In 2008, Klevecz et al. introduced a new concept of a ”cell attractor” which tunes deterministic cellular noises to gate cells into the S phase, to explain how the cell cycle is timed and how stable cellular phenotypes are maintained [9]. The Ca²⁺ oscillation in the cell may be the most likely candidate of cell attractor, because Ca²⁺ oscillation can regulate both cell cycles and circadian rhythms. In the case of cultured mouse ES cells, the spontaneous Ca²⁺ oscillation is generated by an IP3-mediated Ca²⁺ release [10]. A large portion of Ca²⁺ oscillations was found to continue from the G1 to the S phase of the cell cycle, and a smaller portion was detected in the G2/M phase [10].

Although this Ca²⁺ oscillation during the cell cycle is clear, the triggering signal for this oscillation had not been identified. If the Ca²⁺ oscillation of the cells is triggered by released IP3, the oscillation is quite similar to that of fertilization. In the case of mammalian eggs and ascidian oocytes, it was established that Ca²⁺ oscillation is triggered by IP3 [11]. The difference in the Ca²⁺ oscillation of ES cells and that of fertilized eggs is in the trigger signal. Mammalian eggs and ascidian oocytes can be activated by sperm, an injection of sperm factor, an injection of IP3, an application of calcium ionophores and pricking [6], whereas the cell cycle-dependent Ca²⁺ oscillation of ES cells is always observable. The amplitude and wave shape of Ca²⁺ oscillation are affected to some extent by treatment with Ca²⁺-free medium with 10 μM ryanodine [3]. These findings indicate that ES cells have always been activated by an unidentified stimulation.

What type of stimulation is likely to activate somatic cells to generate pluripotent ES cells from somatic cells? Gilkey examined the roles of Ca²⁺ and pH in activation of Medaka fish eggs [3]. His findings can be summarized as follows, based on the results of the microinjection of pH buffers into the Medaka fish eggs and those of exposing the eggs to reagents that should alter cytoplasmic pH:

1) A transient increase in calcium is necessary for egg activation.

2) Altering the cytoplasmic pH does not affect the egg activation per se.

3) Altering the cytoplasmic pH does change the threshold freecalcium concentration required to activate the egg.

According to his results, Gilkey found that the threshold concentration of free calcium to elicit the Ca²⁺ oscillation/wave was pH dependent. These threshold values are fairly below the estimated value, 30 μM from Ca²⁺ wave; the rate of propagation of the Ca²⁺ wave through the egg’s cytoplasm is a function of cytoplasmic pH. Accordingly, Gilkey concluded that if the proton concentration increases by 10-fold (i.e., pH decreases by 1 unit), the amplitude of the Ca²⁺ wave will be amplified by 10-fold [3].

Why does cytoplasmic pH alter Ca²⁺ wave amplitude? Gilkey proposed that “perhaps the calcium binding site for calcium-induced calcium release (CICR) can also bind one proton, and proton binding competitively inhibits calcium binding to this site”. His model postulated a two-calcium-pool model: Ca²⁺ is released from the IP3 receptors (IP3Rs) and the ryanodine receptors equipped endoplasmic reticulum, for example. This was the first proposal of a two-calciumpool model based on the experimental results. However, Gilkey did not consider the contribution of water properties in Ca²⁺ oscillations.

Only a few biologists have theorized that the intracellular water is different from general bulk (free) waters [8], because the interior of a cell is overcrowded with relatively huge macromolecules, leaving little space for water [12]. Water molecules are in contact with macromolecule patchwork of the macromolecular surface domain, regardless of whether the macromolecules are hydrophobic or hydrophilic [8]. When water molecules bound to consecutive polar amino acids (e.g., Asn, Cys, Gln, Ser, The, Tyr) in a polypeptide chain, they can bind to form H-bonded lines. These lines of water molecules are good conductors of protons, but not good conductors of electrons. In this model, protons can move by a successive flip-flop step from the first water molecule to the last water molecule of the chain, rather than migrating along the water chain [13]. This process was eventually named “the proton wave.”

This domino-like wave motion is faster than diffusion. Only the Ca²⁺ wave observed in living cells is always accompanied by a proton wave signal which is more rapid than diffusion, whereas the proton wave is transferred according to the Coulomb interaction on the water surface. In contrast to positively charged Ca²⁺, negatively charged molecules such as IP3 and/or cyclic nucleotides will move by the diffusion process in the small gap under the plasma membrane. It can thus be concluded that Ca²⁺ is a good second messenger which can reach the mitochondria and other endoplasmic reticula to produce Ca²⁺ oscillations.

The following question remained: “Does the Ca²⁺ oscillations change with cell cycle or not?” Larman et al. addressed this question in a 2004 study [14]. They observed that Ca²⁺ oscillation was obtained in the M phase but not in the G1 or G2 phase. What is the difference between the two (M vs. G1/G2)? A few hypotheses have been put forth [14,15], but in the next section we propose a new model based on the water state in the cell.

Changes in water properties and water content during cell cycle were initially investigated by Beall et al. using nuclear magnetic resonance relaxation time (T1) in synchronized HeLa cells [16]. The T1 in the M phase was found to be the highest, and that in the S phase was the lowest. In the G1 and G2 phases, the T1 values showed a transient state; gradual decreasing in the G1 and recovering in the G2 phase. These data could be understandable if low T1 values correspond to free water and the high T1 values correspond to bound water. This explanation can also be supported by the variation of the water content in the cell cycle [16]. Such changes in water properties according to the cell cycle were further investigated by other groups using fertilized sea urchin eggs [17,18]. This changing of water properties along with the cell cycle completely corresponds to Ca²⁺ oscillations described above.

The residual question is why bound water in the M phase starts to be free water in the G1 phase and vice versa in the G2 phase. Important hints for approaching this problem are in the Hoffmeister series, as was suggested by Wiggins [19]. In brief, bound water can be changed to free water when Na⁺ enters the cytosol. As was observed previously, s study of Xenopus oocytes showed a depolarization of the membrane potential by fertilization, which will induce Na⁺ entry into the cell [20]. It was also found that the proton current can flow through (amiloridesensitive) Na⁺ channels in the hamster taste cells [21]. Taken together, these findings lead us to speculate that a protocol of acidic stimulation (as Obokata et al. did for reprogramming [1]) will induce Na⁺ entry, which will induce an increase in intracellular proton level initially; then the water state will be changed gradually from the bound state to the free state, which will initiate a new cell cycle as described above.

Before an elucidation of the aging effect on the induction of pluripotency, we must understand the two-calcium-pool model of Ca²⁺ oscillation, because Ca²⁺ oscillation is thought to be a necessary and sufficient condition in the creation of pluripotency. In the two-calciumpool model, the functioning of both the IP3R and the ryanodine receptor are essential [22]. Cell cycle-dependent Ca²⁺ oscillation was observed and established not only in theoretical models for intracellular Ca²⁺ oscillation [23], but also in mammalian eggs at fertilization [24], in mouse embryos [14] and in mouse ES cells. Therefore it is reasonable to assume that the IP3R is involved in pluripotential ES cells. If such an assumption is correct, it may explain why aged somatic cells cannot have pluripotency by acidic stimulations.

We recently observed that 24-h H2O2-treated human keratinocyte showed poor Ca²⁺ responses (i.e., < 10% of the normal response) for purinergic G protein-coupled receptor (P2Y receptor) activation by ATP, but the responses completely recovered when we reduced the treatment period with hydrogen gas-saturated phosphate-buffered saline solution for 1 h (Hsu et al., unpublished data). These poor responses were found to be due to the oxidation of the IP3R. Since the IP3R is essential to the Ca²⁺ oscillation in ES cells, the IP3R in aged somatic cells cannot continue Ca²⁺ oscillation, resulting in these cells’ failure to have pluripotency.

A summary of the information presented in this review is given in Figure 1 Unfertilized eggs, a type of somatic cells, will be changed to embryos after fertilization with the Ca²⁺ oscillation-associated cell cycles. Similarly, somatic cells obtained from a young body will have pluripotency by acidic stimulation, whereas somatic cells from an aged body cannot have pluripotency because of the deletion of Ca²⁺ oscillation by oxidized (aged) IP3Rs.

Figure 1: Schematic comparison of ES cells and somatic cells from the standpoint of proton signaling, which regulates Ca²⁺ oscillation and waterproperty changes associated with the cell cycle.

We are grateful to Dr. Kazuaki K. Yokoyama (Graduate Institute of Medicine, Kaohsiung Medical University) for his critical reading and discussion.