This article reviews cell cycle changes that occur during midblastula trans
ition (MET) in Xenopus laevis based on research carried out in the authors'
laboratory. Blastomeres dissociated from the animal cap of blastulae, as w
ell as those in an intact embryo, divide synchronously with a constant cell
cycle duration in vitro, up to the 12th cell cycle regardless of their cel
l sizes. During this synchronous cleavage, cell sizes of blastomeres become
variable because of repeated unequal cleavage. After the 12th cell cycle b
lastomeres require contact with an appropriate protein substrate to continu
e cell division. When nucleocytoplasmic (N/C) ratios of blastomeres reach a
critical value during the 13th cycle, their cell cycle durations lengthen
in proportion to the reciprocal of cell surface areas, and cell divisions b
ecome asynchronous due to variations in cell sizes. The same changes occur
in haploid blastomeres with a delay of one cell cycle. Thus, post-MET cell
cycle control becomes dependent not only on the N/C relation but also on ce
ll surface activities of blastomeres. Unlike cell cycle durations of pre-ME
T blastomeres, which show monomodal frequency distributions with a peak at
about 30 min, those of post-MET blastomeres show polymodal frequency distri
butions with peaks at multiples of about 30 min, suggesting 'quantisement'
of the cell cycle. Thus, we hypothesised that MPF is produced periodically
during its unit cycle with 30 min period, but it titrates, and is neutraliz
ed by, an inhibitor contained in the nucleus in a quantity proportional to
the genome size; however, when all of the inhibitor has been titrated, exce
ss MPF during the last cycle triggers mitosis. At MET, cell cycle checkpoin
t mechanisms begin to operate. While the operation of S phase checkpoint to
monitor DNA replication is initiated by N/C relation, the initiation of M
phase checkpoint operation to monitor chromosome segregation at mitosis is
regulated by an age-dependent mechanism. ((C) Elsevier, Paris).