Journal of Applied Physiology


This study compared human muscles following long-term reduced neuromuscular activity to those with normal functioning regarding single fiber properties. Biopsies were obtained from the vastus lateralis of 5 individuals with chronic (>3 yr) spinal cord injury (SCI) and 10 able-bodied controls (CTRL). Chemically skinned fibers were tested for active and passive mechanical characteristics and subsequently classified according to myosin heavy chain (MHC) content. SCI individuals had smaller proportions of type I (11 ± 7 vs. 34 ± 5%) and IIa fibers (11 ± 6 vs. 31 ± 5%), whereas type IIx fibers were more frequent (40 ± 13 vs. 7 ± 3%) compared with CTRL subjects (P < 0.05). Cross-sectional area and peak force were similar in both groups for all fiber types. Unloaded shortening velocity of fibers from paralyzed muscles was higher in type IIa, IIa/IIx, and IIx fibers (26, 65, and 47%, respectively; P < 0.01). Consequently, absolute peak power was greater in type IIa (46%; P < 0.05) and IIa/IIx fibers (118%; P < 0.01) of the SCI group, whereas normalized peak power was higher in type IIa/IIx fibers (71%; P < 0.001). Ca2+ sensitivity and passive fiber characteristics were not different between the two groups in any fiber type. Composite values (average value across all fibers analyzed within each study participant) showed similar results for cross-sectional area and peak force, whereas maximal contraction velocity and fiber power were more than 100% greater in SCI individuals. These data illustrate that contractile performance is preserved or even higher in the remaining fibers of human muscles following reduced neuromuscular activity.

  • chemically skinned fibers
  • unloaded shortening velocity
  • fiber power
  • passive tension
  • spinal cord injury

muscle unloading occurs in a variety of conditions, such as immobilization, disease, paralysis, or exposure to microgravity. The absence of normal weight-bearing activity induces a rapid decrease in muscle mass and strength, especially of antigravity muscles (1, 25). Muscle atrophy induced by unloading is associated with several structural changes, such as modifications of the myosin heavy chain (MHC) isoform expression, inducing fiber-type transitions toward a higher proportion of fast type II fibers. Experiments based on single fiber models have demonstrated great sensitivity in detecting also a certain degree of functional variability of fibers expressing the same MHC isoforms, especially when the pattern of muscle activity changes. Human studies involving 17 days of bed rest (27, 28) or spaceflight (26) revealed that maximal single fiber force (P0) from the soleus muscle was decreased, mainly as a result of a decline in fiber cross-sectional area (CSA), and that maximal unloaded shortening velocity (V0) was increased. Consequently, single fiber power was either maintained or depressed, depending on the study participant or fiber type (29), with fibers expressing type I MHC being generally more affected. Muscle unloading up to 4 mo in a long-term bed-rest study induced a similar pattern of adaptation in the functional properties of type I fibers from the soleus muscle (31), with changes seemingly proportional to the duration of unloading.

Lower limb muscle paralysis as a consequence of spinal cord injury (SCI) is a typical situation of severe, long-term muscle disuse. However, it differs from spaceflight or bed-rest models, since the muscles concerned are not only unloaded, but their neuromuscular activity is also chronically reduced or eliminated. SCI has been shown to induce a pronounced loss of leg muscle mass, reduced oxidative capacity, and alterations in the microvasculature (21, 22). Muscle atrophy is the result of a reduction of the fiber CSA (hypotrophy) and a loss of muscle fibers (hypoplasia) (13). Additionally, this atrophy is accompanied by a relative increase of perimysial tissue (15, 16). Paralyzed muscles display a fiber-type transition toward a predominance of type II fibers (16, 21), and this process seems to be proportional to the duration of the lesion within the first 2 yr posttrauma (4). The modification of fiber-type composition results not only from a loss of type I fibers but also from a progressive conversion of MHC I expression toward MHC IIx expression, which is believed to be the default MHC isoform in humans (12, 19). As a consequence of these phenotypic changes, electrically stimulated lower limb muscles from individuals with SCI have been demonstrated to develop lower isometric force and to have faster contractile properties and greater fatigability (11). More recently, the dynamic mechanical properties of whole muscles were investigated after a long-term spinal cord transection on rat soleus muscle (23). Despite a great loss of maximal force, whole muscle maximal power output was conserved as a result of an increased maximal shortening velocity. It should be noted that whole muscle functional characteristics could be largely influenced by the fiber-type transition following paralysis. Therefore, the single fiber model associated with post hoc fiber-type identification allows for a more in depth analysis of muscle contractile function.

The purpose of this study was to characterize the contractile properties of chemically skinned quadriceps fibers of individuals with chronic SCI and to compare them to those of able-bodied control subjects. Given the results from previous short-term unloading studies, we specifically hypothesized that, in individuals with SCI, single fiber CSA, P0, power, and Ca2+ sensitivity would be decreased and that V0 would be increased. Furthermore, we tested the hypothesis that passive fiber properties would be altered after several years of muscle disuse.



Ten able-bodied men (CTRL; age 26 ± 2 yr, height 180 ± 3 cm, body mass 73 ± 5 kg; mean ± SE) and five men with SCI (age 33 ± 5 yr, height 181 ± 1 cm, body mass 93 ± 8 kg) volunteered to participate in this investigation. The CTRL subjects were all recreationally active for a mean of 2.6 ± 0.5 h/wk. The SCI subjects were not able to stand or walk and had all been permanent wheelchair users for a minimum of 3 yr (15 ± 5 yr). Two SCI participants (subjects SCI 3 and SCI 4; time since injury 28 and 3 yr, respectively) had motor incomplete lesions (level L1 and L4/5, respectively) and were able to have voluntary although not functionally useful contractions of their quadriceps muscles. Subjects SCI 1, SCI 2, and SCI 5 had a complete motor paralysis as a consequence of C7, D7/8, or C7 lesions, respectively (time since injury 17, 24, and 5 yr, respectively). All participants were informed of the risks associated with the investigation and provided written, informed consent. The protocol of this study had been previously approved by the Faculty Ethical Review Committee and complied with the principles of the Declaration of Helsinki.

Muscle Biopsies

A muscle sample was obtained from the vastus lateralis of the right leg using the needle biopsy technique with suction. Control subjects were asked to refrain from intense physical activities 3 days before the sample extraction to minimize the possibility of studying damaged fibers. Muscle samples were immediately placed in cold (0°C) skinning solutions (cf. composition below) and sectioned longitudinally in small bundles of fibers. The bundles were stored in regularly replaced skinning solution at −20°C for at least 5 days before the first experiment.

Skinning, Activating, and Relaxing Solutions

The skinning solution contained (in mM) 125 propionic acid, 2.0 EGTA, 1 MgCl2, 4.0 ATP, 20 imidazol (pH 7.0), 50% (vol/vol) glycerol and protease inhibitors: 0.5 mM PMSF and 20 μg/ml leupeptin. The composition of the relaxing (pCa 9.0; pCa = −log [Ca2+]) and the activating solutions (pCa 4.5) were based on calculations using an iterative computer program described by Fabiato and Fabiato (6), with apparent stability constants adjusted for temperature, pH, and ionic strength. Both solutions contained (in mM) 7.0 EGTA, 20 imidazol, 14.5 creatine phosphate, 1.0 free Mg2+, and 4.0 MgATP. Calcium was added as CaCl2, ATP was added as a dissodium salt, and Mg2+ was added in the form of MgCl2 with a specified free concentration of 1 mM. In both solutions, pH was adjusted to 7.0 with KOH and total ionic strength to 180 mM with KCl. Submaximal activating solutions were prepared by mixing appropriate volumes of activating and relaxing solutions to obtain a series of different free Ca2+ concentrations ranging from pCa 4.7 to pCa 6.4.

Single Fiber Dynamic Force Measuring Setup

Single muscle fibers were subjected to a series of mechanical tests within a time span of 4–5 wk following the muscle biopsy, but not all experiments were performed on all fibers. In a first series of tests, single fiber segments were evaluated either for P0, V0, and force-velocity relationships or for passive tension characteristics. These experiments were performed on a dynamic force measuring setup, allowing for both controlled fiber length changes and force measurements. On the day of an experiment, a muscle fiber segment of ∼3 mm was isolated from the bundle in the relaxing solution and the preparation was mounted between a force transducer (model 400A, Aurora Scientific) and the arm of a high-speed motor (model 312B, Aurora Scientific) as described by Moss (17). The motor was operated either in length (slack tests and passive stretch tests) or in force mode (isotonic contractions) via a high-speed digital controller (model 600A, Aurora Scientific) consisting of an electronic interface, a 16-bit analog-to-digital converter and custom software. Output signals from the motor and the force were recorded by the controller, and collected data were analyzed offline using custom-made software written in our laboratory (LabView, National Instruments). The setup was built over the stage of an inverted microscope (Axiovert 25C, Zeiss) so that the fiber could be viewed with a magnification of ×400 and could be rapidly transferred between wells of a Teflon plate, containing either the relaxing or the maximally activating solution. To perform all the experiments at 15°C, the microscope stage was cooled using a bath/circulation thermostat (Ecoline RE 106, Lauda). The temperature of the solutions was controlled by a thermocouple inserted into one of the wells.

Single fiber dimensions.

Once sarcomere length was adjusted to 2.5 μm (20) by means of a calibrated eyepiece micrometer (×400), a picture of the fiber was obtained using a digital camera (Camedia C3020 Z, Olympus) connected to the microscope while the segment was briefly suspended in the air (±5 s). Assuming that the fiber takes a circular shape, the CSA was determined as the mean of three diameter measurements along the fiber segment on the calibrated digital picture. Fiber length (FL) was measured on a second picture recorded with a magnification of ×50 and defined as the distance between the two fixation ends.

Single fiber P0

Fiber contraction was induced by rapidly transferring the fiber from relaxing into activating solution. Peak activated force (P0, mN) was determined as the stable maximal force developed by the fiber while submerged in activating solution (pCa 4.5). Peak specific tension (kN·m−2) was defined as P0/CSA.

Single fiber V0.

Unloaded shortening velocity (V0) was measured by the slack test method (5-kHz sampling rate). The fiber was fully activated in the pCa 4.5 solution. Once peak isometric force was reached, the fiber was rapidly released so that the force dropped to zero and redeveloped after a time lapse proportional to the step length (of Fig. 1A, inset). The fiber was then transferred back into the pCa 9.0 solution and slowly reextended to its original length. Four to six different length steps were applied on each fiber, all being ≤20% of initial fiber length. The relationship between the time required for force redevelopment and the step length (Fig. 1A) was fit for each fiber with a first-order least squares regression line, the slope of which corresponds to V0 of the fiber. V0 was expressed in fiber lengths (FL) per second to account for differences in the number of sarcomeres in series between different fiber preparations.

Fig. 1.

Representative example of slack test experiments (A) and isotonic contractions experiment (B) in slow [myosin heavy chain (MHC) type I; •] and fast (MHC type IIa; ○) single fibers. Superimposed position and force records of 2 slack tests performed on the same single fiber are illustrated in the bottom right portion of A. When maximal peak force was reached, the fiber was subjected to a rapid shortening of various steps, in this case 8 and 20% of initial fiber length. Four to six different length steps are plotted against the respective duration of unloaded shortening of the fiber. Fiber maximal shortening velocity was determined as the slope of the least squares linear regression and expressed as fiber length (FL)/s. FL and force records of an isotonic contraction experiment are illustrated in the top right portion of B. After maximal activation, the fiber was subjected to 3 successive load clamps of 100- to 150-ms duration. Velocity and force were evaluated over the last 50 ms of each step. The data points of 6 isotonic contraction experiments from the same fiber were fit with the Hill equation.

Single fiber force-velocity relationship.

Fiber shortening velocity was measured during isotonic contractions performed at different loads (5-kHz sampling rate). After full activation with pCa 4.5 solution, the fiber was subjected to three successive isotonic load clamps (Fig. 1B, inset). Each step was 150 ms in duration for slow-twitch fibers and 100 ms in duration for fast-twitch fibers. Shorter time intervals were required for the faster contracting fibers to limit the total distance shortened to ≤20% of initial fiber length. Shortening velocity and force were measured over the final third of the step (50–30 ms), when the force was constant and the fiber shortening velocity linear. The third isotonic clamp was followed by a length step to slacken the segment to 80% of initial fiber length. The fiber was then relaxed in the pCa 9.0 solution and slowly reextended to its original length. This procedure was repeated 5–6 times at different loads so that each fiber was submitted to a total of 15–18 isotonic contractions. All shortening velocities were normalized with respect to initial fiber length and expressed as fiber lengths per second. The data obtained on a single fiber were fitted using an iterative nonlinear curve-fitting procedure (Marquardt-Levenberg algorithm) based on the following Hill equation: (P + a)(V + b) = (P0 + a)b, where P is force, V is velocity, P0 is peak isometric force developed after full activation with pCa 4.5 solution, and the constants a and b have the dimension of force and velocity, respectively (Fig. 1B). The fitting procedure yields values for a and b, which allows calculation of maximal V (Vmax). Only individual experiments for which this relationship yielded a value for r2 of ≥0.98 were retained for further processing. Fiber power was determined based on the parameters of the fitted force-velocity relationship (P0 and Vmax). Absolute power (μN·FL·s−1) was calculated as the product of force and contraction velocity, and normalized power (W/l) was defined as the product of specific tension and contraction velocity. Average curves from CTRL and SCI groups were determined based on individual power-velocity relationships of single fibers.

Passive tension measurement.

Passive fiber characteristics were evaluated in a separate pool of samples from the five SCI participants and five CTRL subjects. These fibers were tested with a progressive stretch-release protocol while remaining in the pCa 9.0 solution using the same ergometer as described above (10-Hz sampling rate). The fiber was first slackened from the initial fiber length to 76% and then progressively stretched to 140% in successive steps of 8% of initial fiber length (Fig. 2, inset), thus covering an assumed range of sarcomere lengths of 1.90–3.50 μm. Each step was reached within 10 s, followed by a period of 3 min of constant fiber length. Passive force progressively declined toward a plateau after the rapid increase induced by the stretch. Following the step at 140% of initial fiber length, the fiber was progressively relaxed using similar time constraints (Fig. 2, inset). Passive tension (kN/m2) was defined as the CSA-normalized force recorded at the end of each step, expressed as the change compared with the value recorded at 76% of initial fiber segment length. For each experiment, passive tension was expressed as a function of fiber strain, evaluated as the ratio of fiber length for a given stretch (mm) divided by fiber length at 76% of initial length (mm) minus one. The ascending limb of the passive tension-fiber strain relationship was analyzed using the equation Y = E·X2, where X is the fiber strain and E (kN/m2) represents complex Young's modulus reflecting the steepness of the curve and thus the stiffness of the fiber (24). Only individual experiments for which this relationship yielded a r2 of ≥0.98 were included for further analyses. To assess visco-elastic properties of the fiber, hysteresis (kN/m2) was calculated as the area between the ascending and the descending limb of the passive tension-fiber strain relationship (Fig. 2).

Fig. 2.

Progressive passive stretch experiments. The fiber was progressively stretched from 76 to 140% of initial fiber length (sarcomere length of 2.5 μm) and then released again to 76%, using a step length of 8% of initial fiber length and a step duration of 3 min (inset). Passive tension was taken as the force recorded at the end of each step divided by fiber cross-sectional area and expressed as the change compared with the value at 76% of initial fiber length. Fiber strain was defined as any given FL divided by FL at 76% of initial length minus 1. The ascending limb of the passive tension-fiber strain relationship (○) was used to determine complex Young's modulus (E) according to Y = E ·X2 (solid line). The data displayed were acquired from a fiber containing type I MHC.

Isometric Force Measuring Setup

In addition to the previously described mechanical tests, single fiber Ca2+ sensitivity was evaluated on samples of four SCI and four CTRL subjects. Due to the time constraints related to this study, these experiments were performed on a different setup than the one described above.

Single fiber preparation.

After isolation of a single fiber segment (approximate length of 3.5 mm) in ice-cooled relaxing solution, the preparation was attached to an isometric force-measuring setup using two aluminium T clips carefully folded over the fiber ends. The output signal from the force transducer (sensitivity 29.41 mV/mN) was amplified and recorded (200-Hz sampling rate) using a Lab-PC+ 12-bit digitizer (National Instrument) for later offline analysis. Fiber length was adjusted so as to yield a sarcomere length of 2.5 μm, as evaluated by laser diffraction. A Peltier element was fitted underneath the experimental wells and maintained the experimental solutions at a constant temperature of 15°C throughout the tests.

Protocol for isometric force measurements.

During the experiment, the fiber was successively activated, either maximally (pCa 4.5 solution) or submaximally (pCa between 4.7 and 6.4), by rapidly transferring it from a well containing the relaxing solution into different activating Ca2+ solutions. The fiber was maintained in the respective activating solution until the developed force showed a plateau. Maximal contractions were performed at the beginning of the protocol, after baseline measurement, as well as once and again throughout the experiment between submaximal contractions (Fig. 3, top inset). The force developed at submaximal activating levels (Pr) was expressed relative to the preceding and following contraction recorded at pCa 4.5.

Fig. 3.

Representative example of force-pCa relationship experiments. Top left inset: gross recording of the experiment performed on a single fiber (MHC type I in this example). Vertical lines indicate full activations using a pCa 4.5 solution. Pr/P0, force expressed relative to maximal Ca2+-activated force, plotted against the respective Ca2+ concentrations expressed in pCa (−log [Ca2+]). Solid line represents the fit obtained by the Hill equation. Bottom right inset: Hill plot constructed by plotting log[Pr/(1 − Pr)] against the respective pCa.

Calcium sensitivity calculation.

The force-calcium relationship was evaluated for each fiber using an iterative nonlinear curve-fitting procedure (Marquardt-Levenberg algorithm). Experimental data points were fitted with the Hill equation Pr = pCan/(pCa50%n + pCan), where pCa50% is the Ca2+ concentration at which half-maximal activation occurs and n is the Hill coefficient, an indicator of the slope of the relationship (Fig. 3). Separate Hill plots were constructed by plotting log[Pr/(1 − Pr)] against pCa, and the Hill plot coefficients, n1 and n2, were calculated as the slope of the least square regression lines fitted to points above and below half-maximal activation, respectively (Fig. 3, bottom inset). The Ca2+ activation threshold was calculated as the pCa for which log[Pr/(1 − Pr)] = −2.5 using data points below half-maximal activation.

Fiber MHC Isoform Determination

After completion of the mechanical tests, the fiber segment was removed from the respective setup, dissolved in 25 μl of SDS sample buffer, and stored at −20°C. Later, the sample was analyzed for MHC isoform content using SDS-PAGE. The sample was heated at 95°C for 3 min, and 5 μl of this extract was loaded on a electrophoresis system (Bio-Rad, Hercules, CA) with a 4% (wt/vol) acrylamide stacking gel and 8% separating gel (2). The gels were run at 140 V for 17 h at 4°C and thereafter stained using a Silver Stain Plus kit (Bio-Rad). The MHC expression was determined on each fiber segment used for the mechanical tests plus on ∼50 fibers per biopsy to determine the MHC profile on some 100 fibers for each subject. A representative example of an SDS gel showing the MHC bands is illustrated in Fig. 4.

Fig. 4.

MHC isoform determination in single muscle fibers. The figure is a representative example of an 8% silver stained SDS gel. The four samples in the middle stem from single fibers containing (from left to right) the type IIa, I, IIa, and IIx MHC isoform, respectively. STD, standard samples containing MHC I, IIa, and IIx isoforms.

Statistical Analyses

Results are presented as means ± SE. One-way ANOVA were used to determine a disparity between the two groups for MHC isoform profiles. Intergroup differences in the fiber contractile properties were analyzed using a general linear model with a nested design (fibers nested within subjects) and “subject” used as a random factor. Because type I/IIa fibers were underrepresented in the samples, no tests were performed for these fiber types. Differences in the passive characteristics between fiber types (according to MHC isoforms) in control samples were assessed using a two-way ANOVA, with subject and fiber type as main factors.

Because of the extremely dissimilar MHC profiles of the two groups and the high proportion of hybrid fibers (which have different functional properties compared with fibers containing only one MHC isoform), composite values that reflect the average for the entire muscle sample were calculated for each individual (25). For a given variable, this composite value was obtained by adding the values from all fibers expressing the different MHC isoforms and dividing the result by the total number of fibers analyzed. This approach allowed us to take into account the hybrid I/IIa fibers and illustrates the effect of long-term muscle disuse on the contractile function of the muscle, assuming that the sample of fibers studied is representative. This analysis was performed on the variables CSA, P0, P0/CSA, V0, power, and normalized power for each individual and then averaged to represent the CTRL and SCI groups. One-way ANOVA was used to determine a disparity between the two groups. Statistical significance was accepted at P < 0.05.


MHC Isoform Composition

MHC isoform profiles of SCI and CTRL groups were established based on the analysis of 536 and 1,474 single fibers, respectively. As illustrated in Fig. 5, long-term muscle paralysis following SCI seems to induce a general transition toward MHC type II fibers, as shown by the lower proportion of type I fibers in SCI compared with CTRL subjects (11 ± 7% vs. 34 ± 5%; P < 0.05). Additionally, type IIa fibers were also significantly less expressed in the SCI group (11 ± 6% vs. 31 ± 5%; P < 0.05), revealing an all-range transition toward type IIx fibers, the proportion of which was sixfold higher in the SCI group (40 ± 13% vs. 7 ± 3%; P < 0.01). In the same way, individuals with SCI showed a higher level of type IIa/IIx fibers (34 ± 5% vs. 23 ± 3%), but this difference was not statistically significant (P = 0.06). A disparity was observed in MHC isoform expression among SCI individuals. Type IIa fibers were expressed in the muscles of the two subjects with incomplete spinal lesions (SCI 3 and 4) plus one participant with a complete paralysis (SCI 1), whereas type I fibers were expressed only in those with incomplete injuries.

Fig. 5.

MHC isoform content profiles of single fibers from control (CTRL) and spinal cord injury (SCI) groups. Significant difference between CTRL and SCI groups: *P < 0.05; **P < 0.01.

Single Muscle Fiber CSA, P0, and P0/CSA

Figure 6 shows P0 as a function of CSA and illustrates the great disparity observed among SCI subjects. For example, subject SCI 2 had surprisingly big and strong fibers, whereas subject SCI 5 was situated at the other extremity of the data cloud. This distribution was not related to the characteristics of the spinal cord lesion, since both subjects had complete muscle paralysis and obvious muscle atrophy. It should be noted that no statistically significant difference was found between SCI individuals with complete and incomplete lesions for any of the variables under study. Although the mean values of CSA appeared to be higher in the SCI group compared with the CTRL group (14% for type I fibers, 16% for type IIa fibers, 36% for type IIa/IIx fibers, and 4% for type IIx fibers), no statistically significant difference was observed between the two groups for any fiber type (Table 1). Similarly, P0 in the SCI group was not different from that of the CTRL subjects, even if a trend for higher values was observed in SCI group for type I (30%), type IIa (13%), and type IIa/IIx fibers (23%). As a consequence, specific tension (P0/CSA) was similar in the SCI and the CTRL groups. The same conclusions were reached when performing the statistical analyses without the data of subject SCI 2.

Fig. 6.

Relation between fiber cross-sectional area (CSA) and peak Ca2+-activated force (P0) with respect to fiber MHC isoform expression. Each symbol represents the results of a single muscle fiber. •, Fibers from CTRL group (n = 84, 96, 50, and 16, for type I, IIa, IIa/IIx, and IIx fibers, respectively); open symbols, fibers of the different individuals from the SCI group (n = 18, 15, 48, and 49 for type I, IIa, IIa/IIx, and IIx fibers, respectively).

View this table:
Table 1.

CSA, P0, and P0/CSA of single fibers from CTRL and SCI groups

Unloaded Shortening Velocity, V0

Frequency histograms of V0, with respect to fiber types and experimental groups, are presented in Fig. 7. Single fibers from the SCI participants had globally higher values for V0 than those from the CTRL group, regardless of fiber type. Significant differences were found for type IIa (3.84 ± 0.19 vs. 3.05 ± 0.07; P < 0.01), type IIa/IIx (5.92 ± 0.20 vs. 3.58 ± 0.10; P < 0.001), and type IIx fibers (6.60 ± 0.21 vs. 4.48 ± 0.29; P < 0.01) but not for type I fibers (1.21 ± 0.05 vs. 1.03 ± 0.03; not significant).

Fig. 7.

Shortening velocity (V0) histograms of single fibers from CTRL and SCI groups with respect to MHC isoform expression, as determined from the slack test method. Total number of analyzed fibers is similar to those given for CSA and P0 determination.

Power-Velocity Relationships

Power-velocity relationships were evaluated on a lower number of fibers than for slack tests (see Fig. 8 legend for details). The extrapolation of the force-velocity relationship was used to determine the maximal shortening velocity of each fiber (Vmax). The results generally confirmed those obtained with the slack test method, in that Vmax was higher in the SCI group for type I fibers (0.94 ± 0.07 vs. 0.68 ± 0.03 FL/s; P < 0.001), type IIa/IIx (4.69 ± 0.14 vs. 1.94 ± 0.10 FL/s; P < 0.001), and type IIx fibers (4.61 ± 0.16 vs. 2.30 ± 0.27 FL/s; P < 0.01). However, no statistical difference was found between the two groups for Vmax of type IIa fibers (2.69 ± 0.26 vs. 1.84 ± 0.08 FL/s for SCI and CTRL subjects, respectively; P = 0.09).

Fig. 8.

Power-velocity relationships from CTRL and SCI groups with respect to MHC isoform expression. The graphs illustrate average power curves ± SE of the CTRL group (n = 83, 94, 49, and 15 for type I, IIa, IIa/IIx, and IIx fibers, respectively) and SCI group (n = 16, 14, 46, and 49 for type I, IIa, IIa/IIx, and IIx fibers, respectively).

Figure 8 illustrates the average (±SE) curve of absolute power in relation to shortening velocity for each fiber type. Absolute fiber peak power was not different between SCI and CTRL subjects for type I (19.0 ± 1.3 vs. 15.2 ± 0.6 μN·FL/s) and type IIx fibers (90.0 ± 6.4 vs. 58.9 ± 5.6 μN.FL/s). Nevertheless, absolute peak power was higher in the SCI group for type IIa (77.4 ± 5.0 vs. 52.9 ± 1.4 μN·FL/s; P < 0.05) and type IIa/IIx fibers (108.7 ± 7.4 vs. 49.7 ± 2.7 μN·FL/s; P < 0.01). When normalized by fiber volume, fiber peak power of SCI individuals was still greater in type IIa/IIx fibers (16.14 ± 0.88 vs. 9.43 ± 0.48 W/l; P < 0.001), whereas no difference was found for type I (3.34 ± 0.18 vs. 2.87 ± 0.09 W/l), type IIa (11.11 ± 1.41 vs. 8.57 ± 0.25 W/l), and type IIx fibers (16.95 ± 0.79 vs. 11.43 ± 1.46 W/l). Since it appears that peak fiber force was not different between groups, the greater fiber power in SCI individuals was mainly the result of a higher shortening velocity. Accordingly, the velocity at which peak power occurs (Vopt) was significantly higher in the SCI group for type I (0.12 ± 0.01 vs. 0.10 ± 0.01 FL/s; P < 0.01) and type IIa/IIx fibers (0.62 ± 0.02 vs. 0.30 ± 0.01 FL/s; P < 0.001) but not for type IIa (0.40 ± 0.03 vs. 0.29 ± 0.01 FL/s; P = 0.12) and type IIx fibers (0.64 ± 0.02 vs. 0.35 ± 0.04 FL/s; P = 0.05). Fiber force at which peak power occurs was similar in both groups (0.16 ± 0.01 vs. 0.15 ± 0.01 mN in type I, 0.21 ± 0.02 vs. 0.19 ± 0.01 mN in type IIa, 0.18 ± 0.01 vs. 0.18 ± 0.01 mN in type IIa/IIx, and 0.15 ± 0.01 vs. 0.18 ± 0.01 mN in type IIx fibers of SCI and CTRL subjects, respectively).

Passive Tension

Passive characteristics of single fibers were assessed on 93 and 76 fiber segments from CTRL and SCI groups, respectively (Fig. 9). The statistical analyses could only be performed on type I, IIa, and IIa/IIx fibers, because the other fiber types were underrepresented. In the CTRL group, no differences were found regarding complex Young's modulus between type I (20.55 ± 0.86 kN/m2), type IIa (16.61 ± 0.78 kN/m2), and type IIa/IIx fibers (16.14 ± 1.08 kN/m2). However, type I fibers had higher values for complex Young's modulus than all type II fibers combined (16.44 ± 0.62 kN/m2; P < 0.01). Hysteresis was higher in type I fibers (1.28 ± 0.08) compared with type IIa/IIx fibers (0.80 ± 0.09 kN/m2; P < 0.05), whereas the results found in type IIa fibers (0.91 ± 0.07 kN/m2) were not significantly different from any other fiber type. Once again, hysteresis was higher in type I fibers compared with all type II fibers (0.87 ± 0.05 kN/m2; P < 0.001).

Fig. 9.

Complex Young's modulus (A) and hysteresis (B) in type I, IIa, and IIa/IIx fibers from CTRL (n = 43, 26, and 15, respectively) and SCI groups (n = 10, 9, and 32, respectively). Due to lack of data, no statistical analysis was performed on type I/IIa and type IIx fibers.

Single muscle fiber stiffness was not altered following long-term muscle paralysis, as complex Young's modulus was not different in the SCI group for type I (21.18 ± 0.92 kN/m2), type IIa (18.77 ± 1.35 kN/m2), or type IIa/IIx fibers (20.90 ± 1.53 kN/m2). Similarly, no difference was observed for the hysteresis values following SCI (1.36 ± 0.12, 1.02 ± 0.07, and 1.18 ± 0.12 kN/m2 in type I, type IIa, and type IIa/IIx fibers, respectively).

Single Muscle Fiber Composite Values

Composite values of fiber characteristics are given in Table 2. The SCI group showed higher values for V0 (P < 0.001), peak power (P < 0.01), and normalized peak power (P < 0.001), whereas no dissimilarity was observed for P0, CSA, and P0/CSA.

View this table:
Table 2.

Single muscle fiber vastus lateralis composite data from CTRL and SCI groups

Force-pCa Relationships

Force-pCa relationships were evaluated on a total of 74 and 87 single fibers from 4 CTRL and 4 SCI individuals, respectively (Table 3). No difference between groups was observed in the force-pCa relationships for type I and type IIa fibers. Despite the fact that the Ca2+ activation threshold was lower in SCI group for type IIa/IIx fibers (P < 0.05), no difference appeared for pCa50% and Hill plot coefficients n1 and n2. Because of insufficient sample size, no statistical analysis could be performed on type I/IIa and type IIx fibers.

View this table:
Table 3.

Force-pCa relationships in CTRL and SCI groups


The present study is the first to characterize the functional adaptations of single muscle fibers from subjects having experienced a chronic decrease or absence of neuromuscular activity for several years. The participants of the SCI group all had longstanding (>3 yr) spinal lesions that had prevented them from normal use of their quadriceps muscles, confining them to permanent wheelchair use. Contrary to our expectations, many of our hypotheses could not be supported by our results.

Fiber-type Profile

The analysis of MHC isoform expression reveals high proportions of type II fibers in atrophied muscles of individuals with SCI. This implicates either that type I fibers disappear or that long-term muscle unloading induces a general scheme of sequential transitions from slow to fast fibers: MHC I → MHC IIa → MHC IIx. Such fiber-type transitions have been previously demonstrated as a consequence of muscle unloading and denervation or cross-reinnervation (19). In the case of spinal lesion, most of this change in fiber-type profile seems to occur within the first 2 yr following injury (4). Nevertheless, two of our SCI participants with motor incomplete lesions (cf. materials and methods) still expressed >20% of type I fibers. Thus it seems possible to preserve a significant proportion of functionally able (cf. below) type I fibers with only very limited muscle use, even after several years of reduced neuromuscular activity.

Single Fiber CSA and P0

Muscle unloading is generally associated with a decrease in maximal muscle force and muscle mass. At the single fiber level, a decrease in fiber CSA has been reported in paralyzed muscles after spinal cord transection in rats (9) or post-spinal injury in humans (5, 22). Why fiber atrophy was not observed in the present study is surprising and difficult to explain. Prior investigations (5, 22) were performed within a relatively short time period following the spinal cord lesion (between 3 wk and 17 mo). Thus it is possible that these analyses included fibers that were in a degenerative process and destined to disappear after longer term paralysis, as in the present study. This could explain why our results are in opposition to those prior data, because the CSA of “survivor” fibers from our SCI participants were not different from those of the CTRL subjects and, if anything, actually bigger. It is also possible that the seated position adopted by our wheelchair-dependent participants placed their vastus lateralis muscle in a lengthened state, which could have limited fiber atrophy, as suggested by data obtained in rat muscle (18). Finally, it cannot be ruled out that spastic muscle contractions in the lower limbs could have helped to preserve some muscle fiber characteristics, especially in the individuals with complete spinal cord lesions.

The similar CSA between the two study groups was associated with equivalent values for single fiber P0, which lends support to the finding that fiber atrophy was not present in the paralyzed muscles. As a consequence, specific tension was not different between SCI and CTRL groups. These results suggest that the loss of muscle mass and force could be largely related to a loss of muscle fibers and maybe less to a deterioration of the contractile function.

Close inspection of Fig. 6 reveals that the data of the participants of the SCI group are more scattered than those of the CTRL group. Thus it seems that not all individuals react in the same way to a long-term reduction in neuromuscular activity. Some degree of disparity between individual reactions to unloading has been reported in previous work. Widrick and coworkers (28) showed that in one subject out of eight, fiber diameter remained unchanged and P0 was increased after 17 days of bedrest, whereas average values showed significant decreases in both variables. This individual also had increases in V0 and fiber power (27, 28). In another investigation from the same group, one out of four astronauts did not display a decrease in fiber diameter following a 17-day spaceflight mission, in contrast to the other three astronauts (26). Given these very specific adaptations of muscle tissue to unloading in some individuals, future research should focus on such disparities and the possible explanations of these peculiar observations.

Shortening Velocity and Power

A well documented but still insufficiently explained adaptation of the contractile machinery to muscle unloading is the increase of maximal shortening velocity observed on whole muscle (23) as well as at the single fiber level (26, 28, 31). Talmadge et al. (23) demonstrated that V0 and Vmax of rat soleus muscles were increased 3–6 mo after spinal cord transection as a result of a fiber-type transition toward larger proportions of type II fibers. Similarly, electrically stimulated quadriceps muscles from individuals with long-term spinal paralysis had higher contraction rates than those of able-bodied control subjects (11). The present data imply that increased whole muscle contraction velocity could not only be the consequence of fiber-type transition but also result from an increase in contraction velocity (V0 and Vopt) of single fibers. To date, there are no satisfactory explanations for the increase in human single fiber contraction velocity independently of MHC isoform expression (3). More confusing is the fact that certain types of exercise training also promote an increase in fiber V0 (14, 30), despite the contrasting physiological stimulus in terms of muscle activation. Clearly, this aspect deserves further attention from the scientific community to explain the underlying mechanisms.

Our hypothesis that long-term muscle disuse would decrease fiber power was not confirmed by the present data. On the contrary, fiber power from individuals of the SCI group was generally higher than in CTRL subjects (Fig. 8), with significant differences in type IIa and IIa/IIx fibers. These results were mainly due to the large increases in contraction velocity, whereas fiber CSA and P0 were similar in the two groups. This observation was surprising because short-term unloading has been associated with losses in power as a result of decreased fiber CSA and P0, insufficiently compensated by a greater shortening velocity (26, 27, 31). Even normalized peak power showed an increasing trend in our SCI participants, the rise being significant for type IIa/IIx fibers. However, this higher normalized peak power could have been caused by a larger proportion of type IIx MHC within those hybrid fibers.

Whole muscle function is not only influenced by single fiber contractile performance but also by fiber-type distribution within the muscle, as well as the total number of fibers present. Composite values were calculated for different variables related to fiber mechanics (25), in an attempt to give a reflection of the average fiber contractile function. The data presented in Table 2 illustrate that average maximal contraction velocity and peak power were greatly increased in SCI individuals, whereas CSA, P0, and P0/CSA were similar to the CTRL group. However, it is well known that whole muscle function deteriorates after several years of unloading, as electrically stimulated lower limb muscles from individuals with spinal paralysis were shown to develop significantly lower force (11). Thus, taken together, these results indicate that the loss of muscle mass and function after long-term reduced neuromuscular activity results from a decrease in fiber number within the muscle, whereas the functional properties of the remaining fibers are conserved.

Ca2+ Sensitivity

The present study failed to demonstrate a change in Ca2+ sensitivity of muscle fibers after several years of reduced neuromuscular activity. These results are in disagreement with the previously reported decreases of Ca2+ sensitivity in human fibers following shorter periods of unloading (26, 27, 31). It is noteworthy that those prior studies were performed on type I fibers from the soleus muscle and, to the authors' knowledge, no data are available on potential changes in Ca2+ sensitivity of single fibers from the vastus lateralis muscle following unloading. This could be an important feature since no changes were found in type I and IIa fibers from monkey gastrocnemius muscles following spaceflight (7, 8) or in fast fibers from the red and white gastrocnemius of rats after hindlimb suspension (10). It appears clearly that different muscles do not respond in the same way to reduced neuromuscular activity, probably depending on their physiological and biomechanical function.

Passive Tension

Passive characteristics of single muscle fibers have not frequently been explored. Chemically skinned fibers from rat soleus muscles were studied by Toursel and coworkers (24) regarding the adaptations in passive characteristics following 14 days of hindlimb suspension. They found that, after muscle unloading, complex Young's modulus was decreased and that this decrease was associated with a loss of the relative amount of titin present in their samples. Once more, our results are in contrast with these observations. Stiffness of the fibers from the paralyzed muscles was not different compared with the samples from our CTRL group (Fig. 9). The discrepancy with the results of Toursel et al. (24) could mainly stem from the unloading paradigm and duration, as well as the species and muscle investigated.

In conclusion, this study analyzed the mechanical properties of single muscle fibers following several years of reduced neuromuscular activity as a consequence of SCI. Our results consistently show that functional contractile performance and passive properties of these fibers are equivalent to those of fibers from individuals with normal muscle function. These surprising observations are in contrast to earlier work, which generally showed significant changes in most mechanical variables, indicating a clear decline in fiber contractile properties. However, past studies have used short-term unloading protocols and have analyzed atrophied fibers with smaller amounts of contractile protein. Although the present results do not provide explanations as to the underlying mechanisms of preserved fiber function, it can be speculated that the main effect of chronic reduced neuromuscular activity is a loss of muscle cells, leaving a population of surviving fibers that have equal or higher contraction performance compared with those from normally functioning muscles.


This study was supported by a grant (to D. Theisen) from the Fonds Spéciaux de Recherche, Université Catholique de Louvain, and the Fonds de la Recherche Scientifique Médicale, Belgique (convention no. 3.4547.04).


  • The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.


View Abstract