Cell cycle parameters of early mouse embryo

Read Also

Numerous studies have been performed in order to precisely determine the cell cycle parameters during early stages of development and clearly established that these parameters are greatly modified during pre-implantation development. Differences were observed between the values obtained in these studies that stem from differences in experimental procedures as well as influences from the genetic background [3] and the parental origin of the genomes [4,5] (table 1). It is nevertheless possible to synthesize these observations as follows (figure 1 and table 1). The first two divisions last approximately 20h. Four to ten hours after fertilization, replication begins and lasts between 4 and 8h. It should be noted that replication is detected first in the male pronucleus [6,7]. G2/M phase length is estimated to 3-5h. Interestingly, the duration of the first mitosis (120 min) is almost twice longer than the second (70 min) and this increase seems to be due to a transient metaphase arrest independent of the spindle assembly checkpoint (SAC) [8]. The second S phase lasts approximately 6h. Gap phases of the second division are very different since G1 is extremely short (1-2h) [9] and G2 very long (12 to 16h) [3,10-12]. Strikingly, it is during this unusually long G2 phase that occurs the major phase of the zygotic genome activation (ZGA) in the mouse [13]. The following four divisions occurring between st-4 and st-64 are more homogeneous in terms of duration (10-14h; G1:1-2h, S:7h, G2/M:1-5h). Importantly, during the 5^th cleavage (between st-16 and st-32), two cellular populations are formed, polarized external cells and apolar internal cells, which seem to differ in their cell cycle parameters [14,15]. As development proceeds, external cells give rise mainly to trophectoderm (polar and mural) while internal cells contribute to the inner cell mass (ICM) that will then segregate into epiblast and primitive endoderm. Based on mitotic index examination, A.J Copp observed that while the number of cells composing the mural trophectoderm increases considerably in late blastocyst embryos, mural trophectodermal cells divide slower than polar ones [16-18]. This observation leads to the proposal that trophectodermic cells are essentially generated in the polar region and then migrate (actively or passively) in the mural region. In the ICM, mitotic index examination revealed partial synchronization of cell divisions between st-30 and st-150 [16]. After implantation in the uterus, the embryo undergoes gastrulation, a very active phase of development during which the three embryonic layers are committed and organized in three dimensions. Important modulations of cell cycle parameters happen during this key developmental process. One of the most salient changes concerned trophoblast giant cells, which undergo endoreplicative cycles, consisting of repeated rounds of S phase without intervening mitosis, until they acquired a DNA quantity equivalent to 500 haploid genomes [19,20]. Endoreplication can be first detected in late blastocysts where approximately 5% of the cells are polyploid [19]. Important modifications are also observed in the epiblast where cell division pace greatly accelerates (table 2). The fact that in all studied species, gastrulation is preceded by fast cell cycles [21], suggests that rapid amplification of embryonic cells is necessary for proper cell type diversification and embryo patterning. Successful gastrulation requires that cell cycle regulation is tightly coordinated to signaling pathways and cell movements. Studies in mice and rat revealed the existence of a region of remarkably fast cell cycle (2-3h), called the proliferative zone, which lies in close proximity to the primitive streak [22,23]. Interestingly, mesodermal cells that are generated from the primitive streak cell population do not keep proliferation with such a high rate [22-24] (table 2), indicating that the transition between embryonic ectodermal to mesodermal cells implies highly dynamic regulation of cell cycle parameters. How such modifications are controlled and whether they play a direct role in the commitment of the mesoderm and endoderm cell lineages remain unanswered.

Checkpoint activities

Contrary to early mammalian development, rapid cleavage cycles lacking intervening G1 and G2 gap phases are found in early embryos from other major phyla (reviewed for example in [21,25,26]). These rapid cycles either lack or display weak checkpoint activities, a situation which, to some extends, seems to be different to that observed in mouse pre-implantation embryos.

DNA damage checkpoint

Genome integrity maintenance is a key process that requires efficient DNA damage detection and DNA repair processes. In response to DNA damage, different checkpoints are activated leading to a cell cycle delay or arrest. Delayed progression of the cell cycle allows time for either repair or elimination of genetically unstable cells by apoptosis. Such adaptative response seems absent from embryonic cycles of various species. Indeed, inhibition of replication does not prevent mitotic entry in drosophila [27], zebrafish [28] or xenopus [29]. In contrast, similar inhibition induces a strong cell cycle arrest in mouse pre-implantation embryo [30,31]. Early mouse embryos also respond to DNA damages induced by irradiation. While the nature of this response depends largely on the quantity and the type of radiation used, two main conclusions can be drawn from the literature: i) whatever the age of the exposed embryo, radiations provoke changes in cell cycle parameters [32] and induce apoptosis [11,33,34] ii) sensitivity to irradiation is highly dependent on the developmental stage which is exposed [35]. Interestingly, irradiation of early post-implantation embryos with low doses of X-rays does not result in marked cell cycle delay but rather induces a strong p53 and ATM dependant apoptotic response [36]. Thus, it appears that, at that time of development, the main pathway used by embryonic cells to respond to DNA damage is cell elimination, probably because the cell cycle regulation during this period of extreme proliferation is not compatible with cell cycle arrest and accurate repair of DNA damages. Similar conclusions can be drawn from the analyses of genetic invalidation models. Although Atm [37] or Chk2 [38] are dispensable to embryonic development, embryos lacking Atr or Chk1 die soon after implantation exhibiting high degree of chromosomal fragmentation [39-41]. A similar phenotype was observed in embryos lacking proteins involved in DNA double strand break repair such as Rad50 [42] and Nbs1 [43]. Finally, early embryonic lethality was observed following inactivation of several genes encoding for DNA repair machinery components such as Fen1 [44], Rad51 [45], Xab2 [46], or Xpd [47]. Interestingly, while non-homologous end joining (NHEJ) repair mechanism has been shown to be extremely active after fertilization [48], inactivation of several genes involved in this process, like DNA-Pkcs [49], DNA ligase IV [50] or Xrcc4 [51] does not lead to early embryonic lethality. It is important to note that in cases where early embryonic lethality was observed, defects were manifest by the time of implantation at the earliest. Several explanations might account for this observation. First, the presence of maternal stores might compensate for the absence of a zygotic product. Second, errors or DNA damages accumulation over several cell cycles might be necessary in order to induce a patent phenotype. Anyhow, it probably also reflects the transition in the cell cycle regulation and the increased sensitivity towards DNA damages that occurs after implantation.

Mitotic checkpoint

During mitosis, improper attachment of kinetochores to microtubules triggers the spindle assembly checkpoint (SAC), preventing the onset of anaphase and potential incorrect segregation of the genetic material into daughter cells. Several lines of evidences indicate that SAC is operating during mouse early development. First, pre-implantation embryos exposed to drugs interfering with spindle assembly arrest very efficiently in metaphase [52-55]. Second, key components of the SAC such as MAD2 and BUBR1 localize to kinetochores of unattached chromosomes of zygotes and blastocysts (evoked in [8,56,57]). Finally, early lethality of embryos deficient for various component of the checkpoint such as Apc10/Doc1 [58,59], Bub3 [60], BubR1 [61], Emi1 [62] and Mad2 [63] demonstrates that SAC plays a critical role in mitotic progression of early embryonic cells. SAC regulates progression of mitosis by controlling the activity of the APC/C complex, which triggers the degradation of several key mitotic proteins. One of the substrates of APC/C is securin, an inhibitory chaperone of separase, which is the protease triggering sister chromatids separation at the anaphase onset. Not surprisingly, inactivation of Separase leads to an early embryonic lethality associated with polyploidy and abnormal centrosomes number [64,65]. In contrast, Securin is not essential for either mitosis or meiosis [65-67]. Finally, inactivation of genes encoding centromeres or kinetochores structural proteins leads to abnormal mitotic figures and to peri-implantation lethality (CenpA [68]; CenpC [69]; CenpE [70]; Incenp [71]; Survivin [72]).

RB-dependent G1 checkpoint

Cell cycles of the early mammalian embryo not only differ from early mitotic cycles found in other organisms but also from mammalian somatic cell cycles. An important difference concerns cell cycle regulation in G1. In somatic cells, the length of G1 phase can vary considerably in response to environmental stimuli such as for example mitogenic factors that impinge on cell cycle progression through Myc and Rb pathways. Mouse early cleavages are characterized by a short G1 phase. Consistently, pre-implantation development is independent of exogenous growth factors (see for example [73,74]). In addition, although genes taking part in the Rb pathway are expressed dynamically during early mouse development [75-77], they are dispensable for this period of development (for review, see [78,79]). Two non-exclusive explanations might account for the lack of RB-dependent G1 checkpoint activity before implantation. Iwamory and collaborators observed that Rb mRNA and proteins were barely detectable before the late blastocyst stage suggesting that low levels of RB is necessary for shortened G1 phase. Accordingly, they observed that forced expression of RB by plasmid injection into zygotes induced developmental arrest before the morulae stage [75]. In another study, Xie and collaborators observed phosphorylated RB proteins throughout pre-implantation development, suggesting that regulation of RB phosphorylation state rather than level of expression might be responsible for the short G1 phase [77]. Interestingly, mouse embryonic stem cells cells (ESC), which derived from the inner cell mass of blastocyst stage embryo, also lack the RB-dependent control of the G1/S transition that characterizes somatic cells ([80] and reviewed in [81]). Recently, it has been shown that rhesus monkey [82] and human ESC [83,84] share such characteristic. Therefore, studying ESC cycle parameters might be a relevant mean to apprehend cell cycle regulation during human early embryonic development.