AMERICA N JOURNAL Vol. 230, No. 4, April
PHYSIOLOGY 1976. Printed
OF
in U.S.A.
Oxygen uptake for negative work, stretching contractions by in situ dog skeletal muscle WENDELL N. STAINSBY (With the Technical Department of Physiology, University of Florida Medicine, Gainesville, Florida 32610
STAINSBY, WENDELL N. Oxygen uptake for negative work, stretching contractions by in situ dog skeZeta1 muscle. Am. J. Physiol. 230(4):1013-1017. 1976. - Oxygen uptake for negative work, stretching twitch contractions by in situ gastrocnemiusplantaris muscle was calculated from measurements of venous outflow and arterial and venous blood oxygen contents. Contractions were produced by valving air at high pressure into the pneumatic lever lo-50 ms before stimulation of the muscle. The loads produced were up to about 2.5 times isometric. Muscle length was always below optimal isometric length. Oxygen uptake for shortening contractions increased with increasing load up to isometric load. Oxygen uptake for stretching contractions decreased with increasing loads above isometric load. Velocity of shortening decreased with increasing loads up to isometric load whereas velocity of stretching increased with increasing loads above isometric. In shortening contractions external work done by the muscle was greatest at intermediate loads, but in stretching contractions the work done on the muscle increased with increasing loads. In stretching contractions the ratio of the energy equivalent of the work absorbed by the muscle reached 8.0 times the energy equivalent of the oxygen uptake. Since this ratio cannot exceed 1.0 for an engine, muscles must act as brakes during stretching contractions.
muscles as brakes; muscle stretching; muscle lengthening; gastrocnemius muscle; force with stretch; stretching twitch contractions
HAVE BEEN SEVERALINVESTIGATIONS ofheatproduction and ATP breakdown or creatine phosphate breakdown by muscles that were stretched while stimulated to contract (for examples, see 1, 7, 9). Heat production and ATP or creatine phosphate breakdown were less than was expected compared to shortening contractions. The suggestion was made that some of the work done on the muscle might have been converted into chemical energy, but the necessary increase in ATP and/or creatine phosphate has not been found (for detailed bibliography, see 7). There also have been studies of oxygen uptake during negative work exercise by intact man (e.g. 2, 4). In these studies oxygen uptake for negative work was less than that observed for the same levels of positive work. The in situ muscle preparation would allow measurement of oxygen uptake without interference in the measurement by other parts of the organism. There seem to be no studies of such contractions with in situ muscles. Since this preparation seems to show an anom-
THERE
Assistance of Donna College of
T. Dolbier)
alous response to passive stretch compared to isolated muscles (15), it might also be unusual in its response to stretching contractions. The study reported here is of measurement of oxygen uptake for negative work contractions in which the muscles were forced to stretch while trying to contract. The data indicate that in situ muscles show decreasing oxygen uptake for negative work contractions with increasing velocities of stretching. The ratio of the energy that the muscle can absorb to the energy required to activate the muscle is well over 1.0. METHODS
Mongrel dogs of lo-15 kg were used for these experiments. They were anesthetized by intravenous pentobarbital sodium, 30 mg/kg. Additional anesthetic was given when needed. The tracheas were intubated to ensure patent airways. A heat lamp and heating pad were used to maintain normal body temperature. Coagulation of the blood was prevented by intravenous heparin, 25 mg/kg (approximately 4,000 U/kg). The venous circulation from the left gastrocnemiusplantaris muscle was isolated as described previously (13, 15). Briefly, all branches of the popliteal vein that did not come from the muscle group were ligated and all venous connections from the muscle that did not go directly to the popliteal vein were ligated. As a result, all of the venous outflow from the muscle had to go via the popliteal vein. The vein was cannulated and the venous outflow was via a flowmeter to the left jugular vein. The sciatic nerve was located near the muscle and followed centrally for an inch or more; it was doubleligated as centrally as possible and cut between the ties. The distal stump of the .nerve was fitted into a small tubular electrode holder for stimulation. The stimulation was square pulses of 8 V amplitude and 0.2 ms duration at a frequency of l/s. The rate of oxygen uptake each minute by the muscle was calculated from the venous outflow, measured by a 3-mm cannulating-type electromagnetic flowmeter connected between the popliteal and jugular veins, and the arteriovenous blood oxygen content difference. The arterial samples were taken from the contralateral femoral artery and the venous samples were taken from the popliteal vein cannula via a small catheter threaded through the wall of the venous outflow tubing. The blood samples, 0.6 ml, were taken into tuberculin sy-
Downloaded from www.physiology.org/journal/ajplegacy by ${individualUser.givenNames} ${individualUser.surname} (129.081.226.078) on December 18, 2018.
W. N. STAINSBY
1014
ringes, capped, and kept in ice until analyzed for oxygen content. Oxygen content of the blood samples was determined with a Lex 0, Con oxygen analyzer. About 100 duplicate analyses were compared with the traditional manometric technique and the results indicated that in our hands the method agrees very well with the manometric method as reported previously (6, 10). Ninetyfive percent of the time duplicate analyses were well within 0.5 vol/lOO ml of each other. The error range was essentially constant over the range from 1.0 to 25.0 vol/ 100 ml. The flowmeter was calibrated at the end of each experiment with each animal’s own blood. The 95% confidence limits of the flow were determined to be k 1.0 ml/min over the flow range observed. Oxygen uptake was calculated as the product of each arteriovenous pair and their blood flow. The muscle and tendons were freed from surrounding tissue. The tendon was fixed to a small aluminum clamp that was connected in turn to the pneumatic myograph (8, 14). The pneumatic myograph was modified by the addition of a large-bore electrically operated valve between the bellows of the lever and the air-spring bottle. The valve was activated by a stimulator arranged so the valve was opened once each second for about 200 ms. This stimulator also triggered a second stimulator after a lo- to 50-ms delay. The second stimulator stimulated the muscle to contract. Figure 1 is a reproduction of recordings of the events Panel A is a normal isotonic conduring contractions. traction with the air valve. At the beginning, the air pressure is zero and the muscle is at rest length, L,. Then the air pressure increases suddenly and the muscle is stretched until it hits the afterload screw, which had been preset to give an initial length, Li, that was less than optimal length, L,,, for isometric contractions. The tension of the muscle is the preload. The remainder of the tension is supported by the afterload screw. The muscle is stimulated and contracts. Tension rises until the muscle tension is equal to the force of the bellows. After this the muscle shortens against the force from a to b and then relaxes back to Li. Subsequently, the tension falls to zero when the air valve empties the bellows. Panel B shows the response to a force slightly w _ a - -. w greater than isometric force, PO. First, tension rises and B
FIG.
1. Reconstructed
shortening stretching
contraction contraction
n
C
records of force and length during a (A), a stretching contraction (B), and a faster (C).
the muscle begins to stretch. The muscle is stimulated. The delay has been adjusted so the contraction is in the same length range as paneZ A. The muscle contraction merely slows the rate of stretch between a and b. In paneZ C, C, a greater force causes more rapid stretching. The force rise time in C is not quite fast enough for a perfect plateau during the contraction. When this occurred the tension in the middle of the contraction, between a and b, was used for the calculation of work and force-velocity relation. The velocity of the shortening was the maximal velocity early in the positive work contractions. The velocity of lengthening was calculated from the slope in the middle of the S-shaped portion, a6, Fig. 1, of the lengthening contractions. The AL for the positive work calculations was measured as usual, Li minus the minimal length during the contractions, a-b, Fig. lA. This AL times the load in gram per gram of muscle was used to give the work per contraction in gram-centimeters per gram of muscle. In the lengthening contractions the AL from the beginning to the end of the straight part, a-b, of the S curve, Fig. 1, B or C, was used for calculation of work. This was called negative work as opposed to positive work for the usual shortening contractions. The negative work is actually the work done on the muscle by the lever during the contraction phase of the lengthening. The order of events was exactly as done previously (13, 14). The preparation was assembled, isometric L4, was determined, and Li was set 6 mm less than L,,. After a 5-min rest period, blood flow was measured and two arterial and two venous blood samples, 1 min apart, *were taken during continued resting conditions. Next, the contractions at 1 twitch/s were started. The four load conditions were applied in random order. After each was applied by adjusting the bellows air pressure, the delay was adjusted to place the contractions in the same length range, and then conditions were allowed to reach a steady state of 5-7 min. After this period two arterial and two venous blood samples were taken, 1 min apart, the load condition was changed, and so on. After the four load conditions, the contractions were stopped, the flowmeter was calibrated, and the length and tension components were calibrated. The muscle was removed, cleaned of visible fat and connective tissue, and weighed. The longest length with no stretch was measured, L,. Average L, was about 10 cm. The two serial oxygen uptake measurements were averaged for each condition. Average oxygen uptake at rest was subtracted from the average oxygen uptake during each contraction condition to yield the net oxygen uptake for the contractions. This in turn was divided by the weight of the muscle and the number of contractions each minute, 60, and the decimal was adjusted to give Vo, as microliters per gram contraction. The velocity was calculated in centimeters per second. Of the four loads used, one was light, but greater than the preload necessary to reach Li (Fig. IA); one was isometric; one was adequate to give a low-velocity stretch (Fig. U?); and one gave a rapid stretch (Fig. 10. 1C). The loads are expressed as P/P,,, with P,, the isometric tension in that experimental preparation.
Downloaded from www.physiology.org/journal/ajplegacy by ${individualUser.givenNames} ${individualUser.surname} (129.081.226.078) on December 18, 2018.
0, FOR
NEGATIVE
1015
WORK
I5
RESULTS
The basic background data of blood flow and blood pressure for these experiments were in the same range as previous studies. Oxygen uptake of the muscle at rest averaged 5.6 $/g min and ranged from 2.5 to 8.7 pi/g min, the same as reported previously (13, 15). In a pilot series, muscle was stretched from L, to the stop at 0.94 Li/L, once each second as done in the stretching contractions. However, in this series, the muscle was not stimulated. Resting flow and oxygen uptake were not altered by this intermittent passive stretch. The net oxygen uptake in microliters per gram contraction versus load in six experiments is presented in Fig. 2. The points in each experiment are presented as numbers. The points in each experiment are connected by straight lines. As seen before (13), oxygen uptake rises with load up to P/P,, = 1.0. Beyond P/P,, = 1.0, oxygen uptake falls with increasing load. The average arteriovenous 0, difference during contractions was about 10 vol/lOO ml and the average blood flow was about 20 ml/min. The statistical analysis utilized an available computer program (5). It was determined that a quadratic model relating oxygen uptake to the relative load, P/P,), was the best fit for the data. The equation for the quadratic model is:
Vel cm/set IO
5
C
-5
-10
-15 0
Oij (P/PO)= /A + CY~+ B, (P/P,,) + B, (P/P,,)” + Eij Oij (P/P,,) = oxygen uptake for a contraction of the jth measurement of dog i at a given load, P/P,, = overall mean oxygen uptake P = effect of the ith dog ai = fractional load P/P = error of jth measurement of ith dog Eij () The estimates
of the parameters
8, = 2.534; I3z= -1.158;
are:
t = 4.71;
P = < 0.0001 P = < 0.0001
t = -5.18;
and maximal oxygen uptake occurs at approximately P/ P,, = 1.09. Statistically, the parameters are highly significant. The model fits the data very well and accounts for 93% of the variation in oxygen uptake. 1.5 vo* yL/g
1
cont.
1.z .9- ’ .6-
0;
Load 0
.5
1.0
1.5
p/p0
2.0
I
2.5
FIG. 2. Oxygen uptake-load relationships observed in 6 experiments for shortening and stretching contractions. Numbers denote individual experiments.
0.5
IO .
1.5
2.0
2.5
3. Velocity-force relationships for shortening and stretching contractions. Symbols indicate 3 runs on the same muscle. Numbers denote experiments in Fig. 2, where 0, uptake was measured. FIG.
A force-velocity curve from three runs on a single preparation, which was otherwise the same but many points were obtained, is presented in Fig. 3. The first run (circles) began with a light load and progressed in steps to 2.25 x P,,. The second run (triangles) began with 2.4 x P,, and progressed in steps to minimal load. The third run (crosses) is a repeat of the first run. The shortening part of the curve is like that recorded previously using higher fidelity recording techniques (3). The force-velocity data from the 0, uptake experiments are numbered as in Fig. 2 and fall along with the points from the single experiment in which many points were obtained. The displacement to the left at lengthening velocities of experiments 3, 4, and 8 is probably due to their short Li, which produced P,, values of only 140 g/g whereas the other experiments averaged 225 g/g (see 10 for similar displacement). All the experiments were adjusted to Li 6 mm less than L,, Li/L, = 0.94, but the length-tension of this muscle is very steep (13). As a result, a few experiments developed low P,, in spite of the effort to fix Li/Lb. The calculation of the work done by the lever on the muscle during the stretching is not as simple as in the shortening contractions. The problem is one of timing and definition. To illustrate the situation, Fig. 4 was made. Figure 4 is a reconstruction plot of the instantaneous length-tension relations that occurred during the shortening and lengthening contractions illustrated in Fig. 1. Panel 4A illustrates a contraction length and tension plot for a regular shortening contraction such as
Downloaded from www.physiology.org/journal/ajplegacy by ${individualUser.givenNames} ${individualUser.surname} (129.081.226.078) on December 18, 2018.
W. N. STAINSBY
1016 IO.00 Lcm
A
IO.25 IO50 10.75 II.00 1 0
q/q
Lood 1 I
50
I
IO0
0
FIG. 4. Length-force plots of contractions 1: external positive work. Area 2: external extra negative work of relaxing muscle.
1
v
150
300
,
450
A and B in Fig. 1. Area negative work. Area 3:
in Fig. lA. With the pneumatic lever, the muscle starts at 10 cm length, L,. When the air pressure rises, the muscle stretches to Li with a small rest tension. When stimulated, the tension rises along the horizontal line with the arrow to a, where it equals the bellows force. Then the muscle shortens along the vertical arrow to minimal length, b. The muscle then relaxes down the vertical line to a and then the tension falls along the horizontal line to Li. When the bellows pressure is reduced to atmospheric pressure, the muscle shortens back to 10.0 cm L,. This cycle was repeated with each contraction. The positive work half of this cycle is the work done by the muscle and it is equal to area 1. In the nega* tive work contraction, 4B, the cycle begins with the muscle at L,. When the valve opens, the tension starts to rise and the muscle begins to stretch. Before the muscle stretches to Li, the muscle is stimulated. The velocity of lengthening is slowed while tension rises rapidly to a in Fig. 4.E and remains reduced, following the arrow to b. After b. the velocitv of stretching increases rapidly until the ‘lever hits the stop. In terms of time since stimulation, a in the lengthening contraction is comparable to a in the shortening contraction. The point b is comparable to b in the shortening contraction. The work absorbed by the muscle while it is trying to contract is area 2 of Fig. 4B. After b, the muscle continues absorbing more work from the lever until it hits the afterload screw. The additional work done on the muscle after its activity declines, after b, is area 3 of Fig. 4B. Presumably, area 3 is irrelevant to the energetics of the contraction as is the equivalent in the relaxation portion of a shortening contraction. When only area 2 was evaluated from the a-b AL and the tension, it was noted that the muscle could absorb more work in a stretching contraction than the work it could do in a shortening contraction. The ratio of the energy equivalent of the work absorbed to the energy equivalent of the oxygen uptake ranged from zero under isometric conditions to 8.0 under the most rapid stretch conditions. DISCUSSION
The data indicate that for a given stimulus oxygen uptake is decreased from the isometric level when the load applied to a muscle becomes large enough to force the muscle to lengthen when it is trying to contract. A
question that might be asked is whether the decrease is caused by reduced flow due to stretching. There is evidence to suggest reduced flow is not the cause because flow and arteriovenous blood oxygen content differences were both decreased at the very high loads. This is a characteristic response of these preparations to a decrease in oxygen use. A reduction in oxygen uptake due to decreased flow is always associated with a large increase in arteriovenous blood oxygen content difference (15). In addition, the muscle was observed to show stable continuous contractions for at least 20 min, which could not have occurred if a significant nonsteady state existed due to forced decrease in oxygen uptake related to insufficient flow. Creatine phosphate (CP) breakdown has been studied for stretching contractions (6). Whereas the present study shows decreasing oxygen uptake with increasing velocities of stretch, this earlier study shows a sharp decrease in CP breakdown with slow stretch but increasing CP breakdown from that level with increasing velocity of stretch. I know of no explanation for this difference. The present study’s decrease in oxygen uptake would fit more closely predictions based on the decreased heat production in muscle stretched while studies, trying to contract (1, 9). In the heat-production the total heat production, work done on the muscle plus the heat production by the muscle trying to contract, progressively approached values equal to only the work done on the muscle as velocity of stretch increased. This suggested that the heat production by the muscle in trying to contract decreased with more rapid stretching and it is this heat production that should be proportional to oxygen uptake, which also decreased as velocity of stretching increased. The force-velocity relationship for shortening and stretching contractions seems to be similar in shape to that reported for heart and frog muscle (11, 12) and to that estimated for intact human muscles (2, 4). The actual velocities differ among the different muscles, but the basic relationship is similar. The cause of the shape of the force-velocity curve and the decrease in oxygen use lies in the force, number, and turnover of crossbridges in the sliding-filament system, which is beyond the realm of th .is study. However, it appears to m .e that inversely with velocOXY gen uptake 1s simply ’ changing ity regardless of the direction of movement. The observation that the oxygen uptake is approximately the same for similar velocities of shortening and stretching clarifies the reasons oxygen uptake is lower for negative than for positive work conditions in exercising humans. It has been suggested previously (2, 4) that the force-velocity curve makes it possible to develop the same force for negative velocity contractions as for positive velocity contractions with many fewer motor units stimulated and at a lower frequency, and it has been shown that the number of impulses delivered to a muscle in a contraction is the major determinant of oxygen uptake for the contraction (8). Therefore, negative work exercise uses less oxygen because there are fewer stimuli necessary for negative work than for the same level of positive work. The fact that the muscles can absorb up to 8 times as
Downloaded from www.physiology.org/journal/ajplegacy by ${individualUser.givenNames} ${individualUser.surname} (129.081.226.078) on December 18, 2018.
0, FOR
NEGATIVE
1017
WORK
much work energy as the energy they use to activate the system to absorb the work energy generates a problem in terminology. It is possible to consider the ratio of the negative W ork done to the energy required to activate the muscl .e to do that work. The ratio being well over 1.0 implies an efficiency of over 100%. A work engine cannot do work with an efficiency over lOO%, so the model is not correct. In reality, when doing negative work the muscle acts as though it were a viscoelastic brake. It is usual for brakes to be able to absorb many times more energy than is needed to activate the brake. For brakes the ratio is called the coefficient of performance. The maximal coefficient of muscle performance observed in these experiments was about 8.0. As brakes go, this is not unusual. That a muscle can do this is exciting. The ability to do this must be translatable into physical chemistry terms with regard to myofilament crossbridges.
The effect of the high coefficient of muscle performante on energy need can be seen in the high coefficient of performance of negative work in exercise studies on man. In an exemplary ’ study (4) th .e apparent net coefficient of performance from the that can be calculated data for positive work on a bicycle is 23%. The same calculation for negative work gives 166%. It would appear that the body as a whole is able to realize a significant energetic advantage when performing negative work. The author thanks Dr. S. B. Tricky of the Department of Physics and Dr. E. A. Farber of the Department of Mechanical Engineering for their advice and comments on the methods and on the interpretation of the data and Dr. Rose Ray of the Division of Biostatistics of the Department of Statistics for her help in the statistical analysis. This study was supported by National Institutes of Health Grant HL 14806-15. Received
for publication
21 April
1975.
REFERENCES 1. ABBOTT, B. C., X. M. AUBERT, AND A. V. HILL. The absorption of work by a muscle stretched during a single twitch or a short tetanus. Proc. Roy. Sot., London, Ser. B 139: 86-104, 1951. 2. ABBOTT, B. C., B. BIGLAND, AND J. M. RITCHIE. The physiological cost of negative work. J. Physiol., London 117: 380-390, 1952. 3. ALLEN, P. D. The Mechanical Properties of in Situ Dog Skeletal MuscZe (PhD). Gainesville, Fla.: University of Florida, August 1973. 4. ASMUSSEN, E. Positive and negative muscular work. Acta Physiol. Stand. 28: 364-382, 1953. 5. BARR, A. J., AND J. H. GOODNIGHT. Statistical Analysis System. Raleigh, N. C.: North Carolina State University Dept. of Statistics, August 1972. 6. CHAPLER, C. K., AND J. KATRUSIAK. Blood flow and carbohydrate metabolism in dog skeletal muscle in situ following propranolol. Med. Sci. Sports 6: 193-197, 1974. 7. CURTIN, N. A., AND R. E. DAVIES. Chemical and mechanical changes during stretching of activated frog skeletal muscle. CoZd Spring Harbor Symp. Quant. BioZ. 37: 619-626, 1973. 8. FALES, J. T., S. R. HEISEY, AND K. L. ZIERLER. Dependency of oxygen consumption of skeletal muscle on numbers of stimuli during work in the dog. Am. J. PhysioZ. 198: 1333-1342, 1960.
A. V., AND J. V. HOWARTH. The reversal of chemical 9. HILL, reactions in contracting muscle during an applied stretch. Proc. Roy. Sot., London, Ser. B 151: 169-193, 1959. 10. KUSHIMA, F., W. C. BUTTS, AND W. L. RUFF. Superior analytical performance by electrolytic cell analysis of blood oxygen content. J. Appl. Physiol. 35: 299-300, 1973. 11. MASHIMA, H., K. AKAZAWA, H. KUSHIMA, AND K. FUGI. The force load velocity relation and the viscous-like force in the frog skeletal muscle. Japan. J. PhysioZ. 22: 103-120, 1972. 12. MASHIMA, H., AND H. KUSHIMA. Determination of the active state by the graphical experiment and instantaneous methods in the frog ventricle. Japan. Heart J. 12: 545-561, 1971. 13. STAINSBY, W. N. Oxygen uptake for isotonic and isometric twitch contractions of dog skeletal muscle in situ. Am. J. Physiol. 219: 435-439, 1970. 14. STAINSBY, W. N., AND 3. K. BARCLAY. Oxygen uptake for brief tetanic contractions of dog skeletal muscle in situ. Am. J. Physiol. 223: 371-375, 1972. 15. STAINSBY, W. N., J. T. FALES, AND J. L. LILIENTHAL, JR. Effect of stretch on oxygen consumption of dog skeletal muscle in situ. Bull. Johns Hopkins Hosp. 99: 249-261, 1956.
Downloaded from www.physiology.org/journal/ajplegacy by ${individualUser.givenNames} ${individualUser.surname} (129.081.226.078) on December 18, 2018.
PHYSIOLOGY 1976. Printed
OF
in U.S.A.
Oxygen uptake for negative work, stretching contractions by in situ dog skeletal muscle WENDELL N. STAINSBY (With the Technical Department of Physiology, University of Florida Medicine, Gainesville, Florida 32610
STAINSBY, WENDELL N. Oxygen uptake for negative work, stretching contractions by in situ dog skeZeta1 muscle. Am. J. Physiol. 230(4):1013-1017. 1976. - Oxygen uptake for negative work, stretching twitch contractions by in situ gastrocnemiusplantaris muscle was calculated from measurements of venous outflow and arterial and venous blood oxygen contents. Contractions were produced by valving air at high pressure into the pneumatic lever lo-50 ms before stimulation of the muscle. The loads produced were up to about 2.5 times isometric. Muscle length was always below optimal isometric length. Oxygen uptake for shortening contractions increased with increasing load up to isometric load. Oxygen uptake for stretching contractions decreased with increasing loads above isometric load. Velocity of shortening decreased with increasing loads up to isometric load whereas velocity of stretching increased with increasing loads above isometric. In shortening contractions external work done by the muscle was greatest at intermediate loads, but in stretching contractions the work done on the muscle increased with increasing loads. In stretching contractions the ratio of the energy equivalent of the work absorbed by the muscle reached 8.0 times the energy equivalent of the oxygen uptake. Since this ratio cannot exceed 1.0 for an engine, muscles must act as brakes during stretching contractions.
muscles as brakes; muscle stretching; muscle lengthening; gastrocnemius muscle; force with stretch; stretching twitch contractions
HAVE BEEN SEVERALINVESTIGATIONS ofheatproduction and ATP breakdown or creatine phosphate breakdown by muscles that were stretched while stimulated to contract (for examples, see 1, 7, 9). Heat production and ATP or creatine phosphate breakdown were less than was expected compared to shortening contractions. The suggestion was made that some of the work done on the muscle might have been converted into chemical energy, but the necessary increase in ATP and/or creatine phosphate has not been found (for detailed bibliography, see 7). There also have been studies of oxygen uptake during negative work exercise by intact man (e.g. 2, 4). In these studies oxygen uptake for negative work was less than that observed for the same levels of positive work. The in situ muscle preparation would allow measurement of oxygen uptake without interference in the measurement by other parts of the organism. There seem to be no studies of such contractions with in situ muscles. Since this preparation seems to show an anom-
THERE
Assistance of Donna College of
T. Dolbier)
alous response to passive stretch compared to isolated muscles (15), it might also be unusual in its response to stretching contractions. The study reported here is of measurement of oxygen uptake for negative work contractions in which the muscles were forced to stretch while trying to contract. The data indicate that in situ muscles show decreasing oxygen uptake for negative work contractions with increasing velocities of stretching. The ratio of the energy that the muscle can absorb to the energy required to activate the muscle is well over 1.0. METHODS
Mongrel dogs of lo-15 kg were used for these experiments. They were anesthetized by intravenous pentobarbital sodium, 30 mg/kg. Additional anesthetic was given when needed. The tracheas were intubated to ensure patent airways. A heat lamp and heating pad were used to maintain normal body temperature. Coagulation of the blood was prevented by intravenous heparin, 25 mg/kg (approximately 4,000 U/kg). The venous circulation from the left gastrocnemiusplantaris muscle was isolated as described previously (13, 15). Briefly, all branches of the popliteal vein that did not come from the muscle group were ligated and all venous connections from the muscle that did not go directly to the popliteal vein were ligated. As a result, all of the venous outflow from the muscle had to go via the popliteal vein. The vein was cannulated and the venous outflow was via a flowmeter to the left jugular vein. The sciatic nerve was located near the muscle and followed centrally for an inch or more; it was doubleligated as centrally as possible and cut between the ties. The distal stump of the .nerve was fitted into a small tubular electrode holder for stimulation. The stimulation was square pulses of 8 V amplitude and 0.2 ms duration at a frequency of l/s. The rate of oxygen uptake each minute by the muscle was calculated from the venous outflow, measured by a 3-mm cannulating-type electromagnetic flowmeter connected between the popliteal and jugular veins, and the arteriovenous blood oxygen content difference. The arterial samples were taken from the contralateral femoral artery and the venous samples were taken from the popliteal vein cannula via a small catheter threaded through the wall of the venous outflow tubing. The blood samples, 0.6 ml, were taken into tuberculin sy-
Downloaded from www.physiology.org/journal/ajplegacy by ${individualUser.givenNames} ${individualUser.surname} (129.081.226.078) on December 18, 2018.
W. N. STAINSBY
1014
ringes, capped, and kept in ice until analyzed for oxygen content. Oxygen content of the blood samples was determined with a Lex 0, Con oxygen analyzer. About 100 duplicate analyses were compared with the traditional manometric technique and the results indicated that in our hands the method agrees very well with the manometric method as reported previously (6, 10). Ninetyfive percent of the time duplicate analyses were well within 0.5 vol/lOO ml of each other. The error range was essentially constant over the range from 1.0 to 25.0 vol/ 100 ml. The flowmeter was calibrated at the end of each experiment with each animal’s own blood. The 95% confidence limits of the flow were determined to be k 1.0 ml/min over the flow range observed. Oxygen uptake was calculated as the product of each arteriovenous pair and their blood flow. The muscle and tendons were freed from surrounding tissue. The tendon was fixed to a small aluminum clamp that was connected in turn to the pneumatic myograph (8, 14). The pneumatic myograph was modified by the addition of a large-bore electrically operated valve between the bellows of the lever and the air-spring bottle. The valve was activated by a stimulator arranged so the valve was opened once each second for about 200 ms. This stimulator also triggered a second stimulator after a lo- to 50-ms delay. The second stimulator stimulated the muscle to contract. Figure 1 is a reproduction of recordings of the events Panel A is a normal isotonic conduring contractions. traction with the air valve. At the beginning, the air pressure is zero and the muscle is at rest length, L,. Then the air pressure increases suddenly and the muscle is stretched until it hits the afterload screw, which had been preset to give an initial length, Li, that was less than optimal length, L,,, for isometric contractions. The tension of the muscle is the preload. The remainder of the tension is supported by the afterload screw. The muscle is stimulated and contracts. Tension rises until the muscle tension is equal to the force of the bellows. After this the muscle shortens against the force from a to b and then relaxes back to Li. Subsequently, the tension falls to zero when the air valve empties the bellows. Panel B shows the response to a force slightly w _ a - -. w greater than isometric force, PO. First, tension rises and B
FIG.
1. Reconstructed
shortening stretching
contraction contraction
n
C
records of force and length during a (A), a stretching contraction (B), and a faster (C).
the muscle begins to stretch. The muscle is stimulated. The delay has been adjusted so the contraction is in the same length range as paneZ A. The muscle contraction merely slows the rate of stretch between a and b. In paneZ C, C, a greater force causes more rapid stretching. The force rise time in C is not quite fast enough for a perfect plateau during the contraction. When this occurred the tension in the middle of the contraction, between a and b, was used for the calculation of work and force-velocity relation. The velocity of the shortening was the maximal velocity early in the positive work contractions. The velocity of lengthening was calculated from the slope in the middle of the S-shaped portion, a6, Fig. 1, of the lengthening contractions. The AL for the positive work calculations was measured as usual, Li minus the minimal length during the contractions, a-b, Fig. lA. This AL times the load in gram per gram of muscle was used to give the work per contraction in gram-centimeters per gram of muscle. In the lengthening contractions the AL from the beginning to the end of the straight part, a-b, of the S curve, Fig. 1, B or C, was used for calculation of work. This was called negative work as opposed to positive work for the usual shortening contractions. The negative work is actually the work done on the muscle by the lever during the contraction phase of the lengthening. The order of events was exactly as done previously (13, 14). The preparation was assembled, isometric L4, was determined, and Li was set 6 mm less than L,,. After a 5-min rest period, blood flow was measured and two arterial and two venous blood samples, 1 min apart, *were taken during continued resting conditions. Next, the contractions at 1 twitch/s were started. The four load conditions were applied in random order. After each was applied by adjusting the bellows air pressure, the delay was adjusted to place the contractions in the same length range, and then conditions were allowed to reach a steady state of 5-7 min. After this period two arterial and two venous blood samples were taken, 1 min apart, the load condition was changed, and so on. After the four load conditions, the contractions were stopped, the flowmeter was calibrated, and the length and tension components were calibrated. The muscle was removed, cleaned of visible fat and connective tissue, and weighed. The longest length with no stretch was measured, L,. Average L, was about 10 cm. The two serial oxygen uptake measurements were averaged for each condition. Average oxygen uptake at rest was subtracted from the average oxygen uptake during each contraction condition to yield the net oxygen uptake for the contractions. This in turn was divided by the weight of the muscle and the number of contractions each minute, 60, and the decimal was adjusted to give Vo, as microliters per gram contraction. The velocity was calculated in centimeters per second. Of the four loads used, one was light, but greater than the preload necessary to reach Li (Fig. IA); one was isometric; one was adequate to give a low-velocity stretch (Fig. U?); and one gave a rapid stretch (Fig. 10. 1C). The loads are expressed as P/P,,, with P,, the isometric tension in that experimental preparation.
Downloaded from www.physiology.org/journal/ajplegacy by ${individualUser.givenNames} ${individualUser.surname} (129.081.226.078) on December 18, 2018.
0, FOR
NEGATIVE
1015
WORK
I5
RESULTS
The basic background data of blood flow and blood pressure for these experiments were in the same range as previous studies. Oxygen uptake of the muscle at rest averaged 5.6 $/g min and ranged from 2.5 to 8.7 pi/g min, the same as reported previously (13, 15). In a pilot series, muscle was stretched from L, to the stop at 0.94 Li/L, once each second as done in the stretching contractions. However, in this series, the muscle was not stimulated. Resting flow and oxygen uptake were not altered by this intermittent passive stretch. The net oxygen uptake in microliters per gram contraction versus load in six experiments is presented in Fig. 2. The points in each experiment are presented as numbers. The points in each experiment are connected by straight lines. As seen before (13), oxygen uptake rises with load up to P/P,, = 1.0. Beyond P/P,, = 1.0, oxygen uptake falls with increasing load. The average arteriovenous 0, difference during contractions was about 10 vol/lOO ml and the average blood flow was about 20 ml/min. The statistical analysis utilized an available computer program (5). It was determined that a quadratic model relating oxygen uptake to the relative load, P/P,), was the best fit for the data. The equation for the quadratic model is:
Vel cm/set IO
5
C
-5
-10
-15 0
Oij (P/PO)= /A + CY~+ B, (P/P,,) + B, (P/P,,)” + Eij Oij (P/P,,) = oxygen uptake for a contraction of the jth measurement of dog i at a given load, P/P,, = overall mean oxygen uptake P = effect of the ith dog ai = fractional load P/P = error of jth measurement of ith dog Eij () The estimates
of the parameters
8, = 2.534; I3z= -1.158;
are:
t = 4.71;
P = < 0.0001 P = < 0.0001
t = -5.18;
and maximal oxygen uptake occurs at approximately P/ P,, = 1.09. Statistically, the parameters are highly significant. The model fits the data very well and accounts for 93% of the variation in oxygen uptake. 1.5 vo* yL/g
1
cont.
1.z .9- ’ .6-
0;
Load 0
.5
1.0
1.5
p/p0
2.0
I
2.5
FIG. 2. Oxygen uptake-load relationships observed in 6 experiments for shortening and stretching contractions. Numbers denote individual experiments.
0.5
IO .
1.5
2.0
2.5
3. Velocity-force relationships for shortening and stretching contractions. Symbols indicate 3 runs on the same muscle. Numbers denote experiments in Fig. 2, where 0, uptake was measured. FIG.
A force-velocity curve from three runs on a single preparation, which was otherwise the same but many points were obtained, is presented in Fig. 3. The first run (circles) began with a light load and progressed in steps to 2.25 x P,,. The second run (triangles) began with 2.4 x P,, and progressed in steps to minimal load. The third run (crosses) is a repeat of the first run. The shortening part of the curve is like that recorded previously using higher fidelity recording techniques (3). The force-velocity data from the 0, uptake experiments are numbered as in Fig. 2 and fall along with the points from the single experiment in which many points were obtained. The displacement to the left at lengthening velocities of experiments 3, 4, and 8 is probably due to their short Li, which produced P,, values of only 140 g/g whereas the other experiments averaged 225 g/g (see 10 for similar displacement). All the experiments were adjusted to Li 6 mm less than L,, Li/L, = 0.94, but the length-tension of this muscle is very steep (13). As a result, a few experiments developed low P,, in spite of the effort to fix Li/Lb. The calculation of the work done by the lever on the muscle during the stretching is not as simple as in the shortening contractions. The problem is one of timing and definition. To illustrate the situation, Fig. 4 was made. Figure 4 is a reconstruction plot of the instantaneous length-tension relations that occurred during the shortening and lengthening contractions illustrated in Fig. 1. Panel 4A illustrates a contraction length and tension plot for a regular shortening contraction such as
Downloaded from www.physiology.org/journal/ajplegacy by ${individualUser.givenNames} ${individualUser.surname} (129.081.226.078) on December 18, 2018.
W. N. STAINSBY
1016 IO.00 Lcm
A
IO.25 IO50 10.75 II.00 1 0
q/q
Lood 1 I
50
I
IO0
0
FIG. 4. Length-force plots of contractions 1: external positive work. Area 2: external extra negative work of relaxing muscle.
1
v
150
300
,
450
A and B in Fig. 1. Area negative work. Area 3:
in Fig. lA. With the pneumatic lever, the muscle starts at 10 cm length, L,. When the air pressure rises, the muscle stretches to Li with a small rest tension. When stimulated, the tension rises along the horizontal line with the arrow to a, where it equals the bellows force. Then the muscle shortens along the vertical arrow to minimal length, b. The muscle then relaxes down the vertical line to a and then the tension falls along the horizontal line to Li. When the bellows pressure is reduced to atmospheric pressure, the muscle shortens back to 10.0 cm L,. This cycle was repeated with each contraction. The positive work half of this cycle is the work done by the muscle and it is equal to area 1. In the nega* tive work contraction, 4B, the cycle begins with the muscle at L,. When the valve opens, the tension starts to rise and the muscle begins to stretch. Before the muscle stretches to Li, the muscle is stimulated. The velocity of lengthening is slowed while tension rises rapidly to a in Fig. 4.E and remains reduced, following the arrow to b. After b. the velocitv of stretching increases rapidly until the ‘lever hits the stop. In terms of time since stimulation, a in the lengthening contraction is comparable to a in the shortening contraction. The point b is comparable to b in the shortening contraction. The work absorbed by the muscle while it is trying to contract is area 2 of Fig. 4B. After b, the muscle continues absorbing more work from the lever until it hits the afterload screw. The additional work done on the muscle after its activity declines, after b, is area 3 of Fig. 4B. Presumably, area 3 is irrelevant to the energetics of the contraction as is the equivalent in the relaxation portion of a shortening contraction. When only area 2 was evaluated from the a-b AL and the tension, it was noted that the muscle could absorb more work in a stretching contraction than the work it could do in a shortening contraction. The ratio of the energy equivalent of the work absorbed to the energy equivalent of the oxygen uptake ranged from zero under isometric conditions to 8.0 under the most rapid stretch conditions. DISCUSSION
The data indicate that for a given stimulus oxygen uptake is decreased from the isometric level when the load applied to a muscle becomes large enough to force the muscle to lengthen when it is trying to contract. A
question that might be asked is whether the decrease is caused by reduced flow due to stretching. There is evidence to suggest reduced flow is not the cause because flow and arteriovenous blood oxygen content differences were both decreased at the very high loads. This is a characteristic response of these preparations to a decrease in oxygen use. A reduction in oxygen uptake due to decreased flow is always associated with a large increase in arteriovenous blood oxygen content difference (15). In addition, the muscle was observed to show stable continuous contractions for at least 20 min, which could not have occurred if a significant nonsteady state existed due to forced decrease in oxygen uptake related to insufficient flow. Creatine phosphate (CP) breakdown has been studied for stretching contractions (6). Whereas the present study shows decreasing oxygen uptake with increasing velocities of stretch, this earlier study shows a sharp decrease in CP breakdown with slow stretch but increasing CP breakdown from that level with increasing velocity of stretch. I know of no explanation for this difference. The present study’s decrease in oxygen uptake would fit more closely predictions based on the decreased heat production in muscle stretched while studies, trying to contract (1, 9). In the heat-production the total heat production, work done on the muscle plus the heat production by the muscle trying to contract, progressively approached values equal to only the work done on the muscle as velocity of stretch increased. This suggested that the heat production by the muscle in trying to contract decreased with more rapid stretching and it is this heat production that should be proportional to oxygen uptake, which also decreased as velocity of stretching increased. The force-velocity relationship for shortening and stretching contractions seems to be similar in shape to that reported for heart and frog muscle (11, 12) and to that estimated for intact human muscles (2, 4). The actual velocities differ among the different muscles, but the basic relationship is similar. The cause of the shape of the force-velocity curve and the decrease in oxygen use lies in the force, number, and turnover of crossbridges in the sliding-filament system, which is beyond the realm of th .is study. However, it appears to m .e that inversely with velocOXY gen uptake 1s simply ’ changing ity regardless of the direction of movement. The observation that the oxygen uptake is approximately the same for similar velocities of shortening and stretching clarifies the reasons oxygen uptake is lower for negative than for positive work conditions in exercising humans. It has been suggested previously (2, 4) that the force-velocity curve makes it possible to develop the same force for negative velocity contractions as for positive velocity contractions with many fewer motor units stimulated and at a lower frequency, and it has been shown that the number of impulses delivered to a muscle in a contraction is the major determinant of oxygen uptake for the contraction (8). Therefore, negative work exercise uses less oxygen because there are fewer stimuli necessary for negative work than for the same level of positive work. The fact that the muscles can absorb up to 8 times as
Downloaded from www.physiology.org/journal/ajplegacy by ${individualUser.givenNames} ${individualUser.surname} (129.081.226.078) on December 18, 2018.
0, FOR
NEGATIVE
1017
WORK
much work energy as the energy they use to activate the system to absorb the work energy generates a problem in terminology. It is possible to consider the ratio of the negative W ork done to the energy required to activate the muscl .e to do that work. The ratio being well over 1.0 implies an efficiency of over 100%. A work engine cannot do work with an efficiency over lOO%, so the model is not correct. In reality, when doing negative work the muscle acts as though it were a viscoelastic brake. It is usual for brakes to be able to absorb many times more energy than is needed to activate the brake. For brakes the ratio is called the coefficient of performance. The maximal coefficient of muscle performance observed in these experiments was about 8.0. As brakes go, this is not unusual. That a muscle can do this is exciting. The ability to do this must be translatable into physical chemistry terms with regard to myofilament crossbridges.
The effect of the high coefficient of muscle performante on energy need can be seen in the high coefficient of performance of negative work in exercise studies on man. In an exemplary ’ study (4) th .e apparent net coefficient of performance from the that can be calculated data for positive work on a bicycle is 23%. The same calculation for negative work gives 166%. It would appear that the body as a whole is able to realize a significant energetic advantage when performing negative work. The author thanks Dr. S. B. Tricky of the Department of Physics and Dr. E. A. Farber of the Department of Mechanical Engineering for their advice and comments on the methods and on the interpretation of the data and Dr. Rose Ray of the Division of Biostatistics of the Department of Statistics for her help in the statistical analysis. This study was supported by National Institutes of Health Grant HL 14806-15. Received
for publication
21 April
1975.
REFERENCES 1. ABBOTT, B. C., X. M. AUBERT, AND A. V. HILL. The absorption of work by a muscle stretched during a single twitch or a short tetanus. Proc. Roy. Sot., London, Ser. B 139: 86-104, 1951. 2. ABBOTT, B. C., B. BIGLAND, AND J. M. RITCHIE. The physiological cost of negative work. J. Physiol., London 117: 380-390, 1952. 3. ALLEN, P. D. The Mechanical Properties of in Situ Dog Skeletal MuscZe (PhD). Gainesville, Fla.: University of Florida, August 1973. 4. ASMUSSEN, E. Positive and negative muscular work. Acta Physiol. Stand. 28: 364-382, 1953. 5. BARR, A. J., AND J. H. GOODNIGHT. Statistical Analysis System. Raleigh, N. C.: North Carolina State University Dept. of Statistics, August 1972. 6. CHAPLER, C. K., AND J. KATRUSIAK. Blood flow and carbohydrate metabolism in dog skeletal muscle in situ following propranolol. Med. Sci. Sports 6: 193-197, 1974. 7. CURTIN, N. A., AND R. E. DAVIES. Chemical and mechanical changes during stretching of activated frog skeletal muscle. CoZd Spring Harbor Symp. Quant. BioZ. 37: 619-626, 1973. 8. FALES, J. T., S. R. HEISEY, AND K. L. ZIERLER. Dependency of oxygen consumption of skeletal muscle on numbers of stimuli during work in the dog. Am. J. PhysioZ. 198: 1333-1342, 1960.
A. V., AND J. V. HOWARTH. The reversal of chemical 9. HILL, reactions in contracting muscle during an applied stretch. Proc. Roy. Sot., London, Ser. B 151: 169-193, 1959. 10. KUSHIMA, F., W. C. BUTTS, AND W. L. RUFF. Superior analytical performance by electrolytic cell analysis of blood oxygen content. J. Appl. Physiol. 35: 299-300, 1973. 11. MASHIMA, H., K. AKAZAWA, H. KUSHIMA, AND K. FUGI. The force load velocity relation and the viscous-like force in the frog skeletal muscle. Japan. J. PhysioZ. 22: 103-120, 1972. 12. MASHIMA, H., AND H. KUSHIMA. Determination of the active state by the graphical experiment and instantaneous methods in the frog ventricle. Japan. Heart J. 12: 545-561, 1971. 13. STAINSBY, W. N. Oxygen uptake for isotonic and isometric twitch contractions of dog skeletal muscle in situ. Am. J. Physiol. 219: 435-439, 1970. 14. STAINSBY, W. N., AND 3. K. BARCLAY. Oxygen uptake for brief tetanic contractions of dog skeletal muscle in situ. Am. J. Physiol. 223: 371-375, 1972. 15. STAINSBY, W. N., J. T. FALES, AND J. L. LILIENTHAL, JR. Effect of stretch on oxygen consumption of dog skeletal muscle in situ. Bull. Johns Hopkins Hosp. 99: 249-261, 1956.
Downloaded from www.physiology.org/journal/ajplegacy by ${individualUser.givenNames} ${individualUser.surname} (129.081.226.078) on December 18, 2018.
