[Introduction] [Monozygotic Twinning] [X-chromosome Inactivation] [Case Studies] [Counterevidence] [Conclusion]
Twinning has long intrigued the intellect of mankind, and early attempts to explain the phenomenon frequently attributed it to mystical or magical forces. In the African Yoruba tribe, for example, twins were thought to be good spirits who entered the womb of a pregnant woman to be born as humans (Scheinfeld, 1967). They were thus considered good-luck omens and celebrations accompanied the birth of twins. Not all cultures viewed twins in such a benevolent light, however; many Native American tribes believed twins were harbingers of evil brought about by sinister powers and routinely smothered them shortly after birth (Scheinfeld, 1967). While biological advances have since illuminated the true origins of twins, they are still frequently greeted with awe and fascination by many.
Monozygotic twins, frequently called identical twins due to the striking similarity these pairs usually exhibit, inspire a particular interest because of the genetic implications of two individuals who are genetically identical. Many potential reasons for monozygotic twinning events have been proposed, although it is not definitively known which events are primarily responsible. Later monozygotic twinning events seem to favor female twins, thus placing an interesting twist on this question. or the cause of . Medical cases in which female monozygotic twins display discordant phenotypes for X-linked diseases has pointed a finger at X chromosome inactivation as a possible player in the cell splitting that results in such twins.
The principal difference between monozygotic twins and dizygotic twins and singletons is that monozygotic twins result from the splitting of early embryonic cells from a single fertilization event, forming two separate embryos with the same genotype. In dizygotic twinning and singleton births, by contrast, each embryo results from the union of a separate egg and sperm. The location of monozygotic twins within the uterus is indicative of the approximate time that initial cell splitting occurred. Separation of embryonic blastomeres prior to the fifth day of gestation, when the trophoblast is formed, results in twins with entirely separate chorions and accounts for approximately one third of monozygotic twins surviving to birth. If splitting occurs after the separation of inner cell mass and trophoblast on the fifth day of gestation, but prior to the formation of the amnion on the ninth day, the resulting twins will share a common chorion, but have separate amnions. Such monochorionic, diamniotic twins constitute the largest proportion of monozygotic twin births, making up approximately two thirds of the total. A very small fraction of monozygotic twins are monochorionic and monoamniotic, resulting from a splitting of the inner cell mass after the formation of the amnion on the ninth day. Conjoined twinning results from incomplete separation of the inner cell mass, and most likely results from the latest attempt at separation. (Machin, 1996)
Unlike dizygotic twinning, monozygotic twinning is rarely influenced by hereditary factors and not induced by fertility drugs (Hall, 1996). Rather, each incidence of twinning is the result of one of many possible sporadic events that could lead to the splitting of blastomeres from one another. Among possible events that could result in such cell splitting are abnormalities in the zona pellucida which surrounds the blastocyst as it travels down the oviduct to the uterus, abnormalities in the developmental clocks of different cells causing them to be at different stages at the same time, and discordance between adjacent cells which might cause them to repulse one another (Hall, 1996). It has been speculated that splitting events occurring at later stages, such as those that lead to monochorionic as opposed to dichorionic twins, may be the result of different developmental events, a speculation supported by the observation that sex ratio is increasingly skewed toward females when splitting occurs later (Figure 1, Machin et al., 1996). This observation has lead researchers to look for events that might cause twinning in females but not males. X-chromosome inactivation is one such possible event.
Figure 1: Sex ratio (female : male) in twins by zygosity, chorion, and amnion status. A steadily increasing female excess in monozygotic twins arising from later twinning events is evident. (After Machin et al., 1996)
In humans, as in the majority of animals, the feature that differentiates females from males is that they possess two X chromosomes, whereas males posses one X and one Y chromosome. The number of X-chromosomal RNA transcripts are kept approximately equal in both sexes, however, due to a phenomenon called X-chromosome inactivation by which one of the X chromosomes in every female cell is inactivated (Gilbert, 1994). The inactivated X-chromosome, usually referred to as a Barr body, is condensed and remains inactivated in all progeny of the initial cell (Griffiths et al., 1996). In addition to being more condensed, inactivated X-chromosomes also differ from their active counterparts in a number of other ways, including methylation and acetylation differences, making it possible to asses which chromosome has been inactivated (Brown and Willard, 1994). The precise time in development that inactivation occurs is not known, although it is speculated to occur before the 64 cell stage (Migeon et al., 1996; Gilbert, 1994). X inactivation is usually a random process, thus females are actually mosaics of cells containing different X genotypes (Griffiths et al., 1996).
Genes for several recessive traits are located on the X-chromosome, and their expression is therefore usually restricted to hemizygous males who have inherited a mutated X from their mothers, who are non-manifesting carriers. On rare occasions women have been found who express these traits, usually due to genetic events such as chromosome loss or translocation that cause the woman to have a homozygous genome. If, however, rather than undergoing the typically random inactivation, the majority of woman's normal X chromosomes became preferentially inactivated, mutant X-linked traits carried on the active chromosome would become expressed (Jorgensen et al., 1992). This phenomenon has been well documented in several studies of female monozygotic twins heterozygous for X-linked disorders. In these pairs one twin has skewed X-inactivation toward the normal chromosome, and thus expresses the recessive phenotype, while her twin has skewed inactivation toward the mutant chromosome and thus exhibits no symptoms.
Several X-linked conditions exhibiting differential expression in female monozygotic twins have been documented, most notably Duchenne Muscular Dystrophy (DMD) and color blindness. In all studies the karyotypes of the twins were examined to rule out the possibility of expression due to chromosome duplication, deletion, or recombination. In remaining twin pairs heterozygosity for the examined trait was confirmed and X-chromosome inactivation patterns determined. Since nearly 70 percent of monozygotic twins share blood circulation while in the womb, studies based on the analysis of non-hematopoietic such as fibroblast tissues have been favored of studies (Hall, 1996).
A number of cases have been documented in which the patterns of X-inactivation were found to be skewed in opposite directions, with the affected girl exhibiting preferential inactivation of the chromosome containing the wild type copy of the gene, while her unaffected twin exhibited preferential inactivation of the chromosome containing the mutated copy (Table 1).
No cases have been documented in which monozygotic twins known to be heterozygous for an X-linked trait both manifest the trait or both display normal phenotype (Jorgensen et al., 1992, Machin et al., 1996), thus a connection between the twinning and X-chromosome inactivation has been suggested. One hypothesis proposes that random inactivation occurs in the initial cell mass in such a way that two clusters of cells are produced, each with opposite X inactivation patterns (Burn et al., 1986, Richards et al., 1990, Machin et al., 1996). The discrepancy in the inactivation patterns then causes each group of cells to regard the other as a foreign entity, resulting in mutual repulsion of the cell masses and their development into two separate embryos (Figure 2, Burn et al., 1986, Richards et al., 1990, Machin et al., 1996). The observation that two of the famous monozygotic Dionne quintuplets, who were heterozygous for X-linked color-vision deficiency, were colorblind supports this theory (Jorgensen et al., 1992). This hypothesis would also help explain skewed X-inactivation leading to X-linked disease expression in apparently singleton births - these individuals might be monozygotic twins whose normally skewed pair died early in development (Machin et al., 1996).
Figure 2: Random inactivation followed by mutual repulsion of cells with differentially inactivated X-chromosomes resulting in discordant monozygotic twins. Paternal X-chromosome and cells with paternal X inactivated are shown in bold. (After Machin et al., 1996)
Cases have also been documented in which the clinically affected twin displays skewed X-inactivation while the unaffected twin exhibits random inactivation (Table 1). An alternate theory connecting twinning and X inactivation might account for this observation. In this second proposal, rather than X inactivation prompting the initial separation of cells, random inactivation occurs followed by unequal allocation of the cells to each twin in a splitting event triggered by some other occurance. One twin receives relatively few cells, a significant portion of which have the same X inactivated, while the other receives a larger allocation of cells with a random, or near random, inactivation pattern (Figure 3, Machin et al., 1996). Indeed, some twin pairs that exhibit this inactivation pattern show definite evidence of unequal allotment of cells to the two embryos (Machin et al., 1996). This theory does not, however, help account for the increased number of female monozygotic twinning events.
Figure 3: Random inactivation followed by unequal allocation of cells to each twin resulting in discordant monozygotic twins. Paternal X-chromosome and cells with paternal X inactivated are shown in bold. (After Machin et al., 1996)
If there is indeed a connection between female monozygotic twinning and X chromosome inactivation it would be expected that twins not manifesting X-linked traits would also display skewed inactivation patterns similar to those described in clinical studies. In a comparative study of X inactivation in monozygotic and dizygotic twins, Goodship and collegues (1996) found that five of the monozygotic twins sets examined who showed skewed X inactivation were skewed in the same direction, while only one set had reciprocally skewed X inactivation. This study thus fails to support the theory that female monozygotic twinning often results from mutual repulsion of reciprocally inactivated X chromosomes. The fraction of monozygotic twins displaying skewed X inactivation of some kind was, however, found to be significantly higher than the proportions literature values predict for the general population, suggesting that there is indeed a relationship between twinning and inactivation. One explanation offered for this observation is that X inactivation occurs in monozygotic twins when the pool of cells is smaller than in singletons, and thus skewed inactivation is more likely to occur. This would imply that the cells in monozygotic twins continue to develop and differentiate at the same time as in singletons, but with a diminished initial cell volume (Goodship et al., 1996). These observations are consistent with the unequal cell allocation explanation of discordant X inactivation, which is based on one twin froming from a smaller pool of cells.
The excess of females among monozygotic twins resulting from later twinning events has not yet been satisfactorily explained. A causal connection between skewed X inactivation and female monozygotic twinning was not supported by a comparitive study of monozygotic and dizygotic twins, although further studies to confirm this result should be performed before this possibility can be conclusively ruled out. A high incidence of skewed X inactivation has been generally observed for monozygotic twins in comparison with the average female population, possibly due to the effect of a smaller cell mass in the initial embryo. This reduced cell mass may potentially have other developmental effects on the monozygotic twin embryos that are dealt with more easily by female embryos than male embryos, and thus explain the excess of females in monozygotic twin births. The majority of analysis thus far completed has been based on clinical studies of twins who manifest radically different phenotypes as a result of their skewed X inactivation. Studies of X inactivation in phenotypically normal monozygotic twins and singletons might enable more predictions to be made about any possible connections between the two events.