Nondisjunction is the improper segregation of chromosomes during either meiotic or mitotic cell divisions. The result of nondisjunction is dosage imbalance – either one extra or one missing chromosome for some proportion of the resulting cells. Mitotic nondisjunction is commonly observed in cancer cells (see Article 14, Acquired chromosome abnormalities: the cytogenetics of cancer, Volume 1), while nondisjunction during meiosis results in gametes that are chromosomally unbalanced (see Article 11, Human cytogenetics and human chromosome abnormalities, Volume 1). If an unbalanced gamete participates in fertilization, the resulting embryo will be aneuploid, with either one chromosome too many (trisomy) or one chromosome too few (monosomy). Such embryos are generally inviable, and for humans, this has serious clinical consequences. As the most commonly identified chromosome abnormality, identified in at least 5% of all clinically recognized pregnancies, aneuploidy is the leading known cause of pregnancy loss and fetal wastage. Among conceptions that survive to term (primarily trisomy for chromosomes 13, 18, and 21 and various combinations of the sex chromosomes), aneuploidy is the leading genetic cause of mental impairment and developmental disabilities.
Unlike humans, nondisjunction in most organisms is a very rare occurrence. For example, chromosome missegregation is estimated to occur in only 1 of every 10 000 meiotic events for Saccharomyces cerevisiae (Sears et al., 1992). The underlying basis for the increased incidence in humans is unclear as the basic meiotic pathway is highly conserved across species. Meiosis generates haploid gametes through a process that consists of one round of DNA replication followed by two cellular divisions (see Article 13, Meiosis and meiotic errors, Volume 1). The first cell division (known as meiosis I or MI) separates homologous chromosomes (Figure 1), while meiosis II (MII) segregates the sister chromatids of each homolog. Nondisjunction can occur at either of these meiotic divisions, and the alternatives can often be distinguished using polymorphic genetic markers at or near the centromere of the nondisjoined chromosome. If both copies of the nondisjoined chromosome are heterozygous for alleles at these markers, it is likely that the error arose at MI. In contrast, homozygosity at the centromere of the nondisjoined chromosomes suggests an error at MII. In much the same way, marker studies can be used to determine in which parent the nondisjunction error occurred. As a result, numerous aneuploid conditions have been studied to determine the parental origin and meiotic stage of the nondisjunction error. Very little data has been identified for monosomic conditions, which appear to result in early embryonic lethality. In contrast, most trisomic conditions are compatible with at least some fetal development and results are available for several trisomies (reviewed in Hassold and Hunt, 2001). These data show remarkable variation with respect to parental origin and meiotic stage. Nonetheless, it is evident that maternal meiosis I nondisjunction errors predominate among nearly all trisomic conditions. This is perhaps not surprising, given that the first stage of meiosis in females is amazingly protracted: it is initiated prenatally in all oocytes but arrested shortly thereafter and is not resumed until just before the oocyte is ovulated, 15 to 50 years later. In addition, increasing evidence suggests that meiotic disturbances are handled differently between males and females (reviewed in Hunt and Hassold, 2002). Abnormalities that often lead to prophase or metaphase-anaphase arrest in male meiosis appear to escape detection in female meiosis, leading to nondisjunction. Consequently, it is clear that an understanding of the risk factors for human aneuploidy will require a greater understanding of the events of meiosis I in the human female.
Figure 1 During the first stage of meiosis, homologous chromosomes segregate, traveling to opposite poles of the spindle. In humans, one homolog is placed in the polar body, while the other remains in the maturing oocyte. Nondisjunction at the first meiotic division results in both homologs traveling toward the same spindle pole. This is frequently associated with altered patterns of meiotic recombination along the nondisjoined homologs: generally either poor positioning of existing recombination or a lack of recombination altogether
Despite the high frequency and obvious clinical importance of nondisjunction, there is a lack of knowledge about the predisposing genetic and environmental factors. However, there is one factor incontrovertibly linked to human aneuploidy -increasing maternal age. Most, if not all, human trisomies are affected by increasing maternal age, although the magnitude of the effect varies among trisomies (Risch et al., 1986; Morton et al., 1988). Among women under the age of 25 years, ~2% of all clinically recognized pregnancies are trisomic but this frequency approaches 35% for women over the age of 40. Very little is understood regarding the mechanisms that underlie this age effect. It is thought to involve meiosis I, consistent with the studies described above examining the parent and meiotic stage of the nondisjunction event. However, the specific timing of the event is unclear, and numerous models have been advanced to describe when in meiosis I the age effect develops (for review see Gaulden, 1992).
Recently, altered genetic recombination has been identified as the first molecular correlate of human nondisjunction. In model organisms, mutations that affect the amount and location of meiotic recombination are associated with nondisjunction. It appears that altered amounts and placement of recombination in humans also increases the risk for nondisjunction. It is possible to recapitulate the recombination patterns that occurred in meiotic events that ultimately lead to nondisjunction. Studies of this type have identified significant reductions in recombination for all meiosis I trisomies studied to date, including maternally derived cases of trisomies 15, 16, 18, 21, and sex-chromosome trisomies (Hassold et al., 1995; Lamb et al., 1996, 1997; Bugge, 1998; Robinson, 1998). Reduced recombination has also been identified in paternally derived cases of trisomy 21 and Klinefelter syndrome (47, XXY) (Savage, 1998; Thomas, 2000). For many of these trisomies, the reduction in recombination has been due to a proportion of cases that never engaged in genetic recombination. In addition, altered placement of meiotic recombination has been identified for a subset of maternally derived trisomies 21 and 16, with the exchange events occurring more distal than expected. These “distal only” exchange events appear to be less efficient at proper chromosome segregation.
Surprisingly, altered patterns of meiotic recombination are also associated with maternal meiosis II trisomy for chromosome 21 (Lamb, 1997). Specifically, the increase was most pronounced in the pericentromeric region of the chromosome. The connection between recombination (a meiosis I event) and the meiosis II nondisjunction may be explained if the meiosis II errors actually originated in meiosis I by chromosome entanglement due to the pericentromeric exchange or premature separation of sister chromatids at meiosis I followed by comigration of the sister chromatids due to chance at meiosis II. As a result, the nondisjoined chromosomes would have identical centromeres, leading to the “meiosis II” classification even though the event was initiated at meiosis I. Subsequent studies of trisomy 18 or sex-chromosome trisomy arising at maternal meiosis II have not, however, identified a significant effect for recombination. As a result, it appears likely that the liability posed by altered recombination varies from chromosome to chromosome.
Do other risk factors exist for nondisjunction? When compared to the maternal age effect, additional proposed risks have a much weaker or unsubstantiated influence. Many such factors have been proposed, including: the effects of diminished oocyte pool leading to increased risk of nondisjunction and earlier maternal menopause (Kline, 2000); a reduced ovarian complement, due to the congenital absence or surgical removal (Freeman, 2000); maternal smoking and oral contraceptive use around the time of conception (Yang, 1999); and maternal polymorphisms at genes in the folate pathway (van der Put, 1998; Hobbs, 2000). In addition, various chemical, drug, or irradiation exposures have also been evaluated (Hook, 1992). To date, however, none of these potential risk factors have been conclusively proven. This does not mean that there are no relevant risk factors – possibly the impact is so small that each risk escapes detection. On the other hand, the correct risk factors may have simply not been identified. Regardless, an understanding of the molecular mechanisms for nondisjunction and the maternal age effect remains elusive.
