|
|
||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Embryo |
National Institute for Medical Research,3 London NW7 1AA, United Kingdom
School of Medicine,4 Mie University, Tsu 514-8507, Japan
Max-Planck-Institut für Molekulare Genetik,5 14195 Berlin, Germany
| ABSTRACT |
|---|
|
|
|---|
developmental biology, embryo, placenta, pregnancy, trophoblast
| INTRODUCTION |
|---|
|
|
|---|
The ectoplacental cone is thought to give rise predominantly to the spongiotrophoblast of the placenta [7], which constitutes approximately 30% of the volume of the placenta in late pregnancy [8]. We have previously observed an increase in spongiotrophoblast tissue in 13.5 days postcoitum (dpc) XpO placentae [9]. This finding suggests that enlargement of XpO placentae in late gestation could be a consequence of a compensatory response to initial ectoplacental cone deficiency. In this study, we investigated further XO placental size in mice in late pregnancy to determine whether hyperplasia was correlated with early ectoplacental cone deficiency of XpO but not XmO embryos.
| MATERIALS AND METHODS |
|---|
|
|
|---|
We utilized four mouse strain crosses. All mice were bred on a random-bred MF1 albino (National Institute for Medical Research colony) background.
1. In(X)/X x XYRIII The females in this cross are heterozygous for the In(X)1H inversion [1]. Although fully backcrossed to MF1, the inversion retains its C3H strain origin because crossing over within the inversion leads to loss of the recombinant products and the associated production of nullo-X eggs. For reasons unrelated to this study, the males in this cross carry a Y chromosome originating from the RIII inbred strain [10]. This is the cross originally used by Burgoyne et al. [3] to generate XpO conceptuses.
2. XX x XY* The males in this cross carry the variant Y* chromosome that has a compound pseudoautosomal region (PAR) together with a non-Y centromere distal to the PAR [1113]. This cross produces a low frequency of XmO conceptuses together with XX, XY*, XY*X, and XXY* conceptuses (Y*X and XY* are the very small and large products of recombination between the X and Y* [12, 13]). We did not use the more efficient XX x XPafY* cross [14] because the Paf mutation affects placental size (unpublished data).
3. XX x XY*O The males in this cross carry the XY* recombinant product originating from Y*. Although males of this genotype are normally sterile [15], we have been able to establish an exceptional pedigree of these males from which a proportion of fertile XY*O males is generated (unpublished results). This cross produces XmO and XXY* conceptuses.
4. XX x XYTdym1Sry YTdym1 (hereinafter abbreviated as Y-) has a deletion removing the testis determinant Sry, and this deletion has been complemented by an autosomally located Sry transgene [16, 17]. This cross generates XX and XY- female conceptuses and XXSry and XY-Sry male conceptuses.
Genotyping
Where necessary, XO, XY*X, and XXY* conceptuses were identified from metaphase spreads obtained from fetal liver [18]. Other genotypes were established using the following PCR assays on DNA obtained from yolk sac.
1. In(X)1H The In(X)1H inversion originated on a C3H/He X chromosome, and although we have back-crossed this inversion to a random-bred MF1 background, the inverted segment has remained C3H because crossing over within the inverted segment leads to loss of the recombinant products [19]. We have therefore identified inversion carriers by PCR for the DXMit16 microsatellite (lying within the inversion) that is polymorphic between C3H/He (86 base pairs [bp]) and MF1 (118 bp).
2. Y long arm Y-bearing males and females in cross 4 were identified by PCR for a mutiple-copy Y gene family (Ssty, YMT2/B-related subfamily [10, 20]) located on the long arm of the mouse Y chromosome.
3. Sry transgene Sry transgene carriers were identified using the transgene-specific primers Tan3' and Tan 5' [21].
Placental and Fetal Weights
Females were killed at 17.5 dpc by cervical dislocation. Fetuses with attached placentae were removed into PBS, pH 7.4. The placentae were then detached, fetal membranes were removed, and the fetuses and placentae dried on tissues before being weighed on a pan balance (Oertling, London, UK).
Statistical Analysis
Because there are major litter effects on placental and fetal size, genotype comparisons were made within litters. For the primary analysis, we used two statistical approaches. First, we used the Student t-test to compare an estimate of the difference in placental size and the difference in fetal size between two genotypes, with the calculated SEMs of these estimates. Each estimated difference is a mean weighted difference (MWD) obtained by pooling the information from a series of litters as described previously for XpO fetal and placental weights [3]. For each litter, a weighted difference is calculated for the two genotypes to be compared, the weighting (w = n1 x n2/n1 + n2, where n1 and n2 are the genotypes being compared) serving to correct for differences in the number per genotype both within and between litters. The MWD is then calculated by dividing the sum of the weighted differences by the sum of the weights (w). SEMs for these differences are estimated from the variance within genotypes within litters using the appropriate degrees of freedom derived from within genotypes within litters. The MWD divided by the SEM is distributed as t. Second, we used a two-way ANOVA to examine the separate effects of genotype and litter on genotype means. As expected, significant litter effects were seen throughout; there were no significant genotype by litter interactions. The difference between the two approaches is that the comparison of MWDs between genotypes attaches more importance to genotype differences based on larger genotype samples within litters, whereas the comparison of genotype means by ANOVA attaches equal importance to genotype means within litters, irrespective of sample size.
For the data from the XX x XY-Sry cross (cross 4), the separate effects of litter, sex, and sex chromosome constitution were analysed by three-way ANOVA.
| RESULTS |
|---|
|
|
|---|
|
|
XmO and XpO Placentae Are Larger than XX Placentae
Contemporaneous data for XpO placentae were obtained from In(X)/X x XYRIII matings (cross 1) because this cross was used for the previous study that showed that XpO placentae are hyperplastic in late gestation [3]. As a by-product of the present study, we found that a maternally derived copy of the In(X)1H inversion leads to placental hypoplasia in late gestation; thus, for the analysis of the data from this cross we excluded data from inversion carriers. The analysis showed that XpO placentae are significantly larger than XX placentae (MWD = 9.0 ± 3.3 mg, P = 0.012), confirming the previous findings [3].
We next used the XX x XY* cross (cross 2) to compare XmO placentae with XX placentae. This comparison revealed that XmO placentae are also significantly larger than XX placentae (MWD = 14.1 ± 3.8 mg, P = 0.003). Because the frequency of XmOs is very low in this cross, we sought to confirm this result using the XX x XY*O cross (cross 3) to provide a larger sample of XmO placentae. However, all females generated from this cross are XmO, and their placentae can be directly compared only with those of their XXY* brothers. To enable an indirect comparison of this large sample of XmO placentae with XX placentae, we obtained an estimate of the difference between XXY* and XX placentae from the XX x XY* cross: XXY* placentae were estimated to be somewhat smaller than XX placentae (MWD = 4.4 ± 2.1 mg). The indirectly estimated XmO versus XX MWD (19.0 mg) for this large sample of XmO placentae was even greater than that for the XX x XY* cross (14.1 mg), supporting the view that these placentae are hyperplastic. The size of this indirectly estimated MWD in relation to its error suggests that XmO placentae are significantly bigger than XpO placentae (see Fig. 1). In an attempt to test this hypothesis, the XX versus XmO MWD from cross 2 was first compared with the indirectly estimated XX versus XmO MWD from cross 3; these two MWDs were not significantly different (t109 = 1.004, P = 0.318) and were pooled. The new pooled XX versus XmO MWD was then compared with the XX versus XpO MWD from cross 1, and these two MWDs were significantly different (t131 = 2.667, P = 0.009). Because of the indirect nature of this comparison, this indication that XmO placentae are larger than XpO placentae should be treated with caution. Nevertheless, XO placentae are larger than XX placentae irrespective of the parental origin of the X chromosome. Thus, the previously observed XpO placental hyperplasia in late pregnancy is not a consequence of Xp imprinting; rather, it appears to be related to having one rather than two X chromosomes.
|
Fetal weight data are also presented in Figure 1. Despite the fact that fetal weights are inherently variable at this stage of pregnancy because of marked effects of position in the uterus, the previously reported reduced fetal weight of XpO but not XmO fetuses [5] is readily apparent.
Other Fetuses with a Single X Genotype Also Have Enlarged Placentae
Because genotyping is carried out subsequent to the collection of the placental and fetal weights, data were available from the In(X)/X x XYRIII (cross 1) and XX x XY* (cross 2) crosses for three other genotypes with a single X chromosome: XY, XY*, and XY*X. Calculation of the MWDs revealed that the placentae of these three genotypes were also larger than XX placentae: XY: MWD = 7.9 ± 2.2 mg (P = 0.001); XY*: MWD = 7.6 ± 2.0 mg (P < 0.001); XmY*X: MWD = 8.5 ± 3.3 mg (P = 0.012) (see Fig. 1). Although the significances of these additional post hoc comparisons may be questionable, the consistency of the data lends support to the hypothesis of larger placental size for fetuses with a single X chromosome.
Androgen Does Not Increase Placental Size
The problem with using XY and XY* placental weight data as support for an X monosomy effect is that these genotypes are male and are thus producing fetal androgen, which could have an effect on placental size. However, the data from the XX x XY* cross (cross 2) also showed that XY* placentae are markedly larger than XXY* placentae (MWD = 10.6 ± 2.5 mg, P < 0.001), which strongly suggests that the X dosage effect outweighs any potential androgen effect. To test directly for an androgen or other gonadal sex effect, we examined placental size in conceptuses from the XX x XY-Sry cross (cross 4), in which sex is independent of the sex chromosomal complement. We used a three-way ANOVA to asess the separate effects of litter, sex, and sex chromosome complement. No significant effect of sex on placental weights was found, although significant litter effects (P = 0.036) were found, as expected. However, there was a highly significant effect of sex chromosome complement (F1,17 = 52.54, P = 0.000001). The results for these four genotypes are therefore again consistent with a single X leading to placental hyperplasia; this conclusion is also consistent with the calculated MWD data (Fig. 1 and Table 1). There is therefore no evidence that fetal androgen or any other male gonad dependent factor augments placental size.
The three-way ANOVA of the fetal weight data showed that fetal weight differences in cross 4 are also dependent on sex chromosome complement (F1,17 = 11.31, P = 0.004) and are independent of sex, indicating that fetal androgen also does not augment fetal size. The fetal weight MWDs (Fig. 1 and Table 2) indicate that XY- and XY-Sry fetuses are significantly larger than XX fetuses at 17.5 dpc but XYRIII and XY* fetuses are not. In some strains of Y mice (including the Y- strain, which is of 129/SvEv-Gpi1c Y strain origin), fetal size is augmented because of an accelerating effect on cleavage during preimplantation development, whereas in YRIII mice this acceleration does not occur [4, 22]. The greater size of XY- and XY-Sry fetuses at 17.5 dpc may be a consequence of this earlier Y effect.
| DISCUSSION |
|---|
|
|
|---|
XY placentae are also significantly larger than XX placentae, and this difference is independent of androgen effects. Others previously reported that male placentae were larger than female placentae in interspecific crosses, but no difference between male and female placentae was observed in the control intraspecific C57BL/6 (Mus musculus) cross [23, 24]. However, these authors did not carry out their analysis within litters, as we have done here. The additional data we collected on C57BL/6 placental weight at 17.5 dpc indicate that XY placentae are indeed larger than XX placentae in this cross (XY vs. XX MWD = 8.7 ± 2.3 mg, t51 = 3.78, P < 0.001).
Although the possession of one X chromosome rather than two leads to an increase in placental size, the underlying mechanism is still to be determined. A deleterious effect of two PARs is an insufficient explanation because XY placentae are clearly larger than XX placentae, yet both have two PARs. One possibility is that there are X-linked genes expressed in embryo-derived components of the placenta that escape from the normal X dosage compensation process (which occurs via paternal X inactivation), with increased levels of expression reducing placental size. Although the X chromosome is said to be deficient in placentally expressed genes [25], a handful of placentally expressed X-linked genes have been identified [2629]. Although none of these genes has been individually identified as escaping X inactivation, a recent study has suggested that the X chromosome in trophoblast giant cells of the mouse placenta may not be subject to paternal X inactivation, so that all X-linked genes in this tissue may be exempt from X dosage compensation [30].
Closer examination of the placental weight data for the range of sex chromosome variant genotypes in the present study suggests an alternative explanation for the placental size differences. Although XY placentae are larger than XX placentae, XXY* placentae, which have an X chromosome and an X-attached Y chromosome [13], are smaller than XY* placentae and may be smaller than XX placentae, although this possibility needs exploration. These results suggest that increasing amounts of sex chromosome material as a whole, rather than increased X dosage, may lead to reductions in placental size (Fig. 2). If this is true, the question remains as to whether the inhibitory effects on placental growth are due to an increase in combined expression of X and Y genes. In addition to the PAR gene Sts, a number of X-Y homologous genes (Smcx/Smcy, Dbx/Dby, Eif2s3x/Eif2s3y, and Utx/Uty) in the mouse are thought to be ubiquitously expressed, with the X homologues exempt from X dosage compensation [33], and thus are candidates for such an overexpression effect. Alternatively, the critical effect of increasing sex chromosome content (or increasing the amount of sex chromosome heterochromatin) could be in extending DNA replication time. In this regard, it may be pertinent that the trophoblast giant cells uniquely undergo multiple rounds of DNA replication without cell division [34], which could be of importance with respect to placental growth. If an extension of DNA replication time leads to reduced ploidy in the trophoblast giant cells, then a reduction in gene expression in these cells could result, with a consequent overall reduction in placental growth. Further studies are needed to differentiate between these possibilities.
|
| ACKNOWLEDGMENTS |
|---|
| FOOTNOTES |
|---|
2 Correspondence: Paul S. Burgoyne, Division of Developmental Genetics, National Institute for Medical Research, The Ridgeway, Mill Hill, London NW7 1AA, U.K. FAX: 44 208 906 4477; pburgoy{at}nimr.mrc.ac.uk ![]()
Received: 22 October 2002.
First decision: 13 November 2002.
Accepted: 26 March 2003.
| REFERENCES |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
A. R. Isles, W. Davies, D. Burrmann, P. S. Burgoyne, and L. S. Wilkinson Effects on fear reactivity in XO mice are due to haploinsufficiency of a non-PAR X gene: implications for emotional function in Turner's syndrome Hum. Mol. Genet., September 1, 2004; 13(17): 1849 - 1855. [Abstract] [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |