Sarcoplasmic reticulum Ca²⁺ depletion in adult skeletal muscle fibres measured with the biosensor D1ER

Abstract

The endoplasmic/sarcoplasmic reticulum (ER/SR) plays a crucial role in cytoplasmic signalling in a variety of cells. It is particularly relevant to skeletal muscle fibres, where this organelle constitutes the main Ca²⁺ store for essential functions, such as contraction. In this work, we expressed the cameleon biosensor D1ER by in vivo electroporation in the mouse flexor digitorum brevis (FDB) muscle to directly assess SR Ca²⁺ depletion in response to electrical and pharmacological stimulation. The main conclusions are: (1) D1ER is expressed in the SR of FDB fibres according to both di-8-(amino naphthyl ethenyl pyridinium) staining experiments and reductions in the Förster resonance energy transfer signal consequent to SR Ca²⁺ release; (2) the amplitude of D1ER citrine/cyan fluorescent protein (CFP) ratio evoked by either 4-chloro-m-cresol (4-CmC) or electrical stimulation is directly proportional to the basal citrine/CFP ratio, which indicates that SR Ca²⁺ modulates ryanodine-receptor-isoform-1-mediated SR Ca²⁺ release in the intact muscle fibre; (3) SR Ca²⁺ release, measured as D1ER citrine/CFP signal, is voltage-dependent and follows a Boltzmann function; and (4) average SR Ca²⁺ depletion is 20% in response to 4-CmC and 6.4% in response to prolonged sarcolemmal depolarization. These results indicate that significantly depleting SR Ca²⁺ content under physiological conditions is difficult.

Introduction

Understanding the complex regulation of endoplasmic/sarcoplasmic reticulum (ER/SR) Ca²⁺ release and its contribution to cytoplasmic signalling is important in a variety of cells. This organelle is particularly relevant for muscle fibres, where it constitutes the main Ca²⁺ store for basic functions, such as contraction.

In mammals, however, loading only the skeletal muscle SR with Ca²⁺-sensitive indicators presents technical difficulties that seriously hamper our ability to determine its Ca²⁺ concentration and release kinetics. Limitation to measure Ca²⁺ concentration and dynamics had prevented reliable examination of questions relevant to muscle fibre physiology and pathology. Whether ER/SR Ca²⁺ depletion controls SR Ca²⁺ release channel/ryanodine receptor (RyR) and/or SR Ca²⁺ release termination is unknown, whilst its influence on Ca²⁺ influx via activation of a store-operated Ca²⁺ entry (SOCE) is debatable [15, 21, 36]. Putative mechanisms are triggered by a considerable decline in SR Ca²⁺, at least in non-excitable cells [25]. Assessing Ca²⁺ depletion as a consequence of SR Ca²⁺ release is relevant to examining lumenal Ca²⁺ buffer capacity and the mechanisms of Ca²⁺-dependent regulation of RyR function, termination of SR Ca²⁺ release, activation of SOCE and age-dependent decline in SR Ca²⁺ release [18]. The pathogenesis of disease states, such as muscular dystrophy, malignant hyperthermia and fatigue syndromes, will undoubtedly benefit from a better understanding of lumenal SR Ca²⁺ homeostasis.

Although not exempt from limitations [7], measuring Ca²⁺ release using engineered protein expression targeted to the SR is increasingly important [27, 38]. It has been claimed that the genetically encoded Ca²⁺ indicator D1ER, created by redesigning the binding interface of calmodulin and a calmodulin-binding peptide, has better reaction kinetics and exhibits a theoretically appropriate dissociation constant for tracking changes in SR/ER or SR Ca²⁺ concentration compared to previous Förster resonance energy transfer (FRET)-based cameleons. In addition, it is not perturbed by large excesses of native calmodulin [27, 28]. Therefore, D1ER has been proposed as a promising tool to examine Ca²⁺ dynamics in the SR lumen of various cell types [28, 49], including skeletal muscle fibres [38], in the only study in the literature to examine skeletal muscle SR Ca²⁺ release. Although this publication is intriguing, a number of questions remain about: (1) the variability in resting SR Ca²⁺ concentration amongst fibres of the same muscle; (2) the reproducibility of the D1ER signal; (3) the dependence of D1ER signal amplitude on basal SR FRET signal; (4) the magnitude of SR Ca²⁺ depletion in response to maximal fibre stimulation; (5) the voltage dependence of the lumenal SR Ca²⁺ response detected with D1ER; and (6) the timeframe of the SR and cytosolic Ca²⁺ transient.

The present study aims to extend previous work [38] to answer the questions above in voltage-clamped adult mouse flexor digitorum brevis (FDB) muscle fibres monitored using the SR Ca²⁺ biosensor D1ER. Maximal SR Ca²⁺ release is also examined by bypassing the physiological Ca_v1.1/ryanodine receptor-1 (RyR1) coupling and directly stimulating RyR1. For this application, we used the specific RyR1 agonist 4-chloro-m-cresol (4-CmC) based on the reproducibility of its action and the lack of any reported effect on SR Ca²⁺ refilling.

Materials and methods

Flexor digitorum brevis muscle fibres

Single skeletal muscle fibres from the FDB were obtained from 3- to 5-month-old Freund virus B(FVB) mice raised in the Animal Research Program of Wake Forest University School of Medicine (WFUSM). Mice were killed by cervical dislocation. Animal handling followed a protocol approved by the WFUSM Animal Care and Use Committee.

FDB fibre electroporation and D1ER expression location

The pcDNA3 plasmid carrying the cameleon biosensor D1ER [28], kindly provided by Dr. Roger Y. Tsien (Department of Pharmacology and Howard Hughes Medical Institute, University of California at San Diego), was electroporated into the FDB muscle from FVB mice following described methods [8]. To determine D1ER subcellular expression, FDB fibres were enzymatically dissociated 2 weeks after muscle electroporation [45, 46]. Fibres expressing D1ER were stained with 5 μM di-8-amino naphthyl ethenyl pyridinium (di-8-ANEPPS) for 30–45 min, and images were recorded with a Radiance 2100 (Bio-Rad, Zeiss) confocal microscope with a Fluar ×100/1.30 objective (Zeiss). D1ER and di-8-ANEPPS signals were excited by an argon laser at 488 nm, whilst the emission wavelengths were 528 and 600 nm, respectively. The 488-nm laser wavelength, although not optimal for CFP, directly excites citrine as reported [35].

Intracellular Ca²⁺ fluorescence

Enzymatically dissociated FDB fibres were voltage-clamped in the whole-cell configuration of the patch-clamp technique [16], according to described procedures specific for muscle fibres [45]. Potential voltage errors associated with whole-cell recording in large cells were minimised by selecting small FDB fibres and adequately compensating for whole-cell capacitance transients. The pipette electrode was filled with the following solution (mM): 130 Cs aspartate, 2 MgCl₂, 10 Cs₂ ethylene glycol tetraacetic acid (EGTA), 10 HEPES, 5 Na-ATP and 0.5 GTP; pH was adjusted to 7.3 with CsOH. High EGTA was used in the pipette solution to reproduce experimental conditions like those in a series of studies on FDB muscle fibres in which SR Ca²⁺ release was recorded in the cytosol [13, 18, 47, 48]. The electrode resistance ranged from 450 to 650 kΩ. We have not seen any difference in Ca²⁺ uptake kinetics in the presence or absence of 10 mM EGTA in the cytoplasm (data not shown), which may be explained by the slow dissociation of the cameleon/Ca²⁺ complex (see below). The external solution contained (in mM): 100 tetraethylammonium hydroxide (TEA)-OH, 50 Na₂SO₄, 2 MgSO₄, 2 CaSO₄, 2 3-4DAP and 5 Na-HEPES. Solution pH was adjusted to 7.3 with CH₄SO₃ [18, 45]. These solutions ensured good control of resting leakage currents, and the preparation was well preserved throughout the experiment. Methane sulfonic acid is the counterion to TEA. Drift and noise imposed by silver chloride wires are corrected by standard methods. All experiments were recorded at room temperature (21–22°C).

Fluorescence was recorded 2 weeks after FDB electroporation using an Axiovert 200 microscope and a spectrofluorometer (PTI, Birmingham, NJ, USA) with a ×20/0.75 or ×40W/1.2 objective (Zeiss). For Fura-2FF (Invitrogen, Carlsbad, CA, USA), excitation and emission wavelengths were set at 350/380 and 510 nm, respectively. Ratiometric citrine/CFP data were recorded with a dual photomultiplier (see below).

SR Ca²⁺ release and cytosolic Ca²⁺ transients were recorded sequentially (20-s interval). No differences in the amplitude and kinetics of cytosolic Ca²⁺ release monitored with Fura-2FF were apparent in D1ER transfected and untransfected fibres (data not shown). Fura-2FF was loaded into the fibre via the patch pipette. Resting Ca²⁺ concentration was measured in dissociated fibres loaded with 5 μM Fura-2 AM mixed with 10% (v/v) pluronic acid (Invitrogen) in anhydrous dimethyl sulfoxide (DMSO) acid at <4, 5–23 and 24 h after dissociation. DMSO was kept at 0.05% final concentration. Ratiometric traces were converted into Ca²⁺ concentration using the following equation: Ca²⁺ = K_D(R – R_min)/(R_max – R) (β) [14], where R is fluorescence ratio measured at 340 and 380 nm; K_D is the Fura affinity constant for Ca²⁺; and β is the intrinsic ratio of intensities of free vs. Ca²⁺-bound indicator [14]. A K_D value of 51 μM [17] and R_max and R_min of 3.5 and 0.02 were measured.

Statistics

Values are given as mean±SEM with the number of observations (n). Statistical analysis was performed using Student’s unpaired t test and the Mann–Whitney rank sum test when values were not normally distributed. P<0.05 was considered significant.

Results

D1ER expression in the SR of FDB fibres

We first examined whether D1ER is expressed in the SR of FDB fibres. Figure 1a shows that expression of the SR Ca²⁺ biosensor produces a striated pattern when recorded with epifluorescence microscopy at two magnifications. Fibres expressing D1ER were stained with di-8-ANEPPS and examined with confocal microscopy to determine the position of the sarcotubular system (Fig. 1b). As expected, the D1ER and di-8-ANEPPS images show similar dualstriation patterns and almost complete overlap due to the sarcotubular system’s proximity to the SR terminal cisternae at the muscle fibre triad. Figure 1c shows the fluorescence intensity profiles corresponding to the rectangular area spanning three sarcomeric spaces drawn on Fig. 1b.

Fig. 1.

D1ER is expressed in the SR of FDB fibres. a D1ER epifluorescence recorded in the same FDB fibre at two magnifications, ×20 and ×40. b FDB fibres expressing D1ER, stained with di-8-ANEPPS (t-tubules) and imaged with a confocal microscope, show a striation pattern (merged image). The calibration bar is 5 μm. c Normalised fluorescence intensity profiles for D1ER and di-8-ANEPPS, measured in the rectangular area shown in b

RyR1 mediates SR Ca²⁺ release recorded with D1ER

To verify that RyR1 mediates the Ca²⁺ release detected with D1ER, we voltage-clamped FDB fibres at −80 mV and exposed them to the channel blocker ryanodine. Peak SR Ca²⁺ fluorescence decreased with time after applying 5 μM ryanodine extracellularly. Figure 2 illustrates D1ER fluorescence in response to a series of three consecutive 100-ms pulses delivered every 3 s before (a) and 20 min after (b) applying ryanodine. The amplitude of the citrine/CFP ratio at 20 min after applying the drug compared to time zero was 0.09±0.03 (n=3). During ryanodine application, basal R was 1.61±0.27 and 1.49±0.28 at times 0 and 20 min, respectively. This difference was not statistically significant (P>0.05). As RyR1 is the only isoform expressed in adult skeletal muscle [3, 33, 37], these results indicate that this RyR isoform mediates decreases in SR Ca²⁺ in response to sarcolemmal depolarization. Using two pharmacological agents that bind the SR Ca²⁺ release channel/ryanodine receptor with high affinity, an antagonist (ryanodine) and an agonist (4-CmC, see below), allows us to demonstrate that the recorded D1ER signal corresponds to Ca²⁺ released from the SR lumen through the ryanodine receptor. The slow timescale of the effect indicates slow diffusion of the agent. The kinetics of ryanodine binding to its receptor may be a contributing factor.

Fig. 2.

Ryanodine receptor inhibition prevents decline in D1ER FRET. Peak fluorescence amplitude decreased with time after adding 10 μM ryanodine to the bath solution. D1ER fluorescence in response to a series of three consecutive 100-ms pulses before (a) and 20 min after (b) applying the drug. The interval between trains was 3 s

Resting SR Ca²⁺ exhibits great variability

Basal SR Ca²⁺, expressed as the ratio between fluorescence signals attributed to citrine and CFP, was measured in 100 FDB fibres showing a value (mean±SEM) of 1.62±0.10 with a range of 0.39–7.40 (Fig. 3a). This high variability is surprising as all fibres examined were from 19 healthy mice of the same strain (FVB) and same age range (3–5 months). They were confined to cages of the same size and exhibited similar levels of physical activity. This variability does not seem to be an exclusive property of the FDB muscle; a significant but less pronounced variability was reported in the mouse tibialis anterior muscle using D1ER as the SR Ca²⁺ biosensor in vivo (basal citrine/CFP ratio range 2.2–4.0) [38].

Fig. 3.

Basal citrine/CFP ratio and its dependence on time after fibre dissociation. a Basal citrine/CFP ratio in 100 FDB fibres. b Basal citrine/CFP ratio at different times after fibre dissociation. The number of fibres analysed at 0–10, 11–20 and more than 24 h are expressed in parentheses above the bars

Figure 3b shows the stability of the citrine/CFP ratio in FDB fibres under voltage-clamp over 48 h after dissociation. The number of fibres examined, not included in Fig. 3a, was 33, 12 and 12 at 0–10, 11–20 and 24–48 h, respectively. These results indicate that the variability of SR Ca²⁺ loading described above does not depend on the time after dissociation that the fibre was examined as differences between the three time periods are not statistically significant.

The variability of the basal citrine/CFP ratio prevented us and another group [38] from reliably calibrating the biosensor because we could not define a single set of R_min and R_max values to transform citrine/CFP ratios into Ca²⁺ concentrations. Palmer et al. [27] suggested that calibration should be performed in each experimental fibre. We believe this is a reasonable approach for culture cell lines but rather cumbersome for muscle fibre patch-clamp studies.

Fractional SR Ca²⁺ release in response to electrical stimulation

Figure 4a shows SR Ca²⁺ fluorescence in response to a 200-ms pulse to 20 mV at two emission wavelengths, 535 nm (citrine) and 485 nm (CFP), and their ratio (citrine/CFP). The decrease in citrine was larger than the increase in CFP emission, producing a decreased citrine/CFP ratio in the SR lumen as an expression of decreased FRET. We have recorded a decrease in D1ER signal in response to electrical stimulation in only 17 out of 32 fibres (53%) voltage-clamped in the whole-cell configuration of the patch clamp, despite the similar basal citrine/CFP ratio (Fig. 4b). We observed a stereotypical D1ER expression pattern in all fibres inspected with confocal microscopy (see above). Therefore, its accumulation in pre-SR compartments is improbable. FDB cell appearance and electrical properties in responding and non-responding fibres was similar, so it is not obvious why approximately half the fibres examined did not respond. Unfortunately, we cannot find any publications with which to compare our success rate. Figure 4c illustrates the citrine/CFP ratio at four potentials from −40 to 20 mV, with the 20-mV interval corresponding to the steeper part of the curve. The relationship between membrane voltage and the normalised citrine/CFP ratio was examined in 14 fibres. Data points corresponding to the amplitude of the FRET decline were fitted to a Boltzmann equation of the form F = F_max/[1 + exp(V_F1/2 – V_m)/K]. The amplitude of the signal was measured from onset (average of ten points immediately before application of the command pulse) or the basal fluorescence (average of ten baseline points before application of the train of pulses) to nadir (see Fig. 5a). v_F1/2 and K values were −15.7±0.9 and 5.7±0.7, respectively. Ca²⁺ response was detected at −27± 3.3 mV. Figure 4d shows the direct linear relationship between the basal SR citrine/CFP ratio and citrine/CFP response amplitude examined in 17 fibres (r²=0.6; Pearson correlation=0.776, P=0.0025). These results indicate that basal SR Ca²⁺ modulates the amplitude of SR Ca²⁺ release and contradict a previous report in which the normalised drop of the YFP/CFP ratio during twitch/basal YFP/CFP exhibited an inverse relationship [38]. The maximal decrease in D1ER ratio elicited by a 100-ms pulse to 20 mV was (mean±SEM) 6.45±0.8%, with a range of 2.6–11.7% (n=16). Longer pulses have been applied (see below), but the magnitude of SR Ca²⁺ depletion did not increase further. These results indicate that SR Ca²⁺ depletion is almost negligible in response to pulses 100 times longer than those necessary to elicit an action potential.

Fig. 4.

Voltage dependence of D1ER FRET. a Citrine and CFP fluorescence and their ratio. b Average basal citrine/CFP ratios corresponding to fibres responding or not responding to electrical stimulation. The number of fibres per group is expressed between parentheses. c Voltage-normalised citrine/CFP ratio relationship. Data points, measured from the onset of the signal to nadir, were fitted to a Boltzmann equation. See the text for the best-fitting parameters. The upper panel illustrates the citrine/CFP ratio at four potentials. d Linear relationship between basal citrine/CFP ratio and the amplitude of the citrine/CFP response to 100-ms pulses to 20 mV

Fig. 5.

SR D1ER FRET in response to repetitive electrical stimulation. a D1ER response to a sequence of five sequential pulses (100 ms, 20 mV) showing where the basal-nadir and onset–nadir ratios were measured. b Voltage–normalised citrine/CFP ratio relationship. Data points, measured from the basal level of the signal to nadir, were fitted to a Boltzmann equation. See the text for the best-fitting parameters. Normalised citrine/CFP ratio amplitude for each pulse using the basal-to-nadir (c) or onset-to-nadir (d) relationships

Examination of SR Ca²⁺ release elicited by repetitive electrical stimulation shows reproducible responses with cumulative SR Ca²⁺ depletion. Figure 5a shows the citrine/CFP ratio in response to a sequence of five 100-ms command pulses to 20 mV. We measured the normalised citrine/CFP amplitude for the five pulses from basal or onset (asterisk) to nadir. Figure 5b shows the normalised citrine/CFP ratio as a function of membrane voltage. Data were fitted to a Boltzmann equation as described above. v_F1/2 and K values were −14.3±0.6 and 5.8±0.5, respectively, for the basal-to-nadir relationship (n=14). Figure 5c shows the normalised citrine/CFP amplitude, measured as the basal-to-nadir relationship as a function of pulse number. Progressive SR Ca²⁺ depletion for the first three pulses, with no further decrease, is observed, which may indicate that Ca²⁺ accumulation in the cytosol results in a larger cytosolic/SR Ca²⁺ gradient, rendering the SR Ca²⁺ pump more efficient [26]. As expected, the amplitude of the citrine/CFP ratio responses, measured as the onset-to-nadir relationship, did not significantly vary throughout the train of pulses (Fig. 5d). These results indicate that SR Ca²⁺ replenishment is effective but slower than that recorded with low-affinity cytosolic Ca²⁺ indicators [6, 12]. This slow signal is characteristic of cameleon sensors [7].

Further analysis of the D1ER signal elicited by progressively longer command pulses showed a graded increase in the citrine/CFP ratio amplitude with saturation in response to pulses longer than 150 ms. Figure 6a shows a family of Ca²⁺ currents (a, b) and their corresponding decline in SR D1ER FRET (c, d) evoked by 100-, 150- and 200-ms pulses to 20 mV. A more pronounced citrine/CFP ratio decline was induced using this protocol. Increasing the pulse duration to 600-, 1,100-, 1,600- and 2,100-ms achieved steady state. These representative studies of 14 fibres used various pulse durations. Figure 6b shows progressive SR Ca²⁺ depletion for short pulses (<150 ms), but it remained steady with more prolonged depolarization (>150 ms), possibly due to cytosolic Ca²⁺ accumulation, which leads to RyR1 inactivation and inhibition of SR Ca²⁺ release.

Fig. 6.

Effect on D1ER FRET of prolonging electrical stimulation. a Inward Ca²⁺ currents and citrine/CFP ratio in response to 100-, 150- and 200-ms pulses (a, c) or 600-, 1,100-, 1,600- and 2,100-ms pulses (b, d) to 20 mV, respectively. b Citrine/CFP ratio amplitude as a function of pulse duration

Voltage-dependent changes in D1ER lumenal signal and cytoplasmic Ca²⁺ transients

FDB fibres expressing D1ER were loaded with 200 μM Fura-2FF via the patch pipette and depolarized with 600-ms command pulses from v_h=-80 to 20 mV (Fig. 7a). Normalised citrine/CFP ratios for the maximal ratiometric signals (R/R_max) of D1ER (a) and Fura-2FF (b) were sequentially recorded and displayed in the same timescale. The kinetics of SR Ca²⁺ release and uptake is similar to the cytoplasmic Ca²⁺ transient increase and decay but slower than that reported in EDL fibres using furaptra (see above). The amplitude and voltage dependence of cytosolic Ca²⁺ is similar to that reported in adult FDB fibres [18]. Also, thevoltage dependence and amplitude of the cytoplasmic signal is similar to that of SR Ca²⁺ release examined with D1ER. The time-to-peak of the D1ER signal was 114±23 and 186±75 ms at 0 and 10 mV, respectively, whilst for Fura-2FF the time-to-peak was 170±31 and 139±10 ms, respectively. Apparent differences in the voltage dependence of the cytosolic and SR Ca²⁺ and time-to-peak of D1ER and Fura-2FF signals were not statistically significant. Figure 7b (a) shows Ca²⁺ transients recorded in response to increasing fibre depolarization. Figure 7b (b) shows the peak Fura-2FF ratio normalised to the maximal response in each fibre (R/R_max) as a function of membrane voltage. Data points were fitted to a Boltzmann equation. Half-activation potential and curve steepness were −14.9±2.7 mV and K = 7.2 ± 0.9, respectively (n=7 fibres). These values do not differ significantly from those recorded for D1ER using the basal–nadir or onset–nadir, respectively. Elevations in cytoplasmic Ca²⁺ were detected at −20± 3.8 mV, similar to the D1ER signal.

Fig. 7.

Voltage-dependent D1ER FRET signal and cytoplasmic Ca²⁺ transients. a An FDB fibre, expressing D1ER and loaded with 200 μM Fura-2FF via the patch pipette, depolarized with two command pulses from v_h=−80 to 20 mV for 600 ms. Normalised ratiometric signals (R/Rmax) for D1ER (a) and Fura-2FF (b) were sequentially recorded (20-s interval) and displayed in the same timescale. b (a) Fura-2FF fluorescence signals in response to command pulses from −70 to 20 mV. (b) Relationship between the normalised Fura-2FF ratio (R/R_max) and membrane voltage (mV) fitted to a Boltzmann equation. See text for best-fitting parameters. Data points represent the mean± SEM of seven fibres

Fractional SR Ca²⁺ release in response to pharmacological RyR1 activation

We then investigated the magnitude of SR Ca²⁺ release depletion evoked by direct RyR1 activation. Figure 8a shows the time course of D1ER FRET in a FDB fibre exposed to the RyR agonist (4-CmC, top bar), which was removed after peak FRET decline was reached. The top panel illustrates the fibre’s fluorescence, recorded with confocal microscopy at three time points after exposure to 1 mM 4-CmC. Figure 8b shows changes in citrine/CFP amplitude ratio as a function of the resting/basal SR citrine/CFP in response to 4-CmC application. The citrine/CFP ratio values for most fibres are grouped between 1.0 and 3.0, whilst only one exhibits a ratio of ~7. These data indicate a linear relationship between basal citrine/CFP ratio and amplitude of 4-CmC-evoked RyR activation. The decrease in SR D1ER FRET elicited by 4-CmC was (mean±SEM) 20.6±3.6%, with a range of 6.6–57.4% (n=15). These results indicate that prolonged SR Ca²⁺ release depletes only a small fraction of the SR Ca²⁺. The amplitude of SR Ca²⁺ release in response to a 100-ms voltage pulse to 20 mV (see above) is about one third of the Ca²⁺ mobilised by 4-CmC, and neither activating manoeuvre comes close to completely depleting SR Ca²⁺.

Fig. 8.

Ryanodine receptor activation elicits SR Ca²⁺ release. a Time course of D1ER FRET in an FDB fibre exposed to the RyR agonist 4-CmC (top bar=30 s). The top panel illustrates fibre fluorescence before (t=0) and after (t=70 and 90 s) exposure to 1 mM 4-CmC. The dimensions of the three images are 30×60 μm. b Basal citrine/CFP–citrine/CFP amplitude in response to 4-CmC application (n=15 fibres)

Discussion

Two properties of D1ER, its expression at the SR (this work) and the decline in FRET signal in the presence of Ca²⁺ [27], which have been analysed in detail, are crucial to examining SR Ca²⁺ dynamics. Despite the technical limitations described above and in a previous publication [38], analysing the D1ER ratiometric signal recorded in muscle fibre provides useful information about relative SR Ca²⁺ changes and is required to estimate SR Ca²⁺ depletion. Relative changes in citrine/CFP ratio, rather than [Ca²⁺]_er, have been reported using D1ER in arterial smooth muscle [49], HEK-293 cells expressing RyR2 [19] and vascular endothelial cells [31].

This work concludes that: (1) the biosensor D1ER is expressed in the SR of FDB fibres according to both di-8-ANEPPs staining, co-registration with D1ER fluorescence experiments and reductions in the FRET signal consequent to electrical and pharmacological stimulation. (2) The basal citrine/CFP ratio shows great variability in electroporated FDB fibres obtained from a homogeneous mouse population. (3) The decline in SR D1ER signal evoked by electrical or pharmacological stimulation of FDB fibres is directly proportional to the basal D1ER ratio, indicating that SR lumenal Ca²⁺ modulates SR Ca²⁺ release. (4) SR Ca²⁺ release as measured with D1ER is voltage dependent and follows a Boltzmann function. (5) The average SR Ca²⁺ depletion is 20%—with a maximum of 57%—in response to 4-CmC and 6.4%—with a maximum of 11.7%—in response to prolonged or maximal sarcolemmal depolarization, respectively. (6) The time to peak of SR Ca²⁺ release and cytoplasmic Ca²⁺ transient at maximal sarcolemmal depolarization does not differ significantly.

D1ER morphological and functional expression in the FDB fibre SR

The dual-striation pattern of D1ER overlapping the T-tubule projections detected with 8-di-ANEPPS indicates that the Ca²⁺ biosensor is predominantly expressed at the SR terminal cisternae. This conclusion is consistent with the observation that mammalian skeletal muscle fibres display two triads per sarcomere, one on each A–I band interface [10][11, 22]. The fluorescence between these two rows (see Fig. 1b, c) may imply D1ER expression in the longitudinal SR as well.

We rule out the possibility that D1ER is located in the cytosol because, in addition to the morphological evidence, basal fibre fluorescence is intense, which indicates that the biosensor is exposed to a Ca²⁺ environment as high as that reported in the SR by biochemical methods (see below). Thus, under high Ca²⁺ conditions, the cameleon shows a decrease in CFP and an increase in citrine emission, indicating increased FRET. In contrast, the cytosolic compartment is strongly buffered, and basal-free Ca²⁺ is below 50nM, preventing significant D1ER FRET. Resting myoplasmic Ca²⁺ concentration using 5 μM Fura-2 AM <4, 4–23 and 24 h after dissociation was (in nM; mean±SEM): 34.7±0.25, 13.7±0.23 and 16.1±0.22, respectively (n=10 fibres for each time group) in D1ER-expressing FDB fibres.

We do not know whether the D1ER FRET decrease in response to fibre excitation fully represents SR Ca²⁺ release, or the signal is influenced by Ca²⁺ redistribution between bound and unbound conformations of the biosensor and calsequestrin. D1ER FRET measurements in a calsequestrin knockout mouse model can probably answer this question.

SR Ca²⁺ release measured in the whole-cell configuration may underestimate Ca²⁺ depletion at the SR terminal cisternae. We do not know the relative contribution of the longitudinal or junctional SR components to our recordings. Future studies using a spot scanning confocal microscope [9] may help to answer this question.

D1ER compared to other methods used to assess SR Ca²⁺ release in skeletal muscle

Various methods have been used to assess SR Ca²⁺ in skeletal muscle fibre. Fluo-5 N AM was successfully loaded into toad muscle fibre SR [20]. In mammals, it has, to our knowledge, been applied to cardiac [34, 39] but not skeletal muscle. Prolonged tetanic activation decreased basal SR fluorescence 40% [20]. Launikonis et al. [24] measured a 10% SR depletion evoked by brief electrical stimulation using shifted excitation and emission ratioing of fluorescence. A 14% decline in SR Ca²⁺ was reported in frog muscle triggered by individual action potentials using the H⁺ displacement from EGTA loaded into the cytoplasm as an indirect measure of Ca²⁺ release [30]. In summary, taking into account the duration of fibre activation, measures of declined D1ER FRET reported here are lower than the percent of Ca²⁺ decline recorded with any other method.

The use of cameleon Ca²⁺ sensors represents an advance over previous methods. First, they can be selectively targeted to intracellular organelles using intact cells or tissues in vivo or in vitro. Second, their ratiometric property minimises confounding factors, such as mechanical artefacts and probe concentration. The slow dissociation of the cameleon/Ca²⁺ complex is a limitation, but it does not prevent reliable peak D1ER FRET transient recordings [5]. Unfortunately, quantifying Ca²⁺ concentration is limited by difficulties in calibrating D1ER in vivo.

Some studies have reported an SR Ca²⁺ concentration of ~1 mM in frog muscle fibres [40, 44], whilst a threefold to fourfold lower value (~308 μM) was noted using the Ca²⁺ biosensor D1ER [38]. To transform the D1ER FRET signal into Ca²⁺ concentration, this group used a K_d for the Ca²⁺ value of 200 μM and R_min and R_max of 1.5 and 4.0, respectively. This approach is problematic for several reasons. The K_D was measured in HeLa cells, which may not reflect muscle cell ionic composition and viscosity. Also, the D1ER was assumed to be saturated at the highest basal citrine/CFP ratio value recorded. That value was used as the R_max, but, as we have shown here, the D1ER signal expands over a wider fluorescence range, so it may be inaccurate. Additionally, treating muscle fibre with cyclopiazonic acid for 30 min does not seem sufficient to induce a maximal SR Ca²⁺ depletion for R_min estimation. Significant SR Ca²⁺ depletion requires multiple fibre stimulations with a RyR1 agonist in the presence of a SR Ca²⁺ pump inhibitor as demonstrated previously [21]. We believe that the D1ER ratiometric signal can be transformed into Ca²⁺ only if R_max and R_min are measured in the experimental fibre, which is rather difficult for the reasons discussed above. Thus, SR Ca²⁺ concentration remains unknown for intact adult mammalian skeletal muscle. On the other hand, relative changes in D1ER FRET provide information about SR Ca²⁺ dynamics and SR depletion.

Rudolph et al. [38] reported a maximal drop in SR Ca²⁺ concentration of ~17%, detected with D1ER in mouse skeletal muscle, whilst the decrease in the absolute citrine/CFP ratio ranged from 3% to 25% of the basal value. Our data show a similar mean value but for 4-CmC experiments in which the release represented a 20% decline in D1ER FRET, whilst the mean decline in response to voltage pulses was considerably less (6.4%). Perhaps nerve stimulation generates a more synchronic SR Ca²⁺ release along the muscle fibre than that recorded under voltage clamp. Given the amplitude of the D1ER FRET responses to 4-CmC application or electrical stimulation reported here, lower D1ER expression may provide a larger fractional signal. FDB fibres, electroporated with the biosensor, exhibit variable expression levels, and the data shown in Fig. 4 enable us to establish a linear relationship between the basal citrine/CFP ratio and the electrically evoked citrine/CFP amplitude ratio. Therefore, our experiments do not support the possibility that lower D1ER expression leads to a larger fractional FRET signal. Moreover, in the initial stages of this work, recordings earlier than 7–10 days after muscle electroporation resulted in fainter D1ER signals than those reported here (data not shown).

Relevance of quantifying D1ER FRET signal amplitude and dynamics

Quantifying D1ER FRET signal amplitude and dynamics helps to determine Ca²⁺-dependent regulation of RyR1 function, termination of SR Ca²⁺ release and activation of SOCE in the cellular environment.

Ca²⁺-dependent regulation of RyR function

Although cooperative Ca²⁺-dependent RyR activation has been proposed [4, 42], this work shows a direct correlation between basal citrine/CFP and this ratio’s amplitude in response to electrical or pharmacological stimulation of the fibre. Results in the literature disagree on the relationship between Ca²⁺ concentration/loading in the RyR lumenal side and Ca²⁺ flux, channel activity and Ca²⁺ sparks (for a discussion, see [36]). Although not a direct demonstration, our results are the first to support the concept that lumenal SR Ca²⁺ modulates RyR1-meditated SR Ca²⁺ release in intact muscle fibre.

Termination of SR Ca²⁺ release

A fast significant decline in lumenal SR Ca²⁺ led to the termination of Ca²⁺ sparks in cardiac myocytes [50] where a whole-cell calcium transient releases about 50% of stored SR Ca²⁺ [2, 41]. In contrast, the percent of decline in D1ER FRET signal in response to a maximal and a prolonged depolarization in the skeletal muscle fibre reported here is almost ten times less, or about a fifth, when the mean or maximal values are considered, respectively. Our results suggest that SR Ca²⁺ depletion is unlikely to be decisive in stopping SR Ca²⁺ release. Whether RyR1 is more sensitive than RyR2 to changes in SR Ca²⁺ concentration is not known; further experiments in intact whole muscle fibre are needed.

Activation of SOCE

It has been suggested that prolonged muscle stimulation activates SOCE [29, 43]. However, in the present study, 4-CmC applied for as long as 30 s elicits only a 20% decline in D1ER FRET. SR Ca²⁺ release in response to a 100-ms pulse recovers up to ~80% in 3–4s and in minutes in mouse muscle bundles [29]. Although SOCE activation in the seconds range has been reported in skinned fibres [23], whether it plays a physiological role in the intact cell is unknown. Our data show a 6.4% decrease in D1ER FRET in response to electrical stimulation, consistent with previous work on ageing muscle fibres [33]. When single FDB fibres from ageing mice are stimulated repetitively in the absence of extracellular Ca²⁺, tetanic force declines to zero [33]. Under these conditions, 4-CmC elicits a contracture of the same amplitude as that recorded in response to a maximal tetanus in the same fibre [32], indicating that significant decrease in SR Ca²⁺ consequent to prolonged fibre stimulation is unlikely. Previous studies demonstrate that repetitive contractures in the presence of a SERCA inhibitor were needed to stimulate Ca²⁺ entry into mouse skeletal myotubes or adult muscle fibres [21, 43]. These results suggest that significant SR Ca²⁺ depletion is difficult to achieve in adult and mature muscle fibre and raise concerns about the role of SOCE under physiological conditions [1]. Analysis of SR lumenal D1ER FRET combined with cytosolic fluorescence Mn²⁺ quenching evoked by muscle fibre electrical stimulation is required to answer this question.

Abbreviations

ER/SR

Endoplasmic/sarcoplasmic reticulum

FDB

Flexor digitorum brevis

SOCE

Store-operated Ca²⁺ entry

FRET

Förster resonance energy transfer

FVB

Freund virus B

CFP

Cyan fluorescent protein

YFP

Yellow fluorescent protein

RyR1

Ryanodine receptor isoform 1

4-CmC

4-Chloro-m-cresol

TEA

Tetraethylammonium

di-8-ANEPPS

Di-8-amino naphthyl ethenyl pyridinium