| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Regular article |
a The Fels Institute for Cancer Research and Molecular Biology and
b Department of Biochemistry, Temple University School of Medicine, Philadelphia, Pennsylvania 19140
ABSTRACT
Diploid androgenetic mouse embryos, possessing two sets of paternally inherited chromosomes, and control fertilized embryos were used to examine the relative effects of X chromosome number and parental chromosome origin on androgenone viability and X-linked gene expression. A significant difference in efficiency of blastocyst formation was observed between XX and XY androgenones in some experiments, but this difference was not uniformly observed. Significant effects of both X chromosome number and parental origin on X-linked gene expression were observed. Male and female control embryos expressed the Xist RNA initially. This expression was followed by a preferential reduction in Xist RNA abundance in male embryos, indicating that dosage compensation for the X chromosome may normally require the downregulation of Xist RNA expression in male embryos, in conjunction with the production of stable Xist transcripts in female embryos. By the late blastocyst stage, XX control embryos expressed significantly more Xist RNA than did XY embryos. Unlike their normal counterparts, XX androgenones did not express significantly more Xist RNA than did XY androgenones at the late blastocyst stage. Androgenones exhibited severe repression of the Pgk1 gene, but during development to the late blastocyst stage Pgk1 mRNA expression increased in XX androgenones and decreased in XY androgenones. Thus, the initial repression of the Pgk1 gene in XX androgenones was lost as the Xist RNA declined in abundance, and this loss was correlated with a failure of XX androgenones to express significantly more Xist RNA than did XY androgenones. These results indicate that androgenones may lack a factor that is expressed from the maternal genome and required for dosage compensation in preimplantation embryos. The results also indicate that early dosage compensation in preimplantation embryos may normally be reversible, thus providing flexibility to meet different developmental requirements of the embryonic and extraembryonic lineages.
gene regulation, implantation/early development
INTRODUCTION
Genomic imprinting is a process whereby gene function is affected in a parental origin-dependent manner. In mammals, genomic imprinting is manifested as a difference in expression of parental alleles of a number of autosomal genes and of the X chromosome. The effects of genomic imprinting arise from differential epigenetic modification of parental alleles during gametogenesis, followed by additional epigenetic processes that occur after fertilization [1].
For the X chromosome, the effects of genomic imprinting are manifested as a preferential inactivation of the paternally inherited X chromosome in extraembryonic tissues of eutherian mammals, and also in somatic tissues of marsupials [2–9]. X chromosome inactivation (XCI) is initiated around the time of embryo implantation. Imprinting of the X chromosome is evident in preimplantation embryos through the preferential expression of the paternal allele of the Xist gene [10–12] and repression of paternal alleles of specific genes, such as Pgk1 [11].
It was previously reported that diploid androgenetic mouse embryos, which are constructed by nuclear transplantation to contain two paternal sets of chromosomes, exhibit a reduced ability for development to the blastocyst stage if they possess two X chromosomes [12]. It was also reported that Xist RNA abundance declines between the eight-cell and blastocyst stages in androgenones, and it was suggested that an imprinted gene (maternal allele expressed) might be required to maintain Xist RNA expression [12]. Consequently, it was proposed that reduced Xist RNA expression in androgenones could lead to a lack of dosage compensation and thus compromise XX androgenone development by causing excess expression of X-linked gene products [12]. Other data revealing reduced expression of X-linked genes in androgenones, however, led us to propose an alternative model in which XX androgenones could be selectively eliminated because of inactivation of both X chromosomes [11]. We proposed that all paternally derived X chromosomes initially exhibit repression of X-linked genes near the X chromosome inactivation center (Xic) (e.g., Pgk1) because of expression of the Xist gene in cis but that a mechanism exists within the preimplantation embryo for counting the number of X chromosomes present and initiating the spread of XCI along the chromosome [11, 13]. As a result of the counting, XX androgenones would undergo spreading of XCI on both of their X chromosomes during preimplantation development and thus would succumb to lethal effects of insufficient X-linked gene expression, whereas XY androgenones would not undergo the spreading of XCI, thus accounting for their greater apparent viability. Diploid gynogenetic and parthenogenetic embryos (two maternal sets of chromosomes) were proposed to be resistant to undergoing XCI in their extraembryonic membranes because of maintenance of maternal Xist gene imprinting in those tissues, thus contributing to the developmental defects in these tissues after implantation [13].
The above two models make specific and distinct predictions with respect to the anticipated levels of X-linked gene expression in XX androgenones relative to XY androgenones and control embryos. With the first model, in which androgenones lack a factor that is required to maintain Xist RNA expression, X-linked genes are expected to be overexpressed in XX androgenones relative to XY androgenones. In the second model, all androgenones should be deficient in expression of genes near the Xic, and XX androgenones are expected to be deficient in expression of X-linked genes located at a distance from the Xic. Recent discoveries of the role of Xist RNA stability in the XCI process [14, 15] lend new interest to evaluating the possible role of a factor expressed from the maternal genome in controlling XCI in the early embryo, as proposed in the first model. Understanding X chromosome counting, proposed in the second model, is also important. Testing the two models should provide insight into the mechanism by which imprinting and X chromosome number control XCI in the early embryo. The predictions of the two models can be tested by comparing X-linked mRNA abundance between XX and XY androgenones, XX and XY control embryos, and gynogenones during development to the blastocyst stage.
To accomplish this objective, we have undertaken a detailed analysis of expression of several X-linked genes located along the length of the X chromosome in individual genetically defined androgenetic, gynogenetic, and control preimplantation embryos. Our results indicate that XX androgenones may exhibit reduced viability to the blastocyst stage under some circumstances, but contrary to expectations, in a majority of experiments XX androgenones are in fact capable of forming blastocysts as efficiently as XY androgenones. In addition, our results indicate that androgenones lack a factor that is required to accumulate the Xist RNA to the same degree as normal embryos and that early X chromosome dosage compensation in androgenones may thus be incomplete or unstable.
MATERIALS AND METHODS
Embryo Culture and Nuclear Transplantations
Embryos were obtained by mating superovulated C57BL/6 (Harlan Sprague-Dawley, Indianapolis, IN) adult females to AKR/J (Jackson Laboratory, Bar Harbor, ME), DBA/2 (Taconic, Germantown, NY), or (DBA/2 x C57BL/6)F1 males. Additional experiments were conducted employing either C57BL/6 homozygous males or congenic C57BL/6-Pgk1a males (denoted here as B6-Pgk; C57BL/6 background but containing a central portion of the X chromosome derived from Danish wild mice and possessing the Pgk1a and Xcec alleles, obtained from the Roswell Park Cancer Institute, Buffalo, NY [16]). Embryos were isolated and cultured as previously described [11]. Embryos were staged according to the elapsed time (hours) following injection of females with hCG for superovulation. All studies adhered to procedures consistent with the National Research Council Guide for the Care and Use of Laboratory Animals. Nuclear transplantations were performed as previously described [11, 17, 18] except that Hepes-CZB medium [19] was used for embryo manipulation and karyoplast fusion was achieved by electrofusion. Electrofusion was performed using the ECM-2000 embryo manipulation system (BTX Instruments, Genetronics, San Diego, CA) with a 90-V pulse (900 V/cm) delivered for 10 µsec. If electrofusion failed after the first pulse, two or three additional pulses were given with 30–60 min between pulses. The electrofusion was conducted in medium containing 0.3 M mannitol, 0.1 mM MgSO4, 0.05 mM CaCl2, and 0.3% BSA, pH 7.2–7.4. Higher voltages or prolonged duration of pulses typically resulted in androgenetic embryo arrest at the two-cell or morula stages.
The design of these experiments was devised to permit unambiguous determination of the sex chromosome composition of all embryos analyzed and thus to permit direct comparison of X-linked gene mRNA abundances among XX and XY androgenones, XX and XY control embryos, and gynogenones. Diploid androgenones were constructed by combining a paternal pronucleus of a (C57BL/6 x AKR/J) embryo with a paternal pronucleus from a (C57BL/6 x DBA/2) or (C57BL/6 x [D2B6]F1) embryo. The resulting combination thus produces androgenones that have one AKR/J paternal genome and one DBA/2 or (D2B6)F1 paternal genome. These androgenones, therefore, should possess one of four possible genotypes with respect to sex chromosome composition: XAY, XDY, XAXD, or YY (YY androgenones will not form blastocysts). No androgenone should possess a C57BL/6-derived XB chromosome. Gynogenones should possess two XB chromosomes. In other experiments, homozygous C57BL/6 embryos were used, and embryos were sexed on the basis of presence or absence of the Y chromosome-specific Zfy gene. In a third series of experiments, androgenones were constructed by combining paternal genomes from homozygous C57BL/6 embryos and (C57BL/6 x B6-Pgk)F1 embryos. In this last experiment, androgenones will be of the genotypes XPXB, XPY, XBY, or YY.
Single Embryo Genotype Analysis
The objective of these studies was to obtain quantitative gene expression data from single morula or blastocyst stage embryos of known genotype. To meet this objective, we devised a new protocol wherein each individual embryo is lysed in a small amount of buffer, total nucleic acids are isolated, and then the nucleic acids are divided in half, with one part used for genotype analysis and the other used for quantitative reverse transcription polymerase chain reaction (RT-PCR) analysis. Embryos were lysed individually in 20 µl of GT lysis buffer (5 M guanidine thiocyanate, 0.5% sarkosyl, 25 mM sodium citrate, pH 7.0, 20 mM dithiothreitol) [20] supplemented with 2 µl of 10 µg/µl polyC. After lysis, nucleic acids were precipitated by the addition of 10 µl of 7.5 M ammonium acetate and 100 µl of ethanol. The precipitate was recovered by centrifugation at 13 000 x g for 30 min at 4°C, and the pellet was washed three times with 75% ethanol. The pellet was dissolved in 50 µl of Tris-EDTA buffer (pH 7.5) containing Prime RNAse Inhibitor (5 Prime-3 Prime, Boulder, CO) diluted to 3% v/v. Nucleic acids were then reprecipitated by the addition of 5 µl of 3 M sodium acetate and 150 µl of ethanol, followed by centrifugation and three washes as above. The final pellet was dissolved in 14 µl of 3% Prime RNAse inhibitor in water. Seven microliters of the lysate was frozen for genotype analysis, and the remainder was used immediately for RT-PCR as described below. The 7-µl sample for genotype analysis was subjected to primer extension preamplification PCR (PEP) as previously described [21], except that a 20-µl reaction volume was used. After PEP, 4–8 µl of PEP reaction product was subjected to PCR using primers that were diagnostic for the different X chromosomes and for the Zfy gene, and the products were visualized by electrophoresis on 5% denaturing acrylamide DNA sequencing type gels followed by autoradiography. Specifically, DXMit236 was used as a diagnostic marker for the XA chromosome and DXMit110 was used as a diagnostic marker for the XD chromosome. The presence or absence of an XB chromosome could be inferred from the results obtained with these two primer pairs. DXMit210 was used as a diagnostic marker for distinguishing the X chromosome of C57BL/6 inbred mice from the X chromosome of the B6-Pgk congenics. Primers to detect the Zfy gene were 5'-GACTAGACATGTCTTAACATCTGTCC-3' and 5'-CCTATTGCATGGACAGCAGCTTATG-3'.
Quantitative RT-PCR
Quantitative RT-PCR was performed using the quantitative amplification and dot blotting (QADB) method as previously described [22], except that nucleic acid pellets were reverse transcribed in 10-µl reactions that lacked Nonidet P-40. After RT and PCR amplification, dot blots were made as described, except that some blots received 5 µl of PCR product per dot instead of 2.5 µl to facilitate detection of transcripts that were present at lower abundances. The QADB method permits quantitative data to be obtained in situations in which alternative methods, such as semiquantitative RT-PCR [23] or single nucleotide primer extension [16], cannot be easily applied because of limitations in the number of embryos available, number of samples to be processed, and number of mRNAs to be assayed and a need for genotype analysis coupled to expression studies. The QADB method allows the rapid quantitative comparison of abundances for an essentially unlimited number of different mRNAs in a large number of samples using single embryos and using procedures that satisfy requirements for obtaining both genotypic and expression data.
Quantitative Aspects of the QADB Method
The QADB method has been used extensively in our laboratory [11, 22, 24–34]. In addition, the RT-PCR method used to generate the cDNA prior to quantitative QADB analysis has been used extensively in the preparation of representative cDNA libraries and in other quantitative studies based on single-cell analyses [20, 35–37]. Comparisons of published data indicate that the QADB method has yielded expression data that are in agreement with data obtained by alternative methods for a number of transcripts, e.g., tPA, Hprt, G protein 
S, q, i2, 13, 11, and 14 isoforms, c-myc, Pgk1, U2afbp-rs, Bcl-2, Bax, Bcl-x, Bad, Bcl-w, caspase-2, and Na+/K+-ATPase 1 subunit (compare above references with the following: tPA [38–40], Hprt [41], G proteins [42], c-myc [43, 44], Pgk1 [16], U2afbp-rs [45], Bcl-2, Bax, Bcl-x, Bad, Bcl-w, caspase-2 [46, 47], Na+/K+-ATPase 1 subunit [48, 49]). The QADB method has even been able to detect the low-abundance Xist RNA expression at the two-cell stage in mouse embryos [11], consistent with XIST RNA expression in the early (four-cell stage) human embryo [50] and Xist promoter-driven expression of a transgene in mouse two-cell embryos [51]. Another study employing a different RT-PCR method was unable to detect Xist expression in two-cell embryos [12]. The patterns of expression observed for different mRNAs during preimplantation development generally exhibit smooth profiles and comparatively small standard errors, indicating good reproducibility in the data. The magnitude of changes in estimated specific mRNA abundance observed with the QADB method has been similar to those derived by other methods of analysis [22]. In addition, the use of the QADB method to examine expression of imprinted genes has revealed the expected effects of parental chromosome origin on the preimplantation expression of the imprinted Mash-2, U2afbp-rs, H19, Xist, and Snrpn genes, including approximately twofold greater levels of expression in the appropriate uniparental embryos as compared with control fertilized embryos [11, 28, 31, 33]. An analysis of a twofold dilution series of rabbit globin mRNA mixed with total cell RNA revealed that the QADB method is quantitative, producing hybridization signals that are linear over at least three orders of magnitude (R2 = 0.992) and extending to a very low mRNA abundance [52]. Analyses of single embryo equivalents of a lysate of blastocyst stage embryos revealed excellent reproducibility (SD < 5%) in the measurements obtained by the QADB method [52]. These quantitative features combined with the specific methods for sample processing make the QADB method ideal for comparing the expression of multiple transcripts among large numbers of single embryos for which genotype data must also be obtained.
Hybridization Probes and Statistical Analyses
Molecular probes and methods of quantitation were as previously described [11]. Expression data within experiments were normalized according to hybridization signals obtained for the nonimprinted EF1 gene transcript [53]. Expression data from different experiments for different stages of embryos were compared after correcting for differences in hybridization probe specific activity. Student's t-test was used to evaluate the significance of differences between means. Chi-square tests were used to evaluate the significance of differences between observed and expected ratios of embryo genotypic classes. Differences producing P values of <0.05 were judged significant.
RESULTS
Efficiency of the Embryo Genotype Analysis Method
The methods devised for these studies provide a new and powerful means for obtaining quantitative gene expression data for single preimplantation embryos of known genotype. In contrast to previously described methods that yield genotype data for single cells or even single sperm [21, 54], our method allows the embryo to be processed in a manner compatible with the needs for quantitative RT-PCR analysis as well as genotype analysis. Because the genotypic characterization of individual embryos is critical to the success of the analysis, it is worth considering the overall success of the PEP-PCR methods used to obtain the genotypic data. An assessment of the efficiency of successful genotypic characterization of embryos was made possible by the genetic combinations employed in these studies. Four hundred seventy-four embryos were taken for analysis in experiments in which all sex chromosomes were genetically distinct (androgenones, gynogenones, and control embryos). These embryos were lysed between approximately 96 and 141 h after hCG injection and subjected to genotype analysis. Of these 474 embryos, 378 yielded genotypes that were consistent with genetic combinations that could be produced through nuclear transplantation or predicted in diploid control embryos, for an overall success of approximately 80%. The remaining embryos yielded no PCR signals, signals for only a single sex chromosome, or positive signals for three sex chromosomes.
Effects of X Chromosome Number on Androgenone Viability
In a previous study, XX androgenones were reported to have a reduced ability to form blastocysts relative to XY androgenones [12]. That earlier study, however, relied on the detection of the Y chromosome-specific Zfy gene for sex determination. As a result, all sex chromosomes present were not individually assayed with diagnostic markers, raising the possibility that the data might have been affected by some incidence of embryos being scored as falsely positive for a Y chromosome. In addition, the number of embryos assayed was small. To address more fully the question of whether XX androgenetic embryos exhibit reduced viability to the blastocyst stage, we examined the developmental capacities of a large number of androgenetic embryos produced using genetic combinations that permitted detection of all sex chromosomes present and hence unambiguous determination of embryo sex (Table 1).
|
Androgenone viability is typically high at the morula stage (96 h post-hCG). Because approximately one-quarter of androgenones should be YY and therefore lack an X chromosome, the expected viability to advanced cleavage stages would be 75%. Consistent with these expectations, viability of androgenones was approximately 79% at 96 h post-hCG, and only a small number of embryos that were successfully analyzed appeared to be of the YY genotype, indicating that such embryos are eliminated during cleavage as expected. The expected ratio of XY:XX androgenones is 2:1 (with 25% of the embryos being YY). We observed at 96 h a ratio of 36:11 (3.3:1); this ratio was higher than the expected ratio, but the difference was not significant.
Several experiments were performed for analyzing blastocysts at 121 h post-hCG. Typically, approximately half of all androgenones prepared with eggs from C57BL/6 mothers form blastocysts [55]. Two of the 121-h experiments employed androgenones containing DBA/2- and AKR/J-derived paternal pronuclei. In one of these two experiments (experiment 2), 46% of the androgenones formed blastocysts, and the ratio of XY:XX androgenones forming blastocysts was 25:16 (1.6:1), not significantly different from the expected ratio of 2:1. In the other experiment (experiment 3), the overall efficiency of blastocyst formation was 36%, and there was a significant reduction in the number of XX androgenones forming blastocysts. Combining the data for the unambiguously typed androgenones attaining the blastocyst stage for these two experiments yielded an overall ratio of XY:XX androgenones of 49:19 (2.6:1), not significantly different from the expected ratio. Two other experiments performed at 121 h employed androgenones made with C57BL/6 homozygous embryos (with embryo sex ascertained by the presence or absence of the Zfy gene; experiment 4) and androgenones made with one nucleus each from C57BL/6 homozygous and (C57BL/6 x B6-Pgk)F1 embryos (experiment 5), in which the X chromosomes were distinguishable using the DXMit210 marker. For experiment 4, using only C57BL/6 homozygous embryos, the overall efficiency of androgenone blastocyst formation was 41%, and there was no apparent deficiency in the development of XX androgenones to the blastocyst stage (XY:XX ratio of 30:16, or 1.9:1). In experiment 5, employing the C57BL/6 homozygous and (C57BL/6 x B6-Pgk)F1 embryos, 60% of the androgenones formed blastocysts, and the ratio of XY:XX androgenones was significantly greater than 2:1 (
2 = 5.72, P < 0.05). Fertilized embryos obtained for C57BL/6 x B6-Pgk crosses exhibited an XY:XX ratio of 41:49. The overall XY:XX androgenone ratio for all four experiments for 121 h post-hCG was 111:41 (2.7:1), slightly greater than expected but not significantly different (2 = 3.23, P > 0.05). At a later time (141 h post-hCG), 47% of the androgenones formed blastocysts and the XY:XX androgenone ratio was 19:10 (1.9:1), very close to the expected 2:1 ratio.
Although the overall difference in the number of XX androgenones developing to the blastocyst stage was not significantly different from the expected ratio of 2:1 when the data from multiple experiments are combined, the data are not inconsistent with the possibility that selection against XX androgenones can occur, as previously suggested [12]. This effect, however, is not uniform and is likely to be dependent upon additional genetic factors (e.g., presence of the B6-Pgk X chromosome) or other undetermined factors.
Patterns of X-Linked Gene Expression in Androgenetic Embryos
The expression of X-linked genes has been proposed to be different in XX androgenetic embryos either as a result of a failure to undergo XCI [12] or a result of inactivating both X chromosomes [11, 13]. Much of the justification for both of these models was the assumption that XX androgenones would manifest reduced viability to the blastocyst stage. Our previous studies, however, indicated that the Pgk1 gene is likely to be repressed in both XX and XY androgenones relative to control embryos [11]. Thus, although the data presented above indicate that XX androgenones are not selectively eliminated in all cases, we were interested in examining the effects of parental X chromosome origin and X chromosome number on X chromosome expression in preimplantation stage embryos. To do this, we applied the QADB method of quantitative RT-PCR analysis to the embryos that had been subjected to genotypic analysis and compared X-linked gene expression between XX and XY androgenones, between androgenetic, gynogenetic, and normal embryos, and between embryos at the three different times post-hCG injection. Through these comparisons, it was possible to obtain insight into how X-linked gene expression changes during the period analyzed and how these changes differ among embryos of the different genotypic classes and between nuclear transplant and normal embryos. Five X-linked genes were analyzed in this manner.
Effects of X Chromosome Number and Origin on Xist Gene Expression
Our analysis of control embryos between 96 and 141 h (Fig. 1) revealed interesting changes in Xist expression in male and female embryos. Xist gene expression was not significantly different between XX and XY control embryos at 96 h (P > 0.2), with a difference in mean expression value of less than 50%. At 121 h, expression in male embryos declined (Fig. 1B), creating an approximately eightfold difference in means between males and females (Fig. 1A). Xist expression was highly variable in the female control embryos at 121 h, however, so that the difference between means was not significant (P > 0.05). Nevertheless, the range of Xist expression values for control females (28–1673 cpm) extended to more than 20 times the maximum value observed for male control embryos (range, 11–83 cpm), and the median value for XX controls (105 cpm) was more than three times that for XY embryos (30 cpm), indicating that the potential for expression of the Xist RNA was much greater for a significant proportion of XX embryos as compared with any of the XY embryos. Because the QADB method exhibits excellent reproducibility in quantitation and because such variability was not observed at other time points, the variability observed for Xist RNA expression at 121 h most likely reflects actual variation between embryos. Expression in the female and male control embryos declined between 121 and 141 h (Fig. 1B). At 141 h, the mean value for Xist RNA expression in XX control embryos was significantly greater (ninefold difference; P < 0.05) than the mean value in XY control embryos.
|
As with control embryos, there was no significant difference between XX and XY androgenones in Xist RNA expression at 96 h post-hCG. At 121 h, Xist RNA expression appeared to increase in the XX androgenones (Fig. 1B), reaching a mean value that was approximately 5 times that observed for XY androgenones, but this difference was not significant (P > 0.05) (Fig. 1A). The range of expression values in XX androgenones (29–2900 cpm) extended to more than 10 times the maximum value observed in XY androgenones (range, 19–215 cpm), but the median values differed only slightly (96 and 75 cpm, respectively). Thus, although some of the XX androgenones manifested a greatly elevated Xist RNA abundance, the differences in expression between XX and XY androgenones were not as great as those observed for control embryos. Similar to control embryos, Xist expression declined in both XX and XY androgenones between 121 and 141 h post-hCG (Fig. 1B). An important difference in Xist RNA expression existed between androgenones and control embryos at 141 h post-hCG. Whereas Xist RNA expression was significantly greater (ninefold difference in means) between XX and XY control embryos, the difference between XX and XY androgenones at 141 h post-hCG was much less (3.5-fold difference in means) and not significant (P > 0.05). At 121 h, XX androgenones, with two paternally inherited X chromosomes, did not express significantly more Xist RNA than did XX control embryos. In fact, the maximum hybridization signal observed among the XX androgenones was less than half the maximum value observed for control embryos, and the mean value for XX androgenones was only 60% of that observed for control XX embryos. Thus, despite having twice the number of paternally derived X chromosomes (and hence twice the number of unmethylated, potentially active Xist alleles), XX androgenones tended to express less of the Xist RNA than did XX control embryos. Gynogenetic embryos expressed an amount of Xist RNA at 96 h similar to that seen in control XX embryos and then exhibited a sharply reduced Xist RNA abundance at 121 h, comparable to that observed in XY control embryos.
Effects of X Chromosome Number and Origin on Pgk1 Gene Expression
Our earlier studies documented a pronounced reduction in Pgk1 mRNA abundance in pooled androgenetic embryos as compared with control embryos [11]. We proposed that this reduction might be related to the close proximity of the Pgk1 gene to the Xist gene on the X chromosome, which could lead to short-range cis repressive effects arising due to Xist gene transcription from all paternal X chromosomes [11, 13]. To determine whether Pgk1 gene expression was uniformly reduced in both XX and XY androgenones as previously suggested or was greater in one of these two classes of embryos, we compared Pgk1 gene expression between XX and XY androgenetic and control embryos (Fig. 2). Pgk1 mRNA expression was significantly reduced in XX androgenones as compared with XX control embryos at 96, 121, and 141 h post-hCG (23-, 7-, and 2-fold differences in means, respectively; P < 0.05). Pgk1 mRNA expression was also significantly reduced in XY androgenones as compared with XY control embryos at 96 and 141 h post-hCG (fivefold and fourfold differences in means, respectively, P < 0.05) (Fig. 2A). Pgk1 mRNA expression did not differ significantly between XX and XY androgenones at 96 or 121 h post-hCG. Pgk1 mRNA expression in XX androgenones was greater at 141 h than at 121 h, whereas in XY androgenones Pgk1 mRNA expression declined over the same time interval (Fig. 2B). As a result, at 141 h post-hCG Pgk1 mRNA expression in XX androgenones was significantly greater (2.5-fold, P < 0.05) than that observed in XY androgenones (Fig. 2A). Moreover, Pgk1 mRNA expression in XX androgenones was not significantly different from expression observed in XY control embryos at 141 h post-hCG. No substantial decrease in Pgk1 mRNA expression occurred in control embryos between 121 and 141 h, and the mean Pgk1 mRNA expression was approximately 40% and 50% greater in XX control embryos than in XY embryos at these two stages, respectively.
|
Effects of X Chromosome Number and Origin on Expression of Other X-Linked Genes
We previously proposed that other X-linked genes located at a distance from the Xic might be repressed in XX androgenones but not in XY androgenones [11, 13]. To test this possibility, we examined the expression of the Hprt gene, which is located centromeric to the Xist and Pgk1 genes on the X chromosome, and the Prps1 and Pdha1 genes, which are located distal to the Xist and Pgk1 genes. There was no significant difference in mean expression values for any of the three genes between XX and XY androgenones or between XX and XY control embryos at 96 h post-hCG (Fig. 3). At 121 h post-hCG, there was no significant difference in means between XX and XY androgenones or between control embryos in expression of the Hprt gene. For the Pdha1 gene, XX androgenones expressed significantly more (2.5-fold difference in means; P < 0.05) transcript than did XY androgenones but not significantly more than did XX control embryos. The ratio of Pdha1 mRNA mean expression values between XX and XY control embryos was only 1.4, a difference that was not significant. No significant difference in Pdha1 mRNA expression was observed between XX and XY androgenones at 141 h post-hCG. For the Prps1 gene, XX androgenones appeared to express more transcript than did XY androgenones at 121 h post-hCG, but the difference in means was not judged significant (P > 0.2). Control XX embryos expressed more Prps1 mRNA than did control XY embryos (2.6-fold difference in means; P < 0.05). These data thus do not indicate either a consistent selective repression or a consistent selective elevation of expression of these three genes in XX androgenones. Gynogenetic embryos expressed significantly more Hprt and Pdha1 mRNA than did XX control embryos at 96 h but not at later stages.
|
We also compared X-linked gene mRNA abundances between XX and XY androgenones and control embryos in experiment 5, which employed the B6-Pgk1 congenics. In this experiment, the number of XX androgenones obtained was significantly lower than expected (Table 1), providing an opportunity to determine whether genes located away from the Xic were underexpressed or overexpressed when XX embryos appeared to be selected against. The number of (C57BL/6 x B6-Pgk)F1 fertilized control embryos available in this experiment was small, which may have limited our ability to observe significant differences for comparisons between adrogenones and control embryos. Significant differences in X-linked gene expression were nevertheless observed (Table 2). The XX androgenones expressed significantly more Pdha1 and Prps1 mRNA than did XY androgenones. For the Pdha1 mRNA, the ratio of mean expression values between XX and XY androgenones (4.7-fold difference) was well above the two-fold difference expected on the basis of X chromosome number, whereas XX androgenones expressed only about 1.8-fold more Prps1 mRNA than did XY androgenones. The mean expression value for the Pdha1 mRNA also appeared greater in XX androgenones than in XX control embryos, but this difference was not judged significant (P > 0.4). The XX androgenones expressed significantly more Prps1 mRNA than did XX control embryos. No significant differences were observed for the Hprt mRNA for comparisons involving XX androgenones, although XY androgenones expressed significantly more Hprt mRNA than did XY control embryos.
|
DISCUSSION
Embryos constructed by nuclear transplantation to contain exclusively maternal or paternal chromosomes provide a useful and unique system with which to examine the effects of genomic imprinting on gene expression in the preimplantation embryo and hence to explore the regulatory mechanisms that exploit genomic imprints to activate or silence preferentially maternal or paternal alleles during early development. Using such embryos, it is possible to explore the function of imprinted genes of one parental origin in the absence of alleles of the opposite parental origin, thus eliminating possible complications of allelic interactions. One area of genomic imprinting biology that is of special interest is the study of the mechanism by which genomic imprinting affects the process of XCI. The study of X chromosome function in preimplantation and peri-implantation stage embryos is valuable because it is during these early periods when the imprinting of the X chromosome first becomes apparent, when X chromosome counting first occurs, and when XCI first occurs. Moreover, XCI must meet differing needs of the extraembryonic tissues, in which preferential paternal XCI occurs, and the somatic tissues, in which XCI is not regulated by genomic imprinting. In this study, we used nuclear transplant embryos to investigate how genomic imprinting and X chromosome number affect X-linked gene expression during preimplantation development.
Two models have been proposed to describe how parental origin and X chromosome number might affect X chromosome function and blastocyst formation in androgenetic mouse embryos. One model proposed that XCI might not occur in androgenetic embryos because of a failure to accumulate a stable pool of the Xist RNA, which in turn is the result of a lack of expression of an imprinted (maternal allele expressed) gene that is required to sustain Xist RNA expression [12]. An alternative model proposed that because of a lack of Xist promoter methylation on paternal X chromosomes, androgenetic embryos would be affected by the expression of Xist RNA from every X chromosome [13]. The Xist RNA was proposed to cause a repression of genes located near the Xist gene itself (e.g., Pgk1) on all X chromosomes, but its effect on genes located at a distance from the Xist gene would be sensitive to an X chromosome counting mechanism and thus would reduce expression of the more distal (relative to the Xic) genes only in XX androgenones [13]. Both of these models were predicated upon the assumption, derived primarily from an earlier study of a limited number of embryos, that XX androgenones have reduced developmental competence to the blastocyst stage [12].
The data presented here indicate that selective elimination before the blastocyst stage of XX androgenones may occur but that this selection does not occur in all situations and may be sensitive to additional genetic and other factors. One genetic factor may be heterozygosity for the Xce locus or another closely linked locus that affects XCI. We observed a deficiency of XX androgenones made with pronuclei from C57BL/6 homozygous plus (C57BL/6 x B6-Pgk)F1 embryos (experiment 5). In the previous study, Kay et al. [12] also reported a deficiency of XX androgenones developing to the blastocyst stage when such embryos were constructed to be Xce heterozygotes. It might be hypothesized that heterozygosity for the Xce locus could facilitate XCI in XX androgenones and perhaps improve their viability by establishing an inherent difference between the two paternally derived X chromosomes. This difference might then overcome the effects of imprinting and allow XCI to occur. Preferential loss of XX androgenones that are heterozygous for the Xce locus, however, would be inconsistent with this hypothesis and would indicate that the effects of parental genomic imprinting on X chromosome function may be dominant to the effects of heterozygosity at the Xce locus. Instead, heterozygosity for alleles at Xce may enhance selective elimination of XX androgenones.
We previously proposed that the efficiency of androgenone development to the blastocyst stage typically observed (approximately 50% [55]) might reflect the combined loss of YY androgenones, which are expected to succumb to a lack of X-linked gene expression, and the preferential elimination of XX androgenones [13]. The loss of these two classes would leave only XY androgenones, which should comprise about half the population. Our data indicate that the typical 50% efficiency of androgenetic blastocyst formation cannot be explained on this basis. In a majority of experiments, about half of the androgenones formed blastocysts as expected. This frequency of blastocyst formation often occurred, however, without an apparent loss of XX androgenones. This result indicates that the developmental capacity of androgenetic embryos is significantly reduced independently of whether they possess one or two X chromosomes. The basis for this finding has not been determined but may reflect either effects at imprinted autosomal loci or an effect of an exclusively paternal origin of X chromosomes.
The data presented here indicate that genomic imprinting can significantly affect X chromosome function in the preimplantation embryo. Two pronounced effects of imprinting were observed. First, the Pgk1 gene was repressed in androgenetic embryos as previously reported [11]. This observation is consistent with the hypothesis that initial transcription of the Xist gene from all paternal X chromosomes leads to a short-range cis effect on gene expression. This repression of genes near the Xic may contribute to the general reduction in androgenone viability. Second, the ratio of Xist expression between XX and XY androgenones, which lack maternal chromosomes, was much less at 141 h post-hCG as compared with that in control embryos. This finding is consistent with the possibility that a product encoded by an imprinted gene expressed from the maternal genome plays a role in controlling Xist RNA accumulation [12] and that stabilization of the Xist RNA may promote dosage compensation in preimplantation stage embryos as has been observed for embryonic cells at later stages [14, 15]. Further studies will be required to test this possibility and to investigate the nature of factors that may be expressed from maternal chromosomes to modulate Xist RNA expression.
Compensation of X chromosome dosage begins during the late preimplantation/early peri-implantation period [7]. In one study, the ratio of paternal:maternal allele expression for the Hprt and Pgk1 genes declined between the eight-cell and blastocyst stages [16], consistent with onset of dosage compensation during the blastocyst stage (with preferential inactivation of the paternal X chromosome). We observed no significant difference in Pgk1 mRNA expression between XX and XY control blastocysts, indicating that a paternally derived X chromosome does not greatly augment expression of the Pgk1 gene above what is attributable to the maternal allele. This finding, together with the severe repression of the Pgk1 gene in androgenones, is consistent with an early effect of imprinting coupled with the onset of dosage compensation for the Pgk1 gene before the blastocyst stage.
We proposed that selective inactivation of genes located at a distance from the Xist gene on both X chromosomes in XX androgenones could lead to their selective demise [11, 13]. Kay et al. [12] proposed that overexpression of X-linked genes might contribute to selective demise of XX androgenones. With the exception of the Pgk1 gene located near the Xist gene, we failed to observe any significant underexpression of X-linked genes (e.g., Hprt, Pdha1, Prps1) in XX androgenones. The Pdha1 gene was more highly expressed in XX androgenones than in XY androgenones in two different experiments, including one in which XX androgenones appeared to be selectively eliminated. Pdha1 mRNA expression was not significantly elevated relative to XX control embryos, however, in either experiment. No elevation in X-linked gene expression was observed at 141 h post-hCG between XX and XY androgenones. Thus, it is not clear that overexpression of X-linked genes is a general feature of XX androgenones or that it can account for any reduction in viability. It may be that X chromosome imprinting affects more strongly other genes not examined in this study. Thus, if X chromosome dosage significantly affects X-linked gene expression and thereby contributes to differential viability of XX and XY androgenones, such an effect may be limited to selected genes and may be sensitive to genetic or environmental factors.
Our observations indicate that, although some X-linked genes (e.g., Pgk1) may become repressed in the preimplantation embryo through a dosage compensating mechanism, this repression may be somewhat unstable and sensitive to changes in Xist RNA expression. This finding is reminiscent of earlier observations indicating that XCI differs between extraembryonic and somatic tissues and that the inactive X chromosome of extraembryonic tissues can be reactivated more readily than that of somatic cells [56]. Reversible XCI in the preimplantation embryo may be necessary to provide for dosage compensation with preferential paternal XCI during early embryogenesis, while permitting random XCI later in the developing embryo proper. This reversibility could be achieved by developmentally regulated changes in the rate of Xist gene transcription and changes in Xist RNA stability. Further studies employing androgenones and normal embryos possessing different alleles of X-linked genes should be useful in characterizing the molecular mechanisms that promote, maintain, and stabilize XCI in the early embryo in a way that is compatible with later developmental requirements of both embryonic and extraembryonic tissues.
ACKNOWLEDGMENTS
We thank Carmen Sapienza for critical comments on the manuscript.
FOOTNOTES
First decision: 7 January 2000.
1 This work was supported in part by grants from the March of Dimes Birth Defects Foundation and the National Science Foundation (MCB-9807542). 
2 Correspondence: Keith Latham, The Fels Institute for Cancer Research and Molecular Biology and Department of Biochemistry, Temple University School of Medicine, 3307 North Broad St., Rm. 302, Philadelphia, PA 19140. FAX: 215 707 1454; klatham{at}astro.ocis.temple.edu
Accepted: February 8, 2000.
Received: November 24, 1999.
REFERENCES
subunit mRNA abundance in preimplantation mouse embryos using a rapid, quantitative RT-PCR approach. Mol Reprod Dev 1995; 41:314–324.[CrossRef][Medline] and ß subunit genes during preimplantation development of the mouse. Dev Genet 1990; 11:41–48.[CrossRef][Medline]This article has been cited by other articles:
|
|
 |
|
  M. Boiani, L. Gentile, V. V. Gambles, F. Cavaleri, C. A. Redi, and H. R. Scholer Variable Reprogramming of the Pluripotent Stem Cell Marker Oct4 in Mouse Clones: Distinct Developmental Potentials in Different Culture Environments Stem Cells, September 1, 2005; 23(8): 1089 - 1104. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P. Zheng, B. Patel, M. McMenamin, S. E. Reddy, A. M. Paprocki, R. D. Schramm, and K. E. Latham The Primate Embryo Gene Expression Resource: A Novel Resource to Facilitate Rapid Analysis of Gene Expression Patterns in Non-Human Primate Oocytes and Preimplantation Stage Embryos Biol Reprod, May 1, 2004; 70(5): 1411 - 1418. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
I. Lagutina, G. Lazzari, R. Duchi, and C. Galli Developmental Potential of Bovine Androgenetic and Parthenogenetic Embryos: A Comparative Study Biol Reprod, February 1, 2004; 70(2): 400 - 405. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 L. P. O'Neill, T. E. Randall, J. Lavender, H. T. Spotswood, J. T. Lee, and B. M. Turner X-linked genes in female embryonic stem cells carry an epigenetic mark prior to the onset of X inactivation Hum. Mol. Genet., August 1, 2003; 12(15): 1783 - 1790. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. E. Latham, H. Akutsu, B. Patel, and R. Yanagimachi Comparison of Gene Expression During Preimplantation Development Between Diploid and Haploid Mouse Embryos Biol Reprod, August 1, 2002; 67(2): 386 - 392. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. Wrenzycki, A. Lucas-Hahn, D. Herrmann, E. Lemme, K. Korsawe, and H. Niemann In Vitro Production and Nuclear Transfer Affect Dosage Compensation of the X-Linked Gene Transcripts G6PD, PGK, and Xist in Preimplantation Bovine Embryos Biol Reprod, January 1, 2002; 66(1): 127 - 134. [Abstract] [Full Text]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
