Cardiovascular Research 1992;26:865-810
865
Myocardial stretch alters twitch characteristics and Ca2' loading of sarcoplasmic reticulum in rat ventricular muscle James Gamble, Paul Brian Taylor, and Kenji Alan Kenno Objective: The aim was to determine the influence of diastolic muscle length on force development and timing parameters of cardiac muscle twitch contraction and to determine whether a length dependency exists for the calcium loading capacity of the sarcoplasmic reticulum. Methods: Right ventricular papillary muscles and trabeculae were isolated from hearts of female Wistar rats weighing 220-280 g. Papillary muscles were stretched to diastolic lengths of 90, 95, and 100% L,, and paced at 1 .O Hz. Individual twitch profiles were characterised by their peak force and the maximum rate (dF/dt) of the positive and negative force changes. Intrinsic timing was identified through waveform analysis that divided the twitch profile into time domains for the ascending limb (TO-TI; Tl-T2) and the descending limb (T2-T3; T3-T4). Each domain was compared at three muscle lengths. The sarcoplasmic reticular calcium content at short (1.88 p m ) and long (2.11 p,m) sarcomere lengths was characterised by rapid cooling contractures after 1 s and 60 s of diastolic rest. Results: Peak developed force and the maximum rate of positive and negative force development decreased as diastolic muscle length was reduced from L,,, to 90% LmW.The intrinsic timing for the segment that reflects the relaxation phase of the twitch (Tl-T4) was shortened as muscle length was reduced. The time domain that reflects the combined effects of calcium release and the early phase of contraction (TO-TI) was insensitive to diastolic muscle length. The fractional release of sarcoplasmic reticular calcium at different muscle lengths was approximately 32-35% of the total sarcoplasmic reticulum calcium pool. Conclusions: The data on the intrinsic timing of the twitch characteristics coupled with rapid cooling contracture analysis suggests a fractional calcium release that is approximately 32-35% of the total sarcoplasmic reticular capacity at either long or short muscle lengths. However, the loading capacity of the sarcoplasmic reticulum is greater when the muscle operates at a shorter diastolic length. This can be interpreted as meaning that diastolic muscle length differentially influences sarcoplasmic reticular calcium storage and release processes. Cardiovascular Research 1992;26:865-870
T
he Frank-Starling mechanism is one of the most fundamental properties of cardiac muscle for exerting intrinsic control over contractile force on a beat to beat basis. The steep relationship between resting muscle length and the degree of developed force can be shifted to the left using various inotropic interventions.'4 This response has been used to argue that the classical lengthtension relation cannot solely be accounted for by the degree of overlap between the thick and thin filaments and that the activation dependent processes must also contribute to the length-tension relation. At steady state levels of force development there is a well established positive association between the intracellular free calcium concentration and the magnitude of peak force.5 In addition to the peak of the twitch response, the shape of the contraction curve can provide further information about the time components as force evolves during the contraction and decays during the relaxation phase of the twitch. The ascending limb of the contraction curve represents the response of the contractile proteins to the gain in myoplasmic calcium. The descending portion reflects the removal of calcium and the decay in contractile force.
Inferential information about these events is contained within the time components of the contraction curve. However, the lack of information about changes in the intrinsic timing of the force curve at one or more muscle lengths probably reflects the difficulty in accurately analysing the non-linear (tail) portions of the curve. In the rat, the calcium within the sarcoplasmic reticulum is the primary source of activator calcium, as reflected by its high sensitivity to ryanodine treatment' and the large responses to rapid cooling contractures.8 In fact, the total sarcoplasmic reticular calcium capacity exceeds that used during electrical stimulation,"-'* implying that a significant reserve remains in the sarcoplasmic reticulum. The possibility exists that muscle length may not only alter the intrinsic timing during a contraction but also modulate the calcium loading capacity of the sarcoplasmic reticulum. Accordingly, it was the purpose of the present study to examine the influence of muscle length on a number of force and timing parameters used to characterise the muscle twitch and to assess the sarcoplasmic reticular calcium capacity at different sarcomere lengths using the rapid cooling contracture technique.
University of Windsor, Windsor, Ontario, Canada N9B 3P4 - Department of Kinesiology: J Gamble; Department of Biological Sciences: P B Taylor, K A Kenno. Correspondence to Dr Taylor.
866
Gamble, Taylor; Kenno
Methods Animal model Female Wistar rats (190-220 g) (Charles River, Quebec, Canada) were housed in a temperature (23°C) and light controlled (12: 12 h light-dark cycle) animal facility accordng to the guidelines set out by the Canadian Council on Animal Care. Water and Purina lab chow were provided for the animals ad libitum. Muscle preparation The animals were ether anaesthetised and injected with 107 units of heparin (Sigma) per kg body weight via the inferior vena cava to prevent blood clotting. The hearts were then rapidly excised, placed in ice cold saline, and transferred to a dissection chamber. The aprta was cannulated and then perfused at a flow rate of 10 mlmin- with a modified Krebs-Henseleit bicarbonate buffer of the following composition (in mmol.litre-'): NaCl 117.1, KCI 5.0, MgC12.6HzO I .2, Na2S04, 2.4, NaHzP04.2H20 2.0, NaHCOl 27.0, glucose 10, CaClz 1.O. This perfusate was oxygenated at room temperature with 95% 0 ~ 5 % COz to obtain a pH of 7.4. Following a washout period (1-2 min), the heart was qrrested by increasing the potassium concentration to 15 mmol4tre-. Using a dissecting microscope, the right ventricle was opened and the free wall carefully pulled back to expose the papillarj muscles and trabeculae according to the method of ter Keurs et al. Only long thin muscles with uniform sides and no branches were selected for study. The muscles were removed and transferred to a muscle chamber secured on the stage of a dissecting microscope and perfused with the modified Krebs-Henseleit buffer at 2621°C as measured by a thermocouple placed next to the muscle within the bath. The muscles, with a portion of the ventricular wall, were mounted horizontally in a cradle mounted on a micromanipulator (Stoelting, Chicago, USA). The opposite end, with an attached portion of the triscuspid valve, was pulled through the cradle and placed over a tungsten wire hook which was attached to the arm of a force transducer (Grass FT03 for papillary muscles; SensoNor AE801 for trabeculae). The muscles were stimulated (Grass S6 C) at 50% above threshold via two platinum wire electrodes located parallel to the long axis and on either side of the muscle. A viable preparation was characterised by a stable baseline force development, with no aftercontractions and no evidence of spontaneous contraction waves. Those muscles exhibiting any of these characteristics were discarded. A moderate load (approximately 5-10% of peak force) was established and the muscle was allowed to equilibrate for 60 min at a pacing frequency of 1.0 Hz. Data recording Force signals generated by the transducer were amplified by a Grass preamplifier (model 7P122D). The signal was then converted to digital form using a 12 bit analogue-digital (A-D) converter at a rate of 4 kHz. Each twitch contraction curve had a minimum of 250 data points. The data were stored on floppy disks and used for subsequent analysis. Twitch characterisation Following equilibration, the papillary muscles were paced at 0.2 Hz and the length was increased gradually to establish L, (ie, the muscle length that elicits the greatest active force). Approximately 16 consecutive function curves in 1.0 mM Ca" were collected at 100, 95, and 90% Lmm.At each length step, the muscle equilibrated for approximately 30 min before data collection. Each data set was signal averaged and analysed for peak developed force and maximum positive and negative rates of force change. In addition, twitch curves were partitioned into four time domains using a waveform analysis program. For this study each contraction curve was divided into (1) the time from the initiation of force development to the point of maximum rate of force change (TO-TI); (2) the interval from the maximum rate of force development to peak tension (TI-T2); (3) the time segment from the point of peak force to the maximum rate of relaxation (T2-T3); (4) the interval between the maximum rate of force decay to complete relaxation (T3-T4). Traditionally, studies have avoided measuring the time to complete relaxation (T2-T4) because of the difficulties in its precise determination. We have developed a computer based arc-cosine transformation procedure which allows precise identification of all parameters used in the present study.14 In brief, this transformation exploits some special properties of the arccosine transformation in which the areas around the maximum (T2) and the minimum (T4) are expanded so that the region of complete relaxation can be more clearly identified. Figure 1 displays a general twitch function curve and defines the various force and timing parameters used to characterise each function. I
Rapid cooling wntractures Rapid cooling contractures described by Bridge" and Ben9 were used to obtain an index of total Ca2' releasable from sarcoplasmic reticulum. Trabeculae were used because of their small cross sectional area which facilitates the process of cooling the core of the tissue. The muscle
iF/dt
f d
I I
I
TO
I I
T1
I I
T2
I I
T3
-7-
I T4
Figure 1 Trace of an actual twitch dejining the nutnber of force and timing parameters used in the chiiracterisation of the twitch function curve. chamber was designed and constructed in the laboratory to permit rapid buffer exchange (-28 m1.min-I) while eliminating any residual pooling. Solenoid valves (Cole parmer) controlled the flow of either cold or warm buffer (0.5 mM Ca-') to the muscle chamber. The exchange of warm buffer (26°C) to cold (-2°C) was typically completed in less than 2 s as measured by a thermocouple at the surface of the muscle. Sarcomere distance was set at an average of 2.1 and 1.8 pm using a video image technique; these values correspond to L,,,,, and approximately 85% of L,, respectively. Briefly, the image of the muscle and sarcomeres was magnified with a 40X long working distance objective. The image was then captured by a video camera (Panasonic WV-1410) and displayed on a monitor (Panasonic TR-930) which was calibrated for direct measurement of sarcomere length. During rapid cooling the force signal was sampled and displayed in real time (20 Hz) using a powerful data acquisitiodcontrol software package (Unkel Software, Lexington, MA, USA). For these experiments, the peak force of the electrically stimulated contraction immediately prior to the cooling and the peak force of the cold induced contracture were analysed from digital records.
Statistical analvsis An analysis of variance (ANOVA) with repeated measures was used in conjunction with a Tukey's HSD post hoc analysis or paired t tests where appropriate to determine statistically significant differences. The level of significance was set at p
865
Myocardial stretch alters twitch characteristics and Ca2' loading of sarcoplasmic reticulum in rat ventricular muscle James Gamble, Paul Brian Taylor, and Kenji Alan Kenno Objective: The aim was to determine the influence of diastolic muscle length on force development and timing parameters of cardiac muscle twitch contraction and to determine whether a length dependency exists for the calcium loading capacity of the sarcoplasmic reticulum. Methods: Right ventricular papillary muscles and trabeculae were isolated from hearts of female Wistar rats weighing 220-280 g. Papillary muscles were stretched to diastolic lengths of 90, 95, and 100% L,, and paced at 1 .O Hz. Individual twitch profiles were characterised by their peak force and the maximum rate (dF/dt) of the positive and negative force changes. Intrinsic timing was identified through waveform analysis that divided the twitch profile into time domains for the ascending limb (TO-TI; Tl-T2) and the descending limb (T2-T3; T3-T4). Each domain was compared at three muscle lengths. The sarcoplasmic reticular calcium content at short (1.88 p m ) and long (2.11 p,m) sarcomere lengths was characterised by rapid cooling contractures after 1 s and 60 s of diastolic rest. Results: Peak developed force and the maximum rate of positive and negative force development decreased as diastolic muscle length was reduced from L,,, to 90% LmW.The intrinsic timing for the segment that reflects the relaxation phase of the twitch (Tl-T4) was shortened as muscle length was reduced. The time domain that reflects the combined effects of calcium release and the early phase of contraction (TO-TI) was insensitive to diastolic muscle length. The fractional release of sarcoplasmic reticular calcium at different muscle lengths was approximately 32-35% of the total sarcoplasmic reticulum calcium pool. Conclusions: The data on the intrinsic timing of the twitch characteristics coupled with rapid cooling contracture analysis suggests a fractional calcium release that is approximately 32-35% of the total sarcoplasmic reticular capacity at either long or short muscle lengths. However, the loading capacity of the sarcoplasmic reticulum is greater when the muscle operates at a shorter diastolic length. This can be interpreted as meaning that diastolic muscle length differentially influences sarcoplasmic reticular calcium storage and release processes. Cardiovascular Research 1992;26:865-870
T
he Frank-Starling mechanism is one of the most fundamental properties of cardiac muscle for exerting intrinsic control over contractile force on a beat to beat basis. The steep relationship between resting muscle length and the degree of developed force can be shifted to the left using various inotropic interventions.'4 This response has been used to argue that the classical lengthtension relation cannot solely be accounted for by the degree of overlap between the thick and thin filaments and that the activation dependent processes must also contribute to the length-tension relation. At steady state levels of force development there is a well established positive association between the intracellular free calcium concentration and the magnitude of peak force.5 In addition to the peak of the twitch response, the shape of the contraction curve can provide further information about the time components as force evolves during the contraction and decays during the relaxation phase of the twitch. The ascending limb of the contraction curve represents the response of the contractile proteins to the gain in myoplasmic calcium. The descending portion reflects the removal of calcium and the decay in contractile force.
Inferential information about these events is contained within the time components of the contraction curve. However, the lack of information about changes in the intrinsic timing of the force curve at one or more muscle lengths probably reflects the difficulty in accurately analysing the non-linear (tail) portions of the curve. In the rat, the calcium within the sarcoplasmic reticulum is the primary source of activator calcium, as reflected by its high sensitivity to ryanodine treatment' and the large responses to rapid cooling contractures.8 In fact, the total sarcoplasmic reticular calcium capacity exceeds that used during electrical stimulation,"-'* implying that a significant reserve remains in the sarcoplasmic reticulum. The possibility exists that muscle length may not only alter the intrinsic timing during a contraction but also modulate the calcium loading capacity of the sarcoplasmic reticulum. Accordingly, it was the purpose of the present study to examine the influence of muscle length on a number of force and timing parameters used to characterise the muscle twitch and to assess the sarcoplasmic reticular calcium capacity at different sarcomere lengths using the rapid cooling contracture technique.
University of Windsor, Windsor, Ontario, Canada N9B 3P4 - Department of Kinesiology: J Gamble; Department of Biological Sciences: P B Taylor, K A Kenno. Correspondence to Dr Taylor.
866
Gamble, Taylor; Kenno
Methods Animal model Female Wistar rats (190-220 g) (Charles River, Quebec, Canada) were housed in a temperature (23°C) and light controlled (12: 12 h light-dark cycle) animal facility accordng to the guidelines set out by the Canadian Council on Animal Care. Water and Purina lab chow were provided for the animals ad libitum. Muscle preparation The animals were ether anaesthetised and injected with 107 units of heparin (Sigma) per kg body weight via the inferior vena cava to prevent blood clotting. The hearts were then rapidly excised, placed in ice cold saline, and transferred to a dissection chamber. The aprta was cannulated and then perfused at a flow rate of 10 mlmin- with a modified Krebs-Henseleit bicarbonate buffer of the following composition (in mmol.litre-'): NaCl 117.1, KCI 5.0, MgC12.6HzO I .2, Na2S04, 2.4, NaHzP04.2H20 2.0, NaHCOl 27.0, glucose 10, CaClz 1.O. This perfusate was oxygenated at room temperature with 95% 0 ~ 5 % COz to obtain a pH of 7.4. Following a washout period (1-2 min), the heart was qrrested by increasing the potassium concentration to 15 mmol4tre-. Using a dissecting microscope, the right ventricle was opened and the free wall carefully pulled back to expose the papillarj muscles and trabeculae according to the method of ter Keurs et al. Only long thin muscles with uniform sides and no branches were selected for study. The muscles were removed and transferred to a muscle chamber secured on the stage of a dissecting microscope and perfused with the modified Krebs-Henseleit buffer at 2621°C as measured by a thermocouple placed next to the muscle within the bath. The muscles, with a portion of the ventricular wall, were mounted horizontally in a cradle mounted on a micromanipulator (Stoelting, Chicago, USA). The opposite end, with an attached portion of the triscuspid valve, was pulled through the cradle and placed over a tungsten wire hook which was attached to the arm of a force transducer (Grass FT03 for papillary muscles; SensoNor AE801 for trabeculae). The muscles were stimulated (Grass S6 C) at 50% above threshold via two platinum wire electrodes located parallel to the long axis and on either side of the muscle. A viable preparation was characterised by a stable baseline force development, with no aftercontractions and no evidence of spontaneous contraction waves. Those muscles exhibiting any of these characteristics were discarded. A moderate load (approximately 5-10% of peak force) was established and the muscle was allowed to equilibrate for 60 min at a pacing frequency of 1.0 Hz. Data recording Force signals generated by the transducer were amplified by a Grass preamplifier (model 7P122D). The signal was then converted to digital form using a 12 bit analogue-digital (A-D) converter at a rate of 4 kHz. Each twitch contraction curve had a minimum of 250 data points. The data were stored on floppy disks and used for subsequent analysis. Twitch characterisation Following equilibration, the papillary muscles were paced at 0.2 Hz and the length was increased gradually to establish L, (ie, the muscle length that elicits the greatest active force). Approximately 16 consecutive function curves in 1.0 mM Ca" were collected at 100, 95, and 90% Lmm.At each length step, the muscle equilibrated for approximately 30 min before data collection. Each data set was signal averaged and analysed for peak developed force and maximum positive and negative rates of force change. In addition, twitch curves were partitioned into four time domains using a waveform analysis program. For this study each contraction curve was divided into (1) the time from the initiation of force development to the point of maximum rate of force change (TO-TI); (2) the interval from the maximum rate of force development to peak tension (TI-T2); (3) the time segment from the point of peak force to the maximum rate of relaxation (T2-T3); (4) the interval between the maximum rate of force decay to complete relaxation (T3-T4). Traditionally, studies have avoided measuring the time to complete relaxation (T2-T4) because of the difficulties in its precise determination. We have developed a computer based arc-cosine transformation procedure which allows precise identification of all parameters used in the present study.14 In brief, this transformation exploits some special properties of the arccosine transformation in which the areas around the maximum (T2) and the minimum (T4) are expanded so that the region of complete relaxation can be more clearly identified. Figure 1 displays a general twitch function curve and defines the various force and timing parameters used to characterise each function. I
Rapid cooling wntractures Rapid cooling contractures described by Bridge" and Ben9 were used to obtain an index of total Ca2' releasable from sarcoplasmic reticulum. Trabeculae were used because of their small cross sectional area which facilitates the process of cooling the core of the tissue. The muscle
iF/dt
f d
I I
I
TO
I I
T1
I I
T2
I I
T3
-7-
I T4
Figure 1 Trace of an actual twitch dejining the nutnber of force and timing parameters used in the chiiracterisation of the twitch function curve. chamber was designed and constructed in the laboratory to permit rapid buffer exchange (-28 m1.min-I) while eliminating any residual pooling. Solenoid valves (Cole parmer) controlled the flow of either cold or warm buffer (0.5 mM Ca-') to the muscle chamber. The exchange of warm buffer (26°C) to cold (-2°C) was typically completed in less than 2 s as measured by a thermocouple at the surface of the muscle. Sarcomere distance was set at an average of 2.1 and 1.8 pm using a video image technique; these values correspond to L,,,,, and approximately 85% of L,, respectively. Briefly, the image of the muscle and sarcomeres was magnified with a 40X long working distance objective. The image was then captured by a video camera (Panasonic WV-1410) and displayed on a monitor (Panasonic TR-930) which was calibrated for direct measurement of sarcomere length. During rapid cooling the force signal was sampled and displayed in real time (20 Hz) using a powerful data acquisitiodcontrol software package (Unkel Software, Lexington, MA, USA). For these experiments, the peak force of the electrically stimulated contraction immediately prior to the cooling and the peak force of the cold induced contracture were analysed from digital records.
Statistical analvsis An analysis of variance (ANOVA) with repeated measures was used in conjunction with a Tukey's HSD post hoc analysis or paired t tests where appropriate to determine statistically significant differences. The level of significance was set at p
