INTRODUCTION

Top Abstract Introduction Methods Results Discussion References

IN 1960, FALES ET AL. (10) showed that the oxygen uptake (O₂) of the contracting canine gastrocnemius-plantaris (GP) muscle group, in situ, was partly dependent on the number of activating stimuli. Subsequently, our laboratory (3, 7, 8, 26, 29) and work by others (13-15, 18) have shown additional factors that can influence O₂ in situ. Many of these studies were performed by using isometric tetanic contractions at lengths less than optimal whole muscle length (L_o) (12, 14, 16-18), or with twitch contractions (17), which limited the attained O₂. Others were performed with isotonic twitch and tetanic contractions with afterloads significantly less than the maximal load (8), which likewise affected O₂. Some of these experiments, using isotonic contractions (3, 7, 8), indicated that higher O₂ was possible with stimulus and mechanical conditions optimized toward achievement of high muscle blood flow (). However, a prior review of these studies indicated the difficulty with their comparison (9), in part due to the fact that no experiments had been performed with repetitive isometric tetanic contractions at L_o, by using the optimizing protocol (200-ms trains at 50 impulses/s, 1 train/s) that sets the impulse-driven demand high (10), maximizes active tension development (26, 34), and optimizes perfusion time between contractions (8). Furthermore, no recent comparative studies have been performed to evaluate the effect of short GP muscle length on mechanical and metabolic performance, with this pattern of stimulus activation.

The purposes of this study were 1) to determine the relationship of fatigue to repetitive contractions at short GP muscle lengths (L_s), under conditions that optimize (3, 8, 10, 26); and 2) to determine the effects of altered contractility, via reduction, under mechanically optimized conditions at L_o. The hypothesis was that, compared with L_o, L_s would reduce tension, O₂, and fatigue, indicating the role of mechanics in the determination of O₂ and fatigue in this preparation. Furthermore, the reduction in contractility induced by reduced at L_o should be apparent as enhanced fatigue, indicating the role of metabolic control in the fatigue process under these conditions.

	METHODS

Top Abstract Introduction Methods Results Discussion References

Mongrel dogs (10-17 kg) were anesthetized initially with pentobarbital sodium (30 mg/kg iv), followed by maintenance doses of 60 mg intravenously. The animals were ventilated through an endotracheal tube with a Harvard respirator (tidal volume = 25 ml/kg). End-tidal CO₂ was maintained at 4.5-5% by adjustment of pump frequency. Body temperature was monitored with an intraesophageal probe and maintained at 37-39°C with a heating pad on the body.

The muscle group studied was the GP, which has been used in numerous prior studies in this laboratory (2, 3, 7, 8, 24-29). Circulatory isolation of the GP muscle was produced by ligation of all vessels except the major arterial supply and the veins that emptied directly into the popliteal vein. The popliteal vein was then cannulated for measurement of the venous outflow, which was returned to the animal via the jugular vein. Anticoagulation of the blood was achieved with heparin (2,600 U/kg). The sciatic nerve was located and cut, with the distal stump being placed in an electrode holder. The calcaneus tendon was freed and cut close to the calcaneus and then was clamped and attached to a pneumatic lever system specifically designed for the canine GP muscle (11). Muscle temperature and moisture were maintained with saline-soaked gauze and plastic wrap. The origin of the GP muscle was anchored with two bone nails placed in the femur and tibia. A stiff adjustable strut was placed between the lever and distal bone nail to minimize lever flexing during contraction. This strut was typically semiparallel to the muscle. Because electrically stimulated canine GP muscles weighing ~43 g can develop forces 120 lb. at L_o (28), further measures were taken to minimize compliances within the lever-support system. A locking bar was placed between the bone nail support bars. This formed a closed rectangle of bone and metal support anchored to the heavy lever base. The lever base was clamped to the surgical table to avoid slippage of the lever unit during contractions. The orientation of the muscle was made nearly parallel to the tabletop to minimize distortions in the popliteal origin region, which might alter muscle orientation and . The orientation between the lever attachment and the muscle was made perpendicular so that torque artifacts were minimized. These procedures ensured more accurate transduction of both passive and active force directly to the strain-gauge transducer at the tip of the muscle lever.

The mechanical variables of force development and muscle length were measured and set by using the pneumatic lever (11). Calibration of the force transducer was done by hanging known weights on the lever. Calibration of the length transducer was done by moving the pivot arm in 2-mm increments along a ruler. L_o was determined as the whole muscle length at which active tension production (total minus passive tension) of an isometric twitch contraction (0.2-ms pulse, 4 V) was maximal. L_o was measured as the length from the femur origin of the medial head of the muscle to the point at which the visible muscle fibers terminated on the calcaneus tendon. Maximal active isometric tetanic tension at L_o (P_o) was determined by using a train of electrical pulses (0.2-ms pulse duration, 200-ms train duration, 50 impulses/s, 4 V) delivered to the sciatic nerve. L_s was set as the length that produced active tension that was ~30-45% of P_o, and L_s was typically 89-91% of L_o. Tetanic isometric contractions at L_s were produced with the same stimulus variables as those at L_o to hold the stimulus-frequency effects constant across all experiments (10).

Mean arterial blood pressure was monitored in all experiments via a catheter placed in the contralateral femoral artery. was measured as the rate of venous outflow from the popliteal vein, which was measured by a 4-mm cannulating-type electromagnetic flow probe (Narcomatic). The flow probe was calibrated by timed collections of the venous outflow at intervals during every experiment. Control of in the ischemia experiments was produced by a peristaltic pump, which was fed by the contralateral femoral artery and supplied the muscle via a short catheter placed into the popliteal artery. A bypass line was present that allowed spontaneous perfusion of the muscle directly from the contralateral femoral artery. When the bypass was open, the pump was run very slowly to avoid the effects of perfusion with standing blood on initiation of through the pump line.

Blood samples for gas analysis were withdrawn into glass tuberculin syringes simultaneously from the venous effluent line and the arterial catheter in the contralateral femoral artery. One-milliliter samples were withdrawn for determination of both arterial ([O₂]_a) and venous effluent ([O₂]_v) oxygen concentration by the manometric method of Van Slyke and Neill (33). Additional 1-ml samples were withdrawn for blood-gas analyses by using a microanalyzer at 37°C (Instrumentation Laboratories) to determine PO₂, PCO₂, and pH. A small portion of the arterial sample was also used to measure arterial hematocrit. O₂ was calculated by the Fick method as

(1)

where is muscle venous effluent flow.

Protocol

Experiment 1: Control of length. After determination of L_o, blood samples and measurements were taken with the muscle at rest. Approximately 5 min later, the GP muscles of individual animals of two separate groups were stimulated to produce repetitive isometric tetanic contractions at either L_o or L_s, 1 contraction/s, over a period of 20 min. During the trials, blood samples were taken at 3, 5, 10, and 20 min of contractions. After the contractions were finished, the animal was killed with an overdose of pentobarbital sodium, and the muscle was then excised, trimmed of fat and connective tissue, and weighed. Muscle wet weights for L_o and L_s groups were 40 ± 14 and 49 ± 13 g, respectively.

Experiment 2: Control of blood flow. In a separate group of animals (n = 6), similar to the approach above, L_o was determined (8.6 ± 2.0 cm), and blood samples and measurements were taken with the muscle at rest. Approximately 5 min after withdrawal of the blood samples, the muscles were stimulated to produce repetitive isometric tetanic contractions at L_s (7.8 ± 1.8 cm), 1 contraction/s, over a period of 3 min. This time period was chosen because prior experiments have shown that maximal O₂ is attained at 3-5 min in this preparation, at these stimulus rates, by using the methods employed (3, 8). This method also represented the most efficient way of quickly assessing the contraction-induced at L_s, while minimizing possible fatigue and muscle activation history. After this brief trial, the muscle was allowed to rest for 30 min, in which time and O₂ returned to levels measured before the brief trial. The muscle was reset to L_o, resting muscle blood samples were taken again, and then a 20-min repetitive contraction trial was initiated, with set to the level measured during contractions at L_s. During this trial, blood samples were taken at 3, 5, 10, and 20 min of contractions. After the contractions were finished, the animal was killed with an overdose of pentobarbital sodium. The muscle was then excised, trimmed, and weighed; average muscle wet weight for this group was 38 ± 9 g.

Statistics

	RESULTS

Top Abstract Introduction Methods Results Discussion References

Cardiovascular characteristics of the animals are shown in Table 1. No significant differences were noted in any of these variables between groups; therefore, all animals were considered to be the same with respect to systemic arterial blood pressure, oxygenation, and acid-base status.

Figure 1 shows the results of the 20-min contraction trials. With the exception of at minute 20, all three variables (, tension, and O₂) were significantly different between the spontaneous- experiments at L_o and L_s at all times. The initial active tension at L_o was significantly higher than the initial tension at L_s [544 ± 77 vs. 174 ± 16 (SE) g/g, P < 0.05]. Maximal spontaneous- (3-5 min) was significantly higher in the L_o group (2.0 ± 0.2 vs. 1.4 ± 0.1 ml · min¹ · g¹), demonstrating a greater flow response of muscles contracting repetitively at L_o compared with L_s. Maximal O₂ was significantly higher in the L_o group (12 ± 1 vs. 9 ± 1 µmol · g¹ · min¹), demonstrating a higher metabolic demand in muscles contracting at L_o compared with L_s. The patterns of tension changes over time also were different, i.e., not parallel, between the two spontaneous- groups ("tension × time" crossed factor, ANOVA, P < 0.05). Tension over time decreased significantly during repetitive contractions at L_o and by 20 min was 64% of initial tension. Tension over time at L_s did not change significantly, remaining at a level of ~30-40% of P_o.

View larger version (19K): [in this window] [in a new window]   Fig. 1.   Graphs of blood flow (; A), oxygen uptake (O2; B) and isometric tension (P/Po; C) over time during repetitive isometric tetanic contractions performed at optimal length (Lo) with spontaneous (), at short length (Ls) with spontaneous (), and at Lo with reduced to that measured during contractions at Ls (). Ls range mean ± SE (long-dashed horizontal line with horizontal dotted lines above and below), value measured during brief contraction trial at Ls, just before reduced- trial at Lo. Points are means ± SE; n = 6/group; abscissa (time) is identical for all graphs shown. Po, maximum isometric tension at 50-Hz train frequency. * P < 0.05 (Duncan's post hoc test), spontaneous- group vs. reduced- group at time shown. + P < 0.05 (Duncan's post hoc test), both groups compared with Ls range mean.

The average at minute 3 of the brief repetitive contraction trials at L_s [1.1 ml · min¹ · g¹, not significant (NS)] was similar to the spontaneous at L_s. This value also was significantly lower than during spontaneous- experiments at L_o. Figure 1 shows the pump-controlled, reduced- averages attained during the 20-min repetitive contraction trial at L_o, produced in the same muscles after the brief trial at L_s. These values 1) were nearly identical to the mean of 1.1 ml · min¹ · g¹ during the brief contraction trial, 2) did not change over the course of the 20-min period of contractions, and 3) were significantly lower than spontaneous during repetitive contractions at L_o over nearly the duration of the trial. The level of O₂ during the brief contraction trial at L_s (7 ± 1 µmol · g¹ · min¹, P < 0.05) was significantly lower than that during the spontaneous- experiments at L_o (Fig. 1). The O₂ attained during the reduced- experiments at L_o stayed within this low range throughout the contraction trial, nearly identical to that at L_s. Initial active tension of the reduced- muscles at L_o (528 ± 25 g/g, NS) was not different from that of the spontaneous- muscles at L_o. Tension development was significantly higher during spontaneous at L_o, compared with reduced at L_o from minute 3 onward (Fig. 1). The crossed ANOVA factor tension × time was significant (P < 0.05), indicating that the decay pattern of tension was different between the two groups, apparent as a greater early decline of tension in the reduced- group described in Fig. 1. This rapid decline in tension from minutes 0-3 is clearly visible as a reduction in contractility, a distinguishing characteristic response to relative ischemia. Initial active tension measured in the brief trial at L_s was significantly lower than that at L_o (201 ± 15 g/g, P < 0.05) and was not different from initial tension during spontaneous at L_s. Tension development in the reduced- group at L_o was greater than the L_s tension range over minutes 0-5. However, by minute 10 and beyond, tension had fallen to become statistically indistinguishable from the L_s mean, assuming a value of 44% of initial tension by minute 20. At these same times, tension development with spontaneous at L_o remained significantly elevated above the L_s mean, finishing 20% higher than in the reduced- muscles.

Oxygen delivery ( × [O₂]_a) in resting muscle was similar across all groups, averaging 2-4 µmol · min¹ · g¹. During contractions, oxygen delivery rose significantly in both L_o and L_s experiments, being higher in the L_o group (17 ± 1 vs. 13 ± 2 µmol · min¹ · g¹, P < 0.05). Oxygen delivery also increased in the brief-trial L_s group and in the reduced- L_o group. These values (9-10 µmol · min¹ · g¹) were not significantly different from each other and were not statistically different from those measured in the spontaneous- L_s group mentioned above. They were significantly lower than those of the spontaneous- L_o group (P < 0.05). Arteriovenous oxygen content difference of resting muscle also was similar across all groups, averaging 2-4 vol%. It increased significantly in all groups during contractions, averaging 13-14 vol%, and was not different across groups. Muscle venous-effluent PO₂ also was similar across groups in resting muscle, averaging 54-58 Torr. It decreased in all groups during contractions, attaining similar minima of 25-27 Torr at 3-5 min. Muscle venous-effluent pH averaged 7.36 ± 0.01 at rest and attained minima of 7.21 ± 0.01, 7.25 ± 0.01, 7.23 ± 0.02, and 7.20 ± 0.01, in respective L_o, L_s, brief-trial L_s, and reduced- L_o experiments, typically at 3 min of contractions (all NS). At the end of the 20-min trials, the respective final venous-effluent pH values were 7.26 ± 0.01, 7.24 ± 0.01, and 7.28 ± 0.02 in the L_o, L_s, and reduced- L_o experiments.

Figure 2 illustrates mean O₂-to-tension ratio (O₂/tension) data for all experiments. The dashed and dotted lines running through these points are the best-fit lines and 95% confidence limits, respectively, for each case of whole muscle length (L_o or L_s), calculated by using simple linear regression constrained to pass through the origin of the plot. The O₂/tension data obtained during repetitive contractions with spontaneous at L_o falls along, and was used to calculate, the rightmost best-fit line. The reduced- muscles at L_o also lie near this best-fit line, indicating the similarity of the O₂/tension between these L_o groups. To the left of these data fall the O₂/tension means for the spontaneous- L_s muscles, which were used to calculate the best-fit line running through these respective points. The average for the brief trials at L_s falls near that of the spontaneous- L_s muscles, indicating its close similarity, even though it was obtained in a separate group of muscles. These data show a clear offset between the O₂/tension relationships of muscles at L_s and L_o.

View larger version (20K): [in this window] [in a new window]   Fig. 2.   O2 vs. active tension developed (Tact) in respective experiments at Lo (), Ls (), Lo with reduced (), and at Ls during the brief trial just before reduced- experiments at Lo (). Heavy dashed lines, best-fit linear regression lines for Lo and Ls data during trials from minutes 3-20, forced through origin; dotted lines, respective 95% confidence limits. Plotted values are means ± SE; n = 6/group.

	DISCUSSION

Top Abstract Introduction Methods Results Discussion References

The main findings of this study were 1) muscle tension development and O₂ during tetanic isometric contractions at L_s were lower than muscles at L_o; 2) tension at L_s did not decrease over time; i.e., significant fatigue at L_s was not observed; 3) the ratio of O₂ to active tension developed was higher at L_s compared with L_o; and 4) decreased during contractions at L_o, to the level measured at L_s, decreased tension and O₂ to levels similar to those measured at L_s. These findings were observed by using muscle preparations in situ that were electrically stimulated in an identical fashion to perform repeated contractions over a fixed time period.

Comparisons with Other Studies

View larger version (13K): [in this window] [in a new window]   Fig. 3.   O2 vs. for 4 isometric tetanic contraction experimental groups. Note relative linear relationship for data from all conditions studied. Symbols are the same as in Fig. 2.

A shorter initial muscle length results in decreased active force production, thought to be due to nonoptimal actin-myosin overlap within the sarcomeres (24, 34). This suboptimal function likely explains why O₂ is lower at short muscle lengths (24). However, shorter muscle lengths also produce less passive tension within the muscle (3, 5), probably because of less stretching of the elastic components and lower intramuscular pressures (3, 5). Experiments have shown greater through resting muscle at shorter initial muscle lengths (3, 30), thought to be resultant of these lower passive tensions. Slightly shorter initial whole muscle lengths of the canine diaphragm (93% of L_o) (30) and the GP muscle (98-99% of L_o) (3) in situ also have been shown to permit higher during repetitive contractions. Conversely, initial muscle lengths longer than L_o have been shown to decrease O₂ in resting muscle (29) and during repetitive isometric contractions (24, 26), perhaps because of stretching-induced decrements in . Because of this phenomenon, many GP muscle experiments in situ have been conducted at initial lengths less than L_o. In some of those studies (13, 15, 18-20), muscles set to very low preloads (~10 g/g) have been considered previously to be contracting at maximal metabolic rates for this preparation, which is not supported by the present data.

The differences in whole muscle O₂, between muscles at L_o and L_s, provide evidence that metabolic rates were different between these two mechanical situations. This suggests that high-energy phosphate utilization was different at these different lengths (24). This suggestion is in agreement with findings in frog muscles contracting at short lengths, in which ATP utilization was shown to be decreased (23). However, magnetic resonance imaging studies in humans have shown that changes in the levels of phosphocreatine, ATP, inorganic phosphate, and pH were not different between muscles contracting at L_o and a shorter length (6). Authors of those studies suggested that this was because of similarities in high-energy phosphate-dependent cross-bridge cycling at both lengths, with differences in force explained by formation of non-force-producing cross bridges at L_s (6). The possibility for weak or non-force-producing cross bridges at L_s is suggested by electron-micrographic studies of actin and myosin in vitro that demonstrate that the myosin S1 head can bind to actin in different orientations (32). This could result in a situation wherein ATP is hydrolyzed at the same rate as with force-producing power strokes, but nonoptimal binding angles in some sarcomeres may result in little or no force production at L_s. However, this nonoptimal orientation mechanism cannot readily explain the fatigue that we observed at L_o, presumably because potential cross-bridge orientation should have remained optimized with the muscle held at L_o. Thus we interpret our data to suggest that the loss of force over time with reduced at L_o is because of a metabolically controlled decrement, in either the number of active force-producing cross bridges or the amount of force being produced by each cross bridge.

Implications for Fatigue

Perfusion heterogeneity, wherein relative differences in exist in discrete areas of the muscle (18, 19), also may have played a role in the fatigue observed in the present study. The magnitude of perfusion heterogeneity in the canine GP muscle in situ has been shown to be considerable (18, 19) and should be exacerbated by any mechanical factor that decreases , such as a high preload (3) and high active tension development (5). Fibers in low-flow areas are likely exposed to a lower oxygen delivery and would adjust their mechanical performance downward, as observed in the reduced- experiments at L_o. Previous hyperperfusion experiments (7) also suggest that this may be a valid explanation because fatigue was significantly delayed when muscles were pump perfused at high flows (3 ml · min¹ · g¹) and pressures (~200 mmHg), presumably increasing perfusion to minimally perfused areas.

That these studies showed that canine muscle contracting repetitively at L_s did not fatigue significantly over 20 min is in agreement with studies of human muscles at L_s in vivo (6), in which fatigue was absent or minimal. Our results are in partial agreement with those in the rat diaphragm in vitro, which had less fatigue of isometric force at L_s (0.7 · L_o) (12). However, the present data disagree with the general notion that L_s always predispose a muscle toward greater fatigue (22). Implicit in this aspect is our consideration of the maximal force-generating capacity of the muscle at L_o as a reference point for the comparisons of fatigue at L_o and L_s. We purposely avoided the convention of assigning the initial tension at each length to the "100%" value because we felt that this biased the comparison in a way that ignored the normal physiology of the muscle's capacity. By using this approach, because the diaphragm in situ has been shown to have higher resting at shorter lengths (30), one would expect a blood-perfused, electrically stimulated diaphragm preparation to demonstrate less fatigue at a shorter initial length, similar to the GP muscles in the present study. However, this does not address clinical situations in which muscle weakening may occur, wherein muscle length could play a significant role in the exacerbation of dysfunction or fatigue, under specific loading conditions.

To further compare the patterns of fatigue between groups, curve fitting of the tension data of the spontaneous- group (4) indicated that the L_s level of tension (0.39 · P_o) would have been reached at 42 min of repetitive contractions. This calculation and the data of Fig. 1 suggest that there were adjustments in muscle mechanical performance that resolved as "slow fatigue." This slow component of fatigue may be related to changes in sarcoplasmic reticulum calcium release or reuptake with repeated electrical stimulation of the muscle (1). On the other hand, the "fast-fatigue" component, easily observed in the initial minute of the trials, appeared to be a function of the difference in the error signal between oxygen supply and demand. When the error signal was large, as in the reduced- group at L_o, fast fatigue predominated initially, with a major adjustment in metabolic demand apparent as rapidly lowered force production. When the error signal was smaller, the initial decline was less rapid and slow fatigue predominated, as shown in the spontaneous- group at L_o. When the error signal was near zero, likely because of close matching of supply to a low metabolic demand, measurable fatigue as a loss of tension was insignificant, as observed in the spontaneous- L_s group. Finally, when the error signal is made opposite in direction and magnitude, as in hyperperfusion experiments at L_o (7), fast fatigue may be minimized and visible as a much slower decay of mechanical output over time. However, consideration of these fatigue mechanisms must take into account that the canine GP muscle may be particularly susceptible to -induced oxygen delivery changes. This enhanced susceptibility may be because of its complicated multipennate architecture (1), which can severely limit during repetitive tetanic isometric contractions at L_o through creation of high intramuscular pressures toward the center of the muscle (5). It may also be because of its fiber-type distribution (31), which has been shown to be solely high oxidative (45% fast twitch, fatigue resistant; 55% slow twitch, fatigue resistant) (21). Other muscles with different architectures and fiber-type distributions may be differentially susceptible to oxygen delivery limitations (31), displaying differing patterns of fatigue during manipulations such as those tested in the present study.

Implications for O₂

View larger version (19K): [in this window] [in a new window]   Fig. 4.   Maximal O2 vs. oxygen delivery (O2). Scale, regression equation, and regression line (dashed line) are from Ref. 9, shown here for reference, with O2 being independent (x) variable designation for oxygen delivery. Isometric twitch contraction data at Lo () are from Ref. 16, with numerals indicating 1, 2, 4, and 6 twitches/s, assuming arterial oxygen content = 18 vol%, at arterial PO2 = 77 Torr (17), for calculation of O2 values. Solid line, regression of isometric twitch and tetanic data at Lo, only. Symbols for data from present study are the same as in Fig. 2. Note that 2 maxima for isometric tetanic contractions at Ls reside near reduced at Lo datum. For reference, isotonic (P/Po = 0.30) contraction data are also shown: 4 twitches/s () and tetanic contractions from Ref. 8 (); preload-released isotonic tetanic contractions from Ref. 3 (); and isotonic tetanic contractions with "forced flow" from Ref. 7 (). All tetanic contraction points are from studies utilizing optimal stimulus-contraction paradigm (50 impulses/s, 200-ms train duration, 1 train/s). See text for additional details.

Three major features are apparent when the above data are considered. First, the maximal datum for isometric tetanic contractions at L_o is higher than that previously reported for the isotonic normoxia experiments (8). This finding indicates the dependency of O₂ on the afterload under these optimized stimulus conditions and is consistent with prior studies in this muscle group, using other stimulus conditions (25). It also indicates that the capacity for O₂ in this muscle group was underrepresented by the isotonic experiments performed at low afterloads (8). Second, these data show that, for repetitive isometric contractions, O₂ is dependent on oxygen delivery over a range of mechanical conditions, from simple twitches to tetanic contractions. Third, the maximal data of the short-length isometric groups (L_s and brief-trial L_s) fall on or near the regression line, near the reduced- group, contracting at L_o. Consideration of these features indicates that the reduced mechanical capacity data from experiments at L_s, and the reduced contractility datum from experiments with reduced at L_o, overlap with points previously considered as maximal working conditions of isometrically contracting muscles set to low preloads (9, 13, 15, 18-20). These data also show that, although the high preload and active tension development at L_o may result in conditions that inhibit spontaneous through the muscle (3, 5), achievement of this whole muscle length appears necessary to attain high O₂ values that approach maximal metabolic conditions in this preparation.