Involvement of Histone H3 (Ser10) Phosphorylation in Chromosome: DISCUSSION
Pig oocytes undergo GVBD in a similar time course required to activate the Cdc2 kinase. Following the gonadotropic stimulation, pig oocytes start to accumulate cy-clin B1 at 21 h, and Cdc2 kinase activity gradually increases around GVBD at 24 h and reaches a maximal level at the metaphase I at 27 h in our culture conditions. Pig oocytes have MAP kinase in the inactive form before the resumption of meiosis, and the MAP kinase becomes phos-phorylated and activated after GVBD. It has been reported that the phosphorylation of histone H3 in the nu-cleosome is closely linked to mitotic chromosome condensation and occurs at the Ser10 position in diverse organisms. We determined the phosphorylation of histone H3 (Ser10) in maturing pig oocytes. There was no phosphorylation of histone H3 in the oocytes at the GV stage (G2 phase in cell cycle), and phosphorylation started in condensed chromosomes at the diakinesis stage. This observation differs from that of mitotic chromosomes in somatic cells, where phosphorylated histone H3 (Ser10) is first detected in pericentromeric heterochromatin in the late G2 phase, spreads throughout the chromosomes, completes in late prophase and is maintained through metaphase. Thus, the meiotic phosphorylation of histone H3 (Ser10) appears late at the time at which the chromosomes have become highly condensed. During meiosis in maize, the phosphorylation of histone H3 (Ser10) also takes place at the diakinesis-metaphase I transition. In mitosis in animal cells and meiosis in maize, histone H3 (Ser10) phosphorylation is diminished in anaphase (anaphase I in mei-osis) and disappears at telophase (telophase I). On the other hand, the phosphorylation of histone H3 (Ser10) in pig oocytes was maintained until metaphase II, even in anaphase I and telophase I. The activity of histone H3 kinase in pig oocytes showed a change similar to the phosphorylation of histone H3 (Ser10) in the present study. It increased the activity at the diakinesis stage and maintained high activity until metaphase II. This indicates the active histone H3 kinase perhaps phosphorylates histone H3 at the Ser10 residue in condensed chromosomes in pig oocytes during maturation.
PP1/PP2A is implicated in regulating meiosis in starfish and Xenopus oocytes. PP1/PP2A has recently been shown to exist in mouse oocytes, and their relocation and activity change are detected during oocyte maturation. PP1/PP2A has also been suggested to play a role in regulating chromosome condensation in mouse oocytes and mouse mammary cells. Histone H3 phosphorylation at both Ser10 and Ser28 are governed by aurora-B kinase and PP1 in mammalian somatic cells. In somatic cells, PP1 was reported to be phosphorylated and inactivated by Cdc2 kinase. Goto et al. also suggested that Cdc2 kinase may play central roles in mitosis-specific chromosome condensation through the regulation of histone H3 Ser28 phosphorylation by the inactivation of PP1. The present study showed that activation of Cdc2 kinase was not required for either chromosome condensation or histone H3 phosphorylation in pig oocytes because CL-A and OA induced chromosome condensation and histone H3 phosphorylation without the activation of Cdc2 kinase. This finding is consistent with a recent report showing that Cdc2 kinase is not required for chromosome condensation in pig oocytes treated with OA and butyrolactone I (an inhibitor of Cdks).
It has been suggested that the MAP kinase cascade induces histone H3 phosphorylation (Ser10). It has been shown that the ERKs-activated Rsk-2 kinase is directly involved in histone H3 phosphorylation in mouse and human fibroblasts. However, the involvement of MAP kinase in chromosome condensation is controversial. In mouse spermatocyte experiments, it was suggested that ERK1 is specifically activated during G2/M transition and is essential for chromosome condensation. In Xenopus oocyte experiments, it was suggested that histone H3 kinase activation, which concerned chromosome condensation during oocyte meiotic maturation, did not require Cdc2 kinase activation but rather depended on activation of the MAPK/p90Rsk pathway. Experiments in pig oocytes suggested that the inhibition of protein phosphatase promptly activated MAP kinase, which in turn induced premature chromosome condensation. In disagreement with these results, a recent study suggested that activation of the ERK1/p90Rsk2 pathway was not necessary for the phosphorylation of H3 in vivo in mouse spermatocytes. In fact, PP1/PP2A inhibitors induce rapid chromosome condensation concomitantly with MAP kinase activation in mouse and pig oocytes. Numerous observations suggest that PP2A plays an important role in the downregulation of the Ras/MAP kinase pathway. PP2A dephosphorylates and inactivates MAP kinase kinase (MEK) and MAP kinase in vitro. Therefore, OA or CL-A perhaps inhibited PP2A and induced MAP kinase activation in our experiments. In a similar time course, CL-A- and OA-treated pig oocytes showed chromosome condensation accompanied by the activation of both MAP kinase and histone H3 kinase. However, even after the activity of MAP kinase was inhibited in the oocytes, the chromosomes condensed, histone H3 (Ser10) was phosphorylated, and histone H3 kinase was activated. These results suggest that MAP kinase is not required for the chromosome condensation.
The disassembly of nuclear lamina is an essential prerequisite for GVBD. Active Cdc2 kinase phosphorylates lamins on specific serine residues and causes nuclear lamina disassembly. It is thought that oocyte GVBD is induced by active Cdc2 kinase. When pig oocytes were treated with PP1/PP2A inhibitors in our experiment, the oocytes underwent GVBD after 6-8 h. Cdc2 kinase was not activated during the culture period, but MAP kinase was activated in the oocytes after 2 h of treatment. Because MAP kinase has a weak activity to phosphorylate nuclear lamins, MAP kinase is supposed to substitute for the role of Cdc2 kinase in GVBD in the oocytes. Actually, when oocytes were treated with an MEK inhibitor and CL-A, MAP kinase activation was completely inhibited. In the oocytes, CL-A did not induce GVBD at all, while chromosome condensation occurred in the GV. Under our normal maturational conditions, the Cdc2 kinase of oocytes was still inactive at the diakinesis stage, and these oocytes had a nuclear membrane (Fig. 1B). However, the chromosomes were highly condensed with their histone H3 phosphorylation, and histone H3 kinase was activated (Fig. 2). Together with the above results, these data suggest that the activation of Cdc2 kinase is required for nuclear membrane breakdown but not for chromosome condensation in the pig oocytes.
In the present study, the change in histone H3 kinase activity accurately corresponded with histone H3 phosphorylation (Ser10) in pig oocytes and chromosome condensation. Although we did not identify histone H3 kinase in pig oocytes, accumulating data reveal that aurora-B is required for mitotic histone H3 (Ser10) phosphorylation in Drosophila and mammals, and aurora-B is a mitotic histone H3 kinase in humans. Goto et al. reported that aurora-B phosphorylated histone H3 at both Ser10 and Ser28 during mitosis in mammalian somatic cells. In both yeasts and nematodes, the reduction of PP1 activity is responsible for histone H3 phosphorylation interacting with aurora kinase. Sugiyama et al. have suggested that human aurora-B is a mitotic histone H3 kinase that associated protein phosphatases (PP1 or PP2A) as negative regulators of kinase activation. Moreover, chromatin-associated PP1 regulates aurora-B activity and histone H3 phosphorylation. Based on these findings and our present results, we propose that a balance of histone H3 kinase and PP1/PP2A activities regulates the meiotic phosphorylation of histone H3 and that the PP1/PP2A inhibition induces histone H3 phosphorylation, leading to chromosome condensation in pig oocytes (Fig. 6).