipette, and superfusion solution were removed for clarity. C–E: 2-photon fluorescence images (×20 objective) of subepicardial (10–25 μm beneath the pericardium) areas of a heart loaded with di-8-ANEPPS. Variability included highly vascularized regions (C), complex cardiomyocyte arrangements (D), and parallel sheets of cardiomyocytes interspersed between capillaries (E).

Dye Loading and Two-Photon Imaging of Perfused Heart Preparation

Following dual-cannulation procedures, the heart was retrogradely perfused at 30 mmHg with solutions oxygenated with 95% O₂-5% CO₂ at 22–25°C. Di-8-ANEPPS was loaded into cells of the heart by perfusing 10 ml of Ca²⁺-free PSS containing 2 U/ml heparin and 10 μM di-8-ANEPPS, followed by 10 ml of Ca²⁺-free PSS with 2 U/ml heparin (perfusion rate was typically 1.5–2 ml/min). The perfusion solution was then switched to Ca²⁺-free cardioplegia solution (perfusion rate, ∼0.5 ml/min), and the preparation was moved to an inverted Leica SP5/MP microscope for frame scan (xy) two-photon fluorescence imaging (×40 oil objective; HCX PL APA, excitation at 840 nm; emission at 500–700 nm, 1,024 × 1,024 pixels at 0.19 μm/pixel). Laser and photomultiplier gain settings were optimized to pixel intensity histograms obtained from fluorescence signal within cardiomyocyte regions. These settings associated with saturated pixel intensities within the coronary vasculature (Fig. 2, C–E). Subepicardial cardiomyocytes located 10–25 μm beneath the surface of the heart were imaged while ventricular pressure was maintained at 0, 5, 10, 20, 30, and 40 mmHg. Two minutes before each set of measurements, the hydrostatic column perfusing the aorta was repositioned to achieve an aortic pressure of 1 mmHg greater than ventricular pressure. This approach was used to maintain aortic valve closure and ensure that in the case of aortic valve incompetence (associated with advancing age), the ventricular pressure would be within 1 mmHg of the desired pressure. Angle of the longitudinal cell plane relative to the imaging plane, particularly with respect to rotation around the middle (width) axis of the cell, affects apparent L_s as described in detail by Bub et al. (6) and Botcherby et al. (5). To determine this angular component, we performed optical sectioning (z sectioning) of subepicardial cardiomyocytes over a distance of 20–25 μm (section thickness of 1 or 2 μm) to determine the number of optical sections separating longitudinal edges of cardiomyocytes. With the assumption of rectangular prism cardiomyocyte geometry, the edges on opposite sides of the prism will be confined within a z distance of (n sections +1) × section thickness. Maximum angular rotation (around the middle axis) of the longitudinal cell plane relative to the imaging plane (θ_max) was calculated according to the equation: Tan(θ_max) = z distance/length between cell edges. The ratio of true L_s to measured L_s was then calculated as Cosθ_max. In subepicardial cardiomyocytes, θ_max ranged from 0.3 to 4.3° (mean, 1.6 ± 0.2°, n = 24); thus measured L_s values are within 1% of true L_s values.

Atomic Force Microscopy Nanoindentation of Intact Cardiomyocytes

Atomic force microscopy (AFM) nanoindentation was performed as previously described to estimate the elastic modulus (stiffness) of individual myocytes (15, 35, 36). In brief, cardiomyocytes were applied to laminin-coated dishes for 40 min and then washed 3 times in nominally Ca²⁺-free PSS at 22–25°C. Cell elasticity was monitored in contact mode using silicon nitride microlevers (model MLCT with spring constants 10–30 pN/nm; Veeco Mertrology, Santa Barbara, CA) and a Bioscope AFM System (Veeco Mertrology) mounted on an Olympus IX81 Microscope (Olympus, NY). Probes repeatedly approached and retracted the surface of the cardiomyocyte at a frequency of 0.5 Hz (tip speed 800 nm/s). Indentation depth of the AFM probe was ∼150–250 nm. Approach curves were used to calculate cell elasticity, where E is elastic modulus, according to the following equation: F = [Tan(α)] × [(2 × E × δ²)/π(1 − ν²)], where α represents probe shape (conical with an approximate half-angle of 21.25°); δ is indentation depth, calculated as the difference in the AFM piezo z-movement and the measured deflection of the probe; and ν is the Poisson ratio (0.5 for the cell). Indentation force (F) was calculated using Hooke's law (F = kΔx), where k is the AFM probe spring constant and Δx is the measured deflection of the probe.

Data Analysis and Statistics

Measurements of L_s were obtained from cellular regions (>20 μm) with well-defined intensity profiles. Analyses were performed by three separate observers using three independent approaches: 1) measurement of the distance between the first and last sarcomere within a defined series (distance/number of sarcomeres); 2) Fast Fourier transform analysis using ImageJ software (average number of input cycles was 26); and 3) wavelength of damped sinusoid fitted to the autocorrelation of the intensity profile. All three approaches yielded similar values for L_s. Analysis of L_s for young versus aged was performed blinded to experimental pressure or group, with values from 4–19 cardiomyocytes averaged to determine the mean L_s value at a defined pressure for each heart. Resulting values were used to determine the relationship between pressure and L_s for each heart (Fig. 4, A and B) and to compile summary data (Fig. 4, C and D). Two-parameter fits were used for nonlinear regression according to the equations Pressure = [exp^k*(Ls-Ls,0)] − 1 (Fig. 4, A and B) and L_s = [k⁻¹·ln(1 + pressure)] + L_s,0 (Fig. 4C), where adjustable parameters k and L_s,0 correspond to the stiffness constant and L_s at zero pressure, respectively (31). For AFM studies, a single elastic modulus per heart was obtained by averaging values from 1–5 individually sampled cardiomyocytes (n = 15 cardiomyocytes from N = 6 young mice; n = 15 cardiomyocytes from N = 6 aged mice). Data are presented as individual observations or as means ± SE. Statistical comparisons were made using paired t-tests or two-way ANOVA (Bonferroni post hoc) with SigmaPlot 10/SigmaStat 3.5 (Systat, San Jose, CA), with statistical significance set at P < 0.05.

Fig. 4.

L_s and LV pressure in hearts of young and aged mice. A and B: LV pressure vs. L_s curves in hearts from young (A, solid lines, n = 8) and aged (B, dashed lines, n = 7). At each pressure (0, 5, 10, 20, 30, and 40 mmHg), L_s values from 4–19 cardiomyocytes were averaged to obtain a single L_s value per heart, and data were fit according to Pressure = [exp^k·(Ls-Ls,0)] − 1. Note that in A and B, data are presented with pressure (independent variable) on the y-axis to correlate with typical presentation of pressure-volume relationships. C: raw data of L_s vs. LV pressure of young (●, n = 8) and aged (○, n = 7), with separate 2 parameter fits for young (black line) and aged (dashed line). Intermediate gray line corresponds to 2 parameter fit of pooled data across age groups. Data from respective age groups were fit best by separate (vs. pooled) analyses indicative of samples representing separate populations. D: summary data of L_s vs. LV pressure in hearts of young (●, n = 8) and aged (○, n = 7). *P = 0.03 young vs. aged (repeated-measures ANOVA).

RESULTS

The aims of this investigation were to 1) determine the relationship between left ventricular filling pressure and L_s in the intact murine heart and 2) test the hypothesis that sarcomere lengthening in response to ventricular filling is impaired with advanced age. First, to determine fluorescent markers of L_s in cardiac cells, ventricular cardiomyocytes were enzymatically isolated from mouse hearts and then loaded with spectrally distinct markers of the plasma membrane/t-tubule network (di-8-ANEPPS) and mitochondria (MitoTracker Deep Red). Figure 1 illustrates that di-8-ANEPPS and MitoTracker Deep Red, while giving reciprocal fluorescence profiles (Fig. 1B, b and c), resulted in L_s measurements that were well correlated with each other (r² = 0.94) and with L_s monitored using transmitted light (r² = 0.97 and 0.95, respectively; Fig. 1C). In many cases MitoTracker Deep Red fluorescence signal was low at both the cellular Z line and M line, yielding two cycles per sarcomere (Fig. 1B,c). Conversely, di-8-ANEPPS consistently produced a single, distinct peak for each sarcomere (at the Z line) and did so over extended distances. Thus di-8-ANEPPS was used for subsequent L_s measurements in whole heart experiments.

Fig. 1.

Fluorescent measurement of sarcomere length (L_s) in isolated cardiomyocytes. A: transmitted light (a), 4-{2-[6-(dioctylamino)-2-naphthalenyl]ethenyl}1-(3-sulfopropyl)-pyridinium (di-8-ANEPPS) fluorescence (b), and MitoTracker Deep Red fluorescence (c) images of an isolated mouse cardiomyocyte. B: high-magnification transmitted light (a), di-8-ANEPPS (b), and MitoTracker Deep Red (c) images (left) and corresponding intensity profiles (right) of a separate cardiomyocyte. C: 3-dimensional scatter plot (n = 43 measurements from 29 cardiomyocytes from 3 mice) comparing L_s obtained simultaneously with transmitted light (x-axis), di-8-ANEPPS (y-axis), and MitoTracker Deep Red (z-axis).

Two-photon fluorescence imaging of hearts loaded with di-8-ANEPPS primarily revealed indicator accumulation in the endothelial cell membrane of the coronary microvasculature and the sarcolemmal/t-tubular membrane of cardiomyocytes (Fig. 2, C–E). Measurements were typically obtained from sheets of subepicardial cardiomyocytes with well-defined sarcomere striations between capillary networks (Fig. 2E), avoiding regions with larger blood vessels (Fig. 2C) or areas with less-organized cardiomyocyte arrangements (Fig. 2D). Upon left ventricular pressurization, the heart freely slid against the coverslip (Fig. 2, A and B), and in some cases it was possible to relocate tissue regions following pressurization and thus obtain the pressure versus L_s relationship from the same series of sarcomeres within the same cardiomyocyte within the heart (Fig. 3).

Fig. 3.

L_s measurements from the same cardiomyocyte in the intact heart at 3 LV filling pressures. A–C: images of cardiomyocytes within an aged heart at 0 (A), 20 (B), and 40 (C) mmHg. Fluorescence profiles, obtained from regions marked by brackets at left of images, are presented below images. Fast-Fourier transform (FFT) analysis of fluorescence profiles resulted in L_s values of 2.02 (A), 2.25 (B), and 2.28 (C) μm. AU, arbitrary units.

Ventricular stiffness increases with advancing age, yet it remains unknown how this relates to the filling pressure versus L_s relationship. We therefore used our experimental approach to monitor this relationship under passive conditions (Ca²⁺-free cardioplegia) in hearts of young and aged mice. Shown in Fig. 4 are pressure versus L_s relationships from eight young (Fig. 4A) and seven aged (Fig. 4B) hearts. L_s increased with pressure and was more dynamic over the pressure range of 0–20 mmHg than the pressure range of 20–40 mmHg. In young hearts from different mice, there was significant overlap in the pressure-L_s relationship and each heart operated over a similar L_s range (Fig. 4A). In contrast, aged hearts from different mice exhibited greater variability in the shapes of their respective pressure-L_s relationship as well as in their L_s range (Fig. 4B). Summary data illustrate that across pressures, L_s was significantly shorter in aged hearts than in young hearts (Fig. 4, C and D).

Increased passive ventricular stiffness arising from alterations in the extracellular matrix (e.g., fibrosis) is well established in humans and animals of advanced age. In addition, changes in the passive stiffness of individual cardiomyocytes may also contribute to passive stiffness of the heart. AFM nanoindentation was used to assess intrinsic cortical stiffness of intact enzymatically isolated cardiomyocytes under Ca²⁺-free conditions (Fig. 5A). The apparent elastic modulus of cardiomyocytes of aged hearts was approximately twofold greater than that of cardiomyocytes of young hearts (Fig. 5B), indicative of increased cell stiffness with advanced age.

Fig. 5.

Passive stiffness of cardiomyocytes from hearts of young and aged mice. A: example traces of dynamic measurement of the elastic modulus of intact LV cardiomyocytes of young (●) and aged (○) with atomic force microscopy nanoindentation. Note difference in dynamic stiffness properties between young and aged, with cardiomyocytes of aged exhibiting high-amplitude oscillations in elastic modulus. B: elastic modulus of young (black bar, n = 15 cardiomyocytes from 6 mice) and aged (white bar, n = 15 cardiomyocytes from 6 mice).*P = 0.006 aged vs. young (unpaired t-test).

DISCUSSION

Cardiomyocyte L_s is the critical determinant of the Frank-Starling mechanism of the heart (2, 8, 10). The operating range of L_s in cardiomyocytes within the left ventricular wall has been determined using hearts arrested and chemically fixed at experimentally controlled pressures, with spacing between electron dense Z lines measured using electron microscopy (30) or transmitted light microscopy (11). These approaches require careful tissue preparation to avoid fixation artifacts associated with shrinkage resulting in underestimation of L_s [e.g., see Huang et al. (16)]. Furthermore, fixed preparations limit L_s evaluation to a single filling pressure in each heart, effectively preventing evaluation of the pressure versus L_s relationship in a single heart. Fluorescent labeling of the t-tubule network (and by anatomical constraints the cardiac Z lines) has previously enabled L_s measurements in living (unfixed) tissue in perfused rat hearts (6). Here we use fluorescent labeling in freshly excised mouse hearts during ventricular pressurization under Ca²⁺-free conditions to define the pressure versus L_s relationship and quantify passive properties of the intact heart at the level of the sarcomere.

Comparison with Previous L_s Measurements in the Heart

The sarcomere is the highly organized, functional unit of contractile protein organization within striated muscle. Despite such ultrastructural organization, L_s varies considerably within the heart because of complex anatomical arrangements (e.g., see Fig. 2D) and endocardial to epicardial regional differences (11, 28, 30). The mean L_s values determined for subepicardial cardiomyocytes in our preparation at 0 mmHg (young, 2.01 μm; and aged, 2.02 μm) fit well within the range (1.93 to 2.09 μm) determined in unpressurized intact rat hearts (5, 6) or fixed hearts maintained at 0 mmHg filling pressure (11, 28, 30). Upon left ventricular filling, the pressure versus L_s relationship of hearts from young mice studied here (0–40 mmHg L_s range, 2.01 to 2.36 μm) was also similar to that reported in classic measurements of dog hearts fixed ex vivo (0–30 mmHg L_s range, 1.93 to 2.33 μm) (30) and from reconstructions of L_s dynamics in vivo at physiological filling pressures (28). Contrasting with these data are measurements of closed-chest fixed rat hearts where the pressure versus L_s relationship was shifted toward a shorter L_s range (0–35 mmHg L_s range, 1.94 to 2.26 μm). However, the latter studies acknowledge a ∼4% L_s shrinkage due to formaldehyde (11). Mathematical correction of their data increases this range to 2.02 to 2.35 μm, which is in agreement with the present findings. The passive pressure versus L_s relationship steepens at L_s values > 2.3 μm (Fig. 4A), which is consistent with data in isolated cardiomyocyte preparations where passive force increases substantially between 2.3 and 2.4 μm (13). Such myocardial properties help prevent sarcomere overextension and allow cardiomyocytes to operate along the ascending limb or near the plateau of the L_s-tension relationship (9).

Ventricular Stiffness with Advancing Age

Cardiac stiffness increases with advancing age (3) and cardiac disease as a result of alterations in both active (i.e., Ca²⁺ dependent) and passive mechanical properties of the myocardium (20). Our experiments used Ca²⁺-free conditions to prevent cardiomyocyte excitation-contraction coupling (i.e., active stiffness) and to thereby monitor passive physical properties of the ventricular myocardium. In hearts of aged mice, we observed a shift in the relationship between L_s and pressure, reflecting impaired ability of the myocardium to stretch in response to a defined increase in pressure. Increased passive stiffness of the aged heart has been attributed to multiple factors including hypertrophy and wall thickening, cellular disorganization, changes in relative expression of extracellular matrix components (collagen, fibronectin, laminin) and fibrosis, or modification of the extracellular matrix via cross-linking or advanced glycation end-products (25). Passive stiffness of individual cardiomyocytes has also been shown to increase with advancing age and may contribute to altered ventricular compliance. Our data in mice are consistent with data in rats where nanoindentation with AFM revealed a significant increase in cortical stiffness in aged compared with young (23). At the level of the enzymatically isolated cardiomyocyte (i.e., a cell largely devoid of extracellular matrix components), increases in passive cardiomyocyte stiffness often occur concurrent with increases in the stiffness properties of the large sarcomeric protein titin (22, 24), and age-associated augmentation of titin-based cardiomyocyte stiffness has been observed in dog and human subjects (4, 12). Passive cardiomyocyte stiffness may also arise from subsarcolemma-associated proteins (e.g., ankyrin, cadherin, catenin, dystroglycans and dystrophin, integrins, spectrin, talin, vinculin) or cytoskeletal proteins (e.g., actin, desmin, tubulin), and changes in expression or macromolecular architecture of many of these proteins has been reported to occur with cardiac disease (14, 33). Such changes, in addition to established alterations in extracellular matrix components, may underlie the decrease in chamber compliance observed with advancing age.

Study Limitations and Future Directions

Crystalloid-perfused heart preparations associate with tissue edema, which may artificially extend L_s (6). The experimental perfusion solution used for this study included 6% dextran to increase colloid osmotic pressure of the perfusate and limit tissue edema. Nevertheless, initial L_s measurements occurred >40 min post-cardiac excision and time-dependent sarcomere lengthening may have occurred during that interval. However, resulting measurements of L_s were consistent with those in previous investigations using fixed tissue, suggesting that this approach resulted in accurate measurements of L_s. This study used nonphysiological temperatures under passive (Ca²⁺ free) conditions to reduce tissue movement and facilitate acquisition of fluorescence images. Diastolic L_s in vivo is likely to be shorter than that observed in the present investigation because of Ca²⁺-dependent diastolic contracture (i.e., active diastolic stiffness), and this effect may be particularly pronounced in the murine heart where high heart rates lead to diastolic elevation of intracellular Ca²⁺ with incomplete myofilament relaxation. Indeed, a recent report using t-tubule labeling in vivo revealed a diastolic L_s of ∼1.9 μm (1), which is less than the lowest value observed in our investigation under passive conditions (2.01 μm at 0 mmHg). Diastolic Ca²⁺ removal is typically impaired with advancing age (20, 21), leading to excessive active diastolic contracture. Therefore, sarcomere lengthening may be further impaired in aged hearts in vivo because of the combined influence of both active and passive myocardial properties. Data acquisition and analysis focused primarily on the subepicardium as this region of the heart exhibited well-organized cardiomyocyte arrangements that were parallel to the focal plane. Although not applied in the present investigation, two-photon microscopy can be used to image into greater depths (>100 μm) of the heart and examine additional aspects of myocardial architecture including cardiomyocyte dimensions, angle, and organization with respect to vascular networks. In subepicardial cardiomyocytes of young and aged mice, the structure of the t-tubule network was largely maintained permitting assessment of L_s. However, with cardiac disease significant loss or disorganization of t-tubules may occur (29, 34), which may limit the accuracy of L_s measurements using this approach. In aged hearts we observed variability in the pressure versus L_s relationship, which suggests variability in cardiac remodeling processes. Future studies are needed for correlation of extracellular matrix properties (fibrosis) or the stiffness of isolated cardiomyocytes with sarcomere lengthening on an animal-to-animal basis to afford additional insight into the variability in cardiac remodeling and ventricular stiffness associated with advancing age.

Conclusions

The novel approach developed for these experiments provides evaluation of the relationship between ventricular filling pressure and L_s in the intact mouse heart. Our studies reveal an aging-associated shift in the pressure versus L_s relationship, consistent with an increase in passive ventricular stiffness. The aged heart operates at elevated diastolic pressures (27), which under resting (i.e., nonstressed) conditions may extend L_s to maintain stroke volume according to the Frank-Starling relationship. However, this form of compensation will associate with a depressed capability to further increase L_s to augment stroke volume and may limit cardiac reserve.

GRANTS

This work was supported by National Institutes of Health Grants R01HL090749 (to K. S. Campbell), UL1TR000117 (to K. S. Campbell), P01HL095486 (to G. A. Meininger), R01HL057852 (to K. S. McDonald), R01HL086483 and R37HL041026 (to S. S. Segal), and K01AG041208 (to T. L. Domeier).

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the author(s).

AUTHOR CONTRIBUTIONS

M.E.N., Y.Z., and T.L.D. performed experiments; M.E.N., J.T.W., Y.Z., A.K.G., L.M.H., K.S.C., G.A.M., K.S.M., S.S.S., and T.L.D. analyzed data; M.E.N., J.T.W., Y.Z., L.M.H., K.S.C., G.A.M., K.S.M., S.S.S., and T.L.D. interpreted results of experiments; M.E.N., K.S.C., and T.L.D. prepared figures; M.E.N. and T.L.D. drafted manuscript; M.E.N., J.T.W., Y.Z., A.K.G., L.M.H., K.S.C., G.A.M., K.S.M., S.S.S., and T.L.D. edited and revised manuscript; M.E.N., J.T.W., Y.Z., A.K.G., L.M.H., K.S.C., G.A.M., K.S.M., S.S.S., and T.L.D. approved final version of manuscript; T.L.D. conception and design of research.