Mitochondria have a well-established capacity to accumulate calcium ions entering the cytosol from the extracellular space or released from the sarco/endoplasmic reticulum (SR/ER) of many cell types including smooth muscle cells. The contribution of mitochondria to cytosolic calcium ([Ca²⁺]_i) regulation has been observed during SR Ca²⁺ release in both visceral (McCarron & Muir, 1999) and vascular smooth muscle cells (Gurney et al. 2000). In the present study, we have investigated whether release of calcium from mitochondria can replenish the agonist-sensitive SR Ca²⁺ pool.

Cells were enzymatically isolated from the uteri of 19- to 21-day pregnant rats, humanely killed by cervical dislocation after CO₂ anaesthesia. Single cells were co-loaded with mag-fluo-4 AM for the SR Ca²⁺ ([Ca²⁺]_L) measurement and fura-2 AM for [Ca²⁺]_i measurement (Shmigol et al. 2001). Individual mitochondria were visualised using Mito-Tracker Green or rhod-2.

Confocal microscopy of cells co-loaded with mag-fluo-4 and rhod-2 revealed that in uterine myocytes some of the mitochondria are localised close to the SR, which suggests a possible interaction between these two organelles during agonist-induced Ca²⁺ signalling. This was further investigated by simultaneous whole-cell measurements of [Ca²⁺]_L and [Ca²⁺]_i. A decrease in [Ca²⁺]_L accompanied by a rise in [Ca²⁺]_i was observed during application of the purinergic agonist ATP. Repetitive applications of 100 mM ATP in the absence of extracellular Ca²⁺ led to the depletion of the SR. Calcium release from mitochondria by application of 10 mM CCCP in the presence of oligomycin elicited little or no change in [Ca²⁺]_i but produced a restoration of [Ca²⁺]_L to approximately 90 % of its initial value (n = 4). Inhibition of the SR Ca-ATPase with 200 nM thapsigargin prevented the increase in [Ca²⁺]_L during CCCP application (n = 3). In the presence of thapsigargin, CCCP application now caused a substantial rise in [Ca²⁺]_i. These data are compatible with the idea that the SR takes up Ca²⁺ released from mitochondria.

When the agonist-sensitive SR Ca²⁺ pool was depleted by repetitive application of ATP in Ca²⁺-free solution, Ca²⁺ released from mitochondria in the absence of extracellular Ca²⁺ led to a partial (approximately 50 %, n = 3) restoration of the response to ATP, indicating close interaction between the SR and mitochondria.

In conclusion, our data suggest that during agonist-induced Ca²⁺ signalling, mitochondria accumulate some of the Ca²⁺ released from the SR and then, during the restoration phase, feed it back to the agonist-sensitive SR Ca²⁺ pool.

This work was supported by the Medical Research Council.

All procedures accord with current UK legislation.