Calcium oscillations are important for intracellular signalling both in excitable and non-excitable cells. The endoplasmic reticulum (ER) is thought to be a major calcium store and responsible for sustained calcium oscillations. It is not clear, however, how other calcium stores, such as mitochondria, and calcium binding proteins (CBPs) cooperate with the ER in regulating this process.

Besides, many experimental results have shown that calcium always oscillates in a more complex manner: During the active phase the system exhibits bursting [1]. There have been several explanations for this form of complex oscillation [1-4], but none of them considers intracellular calcium stores other than the ER.

In this present paper ([5], BIOMD0000000039), Marhl et al. provide another possible mechanism for this complex oscillation. By setting up a three-compartment mathematical model (Fig. 1), including ER, mitochondria, and cytosol, they concluded that the complex calcium oscillations are the result of interplay between three calcium stores.

Figure 1: Three compartment model used by Marhl et al., figure taken from [5].

Figure 2: Calcium spikes, figure taken from [5].

Figure 3: Phases of calcium release, figure taken from [5].

Firstly, high frequency bursting oscillation and basic simple calcium spikes were observed (Fig. 2). Then, the authors analysed one cycle in detail. The first high amplitude calcium spike is the result of fast calcium release from the ER in phase I, as shown in Fig.3. In phase II, the fast exchange of calcium between ER and cytosol forms small, but fast bursting oscillations. Meanwhile, the slow release of calcium from mitochondria facilitates the steady binding of calcium to cytosolic buffer proteins. Finally, the silent phase (phase III) begins, during which the dissociation of calcium from cytosolic proteins is predominant. To summarise, calcium oscillation is driven by the exchange of calcium among the three calcium stores: ER, mitochondria, and cytosolic proteins. In addition, the kinetic properties of ER and cytosolic calcium binding proteins are crucial for the creation of high frequency calcium bursting oscillations.

In order to support the above conclusion, the authors further examined how the change of k_ch (a kinetic parameter describing calcium efflux from ER through calcium-induced calcium release (CICR)) affects the limit cycle and the frequency of bursting. When k_ch is 4100 (per second), there is only one cycle when plotting calcium in mitochondria against cytosol (Fig. 4); when k_ch is reduced to 4000 (per second), there is a two-fold limit cycle (Fig. 5). As a consequence, two different spiking patterns with different frequencies of bursting repeatedly appear, as shown in Fig. 6 (only two cycles are presented). When k_ch is further reduced to 2000 (per second), there is only one cycle and no bursting oscillations appear (Fig.7, 8). This indicates that when calcium cannot be released from the ER quickly, and thus cannot facilitate fast exchange between ER and cytosolic calcium binding proteins, there is no bursting but only simple oscillations appear (Fig. 4-8 were reproduced using XPPAUT).

Figure 4: Single limit cycle for k_ch=4100 s^-1. Figure generated using XPPAUT

Figure 5: Two-fold limit cycle for k_ch=4000 s^-1. Figure generated using XPPAUT

Figure 6: Time scale shown for k_ch=4000 s^-1. Figure generated using XPPAUT

Figure 7: Single limit cycle for k_ch=2000 s^-1. Figure generated using XPPAUT

Figure 8: Time scale shown for k_ch=2000 s^-1. Figure generated using XPPAUT