Effect of ADP on slow-twitch muscle fibres of the rat: implications for muscle fatigue

Abstract

Slow-twitch mechanically skinned fibres from rat soleus muscle were bathed in solutions mimicking the myoplasmic environment but containing different [ADP] (0.1 μm to 1.0 mm). The effect of ADP on sarcoplasmic reticulum (SR) Ca²⁺-content was determined from the magnitude of caffeine-induced force responses, while temporal changes in SR Ca²⁺-content allowed determination of the effective rates of the SR Ca²⁺-pump and of the SR Ca²⁺-leak. The SR Ca²⁺-pump rate, estimated at pCa (−log₁₀[Ca²⁺]) 7.8, was reduced by 20% as the [ADP] was increased from 0.1 to 40 μm, with no further alteration when the [ADP] was increased to 1.0 mm. The SR Ca²⁺-leak rate constant was not altered by increasing [ADP] from 0.1 to 40 μm, but was increased by 26% when the [ADP] was elevated to 1.0 mm. This ADP-induced SR Ca²⁺-leak was insensitive to ruthenium red but was abolished by 2,5-di(tert-butyl)-1,4-hydroquinone (TBQ), indicating that the leak pathway is via the SR Ca²⁺-pump and not the SR Ca²⁺-release channel. The decrease in SR Ca²⁺-pump rate and SR Ca²⁺-leak rate when [ADP] was increased led to a 40% decrease in SR Ca²⁺-loading capacity. Elevation of [ADP] had only minor direct effects on the contractile apparatus of slow-twitch fibres. These results suggest that ADP has only limited depressing effects on the contractility of slow-twitch muscle fibres. This is in contrast to the marked effects of ADP on force responses in fast-twitch muscle fibres and may contribute to the fatigue-resistant nature of slow-twitch muscle fibres.

The duration of the twitch force response in slow-twitch muscle is approximately four times longer than that of fast-twitch muscle (Close, 1967). This is due partially to differences in contractile properties of the fibre types (Stephenson & Williams, 1981), and also to differences in the myoplasmic Ca²⁺ transient that is responsible for the activation and relaxation of the contractile apparatus (Baylor & Hollingworth, 2003). The contractile apparatus of slow-twitch muscle is more sensitive to Ca²⁺ than fast-twitch muscle, and as a result produces force at lower free Ca²⁺ concentration than fast-twitch muscle (Stephenson & Williams, 1985; Stephenson et al. 1994).

At the level of the SR, there are a number of marked differences between the Ca²⁺-handling properties of slow-twitch and fast-twitch muscle. The overall volume of the SR is two to three times smaller in slow-twitch muscle than in fast-twitch muscle (Eisenberg et al. 1974), whereas the endogenous SR Ca²⁺-content of rat slow-twitch muscle is only slightly larger than that of fast-twitch muscle (Fryer & Stephenson, 1996). Furthermore, the SR Ca²⁺-content in slow-twitch muscle cannot be increased beyond the endogenous level (Fryer & Stephenson, 1996), suggesting that under endogenous conditions the SR is maximally loaded with Ca²⁺. In contrast, in fast-twitch muscle, the SR is endogenously loaded to only 20–30% of the SR capacity (Fryer & Stephenson, 1996).

Expression of SR Ca²⁺-pump isoforms also varies between the two muscle fibre types, with the SERCA 2A isoform expressed in slow-twitch muscle, and the SERCA 1 isoform expressed in fast-twitch muscle (van der Linden et al. 1996). The actual SR Ca²⁺-pump density in slow-twitch muscle is only 30–50% of that in fast-twitch muscle (Ferguson & Franzini-Armstrong, 1988). The content of SR Ca²⁺-release channels is also different between the two muscle fibre types, with slow-twitch muscle containing only 60% of that in fast-twitch muscle (Franzini-Armstrong et al. 1988), although the SR Ca²⁺-release channels are of the same isoform (Meissner, 1994). Furthermore, the charge movement associated with the dihydropyridine receptor is also three- to fourfold smaller in slow-twitch than in fast-twitch muscle fibres (Dulhunty & Gage, 1983), indicating that fewer voltage sensors are activated during excitation. Together, these properties contribute greatly to the smaller and slower Ca²⁺ transients observed between slow-twitch and fast-twitch muscle fibres (Baylor & Hollingworth, 2003) and consequently, to the specific force characteristics of each muscle fibre type.

Slow-twitch muscle is considered to be more fatigue resistant than fast-twitch muscle (Fitts, 1994). However, little is known regarding the possible differences in Ca²⁺-handling properties between slow-twitch and fast-twitch muscles in conditions occurring during fatiguing muscle contractions. Under normal resting conditions, the [ADP] within a muscle is tightly maintained at <10 μm (Dawson et al. 1978), primarily due to the activity of the creatine kinase reaction (ADP + CP ⇌ ATP + creatine, equilibrium constant = 260; Chase & Kushmerick, 1995). However, during fatiguing muscle contraction, creatine phosphate (CP) is consumed and, together with an elevated creatine concentration, displaces the equilibrium of the creatine kinase reaction towards a higher [ADP]. In situations of near complete CP depletion, the [ADP] is thought to reach up to millimolar levels (Nagesser et al. 1993; Westerblad & Lännergren, 1994). Despite the lower resting CP concentration in slow-twitch muscle (Meyer et al. 1985), the increase in [ADP] during fatigue is smaller in slow-twitch muscle than that in fast-twitch muscle. Presumably, this is due to slower myosin ATPase activity (Schiaffino & Reggiani, 1996) and lower SR Ca²⁺-pump ATP utilization (Szentesi et al. 2001), along with the lower creatine concentrations (Meyer et al. 1985).

We have previously shown that an elevation in [ADP] to levels occurring during fatiguing stimulation markedly alters the SR Ca²⁺-handling ability in fast-twitch muscle (Macdonald & Stephenson, 2001). However, the effects of an elevation in [ADP] on the SR Ca²⁺-handling ability in slow-twitch muscle are not known. In this study we used the freshly dissected mechanically skinned muscle fibre preparation, in conjunction with solutions of varied [ADP] to examine the effects of ADP on SR function. In this preparation, the SR is intact and physiological conditions can be accurately mimicked. The results show that an elevation in [ADP] within the physiological range reduces the ability of the SR to store Ca²⁺ by both decreasing the SR Ca²⁺-pump rate, and by increasing the passive SR Ca²⁺-leak. Importantly, the extent of this reduction in SR Ca²⁺-handling ability in slow-twitch muscle is much smaller than that in fast-twitch muscle (Macdonald & Stephenson, 2001), which may contribute to the more fatigue resistant nature of slow-twitch muscle.

Methods

Dissection, preparation of fibres and apparatus

Male rats (Long-Evans; 3 months old) were killed by halothane overdose (2% v/v) in accordance with permits issued by La Trobe University Animal Ethics Committee. Soleus muscles were quickly removed, blotted on filter paper, and then transferred to a dish containing paraffin oil (Macdonald & Stephenson, 2001). Mechanically skinned fibres were then prepared and mounted on a force transducer (AME875; SensoNor, Horton, Norway) whilst under oil, as previously described (Fink et al. 1986). The length and diameter were measured and the preparation then stretched to 120% slack resting length to facilitate measurement of force production (Lamb & Stephenson, 1990). The preparation was finally placed into a 2 ml Perspex bath containing low relaxing₁ solution (Table 1). The apparatus used in these experiments and the procedure of changing solutions has elsewhere been described in detail (Stephenson & Williams, 1981).

Table 1.

Composition of solutions

Solution	EGTAtotal (mm)	HDTA (mm)	CP (mm)	Ca2+ (μm)	Mg2+free (mm)
Max activating1	50	—	10	30	1.0
Max activating2	50	10	—	30	1.0
High relaxing1	50	—	10	<0.001	1.0
High relaxing2	50	10	—	<0.001	1.0
Low relaxing1	0.05	50	10	0.05	1.0
Low relaxing2	0.05	60	—	0.05	1.0
Load1	1.0	50	10	0.015	1.0
Load2	1.0	60	—	0.015	1.0
Release	1.0	50	10	0.005	0.05
Wash1	1.0	50	10	0.002	1.0
Wash2	1.0	60	—	0.002	1.0
Leak1	1.0	50	10	0.002	1.0
Leak2	1.0	60	—	0.002	1.0
pSr 5.4 solution	50	—	10	<0.001	1.0
ADP stock	0.05	30	—	0.05	1.0

All solutions were adjusted to pH 7.10 ± 0.01 at room temperature (22 ± 2°C), with KOH and contained (mm): K⁺, 126; Na⁺, 36; total ATP, 8; Hepes, 90; NaN₃ 1.0. ADP stock solution also contained 20 mm ADP. Total Ca²⁺ and Mg²⁺ in solutions was varied to provide the indicated Ca²⁺_free and Mg²⁺_free concentrations. pSr 5.4 solution contained 4 μm Sr. Note that under our conditions EGTA, CaEGTA, HDTA and CP exist almost exclusively as divalent anions. Release solution also contains 30 mm caffeine. CP, creatine phosphate, HDTA, hexamethylene diamine tetraacetate.

Solutions

Solutions were prepared as described by Stephenson & Williams (1981), with their composition shown in Table 1. The osmolarity of all solutions was 290 ± 10 mosmol kg⁻¹ and unless otherwise stated contained (mm): K⁺, 126; Na⁺, 37; total ATP, 8.0; MgATP, 7.0; Hepes, 90 (pH 7.10 ± 0.01); free Mg²⁺, 1.0; NaN₃, 1.0. To determine the Ca²⁺ sensitivity of the contractile apparatus, solutions were prepared by mixing high relaxing₁ solution and max activating₁ solution to give strongly Ca²⁺-buffered solutions of a Ca²⁺ concentration in the pCa (−log₁₀[Ca²⁺]) range of ∼4.5 to >9. The release solution contained 30 mm caffeine and low ionized Mg²⁺ concentration (0.02 mm) to facilitate the rapid and thorough release of SR Ca²⁺ (Fryer & Stephenson, 1996). The free Ca²⁺ concentration in load solutions was determined with a Ca²⁺ electrode (Orion, Boston, USA). Hexamethylene diamine tetraacetate (HDTA) was obtained from Fluka, Buchs (Switzerland) and all other chemicals were obtained from Sigma (St Louis, MO, USA). All experiments were conducted at room temperature (22 ± 2°C).

[ADP] in solutions

Freshly mechanically skinned fibres retain a high creatine kinase activity (Saks et al. 1978; Ventura-Clapier et al. 1994; Walliman et al. 1997) and therefore, in the presence of CP, exogenously added ADP is rapidly converted to ATP within the myoplasmic space. [ADP] was maintained in solution using two methods: (i) buffering the [ADP] close to equilibrium using the preparations endogenous creatine kinase activity, and (ii) by the concomitant removal of CP and addition of ADP to solutions, as previously described (Macdonald & Stephenson, 2001, 2004).

Measurement of Ca²⁺-activated force

Contractile activation parameters were determined by initially exposing the preparation to high relaxing₁ solution containing 2% v/v Triton X-100 (5 min) to destroy all membrane compartments (Stephenson, 1981), before washing in a separate high relaxing₁ solution (approximately 2 min, two separate washes). The force–pCa relationship and maximum Ca²⁺-activated force were then determined by exposing the fibre to solutions of a sequence of solutions in which the Ca²⁺ concentration was heavily buffered at progressively higher levels as previously described (Stephenson & Williams 1982). To compensate for the small deterioration in force between the initial and final response in max activating₁ solution, responses at each pCa were expressed as a percentage of the interpolated values for the maximum Ca²⁺-activated force (van der Poel & Stephenson, 2002). Force was expressed as a percentage of the corresponding maximum Ca²⁺-activated force for that condition, plotted as a function of pCa and fitted with a Hill equation (described by eqn (1), using the analysis program GraphPad Prism; GraphPad Software, Inc., CA, USA):

(1)

where [Ca²⁺]₅₀ (pCa₅₀) represents the [Ca²⁺] (pCa) at which 50% of maximum Ca²⁺-activated force is produced, and n_H is the Hill coefficient. The oscillatory force response characteristic in mammalian slow-twitch muscle at submaximal Ca²⁺ concentrations (Stephenson & Williams, 1981) was observed in all fibres. The force measurement for oscillatory responses at constant Ca²⁺ concentration was taken as the midpoint between the peak and trough of the oscillation. Increasing the [ADP] did not alter either the amplitude or frequency of the oscillatory response.

Caffeine-induced responses

Caffeine-induced force responses were used to estimate the relative amount of Ca²⁺ in the SR by referring to the relative area under the caffeine-induced force response (Endo & Iino, 1980; Macdonald & Stephenson, 2001). The preparation was initially equilibrated in low relaxing₁ solution (2 min), where the SR remained endogenously loaded, and the endogenous SR Ca²⁺-content determined by placing the preparation in wash₁ solution for 30 s before being transferred to release solution to trigger SR Ca²⁺-release. The fibre remained in the release solution for 2 min to ensure complete SR Ca²⁺-depletion (Fryer & Stephenson, 1996), before washing for 30 s in wash₁ solution. Thereafter, the preparation was placed in load₁ solution to reload the SR with Ca²⁺ before the SR Ca²⁺ was again released in release solution, and the cycle then repeated. Force was continuously recorded on a chart recorder and the relative area under the caffeine-induced force response was measured using the gravimetric method (Fink & Stephenson, 1987).

The areas under the caffeine-induced force responses could be used to estimate SR Ca²⁺-content providing that the areas and loading times were directly proportional (Launikonis & Stephenson, 1997). Under the SR Ca²⁺-loading conditions used in this study, there was a linear relationship between the area under the caffeine-induced force responses and the loading time (R² > 0.95), that intersected the y-intercept at +8.2 ± 8.3% endogenous SR Ca²⁺-content. The relative areas under the caffeine-induced responses were corrected by subtracting the y-intercept for individual fibres as previously described (Launikonis & Stephenson, 1997; Macdonald & Stephenson, 2001), which conferred direct proportionality between the corrected areas and the SR Ca²⁺-content.

Load experiments

The SR Ca²⁺-loading ability was investigated by loading the SR with Ca²⁺, in the load₁ solution or load₂ solution with or without ADP, then washing for 30 s in either wash₁ solution or wash₂ solution before releasing SR Ca²⁺ in release solution. The standard caffeine-induced force response (60 s, in load₁ solution) was monitored throughout, to allow for correction of any small deterioration in SR function, and corrected for, as previously described (Macdonald & Stephenson, 2001).

Leak experiments

To estimate Ca²⁺ lost from the SR due to passive leak of Ca²⁺, the SR was loaded for 60 s in load₁ solution, washed in wash₁ solution for 30 s and the SR Ca²⁺-content released in release solution (control). Thereafter, the preparation was washed in wash₁ solution prior to re-loading for 60 s in load₁ solution, before being transferred to leak₁ solution or leak₂ solution (with or without added ADP) for 60 s. The fibre was then washed in wash₁ solution for 30 s and the remaining SR Ca²⁺ was released in release solution. The control was then repeated and the area under the leak response was divided by the average areas under the control responses before and after the leak response, giving an estimate of the fraction of SR Ca²⁺ leaked over the 60 s in leak₁ or leak₂ solution, and the SR Ca²⁺-leak rate constant β was calculated, assuming that the SR Ca²⁺-content decreased exponentially over the 60 s.

Further experiments were performed to determine the role of the SR Ca²⁺-release channels, and the SR Ca²⁺-pump in the passive SR Ca²⁺-leak. The SR Ca²⁺-release channel was blocked during the leak period with 6 μm ruthenium red (RR) (added from 1.2 mm stock dissolved in low relaxing₁ solution), which is expected to markedly block the SR Ca²⁺-release channels during the 60 s exposure (Lamb & Stephenson, 1991). The preparation was then washed in wash₁ solution for 30 s and then SR Ca²⁺ released with release solution in the absence of RR. Under these conditions the area under the caffeine-induced force response decreases marginally from that corresponding to the release of the same amount of Ca²⁺ from the SR but without exposure to RR (see also Lamb & Cellini, 1999). Alternatively, the SR Ca²⁺-pump was blocked with 20 μm 2,5-di(tert-butyl)-1,4-hydroquinone (TBQ) (Inesi & Sagara, 1994), added from a 20 mm stock solution dissolved in pure dimethyl-d₆sulfoxide (DMSO). After exposure to TBQ, the fibre was dipped in paraffin oil for 30 s to remove TBQ from the skinned fibre without markedly changing the SR Ca²⁺-content (Bakker et al. 1996), washed in wash₁ solution for 30 s, and depleted of Ca²⁺ in release solution. Controls for the TBQ experiments contained an equivalent amount of DMSO, and underwent the same procedure, with no significant alteration to normal control responses.

In all load and leak experiments, the release solution was identical to ensure that any difference in caffeine-induced force response, under the different conditions, was due to differences in SR Ca²⁺-content and not altered SR Ca²⁺-release properties (Herrmann-Frank et al. 1999) or alterations to Ca²⁺ sensitivity of the contractile apparatus.

Muscle fibre typing

All muscles fibres used in this study were confirmed as slow-twitch fibre type based on their greater sensitivity to Sr²⁺ compared to fast-twitch fibres (O'Connell et al. 2004). Briefly, pSr 5.4 solution (Table 1) was used to differentiate between slow-twitch and fast-twitch fibres, in which slow-twitch fibres produce 80–90% maximum Ca²⁺-activated force, while fast-twitch fibres produce zero force, as this Sr²⁺ concentration (pSr (−log₁₀[Sr²⁺] 5.4 corresponds to 4 μm[Sr²⁺]) is below the activation threshold (Bortolotto et al. 2000; O'Connell et al. 2004).

Data analysis

Results are expressed as means ± s.e.m. and curve fitting and statistical analyses were performed using the scientific analysis program GraphPad Prism (GraphPad Software, Inc., San Diego, CA, USA). The statistical significance of any difference between groups was accepted at P < 0.05, as determined using Student's two-tailed t test for non-paired observations or ANOVA, where appropriate.

Results

Effects of elevated [ADP] on contractile properties of slow-twitch fibres

Increasing the [ADP] from the control level (<0.1 μm) to 40 μm did not significantly alter the maximum Ca²⁺-activated force in Triton X-100 treated fibres (Table 2). To allow elevation of ADP to high concentrations, CP was removed from solutions, which itself significantly increased the maximum Ca²⁺-activated force by 4.6 ± 0.8% (P < 0.05). Addition of 1.0 mm ADP to this solution without CP, resulted in a significant reduction in the maximum Ca²⁺-activated force, when compared to the 0 mm CP solution (P < 0.05), but was not significantly different from controls containing CP. The effect of elevating the [ADP] on the force–pCa relationship is also shown in Table 2. Increasing the [ADP] in solution from <0.1–40 μm did not significantly alter either the pCa₅₀ or the n_H of the force–pCa curves. The removal of CP from solutions significantly increased the pCa₅₀ (P < 0.05) but not the n_H. The addition of 1.0 mm ADP to this 0 mm CP solution further significantly increased pCa₅₀ (P < 0.05) but not n_H. Thus, elevation of [ADP] from <0.1 μm to 1.0 mm causes an increase in the Ca²⁺ sensitivity of the contractile apparatus without altering the maximum Ca²⁺-activated force.

Table 2.

Effect of ADP and CP on the activation characteristics of mechanically skinned slow-twitch muscle fibres

	Control	40 μm ADP	0 mm CP	1.0 mm ADP
P/Pmax	1.000	1.015 ± 0.008	1.046 ± 0.008*	1.003 ± 0.017†
pCa50	6.08 ± 0.04	6.09 ± 0.05	6.15 ± 0.05*	6.25 ± 0.04*†
nH	3.57 ± 0.66	4.19 ± 0.92	3.39 ± 0.60	2.66 ± 0.57

P/P_max is the maximum Ca²⁺-activated force under test conditions normalized to the maximum Ca²⁺-activated force under control conditions. pCa₅₀ corresponds to the Ca²⁺ concentration giving half-maximum force under specified conditions. n_H (Hill coefficient) corresponds to the steepness of the force–pCa activation curve. Data are means ± s.e.m. from nine individual fibres. Significant difference from control

^*0 mm CP

^†P < 0.05, Student's t test).

Effects of elevated [ADP] on SR Ca²⁺-loading properties of slow-twitch fibres

Preliminary experiments were initially performed where the SR was loaded at pCa 7.1, which is close to the resting physiological pCa in slow-twitch fibres (Baylor & Hollingworth, 2003). Under these loading conditions the SR was loaded to 70 ± 11% (n = 4) of the endogenous level after 30 s, and could not be loaded more than the endogenous level even after 10 min in the loading solution. The inability of the SR to load more than endogenous Ca²⁺ is consistent with the observation of Fryer & Stephenson (1996) that the SR of slow-twitch muscle fibres is maximally loaded under endogenous conditions. Increasing [ADP] from <0.1 μm (control) to 40 μm, did not significantly alter the SR Ca²⁺-content of the SR (after 30 s loading) (67.3 ± 11% of the endogenous SR Ca²⁺-content, P > 0.05). However, further increasing the [ADP] to 1.0 mm did significantly reduce the SR Ca²⁺-content after 30 s loading at pCa 7.1 by about 25% to 52.0 ± 2.5% of the endogenous SR Ca²⁺-content (P < 0.05).

In order to slow down the rate of Ca²⁺ loading in the SR, the pCa of the SR loading solution was increased from 7.1 to 7.8 (load₁ solution). This allowed a similar analysis of [ADP] effects on the SR Ca²⁺-handling properties to those in fast-twitch fibres (Macdonald & Stephenson, 2001), and consequently a comparison of these effects between slow-twitch and fast-twitch fibres. At pCa 7.8, increasing the SR Ca²⁺-loading time resulted in an increase in the area under the caffeine-induced force response, indicating increased loading of Ca²⁺ into the SR (Fig. 1). Following a 60 s load in load₁ solution, the ensuing SR Ca²⁺-release was only about 30% of the response corresponding to the endogenous (maximal) level. This was equivalent to the SR Ca²⁺-level under standard loading conditions used previously by our laboratory for fast-twitch muscle (Macdonald & Stephenson, 2001), and therefore allows for appropriate comparison between ADP-dependent SR properties in slow-twitch and fast-twitch muscle. For all subsequent experiments in the present study, a 60 s load in load₁ solution was considered to be the standard loading condition.

Figure 1. Representative trace showing 30 mm caffeine-induced force responses following SR Ca²⁺-loading at pCa 7.8 in a slow-twitch soleus fibre for different periods of time.

The first response corresponds to the endogenous SR Ca²⁺-level for this individual fibre, with smaller responses as the loading time decreased at 120, 60, 30, 20 and 10 s. Between responses the preparation was washed for 30 s in wash₁ solution, then reloaded in loads₂ solution (pCa 7.8, time indicated in trace) and then washed again in wash₁ solution before the next SR Ca²⁺-release in release₁ solution. The two responses to the right of the panel are the maximum Ca²⁺-activated force response (Max) and the pSr 5.4 force response (Sr) confirming the slow-twitch fibre type, respectively.

Figure 2 indicates that as [ADP] increased, the SR was able to load, and subsequently release, significantly less Ca²⁺ over the entire range of loading times (P < 0.05), indicating that elevated [ADP] reduces the SR Ca²⁺-loading ability in slow-twitch muscle. After a 60 s loading period, there was a 20% (P < 0.05) and 35% (P < 0.05) decrease in SR Ca²⁺-content, when the [ADP] was increased from 0.1 μm to 40 μm and 1.0 mm, respectively.

Figure 2. The effect of elevated ADP on the SR Ca²⁺-loading ability in slow-twitch fibres.

All fibres were loaded under similar conditions for Ca²⁺ (pCa 7.8), pH ionic strength, cationic and anionic composition, except for CP, which was substituted with hexamethylene diamine tetraacetate (HDTA), and ADP, which were <0.1 μm (control, black square), 40 μm (blue triangle) and 1.0 mm (red circle). The SR Ca²⁺ was always released in the same release solution after the preparation had been washed in wash solution to ensure an accurate comparison between results. The data points for individual fibres were normalized to the area obtained under the standard loading condition (60 s) for that fibre. Data points are means ± s.e.m., n = 6, and fitted eqn (2), with an R² > 0.99.

Effects of elevated [ADP] on SR Ca²⁺-pump rate and SR Ca²⁺-leak rate

The results in Fig. 2 and can be accurately described by the following equation

(2)

where t is the loading time; [Ca]_SR(t) is the SR Ca²⁺-content at time t; pump rate is the rate of Ca²⁺ loading into the SR by the SR Ca²⁺-pump (and in first approximation is considered constant for the period of loading), and β is the SR Ca²⁺-leak rate constant, which is proportional to the overall SR permeability to Ca²⁺. Considering that in all experiments the SR was submaximally (<35%) loaded with Ca²⁺, the free Ca²⁺ in the SR was assumed to be proportional to the total SR Ca²⁺, [Ca]_SR. (For details of derivation of eqn (2) see Macdonald & Stephenson, 2001.) Therefore, by fitting the data points from Fig. 2 to eqn (2), in principle the SR Ca²⁺-pump rate and the SR Ca²⁺-leak rate constant (β) could be estimated for each condition. It is also possible to calculate the SR Ca²⁺-loading capacity for t = ∞ (pump rate/β) when the Ca²⁺ uptake equals the Ca²⁺ leak and no further accumulation of Ca²⁺ into the SR occurs.

As shown in Fig. 3, under control conditions (<0.1 μm ADP), the maximum SR Ca²⁺-loading capacity at pCa 7.8 was 36.8 ± 1.4% of the endogenous SR Ca²⁺-level, while the SR Ca²⁺-pump rate was 84.0 ± 3.3% endogenous SR Ca²⁺ min⁻¹, where 100% corresponds to the physiological endogenous Ca²⁺ level in the SR. The SR Ca²⁺-leak rate constant β was 2.16 ± 0.12 min⁻¹. Increasing the [ADP] from <0.1 μm to 40 μm significantly reduced the maximum SR Ca²⁺-capacity by about 12% (32.2 ± 3.0% endogenous SR Ca²⁺, P < 0.05), by significantly decreasing the SR Ca²⁺-pump rate to 71.4 ± 3.6% endogenous SR Ca²⁺ min⁻¹ (P < 0.05), but did not alter the SR Ca²⁺-leak rate constant (β) (2.37 ± 0.16 min⁻¹, P > 0.05). Increasing the [ADP] to 1.0 mm, further significantly decreased the maximum SR Ca²⁺-capacity by another 25% to 23.0 ± 5.0% endogenous SR Ca²⁺ (P < 0.05) but did not further significantly alter either the SR Ca²⁺-leak rate constant (β) (2.89 ± 0.45 min⁻¹, P > 0.05), or the SR Ca²⁺-pump rate (67.0 ± 8.0% endogenous SR Ca²⁺ min⁻¹, P > 0.05). However, compared with controls (<0.1 μm ADP), the SR Ca²⁺-leak rate constant (β) was significantly increased (P < 0.05), while the SR Ca²⁺-pump rate was significantly reduced (P < 0.05).

Figure 3. Effects of ADP on key functional parameters of the SR from mechanically skinned slow-twitch fibres.

A, maximum SR Ca²⁺-capacity; B, SR Ca²⁺-pump rate; and C, SR Ca²⁺-leak rate constant (β). The data points are obtained from eqn (2). Data points are means ± s.e.m., n = 6.

Effects of elevated [ADP] on SR Ca²⁺-leak rate constant (β) at pCa > 8.5

For further comparison of the ADP-induced SR Ca²⁺-leak rate with fast-twitch muscle (Macdonald & Stephenson, 2001), experiments were performed to assess the SR Ca²⁺-leak rate constant (β) at pCa > 8.5. Figure 4 shows that the Ca²⁺ remaining in the SR after exposure to leak₁ solution for 60 s was markedly smaller than that for control responses without exposure to the leak₁ solution (62.9 ± 10.2% of control), indicating that a significant (P < 0.05) amount of Ca²⁺ had been lost from the SR during the leak period, with a SR Ca²⁺-leak rate constant (β) of 0.46 ± 0.08 min⁻¹. Increasing the [ADP] in leak₁ solution to 40 μm, significantly increased (P < 0.05) the SR Ca²⁺-leak rate constant (β) to 0.93 ± 0.10 min⁻¹, with only 39.6 ± 4.4% of initial Ca²⁺ remaining in the SR after 60 s leak. Elevating the [ADP] to 1.0 mm in leak₂ solution, further significantly increased (P < 0.05) the SR Ca²⁺-leak rate constant (β) to 1.26 ± 0.17 min⁻¹, with only 28.4 ± 4.0% of initial Ca²⁺ remaining in the SR. Note that in absolute terms, the values of the SR Ca²⁺-leak rate constant (β) at pCa > 8.5 were smaller than those derived from the SR Ca²⁺-loading curves in Fig. 3 at pCa 7.8. However, the ADP-dependent relative increase in passive SR Ca²⁺-leak rate constant at pCa > 8.5 was markedly greater than that estimated from the SR Ca²⁺-loading curves in Fig. 3 at pCa 7.8.

Figure 4. The effect of elevated ADP on SR Ca²⁺-leak.

The amount of Ca²⁺ remaining in the SR of slow-twitch fibres loaded under standard control conditions (60 s load₁ solution) after a 60-s leak under control (<0.1 μm), 40 μm and 1.0 mm[ADP]. The amount of Ca²⁺ remaining in the SR was determined by normalizing the leak response to a standard control response performed before and after the leak response. Data points are means ± s.e.m., n = 7 and 4 for controls, and both 40 μm and 1.0 mm[ADP], respectively. Significant difference from control (*) and 40 μm ADP (#) (P < 0.05, ANOVA).

Effects of blocking SR Ca²⁺-release channel and SR Ca²⁺-pump on SR Ca²⁺-leak rate

In order to determine whether the ADP-induced increase in the SR Ca²⁺-leak in slow-twitch muscle was through the SR Ca²⁺-release channel or through the SR Ca²⁺-pump, the SR Ca²⁺-leak was measured in the presence of 1.0 mm[ADP] with and without either 6 μm RR to block the SR Ca²⁺-release channels, or 20 μm TBQ to block the SR Ca²⁺-pump. The results (Fig. 5) show that following the leak period performed under control conditions (leak₁ solution), the SR Ca²⁺-content was significantly reduced to 55.6 ± 9.5% of initial SR Ca²⁺-content (P < 0.05). Increasing the [ADP] to 1.0 mm during the leak period (leak₂ solution), reduced the SR Ca²⁺-content to 35.1 ± 6.5% of initial SR Ca²⁺ content (P < 0.05), which was not altered by adding RR to the 1.0 mm ADP leak solution (leak₂ solution) (37.4 ± 5.6% of initial SR Ca²⁺-content, P > 0.05). However, when 20 μm TBQ was added to the 1.0 mm ADP leak solution (leak₂ solution), the 1.0 mm[ADP]-induced SR Ca²⁺-leak was abolished, with SR Ca²⁺ restored back to control levels (49.1 ± 7.7% of initial SR Ca²⁺-content; P < 0.05). Together these results demonstrate that the SR Ca²⁺-pump, and not the SR Ca²⁺-release channel, was responsible for the elevated ADP-induced increase in SR Ca²⁺-leak in slow-twitch muscle.

Figure 5. The pathway of the ADP-induced increase in SR Ca²⁺-leak.

The amount of Ca²⁺ remaining in the SR of slow-twitch fibres loaded under standard control conditions (60 s load₁ solution) after a 60 s leak in leak solutions under control (<0.1 μm ADP), 1.0 mm[ADP] (ADP), 1.0 mm ADP + 6 μm ruthenium red (ADP + RR) and 1.0 mm ADP + 20 μm 2,5-di(tert-butyl)-1,4-hydroquinone (ADP + TBQ). The amount of Ca²⁺ remaining in the SR was determined by normalizing the leak response to a standard control response performed before and after the leak response. Data points are means ± s.e.m., n = 4. *Significant difference (P < 0.05, ANOVA) from control.

Discussion

The present results indicate that in rat slow-twitch muscle, the elevation of [ADP] to levels expected to occur during skeletal muscle fatigue (Karatzaferi et al. 2003; Hancock et al. 2005) had only a minor effect directly on the contractile apparatus, but produced a marked inhibition in the Ca²⁺-handling ability of the SR, by both decreasing the SR Ca²⁺-pump rate and by increasing the leak of Ca²⁺ from the SR.

Effects of elevated [ADP] on the contractile apparatus of slow-twitch fibres

An elevation in the [ADP] to 40 μm, in the presence of 8 mm ATP, had no effect on the steady-state contractile activation characteristics (maximum Ca²⁺-activated force, Ca²⁺ sensitivity or Hill coefficient) of slow-twitch muscle fibres in the present study. This is in agreement with Chase & Kushmerick (1995), who used similar conditions to show no effect of elevated [ADP] up to 50 μm ADP. Removing CP from solutions resulted in a small increase in maximum Ca²⁺-activated force and in Ca²⁺ sensitivity of the contractile apparatus, as previously shown by Fryer et al. (1995). Addition of 1.0 mm ADP to this solution, in the presence of 8 mm ATP, only slightly altered the maximum Ca²⁺-activated force and the Ca²⁺ sensitivity of the contractile apparatus. Small or no changes in maximum Ca²⁺-activated force and Ca²⁺ sensitivity in the presence of ADP were also reported in previous investigations on fast-twitch muscle fibres when ATP was 3–8 mm (Cooke & Pate, 1985; Karatzaferi et al. 2003; Macdonald & Stephenson, 2004). However, at low ATP concentrations there is a marked ADP-dependent increase in maximum Ca²⁺-activated force and Ca²⁺ sensitivity arising from ADP inhibiting the dissociation of (i) ADP from the actomyosin–ADP complex (Cooke & Pate, 1985; Dantzig et al. 1991) and (ii) actomyosin to actin and myosin in the presence of ATP to form myosinATP. This would lead to crossbridges being attached for longer, thereby causing an increase in the average force produced per crossbridge and in the level of activation at given Ca²⁺. The increase in Ca²⁺ sensitivity of the contractile apparatus at elevated [ADP], may in fact help to compensate for the lower SR Ca²⁺-release that occurs in metabolically fatigued skeletal muscle (Fryer et al. 1995; Westerblad et al. 1998). However, it must be noted that in terms of dynamic contractions, the longer crossbridge attachment due to elevated [ADP] also reduces shortening velocity (Godt & Nosek, 1989; Chase & Kushmerick, 1995) and therefore power output.

Effects of elevated [ADP] on SR Ca²⁺-handling ability of slow-twitch fibres

In slow-twitch muscle, elevation of [ADP] to 1.0 mm, resulted in a marked reduction in the ability of the SR to load, retain and release Ca²⁺. At pCa 7.1, the SR Ca²⁺-content was reduced by approximately 20%. At pCa 7.8, the SR Ca²⁺-content was reduced by approximately 40%, and was the result of a combination of a 20% decrease in the SR Ca²⁺-pump rate and a 30% increase in the leak of Ca²⁺ from the SR (Fig. 3). The small reduction in the SR Ca²⁺-pump rate was despite a 1000-fold increase in [ADP], which would reduce the ATP free energy available to drive the SR Ca²⁺-pump by 22.5 kJ mol⁻¹ ATP (RTln10000 = 8.3 × 295 × 9.2 J mol⁻¹). This is a sizeable fraction of the total ATP free energy under our conditions, suggesting that the reduction in ATP free energy is unlikely to play a significant role in causing this effect, because then a much larger effect would have been expected. Interestingly, at pCa > 8.5, the rate of Ca²⁺ loss from the SR increased threefold as the [ADP] was increased from <0.1 μm to 1.0 mm, i.e. by a factor that was 10-fold greater than that at pCa 7.8. However, the increase was from a smaller base (0.46 min⁻¹ at pCa > 8.5 and [ADP] < 0.1 μmversus 2.37 min⁻¹ at pCa 7.8 and [ADP] < 0.1 μm). The markedly larger SR Ca²⁺-leak rates from the SR at pCa 7.8 than at pCa > 8.5 at all [ADP], indicates that the SR Ca²⁺-leak rate is sensitive to myoplasmic [Ca²⁺]. For example, at very low [ADP] (<0.1 μm), the SR Ca²⁺-leak rate was fivefold greater at pCa 7.8 than at pCa > 8.5, suggesting that at pCa 7.8 some Ca²⁺ loss is likely to occur through Ca²⁺-release channels which display strong Ca²⁺ dependence. This was not the case for fast-twitch fibres, where the SR Ca²⁺-leak rate was greater at pCa > 8 than at pCa 6.7 (Macdonald & Stephenson, 2001) and where Ca²⁺-induced Ca²⁺ release does not play an important role in the excitation–contraction (E–C) coupling (Launikonis & Stephenson, 2000). In view of these differences between the SR in slow- and fast-twitch fibres, it is possible that Ca²⁺-induced Ca²⁺ release may play a role in the E–C coupling of slow-twitch fibres.

The ADP-induced leak of Ca²⁺ from the SR at pCa > 8.5 does not appear to occur via the SR Ca²⁺-release channel, as it is insensitive to RR (Fig. 5). However, the SR Ca²⁺-pump blocker TBQ abolished the ADP-induced increase in SR Ca²⁺-leak, indicating that at pCa > 8.5 the SR Ca²⁺-pump is the site of the markedly augmented leak in the presence of high [ADP]. A SR Ca²⁺-pump mediated SR Ca²⁺-leak, has been observed in vesicle studies (Inesi & de Meis, 1989) and in fast-twitch muscle fibres (Duke & Steele, 2000; Macdonald & Stephenson, 2001, 2004). As argued previously (Macdonald & Stephenson, 2001), the mechanism of the ADP-induced SR Ca²⁺-leak is unlikely to be via reversal of the SR Ca²⁺-pump, as under conditions using similar ATP, ADP, inorganic phosphate (P_i) and Ca²⁺ concentrations, it is thermodynamically not possible to drive the pump into reverse mode. It is therefore most likely that the ADP-induced SR Ca²⁺-leak mechanism in slow-twitch muscle is via the ‘slippage’ mechanism described by Inesi & de Meis (1989) and proposed to exist in mechanically skinned fast-twitch muscle, under similar conditions to those used in the current study (Macdonald & Stephenson, 2001, 2004). Essentially, the Ca²⁺-binding sites of the ADP-sensitive state of the SR Ca²⁺-pump can face either the luminal or myoplasmic side of the SR membrane. Thus, the ADP-sensitive state may transport Ca²⁺ in either direction by slipping or flicking from one side of the SR membrane to the other. With an elevation in [ADP], the proportion of SR Ca²⁺-pumps in the ADP-sensitive state would be greater, therefore increasing the amount of Ca²⁺ that can leak from the SR via this slippage mechanism.

Interestingly, the rate constants of SR Ca²⁺-leak at pCa > 8.5 were very similar in slow-twitch and fast-twitch fibres and increased by a similar factor when [ADP] was elevated from 0.1 μm to 1.0 mm. This suggests that an elevated [ADP] affects the two major isoforms of the SR Ca²⁺-pump, SERCA 1 in fast-twitch fibres and SERCA 2A in slow-twitch fibres in a similar manner, with respect to the slippage mechanism. Nevertheless, there are major functional differences between the two SR Ca²⁺-pump isoforms with respect to the pCa values where the pump becomes activated, with the SERCA 2A isoform activated at pCa values that are at least one logarithmic unit higher (lower [Ca²⁺]) than that for SERCA 1. This observation is contrary to that of Lytton et al. (1992) who found these two SERCA isoforms expressed in COS cells, displayed no differences in Ca²⁺, ATP or pH dependency; or in peak Ca²-ATPase and Ca²⁺-uptake activity. The results clearly show that when expressed in slow- and fast-twitch fibres, respectively, the two isoforms display major differences with respect to the level of Ca²⁺ at which they become activated. It is noteworthy that in slow-twitch rat fibres, no phospholamban is present that can modulate SERCA 2A (see Bortolotto et al. 2001). The much lower ADP effect on the rate constant of Ca²⁺ loss from the SR at pCa < 8 than at pCa > 8.5 can be directly explained by the presence of a Ca²⁺-induced Ca²⁺-release mechanism operating in slow-twitch fibres which provides a high background for the Ca²⁺ leakage from the SR at [ADP] < 0.1 μm.

Slow-twitch versus fast-twitch fibres

Slow-twitch muscle has long been considered to be more fatigue resistant in nature than fast-twitch muscle. The lower ATP utilization during contraction (Szentesi et al. 2001) and the greater mitochondrial content of slow-twitch muscle result in a smaller build up of ADP, P_i, and H⁺ and on a slower time scale than in fast-twitch muscle. At physiological pCa, an elevation in [ADP] affects slow-twitch SR Ca²⁺-handling ability to a smaller degree but by a similar mechanism to that of fast-twitch muscle fibres (Macdonald & Stephenson, 2001). To allow a clearer presentation of this comparison, Fig. 6 summarizes the effects of elevated [ADP] on the SR Ca²⁺-capacity, SR Ca²⁺-pump rate and SR Ca²⁺-leak rate (β) for both slow-twitch and fast-twitch muscle. The data for fast-twitch muscle was taken from Macdonald & Stephenson (2001), where experiments were performed under the similar conditions. In slow-twitch muscle, increasing the [ADP] from <0.1 μm to 40 μm and 1.0 mm decreased the maximum SR Ca²⁺-capacity by 20 and 55%, respectively, whereas in fast-twitch muscle, the same range of [ADP] reduced maximum SR Ca²⁺-capacity by four and six times, respectively. In both muscle fibre types, the reduction in SR Ca²⁺-capacity was due to a concomitant moderate slowing of the SR Ca²⁺-pump rate, and an increase in the leak of Ca²⁺ from the SR. The SR Ca²⁺-leak rate was not altered in slow-twitch muscle until the [ADP] had risen to 1.0 mm, and then only by 25%, while in fast-twitch muscle, the SR Ca²⁺-leak rate was extremely sensitive to [ADP], increasing by 3 and 4.5 times as the [ADP] was increased to 40 μm and 1.0 mm, respectively.

Figure 6. Comparison of the effects of elevated [ADP] on SR Ca²⁺-handling parameters of mechanically skinned slow-twitch and fast-twitch muscle fibres.

A, maximum SR Ca²⁺-capacity; B, SR Ca²⁺-pump rate; and C, SR Ca²⁺-leak rate constant (β), for fast-twitch (□) and slow-twitch (▪) muscle fibres, respectively. The data for fast-twitch muscle is taken from Macdonald & Stephenson (2001).

Thus, the effects of elevated [ADP] on SR Ca²⁺ handling in slow-twitch muscle are much less marked than in fast-twitch muscle (Macdonald & Stephenson, 2001), suggesting that slow-twitch muscle is less sensitive to changes in [ADP]. Indeed, others have shown that slow-twitch SR is less sensitive to elevations in intracellular Mg²⁺ than fast-twitch muscle (Stephenson et al. 1998), allowing normal SR Ca²⁺-release to occur at Mg²⁺ concentrations that would abolish such events in fast-twitch muscle.

Relevance to skeletal muscle fatigue

The role of altered SR Ca²⁺-handling is considered a significant contributor in skeletal muscle fatigue. Indeed, a decrease in SR Ca²⁺-release has been implicated in skeletal muscle fatigue (Fitts, 1994; Westerblad et al. 1998), with elevated Mg²⁺ (Stephenson et al. 1998) and low ATP (Dutka & Lamb, 2004) inhibiting the SR Ca²⁺-release channel. A depressed SR Ca²⁺-release may also be due to a reduced SR Ca²⁺-content, either via precipitation with P_i to form calcium phosphate (Fryer et al. 1995) or by an inability of the SR Ca²⁺-pump to re-sequester Ca²⁺ back into the SR (Tupling, 2004). A reduced SR Ca²⁺-pump would contribute to the elevated myoplasmic [Ca²⁺] and prolonged time to relaxation of force observed during skeletal muscle fatigue (Allen et al. 1995).

An elevation in [ADP] may explain the mechanisms responsible for each of these processes. The ADP-induced decrease in SR Ca²⁺-pump rate and increase in SR Ca²⁺-leak rate would contribute to the elevated resting myoplasmic [Ca²⁺]. This would also slow the time to relaxation of force, as Ca²⁺ would be bound to troponin C for a longer period of time. Furthermore, the reduced SR Ca²⁺-capacity at elevated [ADP] would result in a smaller pool of Ca²⁺ in the SR available for release, contributing to the depressed SR Ca²⁺-release. It therefore appears that elevated [ADP] may explain many of the observations associated with skeletal muscle fatigue, however, it must be emphasized that skeletal muscle fatigue is a multifactorial phenomenon (Stephenson et al. 1998).

The smaller ADP-induced reduction in SR Ca²⁺-pump rate and increase in SR Ca²⁺-leak rate in slow-twitch compared to fast-twitch muscle, together with the lower dynamic range of [ADP] in slow-twitch muscle, would lead to a greater stability of SR Ca²⁺-handling in slow-twitch muscle, and may significantly contribute to the fatigue resistant nature of slow-twitch muscle.