The phenomenon of chromosomal mosaicism in the human embryo has been known for some time. It has been estimated that the placenta has a different karyotype from the fetus (confined placental mosaicism—CPM) in 1% of conceptions. CPM was initially detected when first-trimester fetal karyotyping after CVS showed discrepancies between chorionic cells and the embryo proper.
All of the studies performed in recent years to analyze chromosomes in embryos have been carried out using embryos generated by IVF, which may not be representative of in-vivo development. However, some showed that embryos from natural cycles also showed high levels of nuclear abnormalities, and studies of normally conceived pregnancies have revealed that 60% of abortions are chromosomally abnormal.
The majority of human preimplantation embryos are mosaic in relation to chromosome numbers, with the embryo containing a number of both euploid and aneuploid cells. Errors in mitosis during the first cleavage divisions lead to this mosaic pattern. It was thought that the high rate of human pregnancy loss in both spontaneous and IVF cycles was due to this chromosomal inaccuracy. However, recent studies have shown that mosaic human embryos can develop into normal healthy babies.
Studies on embryos donated for research and those with abnormal fertilization (such as polyspermic embryos) have shown that the latter were highly abnormal, in agreement with karyotype data. Chromosomal mosaicism was observed in the majority of cases, but normal diploid embryos from supposedly polyspermic embryos were sometimes identified. This may have been due to misidentification of a vacuole as a pronucleus. As expected, embryos from older women show high levels of chromosome abnormalities, but, interestingly, normally fertilized, normally developing embryos also show high levels of chromosomal abnormalities.
Patterns have been categorized into four major groups:
Uniformly abnormal, such as Down’s syndrome or Turner’s,
Mosaic, where usually both diploid cells and aneuploid, haploid or polyploid nuclei are present,
Chaotic embryos, where every nucleus shows a different chromosome complement.
The data from FISH and CGH analysis show a higher rate of abnormalities than has previously been reported from karyotyping data. However, since mosaic and chaotic embryos are common, if only one or two cells are analyzable from an embryo, then karyotyping would underestimate the level of chromosome abnormalities.
Normal, abnormal and mosaic embryos have all been observed in fetal development. As mentioned previously, confined placental mosaicism has been found in an estimated 1% of conceptions. The presence of two cell lines could arise due to an abnormal chromosome arrangement caused by a post-zygotic event, or the chromosome loss from a trisomic embryo, which restores the diploid state (trisomic rescue). Several mechanisms would indicate that these abnormal cells are more likely to be found in the TE, and hence the placenta.
First, only a few cells from a blastocyst give rise to the embryo, and it would be unlikely that the abnormal cells would be found in the embryo. Second, in most cases a fetus with abnormal chromosomes will not be compatible with life. The chaotic group of embryos was an unexpected finding, as such embryos have not been observed in later stages of embryonic development, probably because these embryos would arrest and fail to implant.
Multi-nucleated blastomeres have been reported and confirmed from both karyotyping and FISH analysis. The presence of such blastomeres maybe more embryos and may occur more readily in some patients. Binucleate blastomeres have been observed in mouse embryos at the morula stage, and it has been suggested that these blastomeres might be the precursors for mural TE giant cells. However, in human embryos the binucleate cells appear at cleavage stages before TE differentiation.
Binucleate cells may arise from asymmetrical cytokinesis so that one daughter cell contains two nuclei and the other is a nucleate. In addition, aneuploidy can activate a spindle-apparatus checkpoint in different types of cells, causing multi-nucleation in that cell. It is possible that blastomere multi-nucleation may be the equivalent of a cell cycle checkpoint, which can convert a mosaic embryo to one that is euploid.
Embryos containing tetraploid cells may be a normal part of development of the TE. Such cells have also been found in cattle, pig and sheep. Overall, extrapolation of this data would suggest that few embryos be completely chromosomally normal at early cleavage stage. However, various models for which there are experimental data may help to explain the observation that pregnancies following IVF do not result in an increased incidence of chromosomally abnormal infants. Few cells (possibly a single cell from an eight-cell embryo) differentiate to the embryo proper—the majority contributes to the cytotrophoblast and fetal membranes.
Data accumulated on the chromosomal constitution of surplus non-transferred embryos from PGD cycles have revealed that, despite the fact that these embryos are from women of proven fertility, the incidence of post-zygotic chromosomal anomalies is similar to that in embryos from routine IVF patients. This finding may provide one explanation for the apparently poor success rate of IVF procedures. A second significant finding is that the incidence of the most bizarre type of anomaly, chaotically dividing embryos, is strongly patient-related. In repeated cycles, certain women regularly produced “chaotic” embryos while others did not, although the frequency of diploid mosaics was similar in both groups.
Interesting evidence of self-correction during preimplantation development in human embryos exists. These data analyzed blastocyst TE biopsies by array CGH after blastocyst TE biopsy, as well as excluded cells of 18 blastocysts that developed from partially compacted morulas. These results were compared with a retrospective time-lapse morpho-kinetic data for the same embryos, in order to identify whether embryos from irregular cleavage divisions, which are presumed to have an abnormal chromosome complement, could develop into euploid blastocysts.
Although embryos identified as having irregular cleavage were at increased risk of developmental arrest, some of them did reach the blastocyst stage, producing chromosomally normal embryos. Some of these embryos were also observed to have incomplete compaction at the morula stage, with the exclusion of some cells, and the authors hypothesize that this cell exclusion might represent a “correction” mechanism, rescuing embryos from chromosomal aneuploidy after abnormal cleavage by preferentially eliminating anomalous cells. Their results further suggested that this hypothetical self-correction mechanism is less efficient in embryos from older women (age>39 years).
Chromosomal Mosaicism and PGD
Chromosomal mosaicism may confound PGD for some diseases, namely dominant disorders and chromosome abnormalities. A misdiagnosis of embryo sex would be unlikely to occur as a XX cell would have to be found in a XY embryo. XO cells have been identified in male embryos, but XO embryos should never be considered for transfer. If the offspring have Turner’s syndrome, they would have the same risk of suffering the X-linked disease as would a male. For recessive disorders, the presence of extra chromosomes or a haploid cell would not lead to a misdiagnosis.
A carrier embryo with a haploid cell would be diagnosed as normal or affected depending on which gene was present in the cell: this would be the same situation if allele dropout had occurred. For dominant disorders, a haploid cell could lead to a misdiagnosis. If a cell from an affected embryo carried only the unaffected gene, the cell would be diagnosed as normal. Therefore, the same precautions as for allele dropout would have to be applied.
For chromosome abnormalities, mosaic embryos containing some normal and some abnormal cells have been identified, such as in cases where a few normal cells arise in an embryo which otherwise carried trisomy 21. If the normal cells are biopsied, the embryo would be diagnosed as normal, resulting in misdiagnosis. As with confined placental mosaicism, this problem cannot be solved. Patients undergoing PGD for chromosomal abnormalities have to be aware that chromosomal mosaicism can lead to a misdiagnosis, but that this is a rare event.