Apoptosis involves redistribution of membrane phospholipids within the lipid bilayer, nuclear fragmentation, cytoplasmic shrinkage, and plasma membrane protuberans known as blebs
[63][64][63,64]. One of the early events of the apoptotic process, before the loss of cell membrane integrity, is the phospholipid phosphatidylserine translocation to the outer leaflet of the membrane bilayer
[65]. Phospholipid phosphatidylserine externalization can be easily detected using annexin V, a phosphatidylserine-binding protein. This apoptosis stage is strongly associated with chromatin condensation events on the inner nuclear membrane
[66], a process that can be detected by labelling DNA with specific fluorochromes such as propidium iodide (PI) and 4′,6-diamidino-2-phenylindole (DAPI). Another important feature of late phase apoptosis is the DNA fragmentation that can be detected by the terminal deoxynucleotidyl transferase dUTP nick end labeling (TUNEL) assay
[67]. Correlation of cellular fragmentation with apoptosis in fragmented or normally developing human cleavage stage embryos represents a controversial finding. Yang and colleagues reported a TUNEL signal in most fragmented cleavage stage embryos (from the two-cell to eight-cell stage) but not in non-fragmented ones
[68]. Antczak and Van Blerkom reported no TUNEL signal or annexin V fluorescence in both fragments and in intact blastomeres of living fragmented embryos between 2- and 8-cell stages
[69]. Levy and colleagues reported increased annexin V staining and TUNEL assay labelling in arrested and fragmented day 2 embryos but no annexin V staining in cleavage stage embryos normally developing after thawing
[70]. Jurisicova and colleagues also proposed cellular fragmentation as a consequence of embryo programmed cell death of blastomeres in human cleavage embryos arrested at different stage of development. Several cellular fragments containing organelles and condensed chromatin within the ZP, associated with apoptosis markers (TUNEL staining) as well as caspase-2 and caspase-3 mRNA expression, were observed in these embryos
[71][72][71,72]. Studies investigating apoptosis pathways reported that some markers such as BAX and BCL (mRNA and proteins) were even expressed from unfertilized oocyte, while others such as PDCD5, BAD (mRNA), caspases, and Harakiri were expressed mainly at the blastocyst stage
[71]. Martinez and colleagues frequently observed positivity for caspase activity in fragments but rarely in normal blastomeres of arrested embryos. No differences were detected in the proportion of caspase-positive cellular fragments between 2-cell and 12-cell stage embryos, thus before and after embryonic gene activation
[73]. In 2001, with an integrated approach between retrospective data and mathematical modeling, apoptosis episodes were demonstrated from morula to blastocyst stage in viable embryos of good morphology
[74]. Hardy and colleagues assessed morphological and biochemical markers of apoptosis in fixed zona-free embryos at different developmental stages until blastocyst stage by using confocal microscopy. Nuclear morphology was evaluated after treatment of samples with DAPI, and fragmented DNA detection was evaluated by TUNEL. The levels of TUNEL-labelled cells substantially increased at blastocyst stage, while apoptosis markers were absent in cleavage stage embryos. The appearance of apoptotic markers has been associated with important steps of pre-implantation embryogenesis: the activation of the embryonic genome, the development of gap junctions, and the maturation of mitochondria. Cell–cell communication via gap-junctions, in particular, rarely present in cleavage stage embryos, was supported as a molecular requirement for apoptotic signal propagation
[75][76][75,76]. In line with the idea that selection/correction of aneuploidies are one of the mechanisms for fragmentation of the embryos, Santos and colleagues observed aneuploid blastomeres leaving the blastocyst following the activation of apoptotic pathways
[77][78][77,78]. Recently, also, the blastocoel fluid was analyzed for the presence of apoptosis markers
[43][79][43,79]. Caspase-3 protease activity has been detected in this compartment, supporting the idea that a fraction of molecules in blastocoel fluid are products of apoptotic embryonic cells
[80]. In general, however, several aspects still need to be elucidated, i.e., factors affecting blastomere apoptosis and the entity of the phenomenon in the human pre-implantation embryo at different stages.
3.2 Reactive oxygen species effect
- 2.
-
Reactive oxygen species effect-
Generated during the physiological consumption of oxygen, reactive oxygen species (ROS) can be the product of the embryo metabolism, but they may also originate from embryo surroundings
[81]. High levels of ROS along with an imbalanced formation of antioxidants is thought to result in oxidative stress, resulting in suboptimal embryos competence
[82][83][82,83]. Indeed, differently from what occurs in vivo, in which the presence of antioxidants or antioxidative enzymes in the fluid and epithelium of oviduct protects the embryo from ROS, in an in vitro culture system, levels of these compound have been demonstrated to inversely correlate to embryo developmental competence
[81][84][85][81,84,85]. Interestingly, several studies have reported a positive correlation between ROS levels in the spent medium and the fragmentation rate in human embryos at cleavage and blastocyst stage
[84][86][84,86]. Thanks to the use of imaging techniques, such as TEM and other fluorescence assays, ROS have been detected at a higher concentration in embryos with a higher rate of cellular fragmentation
[68]. While a certain amount of ROS may benefit the embryo development, as mitochondrial oxidative phosphorylation is an efficient way to produce ATP but at a cost of ROS generation, elevated ROS levels have harmful effects, including DNA damage and alteration of most types of cellular molecules
[86]. Nevertheless, a recent study reported a lack of association between ROS levels in media of cultured individually embryos (as evaluated by a chemiluminescence assay using luminol) and embryonic development or high embryo fragmentation
[87].
3.3. Membrane compartmentalization of DNA
- 3.
-
Membrane compartmentalization of DNA-
As mentioned above, micronuclei have been detected in cleavage stage human embryos as a whole chromosome or a fragment of a chromosome that is not incorporated into one of the daughter nuclei during cell division. There is no evidence of a preferential association of aneuploidy with a subtype of chromosomes: both large and small chromosomes can be sequestered
[88]. In addition, mis-segregated chromosomes and chromatid fragments encapsulated within micronuclei are dynamic entities: they may persist, rejoin the primary nucleus, or might be definitely eliminated from the embryo, in line with the theory of embryo “self-correction”. Moreover, chromosomes can undergo a specific phenomenon of “chromosome pulverization”, known as chromothripsis, that allows the reduction of one or a few chromosomal fragments into many pieces, randomly reassembled in one unique cellular event during a single-cell division
[59]. Not all the fragments are characterized by the presence of sequestered micronuclei; thus, the rearrangement of fragments does not always result in the alteration of the ploidy
[5][19][89][5,19,89].
3.4. Abnormal cytokinesis and cytoskeletal disorder
- 4.
-
Abnormal cytokinesis and cytoskeletal disorder-
It is well documented that embryos with abnormal duration of cell cycles and cytokinesis (generally, with a delayed first mitosis, an earlier start of the second mitosis, and a longer duration of the third mitosis) are more likely to be fragmented
[13]. An incorrect cell cycle may result in genomic alterations because the cell does not have enough time to correct any eventual error during DNA replication. Alikani and colleagues reported that loss of interplay between the spindle complex and cortical microfilaments was associated with blebs and cellular fragments formation. In addition, the authors demonstrated that treatment with cytokinesis inhibitors prevented cytokinesis as well as fragment formation, supporting a cause–effect relationship
[13]. Stensen and colleagues also linked the rate of embryo fragmentation with the duration of meiotic process. A delay in the oocyte meiotic division (formation of the meiotic spindle 36.2 h after human chorionic gonadotropin injection) was associated with higher fragmentation rates [50–100%] in resultant embryos
[90]. The reason underlying this observation may be related to cell cycle defects implicated in oocyte aneuploidy involving alterations in chromosome pairing, recombination, and spindle assembly, resulting in a delayed meiotic cell cycle
[91][92][93][91,92,93]. No correlation, instead, was found between fragmentation and other spindle characteristics (i.e., a delay in its formation and the angle calculated between the first polar body and the meiotic spindle). Lastly, according to the same group, the process of fragmentation was more pronounced during the early phases of cell division, when the maternal genome is still active. After the activation of the embryonic genome, the tendency of human blastomeres to fragment would be lost. Extruded blastomeres from these embryos would express maternal instead of embryonic transcripts, during an inappropriate timing for the developmental stage
[90].
3.5. Extracellular vesicles formation
- 5.
-
Extracellular vesicles formation-
Human embryos can secrete EVs in their culture media that can be easily taken up by endometrial cells
[15]. They are formed through multiple biogenetic pathways: (i) Exosomes are generated from the endosomal system by the formation of late endosomes, which are formed by inward budding of the multivesicular body (MVB) membrane. Invagination of late endosomal membranes results in the formation of intraluminal vesicles (ILVs) within large MVBs
[94]. In the next step, MVBs have two fates: most ILVs are released into the extracellular space upon fusion with the plasma membrane or, alternatively, these components are trafficked to lysosomes for degradation
[95][96][95,96]. (ii) Microvesicles, instead, are formed through the outward budding and fission from plasma membranes. In contrast to exosome formation, the secretion of microvesicles requires the lipid microdomains at the membrane and a reorganization of the actin–myosin cytoskeletal network
[97][98][97,98]. A possible association between the EV quantity in the spent culture media and embryo quality and competence has been suggested
[99][100][101][99,100,101]. Specifically, fewer EVs have been reported in spent culture media of embryos leading to successful pregnancy than in those who failed, suggesting that a good quality and competent embryo releases different amounts/types of EVs compared to a low-quality embryo
[102][103][104][105][102,103,104,105].