Mechanics of dystrophin deficient skeletal muscles in very young mice and effects of age

Abstract

The MDX mouse is an animal model of Duchenne muscular dystrophy, a human disease marked by an absence of the cytoskeletal protein, dystrophin. We hypothesized that 1) dystrophin serves a complex mechanical role in skeletal muscles by contributing to passive compliance, viscoelastic properties, and contractile force production and 2) age is a modulator of passive mechanics of skeletal muscles of the MDX mouse. Using an in vitro biaxial mechanical testing apparatus, we measured passive length-tension relationships in the muscle fiber direction as well as transverse to the fibers, viscoelastic stress-relaxation curves, and isometric contractile properties. To avoid confounding secondary effects of muscle necrosis, inflammation, and fibrosis, we used very young 3-wk-old mice whose muscles reflected the prefibrotic and prenecrotic state. Compared with controls, 1) muscle extensibility and compliance were greater in both along fiber direction and transverse to fiber direction in MDX mice and 2) the relaxed elastic modulus was greater in dystrophin-deficient diaphragms. Furthermore, isometric contractile muscle stress was reduced in the presence and absence of transverse fiber passive stress. We also examined the effect of age on the diaphragm length-tension relationships and found that diaphragm muscles from 9-mo-old MDX mice were significantly less compliant and less extensible than those of muscles from very young MDX mice. Our data suggest that the age of the MDX mouse is a determinant of the passive mechanics of the diaphragm; in the prefibrotic/prenecrotic stage, muscle extensibility and compliance, as well as viscoelasticity, and muscle contractility are altered by loss of dystrophin.

INTRODUCTION

Dystrophin is an intracellular element of the transmembrane protein network, and the lack of dystrophin in MDX (X chromosome-linked Duchenne muscular dystrophy) mice causes muscular dystrophy (1). The 427 kDa cytoskeletal protein, dystrophin, is localized in the subsarcolemmal region of skeletal and cardiac muscle (2). Dystrophin forms part of the link from the cytoskeleton, through the dystroglycan complex, to the extracellular matrix (1). Dystrophin was shown to maintain mechanical stability of muscle fiber membrane during muscle contraction and relaxation (3, 4). It is well recognized that the dystrophin protein functions as a large molecular spring that connects the cytoskeleton to the transmembrane dystrophin glycoprotein complex (3, 4). The MDX mouse was the first recognized and is the most widely used animal model in the study of Duchenne muscular dystrophy (DMD), a progressive X-linked human disease (5–7). DMD is caused by mutations in dystrophin coding or the DMD gene, resulting in the absence of the functional protein product, dystrophin (1). It is important however to recognize that the MDX mouse model is not useful for clinical trials because of its very mild phenotype (8). Nevertheless, the disease progresses through various stages in the MDX mouse. The disease divergently affects skeletal muscles where there is effective functional repair of limb muscles with mild persistent contractile weakness (9–12). This is compared with the observed severe disease and weakness in the MDX mouse diaphragm (12, 13). During the postnatal period, skeletal muscles appear histologically normal, but at ∼3 wk of age, a severe phase of muscle necrosis and muscle degeneration occurs (12). Necrosis is an acute injury of either intrinsic or extrinsic origin that causes irreversible damage and death to the muscle cells. After this degenerative period, the unaffected muscle cells regenerate around the damaged cells. This regeneration of muscle cells occurs in hindlimb muscles such as biceps femoris, however in the diaphragm the level of regeneration of muscle fibers is insufficient to keep pace with the damage/necrosis (12). As a result, the diaphragm shows more progressive cycles of fiber degeneration and muscle loss throughout the lifespan of the MDX mouse. The proportion of viable fibers in the diaphragm decreases to nearly 40% in 2-yr-old MDX mice (13, 14).

Assessing muscle mechanics of skeletal muscle in the prenecrotic state under mechanical loading conditions that mimic the in vivo mechanical environment could remove the confounding effects of necrosis and resultant fibrosis secondary to the loss of dystrophin. The diaphragm and biceps femoris muscles differ in the way they are mechanically loaded in vivo. The diaphragm is loaded with transdiaphragmatic pressure, and therefore, it is subjected to mechanical stresses not only in the direction of its muscle fibers but also in the transverse direction to the fibers (15–18). The in vitro data on the length-tension relationships in the direction of the muscle fibers and transverse to the fibers are essential in assessing the mechanical function of the diaphragm (17). The biceps femoris is one of many skeletal muscles of the hindlimb that is mechanically loaded in vivo only in the direction of its muscle fibers.

A well-designed study by Law et al. (19) tested the hypothesis that dystrophin functions as a structural link between the muscle cytoskeleton and the cell membrane. The authors studied structural alterations in the myotendinous junction (MTJ) in dystrophin-deficient MDX mice during prenecrotic, necrotic, and regenerative phases of postnatal muscle development in MDX mice. The results from this study showed defects in normal, lateral, thin filament-membrane associations in MDX muscle, regardless of age. The authors concluded that these data support the hypothesis that dystrophin functions as a structural link between thin filaments and the cell membrane. Interestingly, Cox et al. (20) have shown that overexpression of dystrophin prevents the development of abnormal mechanical properties associated with dystrophic muscle from a transgenic MDX mouse.

It is well established that deficiency of dystrophin in MDX mice causes skeletal muscles to be more susceptible to mechanical injury (21–24). Skeletal muscles of very young (9–12 days) MDX mice were shown to be resistant to injury from acute mechanical loading. The authors found that the extent of membrane damage of skeletal muscle subjected to lengthening contraction protocols was similar across dystrophic and control animals. Their data support a lack of a role of dystrophin in modulating strength of the membrane. The authors concluded that these data are consistent with the hypothesis that dystrophic sarcolemmal membranes are resistant to stretch injury during early maturation (25). It is well recognized that disruption of dystroglycan causes detachment of the basal lamina membrane from the sarcolemma and renders muscle prone to contraction-induced injury (26). It is important to recognize that the MDX diaphragm has been subjected to in vivo physiologic pressure loading at 3 wk of age. Yet our data show that the MDX diaphragm muscle at that age is healthy. This is consistent with the potential possibility that despite the dystrophin deficiency and secondary disruption of the cytoskeleton there is an age-dependent shift in the role of the dystrophin-dystroglycan cytoskeleton in providing resistance to contraction-induced injury even under conditions of biaxial loading.

We know of one study that focused on measuring passive mechanics of old MDX diaphragms (10). Our earlier study in very young MDX mice focused on uncovering a role of dystrophin in modulating stretch-induced mechanotransduction in skeletal muscle fibers of the diaphragm (13). These studies were limited to longitudinal passive mechanical stretch of the diaphragm muscle. Here, we focused our current study on assessing the anisotropic mechanics of skeletal muscles including the diaphragm and biceps femoris muscle in very young MDX mice. In addition, we assessed the effect of age on the passive mechanics of the diaphragm muscle in the MDX mouse.

We tested the hypotheses that 1) dystrophin serves a complex mechanical role in skeletal muscles by contributing to the passive compliance, viscoelastic properties, and contractile force production and 2) age is a modulator of passive mechanics of skeletal muscles of the MDX mouse. To test these hypotheses, we examined the anisotropic length-tension relationships of the diaphragm muscle sheet and isotropic length-tension relationships of the biceps femoris muscle from very young MDX mice. Our rationale for using very young MDX mice was to avoid the secondary potential confounding effects of muscle necrosis, inflammation, and fibrosis, which have been reported to commence in these mice as early as 21 days of age (6, 11, 27). We also report on how aging affects extensibility and compliance of the MDX mouse diaphragm through measurements of its passive nonlinear and viscoelastic length-tension relationships.

METHODS

Animals and Tissue Preparation

Our study protocols required a total of 70 Dmd^mdx mutant mice (strain: C57/BL10ScSn-DMD^mdx/J, Stock No. 001801, The Jackson Laboratory) and control mice (wild-type; WT) (strain: C57/BL10 SNJ, Stock No. 000666, The Jackson Laboratory). The DMD^mdx mice have a nonsense mutation, which leads to a truncated dystrophin protein. For simplicity, we will refer to the DMD^mdx mice as MDX mice. These mice were maintained in accordance with the National Institutes of Health Guide for the Care and Use of Laboratory Animals, and procedures were approved in advance by the Institutional Animal Care and Use Committee of Baylor College of Medicine. Mice were deeply anesthetized with an intravenous injection of pentobarbital sodium, and either the diaphragm or the long head of the biceps femoris muscles were excised and placed immediately in an oxygenated muscle bath. The diaphragm was excised by making an incision in the thoracic cavity and removing the diaphragm muscle, including its insertions on the spine and ribs as well as its insertions on the central tendon. The biceps femoris muscles were excised by making incisions near the muscle insertions on the lower spine and knee. Excised muscles were immediately submerged in Krebs–Ringer solution [pH 7.4 at 25°C, containing (in mM): 137 NaCl, 5 KCl, 2 CaCl₂, 1 MgSO₄, 1 NaH₂PO₄, and 24 NaHCO₃] equilibrated with a mixture of 95% O₂–5% CO₂. For either the diaphragm or the biceps femoris, two pairs of markers, made of surgical silk thread (0.2 mm in diameter), were sutured on the membrane of the muscle to avoid muscle injury. The markers were placed in a centrally located position on either the surface of the diaphragm muscle sheet, away from either the insertion on the rib cage or the origin at the central tendon or the surface of the biceps femoris muscles. To avoid the effects of any potential mechanical stress concentration at the clamp edges, care was taken in placing the markers far from the clamps. The markers on the surface of the membrane of either muscle were placed in a nearly square pattern, at 1-mm intervals, onto the membrane of two neighboring muscle fascicles. This would ensure that the orientation of the muscle fiber direction as well as the orthogonal direction to the long axis of the muscle fibers are well defined.

Histology

Three 3-wk-old MDX mice (weight: 8.88 ± 0.28 g; age: 21 ± 0.0 days old), three 3-wk-old WT mice (weight: 9.73 ± 1.43 g, age: 21 ± 0.0 days old), three 9-mo-old MDX mice (weight 32.7 ± 2.1 g), and three 9-mo-old WT mice (weight: 32.3 ± 1.3 g) were used for this analysis. The diaphragm and biceps femoris muscles were excised and analyzed by hematoxylin and eosin (H&E) and trichrome staining. Diaphragm and biceps femoris muscles were harvested from three separate mice and fixed in 10% formalin overnight. Samples were processed using standard protocols for H&E and trichrome staining. Tissues were sectioned at 5 μm with three sections per level. The fibrosis and fraction of internal nuclei were blindly calculated using ImageJ analysis of three random cross sections per mouse muscle. The number of myofibers counted for the 3-wk-old WT and MDX diaphragms was 1,341 and 1,035, respectively. Similarly, the number of myofibers counted for the 3-wk-old WT and MDX biceps femoris was 551 and 535, respectively. The number of myofibers counted for the 9-mo-old WT and MDX diaphragms was 1,231 and 2,048, respectively. Similarly, the number of myofibers counted for the 9-mo-old WT and MDX biceps femoris was 1,068 and 858, respectively. The fraction of fibrosis was quantified from trichrome stained images of skeletal muscle cross sections in both the 3-wk- and 9-mo-old WT and MDX diaphragms. Fraction of fibrosis was plotted as percent fibrosis from randomly selected cross-sectional areas. Percent fibrosis was calculated by analyzing trichrome stained cross sections using ImageJ to quantify endomysial connective tissue.

In Vitro Mechanical Stretching Apparatus

An in vitro stretching system was used to measure passive mechanical properties of uniaxially and biaxially loaded moushemidiaphragms as well as uniaxially loaded biceps femoris muscles. The mechanical apparatus consists of two orthogonal axes that are driven by micrometers to stretch the tissue in the plane of the muscle sheet at a constant strain rate. Two small identical alligator clamps (width 0.2 cm, Newark Electronics) were used to hold the muscle at the insertions. For the diaphragm, the insertion of the midcostal region of the left hemidiaphgram muscle at the central tendon was attached to one clamp whereas the insertion of the muscle at the rib cage was attached to the opposite clamp. For the biceps femoris, each clamp was attached to the muscle’s MTJs. The biaxial mechanical loading was only used for the left hemidiaphragm muscle. Four clamps were used for biaxial loading; two opposing clamps were attached to the muscle in the direction transverse to the fibers and two clamps were attached to the rib-cage insertion and central tendon origin of the muscle, as described for uniaxial loading. The muscle was mounted so that during passive muscle lengthening and passive muscle shortening maneuvers, the marker positions on the surface of the membrane of the diaphragm could be viewed and recorded via video camera. The placed markers on the muscle surface were far away from the clamps. This is to avoid any possible effects of potential stress concentration on the strain field of the markers during either the passive stretch or passive shortening protocols.

Experimental Data Collection

Marker displacement on the surface of either the diaphragm or the biceps femoris was monitored using a closed-circuit television (CCTV) video black-and-white camera (Hitachi HV-720U, Edmund Scientific) and recorded on tape via videocassette recorder (Sony SLV-620HF). Two force transducers (FORT250, 250 g differential bridge type, World Precision Instruments) located on each axis were used to measure applied mechanical forces. For passive muscle lengthening and passive muscle shortening cycles in the physiological range, a more sensitive force transducer (LQB 630, 50 g differential bridge type, Cooper Instruments) replaced one of the FORT250 transducers. The force data were amplified (Validyne C019A system), collected at 10 Hz using a data acquisition board (Lab-PC-1200/AI, National Instruments), and recorded using LabVIEW software, v5.0. The video data were digitally captured with a video capture card (Captivator PCI, v9.0; VideoLogic) by using frame grabber software at a sampling rate of 1 Hz (VideoWork, v1.5 and VidEdit, v1.1). The markers on the captured video images were then digitized with ImageTool, v3.0 (http://en.bio-soft.net/draw/ImageTool.html). All markers were assumed to lie within the same plane; the precise marker position was determined on Cartesian coordinate axes. After capturing and digitizing marker displacements, a triangulation algorithm was used to calculate tensile and compressive strains, orthogonal strains, and shear strains (16). The algorithm was coded, forces were decimated, and the relationships between mechanical strain and applied muscle force were generated using MATLAB software, v. R2020b.

Surface Mechanical Strain Calculations

With modifications, mechanical strains were calculated according to methods described in our previous work (16). Briefly, the region enclosed by the four markers on the surface of either diaphragm membrane or the biceps femoris muscle was divided into adjacent triangles, with the markers forming the apexes of such triangles. The three points that define a triangle determine a plane. At the reference position, or unstressed state, the two-dimensional coordinates of these points in the plane were denoted as x_i and y_i (i = 1, 2, 3). In the local coordinate system of the markers, displacement of the fibers from the unstressed state to positions at a stressed state were denoted by u_i, for displacement along the muscle fibers, and v_i, for in-plane displacement transverse to the long axis of the muscle fibers. These variables were assumed to be a linear function of position in the plane of the triangle as described by the following equation:

$${\text{u}}_i={\text{a}}_1+{\text{a}}_2{\text{x}}_{\text{i}}+{\text{a}}_3{\text{y}}_{\text{i}},$$ (1)

with measured values for the displacement and position of any three adjacent markers substituted for u_i, x_i, and y_i, provides a set of three equations for three coefficients a₁, a₂, and a₃. The displacement of the markers relative to the reference state would also provide the information required to determine the coefficients in the following equation, where a₄, a₅, and a₆ are constants:

$${\text{v}}_{\text{i}}={\text{a}}_4+{\text{a}}_5{\text{x}}_{\text{i}}+{\text{a}}_6{\text{y}}_{\text{i}}.$$ (2)

Mechanical strains in the longitudinal direction of the muscle fibers (ε_x) and mechanical strains in the direction transverse to the fibers (ε_y), which occurred during passive lengthening and passive shortening of the muscles, were computed according to the following equations:

$${\text{ε}}_{\text{x}}=\text{δu}/\text{δx},$$ (3)

$${\text{ε}}_{\text{y}}=\text{δv}/\text{δy},$$ (4)

where δu denotes the marker’s displacement in the x (along fibers) whereas δv denotes the marker’s displacement in the y direction (transverse to fibers). The values of the coefficients were substituted in Eqs. 3 and 4 to obtain the values of δu/δx and δv/δy. The values of ε_x and ε_y reflect fractional changes in muscle length in the directions along the muscle fibers and transverse to the fibers, in the plane of the muscle, respectively. As the muscle is stretched either along or transverse to the fibers in the direction of mechanical loading, the dimension of the muscle orthogonal to loading will be reduced in length leading to an extension ratio of less than 1. This behavior is known as the Poisson effect. Calculations of strains were made relative to the unstressed length of the muscle. Unstressed length was defined as the length of the muscle—placed in the oxygenated tissue bath—in the absence of applied mechanical forces. Unstressed length or length of excised muscle is the shortest muscle length and is equivalent to the length of the muscle at a lung volume equivalent to total lung capacity. Optimal muscle length is nearly 125% of this unstressed length. Therefore, a stretch of a 50% would place the diaphragm at a length equivalent to 120% of optimal muscle length. We computed the muscle length as a fraction of the unstressed length by adding 1 to the measured strains either in the direction of, or transverse to the fibers in the plane of the muscle sheet (λ), as described in Eqs. 5 and 6.

$${\mathrm{\lambda }}_{\text{t}}=1+{\text{ε}}_{\text{tensile}},$$ (5)

$${\mathrm{\lambda }}_{\text{c}}=1+{\text{ε}}_{\text{compressive}},$$ (6)

where λ_t is the muscle length as a fraction of the unstressed length in response to the tensile passive strain (ε_tensile) in the direction of loading and is greater than 1, whereas λ_c is the muscle length as a fraction of the unstressed length in response to the compressive passive strain (ε_compressive) in the direction orthogonal to loading, and is less than 1.

$$\mathrm{Compliance}\mathrm{ratio}=\frac{{\mathrm{\lambda }}_{\text{t}/\text{MDX}}}{{\mathrm{\lambda }}_{\text{t}/\text{Control}}}.$$ (7)

We then computed the compliance ratio by dividing the λ_t for the MDX mice (by λ_t for the control mice) as shown in Eq. 7. The compliance ratio was not computed for λ_c. Note that λ_c was computed for the purposes of plotting the stress/strain relationships during loading.

Passive Muscle Stretching

We measured optimal muscle length, L_o, by determining the length at which the muscle produced maximal contractile twitch force in response to electrical stimulation during isometric muscle contraction. The range of strain rate was between 0.1% and 1.0% change in length/s. Stretch ratios are generally insensitive with respect to strain rate, as long as the change in strain rate does not exceed an order of 10³ (28). Muscle passive stretching is sensitive to the history of deformation; therefore, to establish a constant history of mechanical loading, each muscle sample was mechanically preconditioned with five cycles of passive lengthening along the muscle fibers from the unstressed length, ∼60% L_o, to L_o, and then passive shortening back to a muscle length that was closest to the unstressed length. Length-tension curves were obtained by passively lengthening the muscle in the direction of the muscle fibers, from the unstressed length to a muscle length at a stress of ∼5 N/cm². Then, the muscle was passively shortened to the unstressed length. Passive muscle length-tension curves were also measured in the direction transverse to the muscle fibers. According our data, the optimal muscle length of the WT mouse diaphragm muscle is at 1.25 of the unstressed length, therefore we determined that the stress at which this optimal length occurs is 0.8 N/cm². The fraction of the unstressed length for the MDX mice was then determined at the same stress of 0.8 N/cm². A compliance ratio was computed as the fraction of the unstressed length of the MDX mouse diaphragm divided by the fraction of the unstressed length of the WT mouse diaphragm. A compliance ratio of greater than one implies that the MDX muscle is more compliant compared to its WT counterpart. When the mechanical properties of a muscle vary relative to the direction of applied stretch, the muscle is considered anisotropic. Alternately, when the mechanical properties of a muscle sheet along the muscle fibers are essentially the same as that in the transverse to the muscle fiber direction, the mechanical properties of a muscle sheet are considered isotropic. The anisotropy ratio is computed as the ratio between the compliance in the transverse muscle fiber direction to that of the compliance in the direction along the muscle fibers as determined from the muscle’s length-tension relationship. A ratio of 1 would indicate that the muscle is isotropic.

Mice Used in the in Vitro Passive Length-Tension Relationships Protocol

The protocol used diaphragms from nine WT mice (weight: 15.8 ± 3.6 g; age: 31.9 ± 8.4 days) and 12 MDX mice (weight: 6.6 ± 1.2 g; age: 20.0 ± 1.4 days). To observe the age effect on diaphragm muscle compliance in MDX mice, three 9-mo-old MDX mice (weight: 32.7 ± 2.1 g) were used, along with four 9-mo-old WT mice (weight 32.0 ± 1.1 g). In addition, the protocol used the biceps femoris from 12 WT 3-wk-old mice (weight: 10.8 ± 3.2 g; age: 24.8 ± 10.8 days) and 10 MDX mice (weight: 7.6 ± 0.5 g; age: 19.8 ± 1.7 days).

Measurements and Modeling of Viscoelastic Properties of the Diaphragm

The protocol used diaphragms from five WT mice (weight: 14.3 ± 4.3 g; age: 31.6 ± 14.9 days) and six MDX mice (weight: 6.2 ± 1.2 g; age: 20.5 ± 1.4 days). As before, we preconditioned the muscle by stretching the muscle tissue to L_o for five cycles. In the axial direction, the muscle was stretched to L_o and maintained at that length. In the transverse to the muscle fiber direction, the diaphragm muscle was stretched by a passive force equivalent to that applied in the axial direction and maintained at that length. Viscoelastic stress relaxation is a decline in muscle resistance to stretch. Passive muscle force was allowed to relax asymptotically, until it essentially reached a plateau. Passive muscle force was collected at 10 Hz by using a force transducer. Using nonlinear least-squares algorithms created in the MATLAB environment, we fitted the stress-relaxation data to the standard linear solid model of viscoelasticity (28). This simple model describes the muscle as the parallel combination of a dashpot with a coefficient of viscosity η₁ and a linear spring having a spring constant of µ₁, with a second linear spring parallel to the first spring and dashpot whose constant is µ₀. The relaxation function based on this model is shown in Eq. 8:

$$\text{F}={\text{E}}_{\text{R}}[1-[1-({\text{τ}}_{\text{σ}}/{\text{τ}}_{\text{ε}})]e^{^\hspace{0.17em}-(\text{t}/\text{τ})}\text{ε}],$$ (8)

where F is the relaxation force, t is time, E_R is the relaxed elastic modulus, τ_ε is the relaxation time for constant applied mechanical strain, and τ_σ is the relaxation time for constant applied mechanical stress. We calculated the viscoelastic coefficients, η₁, µ₀, and µ₁, based on the force-displacement relationships of the model. The ratio η₁/µ₁ is a relaxation time that characterizes the rate of relaxation of the dashpot. We reported relaxation time. A linear spring instantaneously produces a deformation proportional to the load. A dashpot produces a velocity proportional to the load at any instant. Furthermore, a dashpot represents a damping term that is proportional to velocity. The dashpot relaxation time is the length of time necessary for the dashpot, or shock-absorbing quality of muscle, to be essentially relaxed. That is the completely relaxed point at which the muscle is characterized as elastic. We also reported the relaxation elastic modulus, E_R. When the muscle is stretched to optimal length and kept at that constant length, the passive force or passive stress begins to decrease. The rate of mechanical relaxation is measured by the time dependent stress divided by a fixed strain, called the relaxation modulus, E_R. Another measure of viscoelastic behavior of the muscle is hysteresis and that is, at the same tension, the muscle would exhibit lower mechanical strain during passive lengthening than during passive shortening of the muscle. We reported hysteresis values for both MDX and its control mice for both the diaphragm and biceps femoris muscles that were stretched uniaxially along the muscle fibers or uniaxially stretched in the direction transverse to the fibers.

Measurement of Isometric Biaxial Contractile Properties of the Diaphragm

We used two different in vitro experimental protocols. In the first protocol, diaphragms from eight MDX mice (weight: 7.4 ± 0.7g; age: 18.7 ± 1.0 days) and six WT mice (weight: 7.6 ± 0.4 g; age: 19.8 ± 1.6 days) were used. After animals were anesthetized, we excised the left hemidiaphragm and immediately placed it in a continuously circulating, oxygenated 95% O₂–5% CO₂ Krebs–Ringer solution. The muscle was clamped on both the central tendon and ribs, and one clamp was connected to a Cooper instrument, LQB 630, force transducer. Once the muscle was positioned between two stainless steel mesh electrodes, it was stimulated by a Grass S88 stimulator. L_o was determined by twitch responses (0.1 ms stimulus duration, supramaximal voltage) and we tetanically stimulated the diaphragm muscle at 5, 10, 30, 50, 60, and 100 Hz, with 90 s recovery (supramaximal voltage, 0.5 ms pulses, tetanic train duration of 500 ms). The muscle was then clamped transverse to the long axis of the muscle fibers, with one clamp attached to a World Precision Instruments FORT250 force transducer and the other attached to a force carriage. Contractile force frequency data were obtained during biaxial loading by applying isometric tetanic stimulation sequences at 5, 10, 30, 50, 60, and 100 Hz in the presence of either 1 g or 2 g passive force applied in the direction transverse to the long axis of the muscle fibers. These measurements were collected at room temperature (25°C), which minimized temperature-dependent deterioration of the in vitro muscle preparation. The tetanic force data for both axes were collected at a sample rate of 300 Hz, using a data acquisition board (Lab-PC-1200/AI, National Instruments) and LabVIEW software, v5.0. The force data were stored in an ASCII file for postanalysis. We show in Fig. 1, A–C, how the diaphragm muscle was subjected to a biaxial load before the muscle was tetanically stimulated. In the second in vitro experimental protocol for measuring isometric contractile muscle force during biaxial loading, we used eight WT mice (weight: 10.1 ± 0.5 g; age: 22.0 ± 3.0 days) and 10 MDX mice (weight: 7.1 ± 0.9 g; age: 20.6 ± 1.4 days). We tetanically stimulated the muscle at 100 Hz, with 120 s recovery (supramaximal voltage, 0.5 ms pulses, tetanic train duration of 500 ms), and repeated the stimulation in the presence of either 1 g or 2 g passive force (Fig. 1B), applied in the direction transverse to muscle fibers of the diaphragm. This series of tetanic stimulations was repeated twice. Twitch responses (0.1 ms stimulus duration, supramaximal voltage) were collected for all loading sequences before tetanic stimulation. All contractile data were acquired at 300 Hz.

Figure 1.

Schematic showing how the diaphragm muscle was subjected to biaxial loading during maximal tetanic stimulation. A: the muscle sheet adjusted to optimal length. B: transverse force of either 1 g or 2 g was applied in the direction transverse muscle fibers. C: muscle is under a biaxial load where forces are applied in the direction of the muscle fibers and in the presence of a transverse load of either 1 g or 2 g.

Muscle Thickness Measurements

We adapted tissue thickness measurement methods designed by our group and published elsewhere (29). Briefly, each muscle was trimmed from bone and connective tissue. Unstressed muscle thickness measurements were obtained from the excised muscles. We generated a digital image of muscle surface and the surface area of the muscle sheet was determined by using ImageTool, v3.0. Immediately at the end of every ex vivo mechanics experimental protocol, the muscle was gently blotted dry with a cotton-tipped swab and immediately weighed. Muscle thickness was computed as t = m/Ad, where t is muscle thickness (cm), m is muscle mass (g), A is the surface area of the sample (cm²), and d is the density of the muscle (1.059 g/cm³) (16). Computed values of the cross-sectional area were used to compute passive and contractile stresses. Stress in N/cm² was computed as the ratio of either measured contractile force or passive applied force in grams divided by the product of unstressed muscle length in cm and muscle thickness in cm.

Statistical Analysis

We have computed differences between groups by performing ANOVA with the use of the SAS software program (30). The statistical model that was used was a two-factor effects model for two groups of mice (mdx vs. WT) and two treatments (passive mechanics: uniaxial stretch in the direction of the muscle fibers versus uniaxial stretch in the direction transverse to fibers; isometric muscle contractile stress: uniaxial loading versus either form of the biaxial loading). Muscle samples were handled as a random effect; force, treatment, and force-by-treatment interaction were handled as fixed effects. Pairwise comparisons were performed to test a priori hypotheses using linear contrast. Data are expressed as means ± SE, where SE is the standard error unless otherwise indicated. We chose a P value of 0.05 as an acceptable level of significance throughout the analysis of all data. For example, comparisons between groups were considered significant at P < 0.05.

RESULTS

Lack of Muscle Pathology in Very Young 21-Day-Old MDX Mice

A representative set of the 3-wk-old MDX mouse and the 3-wk-old WT mouse are shown in Fig. 2A. For the most part, these mice appear healthy with no significant difference in their weight. Overall, the 3-wk-old WT mice weighed 9.73 ± 1.43 g whereas the 3-wk-old MDX mice weighed 8.88 ± 0.28 g. The histological images shown in Fig. 3, A and B are for the diaphragms and biceps femoris from 21-day-old MDX and WT mice. The diaphragms and biceps femoris of 21-day-old WT and MDX mice do not show significant increased endomysial thickening, fibrosis, or degeneration/regeneration. Quantitative histological analysis of the fraction of fibrosis and fraction of internal nuclei demonstrates that there are no significant differences between the diaphragm or biceps femoris muscles of 21-day-old MDX mice when compared with 21-day-old WT mice (Fig. 4, A–D). The fraction of internal nuclei was counted from three independent WT mice representing 1,450 myofibers counted from at least three random diaphragm cross sections. The fraction of internal nuclei was calculated from three independent MDX mice representing 1,200 myofibers counted from at least three random diaphragm cross sections. The fraction of central nuclei (means ± SE) of the WT and MDX diaphragms (DIA) was 0.002 ± 0.001 versus 0.003 ± 0.002, respectively (P = 0.44). The fraction of internal nuclei was counted from one WT mouse biceps femoris representing 500 myofibers counted from at least three random cross sections. The fraction of internal nuclei was calculated from an MDX mouse biceps femoris representing 600 myofibers counted from at least three random cross sections. The fraction of internal nuclei (means ± SE) of WT versus MDM biceps femoris myofibers was 0.01 ± 0.005 versus 0.008 ± 0.003, respectively (P = 0.76). The fraction of fibrosis was quantified from trichrome-stained images of skeletal muscle cross sections. The WT diaphragm (n = 3), MDX diaphragm (n = 3), WT biceps femoris (n = 1), and MDX biceps femoris (n = 1) are plotted as % fibrosis from randomly selected cross-sectional areas. The percent fibrosis (means ± SE) of WT versus MDX diaphragm was 1.77 ± 0.42 versus 1.80 ± 0.18, respectively (P = 0.08). The percent fibrosis (means ± SE) of WT versus MDX biceps femoris myofibers was 0.85 ± 0.26 versus 1.22 ± 0.30, respectively (P = 0.45).

Figure 2.

A: left, 3-wk-old wild-type (WT) mouse that is a representative of three WT mice that were used for histological study. For the most part the MDX mice appear healthy with no significant difference in their weight from the WT mice. Right, 3-wk-old MDX mouse. This is a representative mouse from a group of three mice used for the histological experiment B: left, 9-mo-old WT mouse that is a representative of three WT mice that were used for histological study. Right, 9-mo-old MDX mouse. The weight of the MDX mice were not significantly different from that of the corresponding WT mice at 9 mo old.

Figure 3.

A: hemotoxylin and eosin (H&E) and trichrome staining of a 3-wk- and 9-mo-old MDX and wild-type (WT) diaphragm muscles. Representative sections from three independent mice for each group show that MDX muscles at 3 wk old is not significantly different to WT without significant fibrosis, endomysial thickening, or myofiber regeneration/degeneration. Representative sections from three independent mice for each group show that there is significant fibrosis in the 9-mo-old MDX diaphragms compared with the same aged WT mice. B: H&E and trichrome staining of a 3-wk- and 9-mo-old MDX and WT biceps femoris muscles. Representative sections from three independent mice for each group show that MDX muscles at 3 wk old is not significantly different to WT without significant fibrosis, endomysial thickening, or myofiber regeneration/degeneration. Representative sections from three independent mice for each group show that there is significant fibrosis in the 9-mo-old MDX diaphragms compared with the same aged WT mice.

Figure 4.

A: percent central nuclei in 3-wk-old (3WK) and 9-mo-old (9 MO) WT and MD diaphragm muscles. At 3WK, WT (n = 3) vs. MDX (n = 3) was 0.53 ± 0.26 vs. 2.1 ± 0.67 (means ± SE), respectively. At 9 MO, WT (n = 2) vs. MDX (n = 5) was 1.57 ± 0.57 vs. 12.67 ± 0.57 (means ± SE), respectively. B: percent fibrosis in 3-wk-old (3WK) and 9-mo-old (9 MO) (WT) and MDX diaphragm muscles. At 3WK, WT (n = 3) vs. MDX (n = 3) was 4.38 ± 1.13 vs. 3.54 ± 0.66 (means ± SE), respectively. At 9 MO, WT (n = 2) vs. MDX (n = 5) was 4.88 ± 0.38 vs. 33.00 ± 2.38 (means ± SE), respectively. C: percent central nuclei in 3-wk-old (3WK) and 9-mo-old (9 MO) WT and MDX biceps femoris muscles. At 3WK, WT (n = 1) vs. MDX (n = 1) was 2.00 ± 0.44 vs. 4.17 ± 1.74 (means ± SE), respectively. At 9 MO, WT (n = 2) vs. MDX (n = 5) was 0.33 ± 0.33 vs. 22.44 ± 1.77 (means ± SE), respectively. At least 500 myofibers were counted for each group. NS = not significant, ***P < 0.001, and ****P < 0.0001. D: percent fibrosis in 3-wk-old (3WK) and 9-mo-old (9 MO) WT and MDX biceps femoris muscles. At 3WK, WT (n = 1) vs. MDX (n = 1) was 0.68 ± 0.16 vs. 1.79 ± 0.34 (means ± SE), respectively. At 9 MO, WT (n = 2) vs. MDX (n = 5) was 1.26 ± 0.44 vs. 6.13 ± 0.75 (means ± SE), respectively. NS = not significant, **P < 0.01, and ****P < 0.0001. WT, wild type.

Significant Muscle Pathology in 9-mo-Old MDX Mice

A representative set of the 9-mo-old MDX mouse and the 9-mo-old WT mouse are shown in Fig. 2B. Overall, the 9-mo-old WT mice weighed 32.3 ± 1.3 g whereas the 9-mo-old MDX mice weighed 32.7 ± 2.1 g. Both the WT and MDX mice appeared healthy despite their significant muscle pathology. The histological images shown in Fig. 3, A and B are for the diaphragms and biceps femoris from 9-mo-old MDX and WT mice. The diaphragms of the 9-mo-old MDX mice have significant endomysial thickening and fibrosis compared with the diaphragms 9-mo-old WT mice. Quantitative histological analysis of the fraction of fibrosis and fraction of internal nuclei demonstrates that there are no significant differences between the diaphragm or biceps femoris muscles of 21-day-old MDX mice when compared with 21-day-old WT mice (Fig. 4, A–D). The fraction of internal nuclei was counted from three independent WT mice representing 1,450 myofibers counted from at least three random diaphragm cross sections. The fraction of internal nuclei was calculated from three independent MDX mice representing 1,200 myofibers counted from at least three random diaphragm cross sections. The fraction of central nuclei (means ± SE) of the WT and MDX diaphragms (DIA) was 0.002 ± 0.001 versus 0.003 ± 0.002, respectively (P = 0.44). The fraction of internal nuclei was counted from one WT mouse biceps femoris representing 500 myofibers counted from at least three random cross sections. The fraction of internal nuclei was calculated from an MDX mouse biceps femoris representing 600 myofibers counted from at least three random cross sections. The fraction of internal nuclei (means ± SE) of WT versus MDM biceps femoris myofibers was 0.01 ± 0.005 versus 0.008 ± 0.003, respectively (P = 0.76). The fraction of fibrosis was quantified from trichrome stained images of skeletal muscle cross sections. The WT diaphragm (n = 3), MDX diaphragm (n = 3), WT biceps femoris (n = 1), and MDX biceps femoris (n = 1) are plotted as % fibrosis from randomly selected cross-sectional areas. The percent fibrosis (means ± SE) of WT versus MDX diaphragm was 1.77 ± 0.42 versus 1.80 ± 0.18, respectively (P = 0.08). The percent fibrosis (means ± SE) of WT versus MDX biceps femoris myofibers was 0.85 ± 0.26 versus 1.22 ± 0.30, respectively (P = 0.45).

Increased Longitudinal and Transverse Compliances and Extensibility in Diaphragm Muscles from Very Young MDX Mice

Muscle length-tension relationships in the direction of the muscle fibers from a representative 3-wk-old MDX mouse and a representative 3-wk-old WT mouse are shown in Fig. 5A. These muscles were passively lengthened and passively shortened in the direction of the muscle fibers. Our data show that stretching and shortening loops are clockwise, in which the shortening curve falls below the lengthening curve. The abscissa is a ratio of the length of the muscle fibers at a stressed state compared to the length of the muscle fibers at the unstressed state. The ordinate is applied muscle passive tension. It appears that for the same applied mechanical tension the magnitude of strain in the MDX diaphragms is greater than that of WT diaphragms in the direction of loading as well as in the orthogonal direction. The data demonstrate slow and continuous increase in muscle passive tension over the range of imposed stretch. At a tension of 0.8 N/cm², muscle lengths as a fraction of their unstressed length in the direction along the fibers are 151% and 125% in the MDX and WT mice, respectively, yielding a compliance ratio of 1.21. At the same level of tension along the fibers, the muscle length as a fraction of unstressed length in the orthogonal direction of loading is ∼122% and 111% in the MDX and WT mice respectively, yielding a compliance ratio of 1.10. There is a rightward shift of the length-tension curves of the MDX diaphragm muscles compared to the WT muscles. These data suggest that the diaphragm muscle of MDX mice is more compliant and more extensible than the diaphragm muscle of WT mice. Therefore, dystrophin deficiency may increase muscle compliance and extensibility of the diaphragm. A representative set of transverse muscle length-tension relationships of the MDX and control mice diaphragms are also shown in Fig. 5B. Muscle was passively lengthened and shortened in the transverse direction to the long axis of the muscle fibers. Muscle length transverse to fibers was computed as a fraction of the unstressed muscle length in the transverse-fiber direction. The mechanical strains in the MDX diaphragm are greater than that of controls in both the direction of loading and in the orthogonal direction. At tension of 0.8 N/cm², muscle length as a fraction of unstressed length (λ) in the direction transverse to the muscle fibers were 122% and 110% in the MDX and WT mice, respectively. This results in a compliance ratio of 1.11. At the same level of tension, the muscle length as a fraction of the unstressed length (λ) in the orthogonal direction of loading were 112% and 105% in the MDX and WT mice, respectively. This results in a compliance ratio of 1.07. These data suggest that deficiency of dystrophin may increase muscle compliance and extensibility in the direction transverse to the muscle fibers.

Figure 5.

Representative length-tension relationships along the diaphragm muscle fibers and transverse to fibers of a 3-wk-old MDX mouse and a 3-wk-old wild-type (WT) mouse. A: loading and unloading curves in the longitudinal direction of the muscle fibers are shown by the symbols ● and ○ for the MDX mice and their WT, respectively. The corresponding passive loading and unloading orthogonal to the loading in the longitudinal direction of the muscle fibers are shown by the symbols ■ and □ for the MDX mice and their WT, respectively. B: loading and unloading curves in the transverse direction of the muscle fibers are shown by the symbols ● and ○ for the MDX mice and their WT, respectively. The corresponding passive loading and unloading orthogonal to the loading in the transverse direction of the muscle fibers are shown by the symbols ■ and □ for the MDX mice and their WT, respectively.

Increased Compliance in Skeletal Muscles of the Biceps Femoris of Very Young MDX Mice

Data in Fig. 6A show representative length-tension curves for the biceps femoris muscles of the MDX and WT mice in response to passive lengthening and passive shortening in the longitudinal direction along the muscle fibers. In addition, we show the length-tension curves in response to lengthening and shortening in the orthogonal direction of loading. Note that compressive strains occur in the direction orthogonal to loading in response to the imposed longitudinal stretch. The data demonstrate that during passive lengthening along the muscle fibers there is a slow and continuous increase in tension over the range of imposed strains. In addition, the length-tension relationship in long the muscle fiber directions of the MDX mouse biceps femoris are shifted to the right compared to the corresponding curve of the WT muscle. In the direction along the fibers of the biceps femoris muscle, and at a tension of 0.5 N/cm², λ was greater in the MDX mice, compared with WT mice (MDX: λ = 1.21; control: λ = 1.14), yielding a compliance ratio of 1.07. There was minimal hysteresis for WT biceps femoris muscles when lengthened to tensions under 2 N/cm² along the fibers (χ ± SE: 0.097 ± 0.023); however, hysteresis was greater for MDX biceps femoris muscles when lengthened to tensions under 2 N/cm² along the fibers (χ ± SE: 0.271 ± 0.085). These data show that biceps femoris muscles lacking dystrophin are more extensible and more compliant than those of the WT muscles when loaded in the longitudinal direction.

Figure 6.

Representative length-tension relationships along the biceps femoris muscle fibers and transverse to fibers of a 3-wk-old MDX mouse and a 3-wk-old WT mouse. A: loading and unloading curves in the longitudinal direction of the muscle fibers are shown by the symbols ● and ○ for the MDX mice and their WT, respectively. Note that the hysteresis for the MDX mouse biceps femoris for loading in the longitudinal direction was 0.23 N/cm², whereas the hysteresis for the WT mouse biceps femoris was 0.005 N/cm². The corresponding passive loading and unloading orthogonal to the loading in the longitudinal direction of the muscle fibers are shown by the symbols ■ and □ for the MDX mice and their WT, respectively. B: loading and unloading curves in the transverse direction of the muscle fibers are shown by the symbols ● and ○ for the MDX mice and their WT, respectively. Note that the hysteresis for the MDX mouse biceps femoris for loading transverse to the muscle fibers was 0.10 N/cm², whereas the hysteresis for the WT mouse biceps femoris was 0.007 N/cm².The corresponding passive loading and unloading orthogonal to the loading in the transverse direction of the muscle fibers are shown by the symbols ■ and □ for the MDX mice and their WT, respectively. WT, wild type.

Data in Fig. 6B show representative length-tension curves for the biceps femoris muscles of the MDX and WT mice in response to passive lengthening and passive shortening in the direction transverse to the muscle fibers. In addition, we show the length-tension curves in response to lengthening and shortening in the orthogonal direction of loading. Note that compressive strains occur in the direction orthogonal to loading in response to the imposed transverse stretch . In the direction transverse to the fibers of the biceps femoris muscle at the same level of tension, λ was again greater in the MDX mice, compared with WT mice (MDX: λ = 1.24; WT: λ = 1.18), yielding a compliance ratio of 1.05. When the biceps femoris muscle was loaded transverse to fiber direction, there was minimal hysteresis for WT biceps femoris (χ ± SE: 0.062 ± 0.022). Similarly, when lengthened in the direction transvers to the fibers, the MDX biceps femoris muscle had minimal hysteresis (χ ± SE: 0.048 ± 0.035). The axial length-tension curves of the MDX and control biceps femoris muscles appear superimposed on their respective transverse length-tension curves. These data show that biceps femoris muscles lacking dystrophin are more extensible and more compliant than those of the WT muscles when loaded in the transverse direction. In contrast to the length-tension relationships in the direction of loading, where the MDX biceps femoris is more extensible than the WT, the length-tension relationships orthogonal to loading are essentially indistinguishable.

Length-Length Relationship for Loading Longitudinal Muscle Fiber Length and Loading Transverse to Fiber Length in MDX and WT Diaphragm and Biceps Femoris

Muscle length was normalized to its unstretched length either along or transverse to the fibers. The normalized muscle length in the direction along the fibers was compared to normalized muscle length in the direction transverse to the fibers for representative MDX and WT mouse diaphragms in Fig. 7A. Both MDX length-length relationships show greater extensibility in the direction of loading compared with their respective WT length-length relationship. The dashed and dotted-dashed lines represent an ideal linear incompressible isotropic material (Poisson’s ratio of 0.5). Poisson’s ratio is a measure of the Poisson effect, the phenomenon in which a material tends to compress in directions perpendicular to the direction of stretch. A perfectly incompressible linear elastic material when deformed at small strains would have a Poisson's ratio of exactly 0.5. Any deviation from these lines would indicate that the muscle is anisotropic. Both control and MDX length-length curves deviate from the lines of isotropy. Note that the length-length relationship in the direction transverse to the muscle fibers, both the MDX and WT relationships deviate from the line of isotropy, indicating an anisotropic behavior of the diaphragm muscle.

Figure 7.

A: a representative diaphragm muscle length (normalized as a fraction of the unstressed length) along the muscle fibers is plotted against muscle length transverse to the fibers, and muscle length in either directions is shown for a representative WT mouse of eight 3-wk-old mice and one MDX mouse of eight 3-wk-old MDX mice. When loading was applied along the fibers, the control is shown by ○ and MDX by ●; and transverse to the fibers, control is shown by □ and MDX by ■. The abscissa is the relative degree of muscle extensibility in the direction along the muscle fibers whereas the ordinate is extensibility in the direction transverse to the muscle fibers. When loads were applied along the fibers, tensile strains were plotted on the abscissa and orthogonal strains on the ordinate. When loads were applied transverse to the muscle fibers, tensile strains were plotted on the ordinate and orthogonal strains on the abscissa. B: a representative biceps femoris muscle length (normalized as a fraction of the unstressed length) along the muscle fibers is plotted against muscle length transverse to the fibers, and muscle length in either directions is shown for one control of eight 3-wk-old mice and one MDX of eight 3-wk-old MDX mice. When loading was applied along the fibers, the control is shown by ○ and MDX by ●; and transverse to the fibers, control is shown by □ and MDX by ■. The abscissa is the relative degree of muscle extensibility in the direction along the muscle fibers whereas the ordinate is extensibility in the direction transverse to the muscle fibers. When mechanical forces were applied along the fibers, tensile strains were plotted on the abscissa and orthogonal strains on the ordinate. When mechanical forces were applied transverse to the muscle fibers, tensile strains were plotted on the ordinate and orthogonal strains on the abscissa. Note that the dashed lines shown in A and B extending from 1.2 to 1.8 on either the x-axis or the y-axis are two lines that are used as references in terms of a linear incompressible isotropic two-dimensional material. WT, wild type.

Normalized muscle length in the direction along the fibers was compared to the normalized muscle length in the direction transverse to the fibers for representative MDX and control mouse biceps femoris in Fig. 7B. Both MDX curves show greater extensibility in the direction they were lengthened and shortened compared to their respective controls. Both the MDX and WT biceps femoris muscle are superimposed on the line of line of isotropy for both the loading along the fibers and transverse to the fibers (up to 20% stretch). The WT biceps femoris muscle slightly deviates from the isotropy lines. During uniaxial loading along the muscle fibers, the MDX length-length curve deviates away from the line of isotropy. During transverse loading, however, the MDX length-length curve is essentially superimposed on the line of isotropy.

Decreased Muscle Compliance and Decrease Extensibility of the Diaphragm, as a Result of Aging of the MDX and Aged-Matched Control Mice

Data in Fig. 8 show representative length-tension relationships along the muscle fibers for MDX diaphragms and age-matched controls, at 1.5 yr, 9 mo, and 3 wk.

Figure 8.

Representative length-force relationships along the direction of muscle fibers in diaphragms of MDX and age-matched WT mice at 1.5 yr, 9 mo, and 3 wk of age. The symbols ○ and ◊ indicate lengthening and shortening curves for 1.5-yr-old WT and MDX mice, respectively [replotted from the previous work of Stedman et al. (12)]. The symbols □ and X indicate lengthening and shortening curves for the diaphragm muscle of a 9-mo-old WT mouse and an MDX mouse, representative of n = 3 WT and n = 3 MDX mice respectively. The symbols ▵ and * indicate lengthening and shortening curves for the diaphragm muscle of a 3-wk-old WT mouse and an MDX mouse, representative of n = 8 WT and n = 8 MDX mice, respectively. Stress is computed as force per cross-sectional area (A). Data for A are as follows: A_{1.5-yr-old MDX and WT} = 0.005 cm² (12); A_{9-mo-old MDX} = 0.01 cm²; A_{9-mo-old WT} = 0.007 cm²; A_{3-wk-old MDX} = 0.008 cm²; A_{3-wk -old WT} = 0.008 cm². WT, wild type.

Lengthening and shortening curves for the 1.5-yr-old control and MDX mice were plotted from Stedman’s et al. previous work (12). All muscles were passively lengthened and shortened in the direction along the muscle fibers. Muscle length was computed as a percentage of optimal length and unstressed length was 93% of optimal length. For the 1.5-yr-old data from Stedman, we computed stress using the cross-sectional area reported for both MDX and age-matched controls (0.005 cm²). The cross-sectional area for the 9 mo and 3-wk-old control mice was 0.007 cm² and 0.008 cm², respectively. The cross-sectional area for the 9-mo- and 3-wk-old MDX mice was 0.01 cm² and 0.008 cm², respectively. Interestingly, the thickness of the 9-mo-old MDX diaphragm was significantly thicker than its age-matched control (P = 0.003). Compared with controls, the length-tension curve of the 3-wk-old MDX mouse diaphragm is shifted rightward. Whereas, the curves for 9-mo-old and 1.5-yr-old MDX mice are shifted leftward, compared with controls. The length-tension curve for the 9-mo-old control mouse lies in between those of the 3-wk-old and 1.5-yr-old control mice. At a tension of 0.5 N/cm² (1 g force/0.2 cm width of clamp), the λ of the MDX diaphragm was greater than the control at 3 wk of age, but less than the control at 9 mo of age (MDX at 3 wk: λ = 1.49, 9 mo λ = 1.06; control at 3 wk: λ = 1.36, 9 mo λ = 1.25). At tension of 0.5 N/cm², the tensile strain in the direction along the fibers for 3-wk-old mice is 49% and 36%, respectively, for MDX and control mice, yielding a compliance ratio of 1.10. The tensile strain at the same level of tension for 9-mo-old mice is 6% and 25%, respectively, yielding a compliance ratio of 0.8. These data demonstrate that the passive extensibility of the MDX mouse diaphragm decreases with age.

Altered Diaphragm Muscle Viscoelasticity in Dystrophin Deficient Muscles from Very Young MDX Mice

The dashpot relaxation times for dystrophin-deficient diaphragms were not statistically different than the controls under the three loading conditions: along the fibers, transverse to the fibers, or biaxial loading. Data in Fig. 9A show E_R, relaxed elastic modulus, values for MDX 3-wk-old mouse diaphragms, compared with age-matched MDX controls. E_R values during stress relaxation in response to stretch in the direction of the muscle fibers were about the same for MDX and control diaphragms. During transverse loading conditions, MDX muscles exhibited a higher value of E_R than the controls (P < 0.05). During biaxial loading conditions, E_R values were about the same for control muscles (data not shown). The data in Fig. 9A show that the MDX mouse diaphragms exhibit altered viscoelastic properties with increased relaxed stiffness, most pronounced in the transverse direction. These data suggest that relaxed muscle stiffness is greater in the transverse plane of dystrophin-deficient diaphragms, compared to controls. Fitted viscoelastic stress-relaxation curves with corresponding representative raw data are shown in Fig. 9B during the holding phase of stretch for an MDX and a WT mouse diaphragms. The two curves represent the loading condition transverse to the muscle fibers for MDX and WT mice. The E_R values were significantly greater for MDX mice diaphragms, compared with controls under transverse loading (χ_ER ± SE = 0.071 ± 0.02 for MDX and 0.64 ± 0.03 for controls; P < 0.05). These data suggest that dystrophin contributes to the viscoelastic behavior in the transverse direction of the diaphragm muscle.

Figure 9.

A: relaxed elastic modulus (E_R) values for diaphragms of five WT mice and six MDX mice at 3 wk of age. During longitudinal viscoelastic stress relaxation in the direction transverse to the muscle fibers, E_R in the MDX diaphragm was significantly higher values than the WT (P < 0.05) as indicated by (*). Error bars represent the standard error of the mean. B: representative viscoelastic stress-relaxation curves for the diaphragm muscle during static stretch to optimal length. The passive muscle forces were measured during the static stretch holding phase. These representative curves were fitted to raw data for the diaphragms of an MDX mouse (of 6 mice) and an age-matched WT mouse (of 5 mice) at 3 wk of age. The two curves represent the stress-relaxation data for the diaphragm muscle in the transverse-fiber direction. The symbols O and ● represent raw data for the WT mouse and the MDX mouse, respectively. Fitted data are represented by dashed lines. Stress relaxation is the decrease in force and can be expressed as a percentage of peak force. E_R values were higher in MDX diaphragms, compared with WT (P < 0.05). WT, wild type.

Loss of Isomteric Contractile Muscle Force under Uniaxial and Biaxial Loading in Very Young MDX Mice

The force frequency curves presented in Fig. 10 demonstrate, under uniaxial loading conditions, that muscle tetanic stress is significantly reduced in MDX mouse diaphragms, compared with controls. Data are derived from eight MDX mice (weight: 7.4 ± 0.7 g; age: 18.7 ± 1.0 days) and six WT mice (weight: 7.6 ± 0.4 g; age: 19.8 ± 1.6 days). The force frequency curves for control and MDX mice reached maximum levels of relative contractile force at a stimulation frequency of ∼50 Hz. Under uniaxial loading, muscle contractile force appears depressed in MDX diaphragms, compared with WT diaphragms—especially at high stimulation frequencies. These data demonstrate that for 3-wk-old MDX mice, the effect of passive transverse stress on contractile properties is negligible. Data in Fig. 11 show the means ± SD of maximal muscle isometric contractile stress in WT and MDX diaphragm muscles under uniaxial and biaxial loading conditions, at a maximal stimulation frequency of 100 Hz. Data were obtained from eight WT mice (weight: 10.1 ± 0.5 g; age: 22.0 ± 3.0 days) and ten MDX mice (weight: 7.1 ± 0.9 g; age: 20.6 ± 1.4 days). In MDX muscles, muscle contractile stress was depressed during both uniaxial and biaxial loading conditions (P < 0.0001), compared with controls. These data demonstrate that dystrophin-deficient diaphragm muscles have reduced contractile force-generating capacity.

Figure 10.

Force frequency curves for WT and MDX mice diaphragms under uniaxial and biaxial loading conditions. The open symbols are for wild-type mice (n = 6), and the closed symbols are for MDX mice (n = 8). Both ▲ and ^▵ represent data during uniaxial loading. The □ and ■ are data during biaxial loading with passive transverse force of 1 g, whereas the ○ and ● are data during biaxial loading with passive transverse force of 2 g. Error bars represent the standard error of the mean.

Figure 11.

Maximal tetanic stress (N/cm²) in WT and dystrophin-deficient diaphragms under uniaxial and biaxial loading conditions, at a stimulation frequency of 100 Hz. Error bars indicate the standard deviation of the mean. *Significant differences between WT and MDX diaphragm muscle maximal tetanic stress (P < 0.0001). WT, wild type.

DISCUSSION

Our study is the first to use biaxial mechanical loading of the MDX diaphragm for both passive and contractile muscle properties in very young mice. It demonstrates that, under these conditions, dystrophin is a mechanical load-bearing protein by contributing to the passive compliance, viscoelastic properties, and contractile force production. Our study also demonstrated that age is a modulator of passive mechanics of skeletal muscles of the MDX mouse.

We report contractile muscle function of very young MDX mouse diaphragm in the pre-necrotic state during uniaxial and biaxial loading conditions. In addition, we report data on viscoelastic properties of diaphragm muscle at a very young age. Our data show increased muscle extensibility and compliance in the dystrophin-deficient diaphragm and dystrophin- deficient biceps femoris as shown in Figs. 5 and 6. Our current findings are consistent with the previous report by Pasternak about membrane stiffness in cultured MDX myotubes produced by fusion of single MDX mouse myoblasts (31). Our findings on the passive mechanics of the three-3-wk-old MDX mice and their controls are also consistent with that of Grange et al. (25) demonstrating an increased compliance of dystrophin-deficient membranes at ages of 9–12 days old compared with those of age-matched wild-type mice. As the diaphragm in vivo is mechanically loaded biaxially by transdiaphragmatic pressure gradient, measuring the length-tension relationships along the muscle fibers direction and transverse to fibers is essential to understanding diaphragm mechanical function. It is important to recognize that length-tension relationship of a homogeneous elastic material is stiffer when subjected to a biaxial mechanical stretch in comparison to a uniaxial stretch. This is important when comparing length-tension relationship measured in isolated muscle preparations to in vivo mechanical loading of the diaphragm. An increase in muscle compliance of the dystrophin-deficient diaphragm would be less pronounced in a biaxial state of mechanical loading, compared with muscle compliance in a longitudinal loading for which the muscle is stretched only in the muscle fiber direction.

Both our in vivo (32) and our in vitro (15, 17, 33, 34) animal studies in the dog, rat, and mouse have shown that the mechanical properties of the diaphragm are anisotropic. The diaphragm muscle has a smaller muscle compliance in the direction transverse to the muscle fibers compared with that in the longitudinal direction of the fibers. Our ex vivo mechanics data of the MDX mouse diaphragm show anisotropic behavior and appear consistent with those of the WT diaphragm of dog, rat, and mouse (15, 17, 34). In contrast to the diaphragm, however, length-tension curves along the muscle fibers of the biceps femoris are nearly superimposed on the length-tension curves transverse to the fibers for MDX and WT mice. That is, regardless of genotype, length-tension relationships of skeletal muscles of the biceps femoris appear to be independent of the direction of stretch. This indicates that the biceps femoris muscle is nearly mechanically isotropic muscle.

We used biceps femoris because it is mechanically loaded uniaxially in vivo whereas the diaphragm is loaded biaxially. Furthermore, in the MDX mouse, the level of regeneration of the diaphragm muscle is insufficient to keep pace with muscle necrosis whereas muscle fibers in the biceps femoris muscle succeed to regenerate. Although the MDX diaphragm muscle fibers regenerate, this rate of regeneration does not lead to recovery as in the limb skeletal muscles and consequently, the diaphragm is a more severely affected muscle in the MDX mouse. In addition, in contrast to bicep femoris, the diaphragm muscle in vivo is constantly under cyclic mechanical stretch.

Interestingly, muscle fibers in the biceps femoris run along the length of the muscle. In contrast, muscle fibers in either the soleus or EDL muscles are oriented at an angle to the long axis of the muscle. Therefore, it would be difficult to apply a transverse fiber load that is orthogonal to the long axis of the muscle fibers in either the soleus or the EDL muscles, whereas it is feasible to apply a transverse load to the biceps femoris. In addition, the biceps femoris is a flat muscle and therefore, it is easy to apply mechanical forces in the plane of the muscle along or transverse to the long axis of its muscle fibers. Differences between the biceps femoris and the diaphragm may shed some light on the identification of force transmission pathways that are unique to the diaphragm. The concentration of some key cytoskeletal proteins is much smaller in the biceps femoris than in the diaphragm muscle. For example, the concentration of desmin in the diaphragm muscle is nearly 40% greater than that in the biceps femoris muscle (34). Interestingly, our data show that muscle stiffness in the direction of, and transverse to, the fiber in the biceps femoris is very similar, whereas the diaphragm muscle is stiffer in the transverse-fiber direction than along the fibers. In addition, the biceps femoris is mechanically loaded uniaxially in vivo, whereas the diaphragm is loaded biaxially in vivo.

Interestingly, Wolff et al. (35) have shown that the MDX and control EDL muscles of young mice at 14–35 days old exhibited similar passive mechanical properties. The investigators used EDL whereas our study used the diaphragm and biceps femoris. It is important to note however, that the protocols of passive stretch of Wolff et al. study was distinctly different from ours. The EDL muscles were subjected to passive stretch to a length of 1.05 L₀ at a rate of 1.5 L₀/s whereas our experimental protocol showed nearly a 50% of passive stretch which placed the diaphragm muscle at 120% of optimal length at strain rate was between 0.1% and 1.0% change in length/s.

In contrast to our earlier work (36–38), our data on muscle contractility of wild-type diaphragms from very young mice did not show a significant effect of transverse passive stress on axial contractile muscle force. It is important to note however, that the mice in the current study are much younger than those utilized in the earlier studies (36–38). Therefore, it is highly likely that the effect of passive transverse stress on muscle contractility is age dependent. Dystrophin can be envisioned as tensile cables that carry tension from actin filaments to the cell membrane; thus, in the absence of dystrophin, this model would predict a loss of contractile muscle force. Furthermore, muscle forces are transmitted in part by dystrophin to the extracellular matrix through its tight linkage with the glycoprotein complex at the sarcolemma, in the directions along and transverse to the myofibers. In the MDX adult mouse, there is an increased level of vinculin, talin, α-actinin, and integrin, as well as some elevation in the dystrophin-related protein, utrophin. However, Law et al. have shown that, in 2-wk-old MDX mice and age-matched controls, there is no systematic difference regarding concentrations of many of the aforementioned structural proteins (e.g., vinculin, talin, utrophin, or α-actinin) (19). This indicates that upregulation of these proteins in adult MDX mice is not a direct response to the absence of dystrophin, because dystrophin is absent in very young MDX mice. This also supports the notion that upregulation of these proteins is not constitutive in MDX; rather, it is a condition that develops later as a consequence of MDX pathology. Therefore, upregulation of such proteins may be a response to muscle injury that begins at ∼25 days of age, during the process of regeneration. It is also of note that there is no change in the expression of desmin, vinculin, costameres, and nebulin in 18-day-old MDX mice (39). An absence of dystrophin does not appear to disrupt the cytoskeleton in the prenecrotic state. Thus, using muscles from MDX mice at very young age would appear to resolve this problem by obviating both muscle injury and the upregulation of other key cytoskeletal or transmembrane proteins. Consequently, we designed our experiments using very young mice (3-wk-old) to detect functional differences between skeletal muscles from MDX and corresponding muscles from control WT mice that could be attributable to the absence of dystrophin. An absence of dystrophin leads to a dramatic reduction in dystrophin-associated proteins in the sarcolemma of patients with Duchenne muscular dystrophy and in MDX mice (40).

The lack of dystrophin does not alter muscle anisotropy in the diaphragm. The length-tension relationships of the diaphragm of the WT mouse suggest that the muscle has two different compliances as muscle extensibility along the fibers and transverse to the fibers are not the same. This is confirmed by the length-length relationships of the diaphragm in Fig. 7A. When the diaphragm of a WT mouse was lengthened and shortened along the muscle fibers beyond physiological range compared to transverse to the fibers, both curves do not have slopes that equal that of the lines of isotropy. In the absence of dystrophin, the slopes of the data do not shift significantly; indicating muscle compliance is not altered in ether directions. Furthermore, the increase in muscle compliance in the diaphragm seen with the length-tension relationships can be noted with the length-length relationships. The biceps femoris muscle, unlike the diaphragm, is regarded as having compliance that is independent of directionality. According to the length-length relationships in Fig. 7B, the biceps femoris muscle of the WT mouse closely behaves nearly like an isotropic muscle both in the longitudinal direction of the muscle fibers and transverse to the fibers’ direction, shown by the small deviations in the slopes of the data from that of the lines of isotropy. In the absence of dystrophin, the increase in muscle compliance in both directions is shown in the length-length graph as these curves extend further away from 1.0 compared to that of the biceps formoris of an age-matched WT mouse. In addition, muscle compliance within the physiological range is not altered by the absence of dystrophin in the longitudinal to the muscle fiber direction or transverse to the fiber direction. However, beyond physiological range, loss of dystrophin appears to alter muscle compliance of the biceps femoris, as the slope of the data in the direction along the fibers deviates from the line of isotropy.

We have shown that loss of dystrophin causes aberrant mechanotransduction in skeletal muscle fibers of the diaphragm in very young MDX mice (41). Mechanical stretch of dystrophin-deficient muscles could potentially lead to upregulation of proinflammatory transcription factors (e.g., NFκB) (42). Our published data showed that persistent stimulation of the MDX diaphragm by passive mechanical forces could implicate pro-inflammatory cytokines (e.g., TNF-α and IL-1β); such activity could result in the over stimulation of NFκB, and could contribute to the development of muscular dystrophy

Altered mechanical properties of the aging diaphragm were shown by others in rats and correlated with increased cross linking and higher content of intramuscular collagen (42, 43).

Other investigators have shown that ECM of skeletal muscles from old mice has a higher modulus than the ECM of adult muscle, potentially due to accumulation of densely packed extensively crosslinked collagen (44). Previous work by our group has shown that aging is an important modulator of passive mechanics of the healthy diaphragm of WT mice (45). In particular, we have shown that aging of the diaphragm from WT mice ranging in age between 2 mo and 2 yr old caused loss of muscle compliance and reduction in stress relaxation in the direction of the muscle fibers (45).

It is not surprising that alterations in muscle compliance of MDX mice are critically dependent on age. The onset of muscle degeneration and fibrosis in the MDX diaphragm and the biceps femoris muscles is rare before 25 days of age, and this is consistent with our data in Figs. 3 and 4. However, muscle necrosis and fibrosis continually affects the MDX diaphragm after onset, as is demonstrated in our histological data in Fig. 3 and morphometry data in Fig. 4 for the 9-mo-old MDX mice. Our data compliment the data presented by Stedman et al. (12) in their histological findings for the 1.5-yr-old MDX mice. It is well recognized that centrally located nuclei are indicative of regenerating muscle fibers (46). Louboutin et al. (47) found that in MDX mice at 9 mo of age, 35% in diaphragm muscle had central nuclei, whereas 70%–80% of fibers in hindlimb muscles had central nuclei. Our data in Fig. 4 indicate that there is a significantly higher percent central nuclei in the 9-mo-old MDX mouse diaphragm compared with the age-matched WT controls, which is consistent with the histological findings by Henry et al. (48), where the percent internal nuclei was associated with a greater number of regenerating fibers in the diaphragms of 10-mo-old MDX mice compared with age-matched WT controls. Previous studies have demonstrated that at 9 mo old, MDX diaphragm muscle was characterized by perimysial and endomysial fibrosis; this latter feature was absent from MDX hindlimb muscles (47). In contrast, our data in Fig. 4 demonstrated that there is significantly greater fibrosis in both the biceps femoris and diaphragm muscles of the 9-mo-old MDX mice.

Our data in Fig. 8 demonstrate the progressive decrease in muscle extensibility and muscle compliance of the MDX mouse diaphragm at the ages of 9 mo compared with that at 3 wk MDX old mouse diaphragm. As shown in Fig. 8, comparing the data of Stedman et al. at 1.5 yr old to our data at 3-wk old and 9 mo old provides an additional experimental evidence of reduced extensibility and compliance of the diaphragm with age. It is known that upregulation of collagen occurs during the degeneration of MDX muscles (12). Stedman et al. (12) reported that, at 1.5 yr of age, there is a sevenfold increase in collagen content in the MDX diaphragm, compared to a tenfold increase in the MDX quadriceps muscle. Compared with wild-type mice, MDX mice showed reduced life spans and were more susceptible to spontaneous rhabdomyosarcoma (49). Previous work by Marshall et al. on the soleus muscle demonstrated that there is a 4.5-fold increase in both endomysial and perimysial connective tissue content in 9-mo-old MDX mice, compared with 3-wk-old MDX mice (50). According to these data, we can expect a significant upregulation of collagen in the 9-mo-old MDX mouse diaphragm, which could potentially contribute to loss of muscle extensibility and reduced muscle compliance compared to muscles from very young mice. Although the inextensibility of collagen fibrils and necrotic muscle cells explains the significant loss of muscle compliance, one must account for the increase in collagen content occurring in WT mice as a result of aging. Data presented by Marshall et al. (50) on the soleus muscle show increases of ∼200% and 70%, respectively, in the endomysial and perimysial connective tissue content of the 9-mo-old WT mouse, compared with that in the 3-wk-old WT mouse. Although, with age, the proportion of collagen increases in the diaphragm, the marked loss of muscle compliance in older MDX diaphragms is primarily due to muscle pathology. Thus, the aged dystrophin-deficient diaphragm undergoing progressive degeneration with lesser compensating regeneration may be less extensible and less compliant, compared to the aged WT diaphragm muscle. A marked accumulation of collagen in the muscles of DMD patients has been shown, and this could potentially contribute to a loss of skeletal muscle compliance in these patients (51). The excess accumulation of collagen in dystrophic skeletal muscle, compared with healthy muscle, is related to increased gene expression—and is probably triggered by a response to injury that induces muscle fiber damage due to a lack of dystrophin (51). Interestingly, other investigators reported that collagen content does not alter the passive mechanical properties of fibrotic skeletal muscle in the MDX mice (52). The authors concluded that collagen content is not predictive of muscle stiffness. It is important to note however, that the passive mechanics in that study were assessed following the onset of muscle necrosis at ages of 20-wk-old and 1-yr-old MDX mice. In addition, the passive mechanical protocol was conducted following an active mechanical protocol. It is difficult to compare these data with ours because of different ages and different experimental protocols.

Our data show that dystrophin plays a role in modulating viscoelastic behavior of the diaphragm. Interestingly, the MDX diaphragm muscle shows a significant viscoelastic stress relaxation (decline in muscle resistance to stretch) in the transverse direction to the muscle fibers as shown in Fig. 9B. This indicates the interesting possibility that the viscoelastic stress relaxation is anisotropic (a direction dependent) in the MDX mouse diaphragm. Regardless of genotyping, the diaphragm and biceps femoris muscles demonstrate another aspect of viscoelastic behavior, hysteresis or muscle dissipation as shown in Figs. 5 and 6. Curiously, unlike the diaphragm, regardless of the direction of loading, there is no hysteresis within the range of loading of nearly 20% stretch in biceps femoris of the WT mice. However, in the MDX mice, hysteresis exhibited by biceps femoris was dependent on the direction of stretch, showing greater hysteresis in the longitudinal direction of muscle fibers compared to transverse direction of the fibers. This demonstrates a complex role of dystrophin in modulating an anisotropic viscoelastic behavior of skeletal muscles.

Our data show that contractile muscle force is depressed in very young MDX mice diaphragms during isometric contraction in either the absence or presence of transverse fiber passive stress as shown in Fig. 11. Therefore, the cytoskeletal protein dystrophin appears to contribute to both muscle force production and force transmission in the diaphragmatic muscles in very young mice. Dystrophin is therefore, an important mechanical determinant of muscle contractile force production as well as force transmission.

Potential Limitations of the Study

There is a number of potential limitations in the experimental design of our study. We did not establish a causal mechanistic link between the altered muscle mechanics and the progressive muscle damage in the MDX mouse. Our study lacks an implementation of a quantitative approach for correlating muscle fibrosis and increased proportion of collagen to the reduced muscle compliance in the aged MDX muscles. Nevertheless, published work by Stedman et al. (12) documented nicely the progression of the disease and the increased collagen content in the diaphragm and soleus muscles. Regarding the choice of long head of bicep femoris, we recognize that it is one of the least studied mouse skeletal muscles. However, due to the complexity of the in vitro mechanical loading in our study and the unique orientation of muscle fibers of bicep femoris relative to the applied mechanical forces, biceps femoris is ideal to study the mechanics of an in vivo uniaxially mechanically loaded hind limb muscle. One other limitation of our study is that sarcomere length for either the diaphragm or the bicep femoris was not evaluated as function of the applied mechanical loading conditions. There is a potential for a mechanical stress concentration at tissue clamping site and this could potentially generate sarcomere length inhomogeneity at the site of clamping. However, mechanical strain was evaluated at the region enclosed by markers at the center of the diaphragm muscle along the muscle fascicles away from the boundary at the clamping site. Therefore, sarcomere length homogeneity of the central region of the muscles away from the clamping site is not affected. One additional limitation of our study is lack of a quantitative assessment of the mechanics of the MTJ of the diaphragm and biceps femoris in the MDX mouse and WT mice. We have however, discussed the elegant work by Law et al. (19) on the structural changes of the MTJ following the onset of necrosis in skeletal muscles of the MDX mouse. Finally, it is quite challenging to distinguish the effects of aging from those effects of progressive dystrophy and secondary pathology/fibrosis in the MDX mouse.

SUMMARY

Our in vitro mechanical data demonstrated that, in the prenecrotic state, diaphragms and biceps femoris muscles from very young MDX mice are more compliant, compared with those of WT mice. Diaphragm muscle in the older MDX mouse is less compliant, compared with that in the young MDX mouse, largely in part because of the increase in the observed percent fibrosis in old MDX muscles following the onset of the disease. The diaphragm of the very young MDX mice as well as age-matched WT control mice is mechanically anisotropic with greater extensibility and larger compliance in the direction along the muscle fibers compared with that in the transverse direction to the fibers. Regardless, of genotype, biceps femoris, however, is nearly isotropic with little difference between the longitudinal and transverse length-tension relations. Compared with controls, muscle viscoelasticity is altered in the very young MDX mouse diaphragm. Finally, in the prenecrotic MDX mouse, under both uniaxial and biaxial mechanical loading conditions, diaphragm muscle contractile stress is depressed compared with that of age-matched WT mouse. Our data on very young mice calls into question the postulate that progressive muscle damage is the leading cause of skeletal muscle weakness in the MDX mouse. Our data, however, strongly suggest that dystrophin deficiency is responsible at least in part for skeletal muscle weakness.

GRANTS

This work was supported by the National Heart, Lung, and Blood Institute Grant HL-63134 and partly by the National Science Foundation Grant 1714478. M.A.L. is currently funded by an NIH NINDS K08 grant NS120812 and Kaul Pediatric Research Institute.

DISCLAIMERS

A.M.B. is the guarantor of this work and, as such, had full access to all of the data in the study and took responsibility for the integrity of the data and the accuracy of the data analysis. All authors reviewed and take full responsibility for the contents of the manuscript.

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the authors.

AUTHOR CONTRIBUTIONS

A.M.B. conceived and designed research; M.A.L. and A.M.B. performed experiments; M.A., S.W., and A.M.B. analyzed data; M.A.L., S.B., M.A., M.S.A., S.W., and A.M.B. interpreted results of experiments; A.M.B. prepared figures; A.M.B. drafted manuscript; M.A.L., S.B., A.I.S., M.S.A., S.W., and A.M.B. edited and revised manuscript; M.A.L., M.A., A.I.S., M.S.A., S.W., and A.M.B. approved final version of manuscript.