Ca2+-mobility in the sarcoplasmic reticulum of ventricular myocytes is low

Abstract

The sarcoplasmic reticulum (SR) in ventricular myocytes contains releasable Ca(2+) for activating cellular contraction. Recent measurements of intra-SR (luminal) Ca(2+) suggest a high diffusive Ca(2+)-mobility constant (D(CaSR)). This could help spatially to unify SR Ca(2+)-content ([Ca(2+)](SRT)) and standardize Ca(2+)-release throughout the cell. But measurements of localized depletions of luminal Ca(2+) (Ca(2+)-blinks), associated with local Ca(2+)-release (Ca(2+)-sparks), suggest D(CaSR) may actually be low. Here we describe a novel method for measuring D(CaSR). Using a cytoplasmic Ca(2+)-fluorophore, we estimate regional [Ca(2+)](SRT) from localized, caffeine-induced SR Ca(2+)-release. Caffeine microperfusion of one end of a guinea pig or rat myocyte diffusively empties the whole SR at a rate indicating D(CaSR) is 8-9 microm(2)/s, up to tenfold lower than previous estimates. Ignoring background SR Ca(2+)-leakage in our measurement protocol produces an artifactually high D(CaSR) (>40 microm(2)/s), which may also explain the previous high values. Diffusion-reaction modeling suggests that a low D(CaSR) would be sufficient to support local SR Ca(2+)-signaling within sarcomeres during excitation-contraction coupling. Low D(CaSR) also implies that [Ca(2+)](SRT) may readily become spatially nonuniform, particularly under pathological conditions of spatially nonuniform Ca(2+)-release. Local control of luminal Ca(2+), imposed by low D(CaSR), may complement the well-established local control of SR Ca(2+)-release by Ca(2+)-channel/ryanodine receptor couplons.

FIGURE 1

Protocol for stabilizing SR Ca²⁺ load. Rat myocytes. (A) Fluo-3 fluorescence (F/F₀) was measured in cells paced at 2 Hz in NT, and then superfused with 0Na0Ca medium (fluorescence normalized to diastolic levels in NT). Ca²⁺-waves were observed in the first 2 min in 0Na0Ca. After waves ceased, diastolic F/F₀ was ∼1. (B) In 0Na0Ca solution with SERCA inhibited by 10 μM CPA (stabilizing solution, SS), RyRs were dose-dependently inhibited by 0.3 or 2 mM Tet. Inclusion of Tet in SS-superfusate blocked Ca²⁺-waves. At both doses of Tet, resting F/F₀ stabilized (at a 25% lower level) with a half-time of ∼36 s. Cells stabilized in SS + Tet were then used for the main experiments. (C) The reduction of resting F/F₀ observed in the presence of SS + Tet was absent in rat cells preloaded with 100 μM BAPTA-AM. (A–C) Mean of 7–10 cells.

FIGURE 2

Mobilizing SR Ca²⁺ with caffeine. (A) Rat myocyte, AM-loaded with Fluo-3 and initially paced at 2 Hz in NT was then superfused with 0Na0Ca medium containing 10 μM CPA (stabilizing solution, SS), interspersed with three 30-s episodes of SS + 10 mM caffeine. (Upper panel) time course for F/F₀ averaged along a longitudinal line scan (inset indicates position of line scan axis); (middle panel) F/F₀ line scan; (lower panel) cell shortening. The first caffeine exposure mobilizes all SR Ca²⁺. (B) Different rat myocyte exposed to SS containing 10 mM caffeine and 10 mM [Ca²⁺]_o; note lack of -recovery, suggesting block of sarcolemmal PMCA. Final recovery in 0Na0Ca (with zero CPA) is attributable to SERCA and PMCA activity.

FIGURE 3

Assessing SR Ca²⁺-leak. Myocytes paced at 2 Hz in NT were then superfused with SS containing 0.3 or 2.0 mM Tet. (A) After 1 min in SS + 0.3 mM Tet, rat myocyte was exposed to Tet-free SS + 10 mM caffeine for 30 s. (Upper panel) longitudinal line scan average for F/F₀; (middle panel) F/F₀ line scan (y axis normalized to cell length); (lower panel) cell shortening. (B) Different rat myocyte, exposed to caffeine after 4 min in SS + 0.3 mM Tet. Smaller Ca²⁺-release is due to longer preceding period of SR-membrane Ca²⁺-leakage. (C) (i) Peak F/F₀ during caffeine-exposure, measured after 1–6 min delay (normalized to F/F₀ at switch into SS) in rat myocytes in 0.3 mM (green) or 2 mM (black) Tet, and in guinea pig myocytes in 2 mM Tet (gray) (n = 5–10 cells/data-point). (ii) Cell shortening responses to caffeine in rat myocytes. Rundown of both F/F₀ and cell shortening gives time course of SR depletion caused by Ca²⁺-leak.

FIGURE 4

Measuring SR Ca²⁺-mobility. Caffeine is used to estimate the decline of distal [Ca²⁺]_SRT, while Ca²⁺ is being drained from the proximal end of SR. (A) Dual-microperfusion apparatus is used to deliver two solutions: SS (without Tet or with 0.3 or 2 mM Tet) and SS + caffeine (10 mM with no Tet). Myocyte is first bathed in SS + Tet (configuration 0). Throughout the experiment, intracellular Fluo-3 fluorescence (F/F₀) is averaged in two ROIs, each 20% of cell length, one positioned proximally (ROI-P) and one distally (ROI-D), at the ends of a line scan (see schematic diagram shown to the right of configuration 0). SS + caffeine is then applied to proximal end of myocyte to start SR-drainage (configuration 1); after a 30–240 s delay, SS + caffeine is applied to whole cell (configuration 2) to assess remaining SR Ca²⁺-content. At the end of the experiment, cells are returned to 0Na0Ca, free of CPA, Tet, and caffeine (not illustrated, denoted by arrow 3 in B and C). (B) SS microstream (but not SS + caffeine microstream) contained 0.3 mM Tet. (i) Rat myocyte; delay of 30 s between configurations 1 and 2. (Upper panel) F/F₀-time course for [Ca²⁺]_i averaged in ROI-P (black trace) and ROI-D (gray trace). Dashed line shows predicted decline of caffeine-mobilizable Ca²⁺-release because of SR Ca²⁺-leak (from Fig. 3). (Middle panel) Longitudinal F/F₀ line scan (y axis normalized to cell length). (Lower panel) cell shortening. (ii) Rat myocyte; delay of 240 s between configurations 1 and 2. (C) SS microstream (but not SS + caffeine microstream) contained 2.0 mM Tet. (i) Rat myocyte; delay of 120 s between configurations 1 and 2. (ii) Guinea pig myocyte; delay of 180 s between configurations 1 and 2. (D) Ratio of distal fluorescence rise (configuration 2) to initial proximal rise (configuration 1), ΔF2/ΔF1, for different delay times. Rat myocytes with 2 mM Tet (black), 0.3 mM Tet (green), and zero Tet (blue). Guinea pig myocytes with 2 mM Tet (gray) (4–12 cells/data point). Time constants (τ) obtained from monoexponential fits.

FIGURE 5

Distal SR depletion with proximal caffeine is unaffected by Fluo-3. (A) Myocyte AM-loaded with carboxy-SNARF-1, a non-Ca²⁺ binding fluorophore. Upper panel: fluorescence line scan (y axis shows cell length) used to monitor cell shortening (carboxy-SNARF-1 excitation at 514 nm, emission at 580 nm). (Lower panel) time course of cell shortening (derived from upper panel) in response to caffeine. Protocol of Fig. 4 was applied. Cell initially exposed to SS + 0.3 mM Tet. At circle 1, the proximal end of the cell exposed to SS + 10 mM caffeine; at circle 2 the whole cell exposed to SS + 10 mM caffeine. (B) Ratio of second/first shortening-event is plotted versus time delay between circles 1 and 2 in the presence of intracellular carboxy-SNARF-1 (black; n = 8, 7) or Fluo-3 (gray; n = 6–13). Time constants (τ) obtained from monoexponential fits. The presence of the Ca²⁺-binding fluorophore, Fluo-3, does not affect the rate of decline of distal contraction (and hence distal SR-depletion) during proximal caffeine exposure.

FIGURE 6

Quantifying D_CaSR. (A) (i) Schematic diagram of the simple Ca²⁺-diffusion model, incorporating SR-membrane leak and proximal SR-exposure to caffeine (see Appendix 1). (ii) Best-fitting of diffusion model (SERCA activity = 0, simulating CPA-addition) to data in Fig. 4 D gives displayed values for D_CaSR (0.3 mM Tet, green; 2 mM Tet, black: rat myocytes; 2 mM Tet, gray: guinea pig myocytes). Note unique value for D_CaSR (8–9 μm²/s). (iii) Best-fitting when SR Ca²⁺-leak is ignored; note lack of unique value for D_CaSR (42, 23, and 10 μm²/s for rat myocyte data (Fig. 4 D) with 0, 0.3, and 2 mM Tet, respectively). (B) (i) Schematic diagram of diffusion-reaction model showing principal reaction elements (see Appendix 1). (ii) Displayed D_CaSR: best-fit values averaged between 0 and 240 s. (Green) rat, 0.3 mM Tet; (black) rat, 2 mM Tet; (gray) guinea pig, 2 mM Tet. Note similarity with D_CaSR values from Aii. (C) Output of cellular diffusion-reaction model simulating experimental protocol (Fig. 4). (i) Time course of F/F₀ from simulated line scan in proximal (black) and distal (gray) ROIs. (ii) Simulated line scans for F/F₀, [Ca²⁺]_i and [Ca²⁺]_SR.

FIGURE 7

Effect of D_CaSR on longitudinal [Ca²⁺]_SR profile during simulated local Ca²⁺-release. Diffusion-reaction model was run to simulate local SR Ca²⁺-release (see Appendix 1). [Ca²⁺]_SR was calculated 250 ms after increasing SR-membrane permeability to 1.35 μM/s in a central 2 μm length of SR. This estimate of D_CaSR (9 μm²/s) supports large and localized SR-depletion, with only small effects on adjacent luminal regions. Higher D_CaSR significantly depletes SR Ca²⁺ from regions beyond the release site. (Dashed line) 5% depletion threshold.

FIGURE 8

Modeling Ca²⁺-diffusion in SR nanoarchitecture. (A) (i) 2-D geometry for a unit of sarcomeric SR (jSR-cisterna connected to four nSR-tubes) triangulated for diffusion modeling. (ii) To simulate jSR Ca²⁺-recovery after local Ca²⁺-release, [Ca²⁺]_jSR was set instantaneously to 0.25 (50% depletion), whereas [Ca²⁺]_nSR was set initially to 0.5 mM and the distal ends of the nSR elements were coupled to an infinite 0.5 mM Ca²⁺ source. The time constant (τ) of recovery of [Ca²⁺]_jSR was then derived for a range of jSR Ca²⁺-mobility values (D_CajSR), with Ca²⁺-mobility in nSR assumed to be 1000 μm²/s. The predicted time courses of change for [Ca²⁺]_jSR and [Ca²⁺]_nSR are shown in the inset. Filled circle denotes τ = 29 ms, which has been determined experimentally for rabbit SR Ca²⁺-blinks. (iii) Spatial [Ca²⁺]_SR map after 29 ms of jSR-refilling for a D_CajSR of 1.8 μm²/s. (B) (i) 63 × 10 SR units based on Ai, coupled into a 2-D network for simulating whole-SR during proximal caffeine exposure (see Appendix 2). (ii) Predicted relationship between global D_CaSR and D_CajSR. D_CajSR of 1.8 μm²/s (derived from normal Ca²⁺-blink recovery rate) predicts a global D_CaSR of 8 μm²/s, similar to the value measured experimentally. In contrast, a high global D_CaSR (60 μm²/s) predicts a relatively high D_CajSR (41 μm²/s), which would then predict an ultrafast Ca²⁺-blink recovery τ of <1 ms (see Aii). (Inset) Close-up of D_CaSR range between 0 and 10 μm²/s. (iii) Model of nSR/jSR unit featuring a minimal length nSR (0.65 μm), divided into snSR and inSR. Diffusion in the inSR is slower than in the snSR (see Appendix 2 for details).