BIOLOGY OF REPRODUCTION 67, 386–392 (2002)
Comparison of Gene Expression During Preimplantation Development Between
Diploid and Haploid Mouse Embryos1
Keith E. Latham,2,3,4 Hidenori Akutsu,5 Bela Patel,3 and Ryuzo Yanagimachi5
The Fels Institute for Cancer Research and Molecular Biology3 and The Department of Biochemistry,4
Temple University School of Medicine, Philadelphia, Pennsylvania 19140
Institute for Biogenesis Research,5 The John A. Burns School of Medicine, University of Hawaii at Manoa,
Honolulu, Hawaii 96822
ABSTRACT
ganisms, is subject to gene regulatory mechanisms that disrupt the ability of a haploid genome to support early development.
The reasons for the severely limited developmental potential of haploid embryos in mammals have not been discerned. Several interesting possibilities have been examined, but none has provided a likely explanation. For example, earlier studies revealed that aneuploidy can occur in
haploid parthenogenones [13], and that this might result in
an absence of essential gene functions through chromosome
loss. Aneuploidy affects only a small fraction of embryos,
however, and is dependent upon the method of oocyte activation employed [13, 14], making this an unlikely explanation. Other studies examined the possible influence of the
DNA:cytoplasm ratio on haploid developmental restrictions
[2, 8]. Haploid embryos made by zygote bisection, or by
pronuclear removal followed by cytoplasmic reduction,
showed some improvement in development, but the overall
rate of development remained severely reduced relative to
diploid embryos. Genomic imprinting offers another possible explanation, as haploid embryos possess exclusively
maternal or paternal chromosomes. This is also unlikely by
itself to provide a satisfactory explanation, because haploid
androgenones, gynogenones, and parthenogenones develop
substantially worse than their diploid counterparts [2, 15,
16]. Effects of genomic imprinting on X chromosome function, however, is a possibility that has not been tested. Imprinting of the X chromosome, attributable to differential
Xist gene methylation, leads to preferential inactivation of
the paternal X chromosome in extraembryonic tissues and
preferential expression of the maternal X chromosome during preimplantation development (reviewed in [17]). Imprinting of the X chromosome could thus lead to potentially
lethal repression of the single X chromosome in haploid X
chromosome-bearing androgenones; the Y chromosomebearing haploid androgenones would obviously also lack
X-linked gene expression. Similarly, dosage compensation
by X chromosome inactivation would be unlikely in haploid parthenogenones, because such embryos would have
to achieve a 50% expression level of their 1 X chromosome
relative to their 1 set of autosomes, and no evidence for
such regulation of the X chromosome exists for mammals.
Thus haploid parthenogenones may overexpress X-linked
genes relative to autosomal genes.
The aim of this study was to examine the effects of
haploid development on gene expression, and in particular
effects on X-linked gene expression, and to evaluate to
what degree newer techniques of producing and culturing
such embryos might affect developmental potential. We
produced haploid and diploid parthenogenetic and androgenetic embryos, and re-evaluated their developmental potential, their genomic integrity, and their relative expression
Haploid development is a normal part of the life cycle for
some animals, but it has not been observed in mammals. Studies
in mice have revealed that the preimplantation developmental
potential of haploid embryos is significantly impaired relative to
diploid embryos. The reasons for the severely limited developmental potential of haploid embryos in mammals have not been
discerned. To examine the effects of haploid development on
gene expression, and in particular on X-linked gene expression,
and to evaluate to what degree newer techniques of producing
and culturing such embryos might affect developmental potential, haploid and diploid parthenogenetic and androgenetic embryos were produced and reevaluated for developmental potential, genomic integrity, and relative expression levels of specific
autosomal and X-linked gene transcripts. Our data confirm the
previously observed restriction in haploid developmental potential, eliminate chromosomal abnormalities as a major factor in
this restriction, and reveal subtle alterations in gene expression.
Haploid parthenogenones display only very subtle alterations in
the expression of most mRNAs but a consistent elevation in Xlinked Bex1 mRNA expression. Haploid androgenones seem to
lack repression of the Pgk1 gene that is seen in diploid androgenones, but this may reflect ongoing loss of those haploid androgenones that experience X chromosome inactivation. The significance and possible explanations for these differences are discussed.
early development, embryo, gene regulation
INTRODUCTION
Haploid development is a normal part of the life cycle
for some animals (e.g., parasitic wasps) [1], but it is not
observed in mammals. Studies in mice reveal that the preimplantation developmental potential of haploid embryos is
significantly impaired relative to diploid embryos [2–9].
Embryonic stem (ES) cells derived from haploid parthenogenones only contribute to the development of chimeric
animals after they become diploidized [10], and ES cell
lines established from haploid parthenogenones likewise
become diploid [11]. Development of haploid rabbit parthenogenones is also inefficient [12]. This indicates that the
mammalian genome, in contrast to genomes of other orSupported in parts by grants from the NIH (RR15253), the Harold K.
Castle Foundation, and the Victoria and Bradley Geist Foundation.
Correspondence: Keith E. Latham, The Fels Institute for Cancer Research,
Temple University School of Medicine, 3307 North Broad St., Philadelphia, PA 19140. FAX: 215 707 1454; e-mail: [email protected]
1
2
Received: 16 November 2001.
First decision: 3 December 2001.
Accepted: 22 February 2002.
Q 2002 by the Society for the Study of Reproduction, Inc.
ISSN: 0006-3363. http://www.biolreprod.org
386
PREIMPLANTATION GENE COMPARISON IN MOUSE EMBRYOS
levels of specific autosomal and X-linked gene transcripts.
Our data confirm the previously observed restriction in haploid developmental potential, eliminate chromosomal abnormalities as a major factor in this restriction, and reveal
subtle alterations in gene expression. The significance and
possible explanations for these differences are discussed.
MATERIALS AND METHODS
Reagents
Polyvinyl alcohol (PVA, cold-water-soluble, molecular weight
;10 000) and polyvinyl pyrrolidone (PVP, molecular weight ;360 000)
were purchased from Sigma Chemical Co. (St. Louis, MO). Bovine testicular hyaluronidase (200 UPS U/mg) was obtained from ICN Biochemicals (Costa Mesa, CA). BSA (fraction V) was purchased from Calbiochem
(La Jolla, CA) and mineral oil from Squibb and Sons (Princeton, NJ). All
other reagents were obtained from Sigma unless otherwise stated.
Media
CZB medium [18] supplemented with 5.56 mM D-glucose was used
for the culture of mouse oocytes after microsurgery. This was called standard CZB. The medium for collection of oocytes from oviducts and subsequent oocyte treatments, including micromanipulation, was a modified
CZB (Hepes-CZB, [19]) containing 20 mM Hepes-HCl, a reduced amount
of NaHCO3 (5 mM), and 0.1 mg/ml PVA instead of BSA. Standard CZB
was used at 378 under 5% CO2 in air, and Hepes-CZB was used under air.
Animals
Spermatozoa were collected from caudae epididymides of (C57BL/6
3 DBA/2)F1 male mice. Females of the same genotype were used as
oocyte donors. These animals were maintained in accordance with the
guidelines of the Laboratory Animal Service at the University of Hawaii
and those prepared by the Committee on Care and Use of Laboratory
Animals of the Institute of Laboratory Resources National Research Council (DHEW publication [NIH] 80-23, revised in 1985). The animal handling and treatment protocol was reviewed and approved by the Animal
Care and Use Committee at the University of Hawaii.
Preparation of Androgenones and Parthenogenones
Haploid and diploid androgenones were prepared using unfertilized
oocytes from 4- to 8-wk-old females injected with 5 IU eCG followed by
5 IU hCG 48 h later. Mature oocytes were collected from oviducts at
approximately 15 h after the hCG injection. They were freed from the
cumulus cells by 5-min treatment with 0.1% bovine testicular hyaluronidase in Hepes-CZB. The oocytes were rinsed and kept in CZB medium
at 378C under 5% CO2 in air for less than 2 h before further treatments.
The oocytes were transferred to a 20-ml droplet of Hepes-CZB containing
5 mg/ml cytochalasin B (CB; 1003 stock in dimethylsulfoxide), which
had previously been placed in the operation chamber on the microscope
stage. They were kept there for 5–10 min before enucleation. The metaphase II chromosome-spindle complex was aspirated into a pipette (8–10
mm in the inner diameter) with a minimal volume of oocyte cytoplasm
[20], and the resulting oocyte cytoplasts were then cultured in cytochalasin
B-free CZB for up to 2 h at 378C before sperm injection. Haploid androgenetic embryos were made by injection of a single spermatozoon into
oocyte cytoplasts according to Kimura and Yanagimachi [19], except that
only the sperm head was injected and the operation was performed at room
temperature. The sperm head was separated from the tail by applying a
few piezo-pulses to the head-tail junction. Because the sperm heads contribute oocyte-activating factors [21], no additional treatment was needed
for activation. After further culture in CZB, embryos with 1 male pronucleus were considered haploid androgenones. Diploid androgenones were
prepared in the same way as haploid androgenones, but 2 isolated sperm
heads, instead of 1, were injected into each oocyte cytoplast. Embryos
with 2 male pronuclei were considered diploid androgenones.
Haploid and diploid parthenogenones were prepared from metaphase
II oocytes by activating for 6 h in Ca21-free CZB containing 10 mM Sr21
at 378C in a humidified atmosphere of 5% CO2 in air [22], with or without
5 mg/ml CB to suppress second polar body extrusion. Eggs that formed
single female pronuclei and second polar bodies were considered haploid
parthenogenones and cultured in CZB. Second polar bodies had sharply
defined membranes and round centrally located nuclei and thus could be
387
distinguished from first polar bodies, which had granular cytoplasm and
scattered chromosomes [23]. For oocytes not treated with CB, eggs with
2 distinct pronuclei with or without a second polar body were discarded.
For eggs treated with CB, those with 2 distinct pronuclei were regarded
as diploid parthenogenones and cultured in CZB.
Chromosome Analysis
Seven to 8 h after the start of incubation in oocyte-activating Sr21
medium or sperm injection, a group of pronuclear eggs was transferred
into another droplet (0.2 ml) of CZB containing 0.006 mg/ml vinblastine,
a microtubule-disrupting agent. Between 19 and 21 h after sperm injection,
eggs that had been arrested at the first mitotic metaphase were treated with
0.25% (w/v) pronase (Kaken Pharmaceuticals, Tokyo, Japan) for 5 min to
remove zonae pellucida, then exposed to a hypotonic solution (1:1 mixture
of 1% Na-citrate and 30% fetal bovine serum) for 10 min at room temperature. Fixation of eggs and spreading of chromosomes were performed
as described [24]. The chromosomes on slides were stained with 2%
Geimsa solution for 8 min. After conventional chromosome analysis, they
were C-banded [25] to detect acentric and dicentric chromosomes [24].
Another group of pronuclear stage eggs was cultured in CZB for 36 h to
allow them to develop into 2-cell embryos. They were transferred into
CZB containing 0.006 mg/ml vinblastine for 10–14 h to arrest them at the
metaphase of the second cleavage. They were processed for chromosome
examination as already described.
In Situ Hybridization
Haploid and diploid androgenones (blastocysts, Day 5 of culture) were
treated with 0.25% (w/v) pronase as above, then exposed to a hypotonic
solution (1:4 mixture of 1.2% Na-citrate and 40:fetal bovine serum) for
10 min at room temperature. Embryos were fixed as described [24]. The
in situ hybridization protocol followed the procedure of the supplier of
biotin-labeled probes for mouse chromosome X and specific subcentromere and whole chromosome Y probes (Applied Genetics Laboratories Inc.,
Melbourne, FL). Fluorescence microscopy was performed using a Nikon
Microscope ECLIPSE E600 with filter module B2A (Nikon Inc., Melville,
NY). Blastomeres having an X chromosome show a small dot of fluorescence, whereas a Y chromosome produces a much larger fluorescence
signal. For diploid embryos, XX embryos have 2 small X chromosome
signals, whereas XY embryos have the small X chromosome signal and
the larger Y chromosome signal. Haploid embryos obviously have either
a small X chromosome signal or a small Y chromosome signal. All embryos analyzed fit into these categories. Because diploid blastocysts contain many cells, which in some cases had overlapping nuclei, analyses
were limited to those nuclei that had no such overlap and could be unambiguously characterized; between 5 and 10 cells were analyzed for diploid androgenetic blastocysts to ensure accurate genotyping. For haploid
androgenetic blastocysts, which contained fewer cells with fewer overlapping nuclei, nearly all cells could be analyzed.
Quantitative RT-PCR Analysis
The abundances of specific mRNAs were measured using the quantitative amplification and dot blotting (QADB) method for quantitative reverse transcription-polymerase chain reaction (RT-PCR) analysis [26, 27].
Briefly, embryos were lysed in guanidine thiocyanate buffer and stored at
2708C, and then nucleic acids prepared by ethanol precipitation as described [17]. After heat denaturation and annealing with oligo(dT), reverse
transcription and polymerase chain reaction were performed as described
[17, 26, 27] to amplify quantitatively the 39 terminal portions of the entire
mRNA population. Quantitation of specific mRNAs was achieved by dot
blotting and cDNA hybridization as described [26, 27]. The sensitivity and
reliability of the QADB assay have been extensively documented [28].
The QADB method is applicable to small amounts of material, even single
embryos; provides the ability to quantify expression of a large number of
mRNAs; and can provide estimates of actual mRNA abundance [26, 27].
These properties make the QADB method ideal for examining mRNA
abundance in nuclear transplant embryos. The QADB method is fully
quantitative and produces hybridization signals that are linear over at least
3 orders of magnitude (R2 5 0.992), extending to a very low mRNA
abundance, and with excellent reproducibility [28]. The QADB method
also exhibits excellent reproducibility between experiments with respect
to qualitative patterns of mRNA expression and quantitative estimates of
mRNA abundance (e.g., [27, 29]).
388
LATHAM ET AL.
TABLE 1. In vitro development of mouse haploid and diploid parthenogenetic and androgenetic eggs.
Total
number
of eggs
cultured
Type of egg (No. exp.)*
Number (%) of eggs developed into
2-cell
4-cell
Morula
Blastocyst
Parthenogenone
Haploid
165 (5)
Diploid
304 (6)
152 (92)
304 (100)
131 (79)
297 (98)
118 (72)
285 (94)
86 (52)a
283 (93)b
Androgenone
Haploid
287 (6)
Diploid
620 (17)
281 (98)
589 (95)
225 (78)
465 (75)
101 (35)
421 (68)
36 (12)c
303 (49)d
* No. exp., Number of experiments.
a,b,c,d a vs. c, b vs. d are significantly different (P , 0.005), and a vs. b,
c vs. d are significantly different (P , 1026).
Statistical Analysis
Developmental and chromosomal data for haploid and diploid embryos
were evaluated using the chi-square test with Yates correction for continuity. For mRNA expression data, the statistical significance of differences
in means was evaluated using the Student t-test.
RESULTS
Developmental Potential of Haploid
and Diploid Embryos
The methods employed here to produce haploid embryos
differed somewhat from those used in earlier studies, most
notably in the use of Sr21 for egg activation and the use of
sperm head injection to produce androgenones. The use of
Sr21 for egg activation has advantages over ethanol activation employed in many earlier studies. Sr21, like spermatozoa, induces repetitive oscillations in intracellular free
Ca21 concentration, whereas other agents like ethanol and
calcium ionophores induce a single Ca21 increase [30].
Thus Sr21 treatment more closely approximates physiological oocyte activation and leads to enhanced development
[30]. In addition, the culture medium employed in this
study (CZB) [18] supports enhanced development as compared with media employed in earlier studies. Given these
differences, it was necessary to re-evaluate the developmental potential of haploid and diploid androgenones and
parthenogenones produced using the methods described
above. As seen in earlier studies, preimplantation development was significantly more efficient in diploid parthenogenones than haploid parthenogenones (P , 1026; Table
1). Haploid androgenones were likewise significantly less
efficient in blastocyst formation than diploid androgenones
(P , 1026). As expected, both types of androgenones developed significantly more poorly than their parthenogenetic counterparts. The efficiency of blastocyst formation
for diploid androgenones (49%) was as expected and consistent with earlier studies, which revealed the demise of
those of the YY genotype (see below) combined with developmental arrest among those of other genotypes due to
other, unidentified factors [17]. The blastomeres of haploid
androgenones were of irregular size starting from the 4-cell
stage. We also observed that the second cleavage division
was either delayed or exhibited greater asynchrony between
blastomeres than is seen with haploid parthenogenones.
This was reflected in a significantly larger fraction of haploid androgenones persisting at the 3-cell stage at 48–50 h
of development, as compared to haploid parthenogenetic
embryos (Table 2). Differences in developmental rates be-
TABLE 2. Appearance of mouse haploid androgenetic and parthenogenetic eggs after 48–50 h of development in vitro.
Type of egg
Parthenogenone
Androgenone
Number (%) of eggs
developed into
Total number of
eggs cultured
(No. exp.)*
3-cell
4-cell
91 (3)
470 (10)
4 (4.4)a
90 (19.1)b
79 (86.8)
324 (68.9)
* No. exp., Number of experiments.
a,b a vs. b are significantly different (P , 0.005).
came more apparent for haploid embryos with continued
incubation so that haploid embryos required approximately
24 h longer than their diploid counterparts to attain the
blastocyst stage. Haploid androgenones were the least efficient in forming blastocysts, and those few that progressed
to the blastocyst stage displayed small blastocoele cavities
and poor morphologies (Fig. 1). Thus the results obtained
here, using these newer methods of haploid embryo construction, confirm the limited developmental potential of
haploid embryos relative to their diploid counterparts.
Chromosome Analysis of Haploid and Diploid Embryos
To test the possibility that haploid embryos are compromised as a consequence of chromosomal abnormalities, we
performed chromosome analyses of embryos on metaphase
spreads obtained at the first and second mitotic divisions
(Table 3). These analyses revealed that chromosomal abnormalities affected only a small fraction (2%–7%) of haploid androgenones or parthenogenones. This did not differ
significantly from the rate of chromosomal abnormalities
seen in control biparental embryos made by intracytoplasmic sperm injection (ICSI). Because such a small fraction
of haploid embryos displayed chromosomal abnormalities,
such abnormalities cannot account for the reduced developmental potential.
To evaluate whether sex chromosome composition affected development, we examined this feature of haploid
and diploid androgenones at the blastocyst stage (Table 4).
As expected based on earlier studies [17], diploid androgenetic blastocysts did not include any YY embryos. There
appeared to be a slightly larger than expected number of
XY androgenetic blastocysts (2:1 XY:XX ratio predicted),
but this difference was not statistically significant. This
FIG. 1. Haploid androgenetic blastocysts. Embryos were photographed
after 4 days of culture.
389
PREIMPLANTATION GENE COMPARISON IN MOUSE EMBRYOS
TABLE 3. Chromosome analyses of the first and second cleavage metaphase in haploid androgenetic and parthogenetic ova.
Embryo
Androgenones
Parthenogenones
ICSI control
Cleavage
number
1
2
1
2
1
Number of
eggs analyzed
(No. exp.)*
123
83
103
111
64
Number (%) of eggs with
chromosomal abnormalities
Number (%) of
eggs with normal
karyotype
(3)
(6)
(3)
(4)
(8)
115
79
101
106
61
Structurala
(94)
(95)
(98)
(96)
(95)
8
4
2
4
3
Aneuploidy
(7)b
0
0
0
0
0
(5)b
(2)b
(4)b
(5)b
(0)
(0)
(0)
(0)
(0)
Mosaic
—
0 (0)
—
0 (0)
0 (0)
* No. exp., Number of experiments.
a Chromosome breaks.
b Not significantly different (P . 0.05).
confirms the earlier observation that XX androgenones are
not significantly impaired relative to XY androgenones
[17], in contrast to another study reporting such a difference
[31]. Not surprisingly, haploid Y chromosome-bearing androgenones also perished before the blastocyst stage. This
loss of Y chromosome-bearing haploid androgenones
should prevent 50% of these embryos from forming blastocysts. The observed rate (12%), however, is far below
50%, indicating an effect of factors other than just an absence of an X chromosome.
Gene Expression Analysis of Haploid
and Diploid Embryos
Gene expression patterns have never been compared between haploid and diploid embryos. In order to evaluate
the degree to which haploid and diploid androgenones and
parthenogenones might differ at the molecular level, we
compared these 4 types of uniparental embryos to fertilized
control embryos at the morula and blastocyst stages for
relative abundance of 11 different mRNAs. Five of these
mRNAs were autosomally encoded transcripts, comprising
a collection of 4 housekeeping mRNAs that are expected
to be constitutively expressed (actin, translation elongation
factor EF1a, ribosomal protein L23, and transcription factors Sp1 and TATA box binding protein TBP), and 1
mRNA encoding another transcription factor (mTEAD2)
that is developmentally up-regulated at the blastocyst stage
and thus may serve as a molecular marker of developmental
progression. Another 6 mRNAs analyzed encompassed Xlinked transcripts that provided the ability to evaluate X
chromosome function in these embryos. The temporal patterns of expression of these genes have been published elsewhere [26, 28, 29, 32, 33].
Among the housekeeping mRNAs, only a few statistically significant differences were observed among the 4
types of uniparental embryos relative to fertilized embryos,
and these were observed only in blastocyst stage embryos
(Fig. 2). Diploid parthenogenones displayed an approximately 3-fold reduced mean expression of the L23 mRNA
compared with fertilized controls. Haploid and diploid parthenogenones displayed a reduced mean expression (34%–
45%) of the EF1a mRNA as compared with fertilized controls (P , 0.05). EF1a mRNA mean expression also appeared reduced in haploid androgenones, but this difference
was just below the level of statistical significance (P 5
0.055). The Sp1 mRNA also appeared reduced in both
types of parthenogenones, but the difference in means was
not statistically significant (P 5 0.088 and 0.104). Actin
mRNA expression was variably elevated in haploid androgenones, but the difference in means did not reach the level
of statistical significance (P 5 0.14).
Among the X-linked transcripts analyzed, a number of
significant differences were observed (Fig. 2). Haploid androgenetic blastocysts exhibited a significant 7-fold reduction in mean Hprt mRNA abundance relative to fertilized
controls (P , 0.02). Expression of the Pdha1 mRNA was
reduced approximately 2-fold (P , 0.03) in diploid androgenetic morulae. The mean Pgk1 mRNA abundance was
significantly reduced in diploid androgenones at both the
morula and blastocyst stages (P , 0.005 and 0.008), consistent with earlier observations [17, 28]. The mean value
for Pgk1 mRNA expression in haploid androgenetic blastocysts was 2-fold lower than for fertilized controls. The
mean value for haploid androgenones was not significantly
different from the mean for fertilized controls, but neither
was it significantly different from the mean for diploid androgenones. Thus it appears that Pgk1 expression in haploid
androgenones is variably reduced. Expression of the Prps1
mRNA was significantly reduced in haploid and diploid
parthenogenones (P , 0.03). The Bex1 mRNA displayed
significantly increased mean expression values in all 4
types of uniparental embryos at the morula stage relative
to fertilized control embryos, but only in haploid parthenogenones at the blastocyst stage (P , 0.03–0.01). The expression of Xist RNA appeared to be elevated in haploid
androgenetic blastocysts relative to diploid androgenetic
blastocysts and fertilized controls, but expression was quite
variable and the difference in means was not statistically
significant.
DISCUSSION
The data presented here confirm earlier reports that haploid mammalian embryos have severely limited developmental potential, even when newer culture media, newer
methods of oocyte activation (parthenogenones), and alternative methods of embryo construction (androgenones) are
employed. It could be suggested that physical damage related to sperm injection reduced the developmental potential of androgenones. However, diploid androgenone development was comparable to what is observed using the karyoplast fusion technique [15, 16], and haploid androgenone
TABLE 4. FISH analysis for haploid and diploid androgenetic blastocysta
embryos.
Number
of eggs
Genotypes analyzed
Haploid
Diploid
32
45
Number (%) of embryos of each genotype
X
Y
XX
XY
YY
32 (100)
—
0 (0)
—
—
11 (24)
—
34 (76)
—
0 (0)
a Diploid blastocysts were obtained after 96–100 h of development. Haploid blastocysts were obtained after 120–126 h of development.
390
LATHAM ET AL.
FIG. 2. Comparison of mRNA abundance
between haploid and diploid embryos. The
expression of 5 constitutively expressed
housekeeping mRNAs is shown in the lefthand set of graphs, and the expression of
6 X-linked transcripts is shown in the
right-hand set. The vertical line demarcates
samples from morulae and samples from
blastocyst stage embryos. Data are expressed in units of mean cpm bound
(6SEM) of 5–6 samples (7–10 embryos
each). Asterisks denote means that were
significantly different from the fertilized
controls. A, Androgenones; P, parthenogenones; 1N, haploid; 2N diploid; Fert, fertilized embryos. Expression data were normalized to amount of DNA bound to each
dot. Diploid morulae and blastocysts were
obtained after 76–80 and 96–100 h of development, respectively. Haploid morulae
and blastocysts were obtained after 96–
100 and 120–126 h of development, respectively.
development was approximately equal to that of 1-pronucleate (presumably haploid) androgenones obtained by insemination under conditions of limiting sperm concentration [34]. Moreover, the production of embryos by intracytoplasmic sperm injection of ovulated oocytes results in
up to 93% blastocyst formation [35], thus revealing no apparent negative effect of the injection procedure alone. The
data also exclude chromosomal abnormalities as a plausible
explanation for the limited development of haploid embryos. Thus it seems most likely that the reduced developmental potential of haploid embryos is related to defects in
gene expression.
For some genes, haploidy could inhibit accumulation of
transcripts to the normal cellular abundance, whereas for
other genes aspects of mRNA stability and the modulation
of transcription rate may permit accumulation to normal
transcript abundance. Because the QADB method employed here measures mRNA abundance as a fraction of
the whole mRNA population, alterations in proportional
mRNA abundance through such effects should be detectable. Our data failed to reveal for the genes analyzed here
simple 2-fold reductions in mRNA abundance that might
be expected to arise through such an effect. Moreover, reduction of cytoplasmic volume fails to improve the developmental potential of haploid embryos [2, 8], arguing
against a simple explanation related to a DNA:cytoplasm
ratio. Our data reveal, however, subtle effects on X-linked
gene expression, which indicate that X chromosome imprinting and inactivation may contribute to developmental
arrest in haploid embryos. Hprt mRNA expression was significantly reduced in the haploid androgenones at the blastocyst stage. This could reflect either a developmental delay
in these embryos, with a resultant delay in accumulation of
this mRNA, or a possible imprinting-related paternal X
chromosome inactivation in these embryos. Xist RNA expression tended to be elevated in haploid androgenones,
which would be consistent with the latter possibility. Although Xist RNA expression was variable and the difference in means was not statistically significant, a similar
degree of variability was observed previously [29], even
among single embryos (androgenetic and fertilized) genetically determined to be of the XX genotype [17]. Diploid
androgenones were previously found to display a consistently reduced abundance of the Pgk1 mRNA regardless of
whether they are XX or XY [17, 29]. Diploid androgenones
displayed that same feature in this study. Pgk1 mRNA expression was variable in haploid androgenones. Although
one explanation for this is that haploid androgenones are
deficient in X chromosome inactivation, another possibility
is that these embryos initiate the X chromosome inactivation process, but then begin to die soon thereafter. Those
haploid androgenones with the greatest degree of Xist RNA
expression, Pgk1 gene repression, and repression of other
X-linked genes may die within a narrow period of time just
before the blastocyst stage, which would be consistent with
the fact that the majority of haploid androgenones fail to
progress to the blastocyst stage. This would create the variability observed in Pgk1 mRNA and Xist RNA expression
and minimize the apparent difference relative to fertilized
control embryos among the surviving haploid androgenones. Diploid androgenones may be able to accumulate larger amounts of X-linked transcripts before repression begins
(particularly Pgk1 mRNA), which may allow them to progress to the blastocyst stage.
Haploid parthenogenones overexpressed the Bex1
mRNA relative to fertilized control embryos at both the
morula and blastocyst stages. The other 3 classes of uniparental embryos overexpressed this mRNA at the morula
PREIMPLANTATION GENE COMPARISON IN MOUSE EMBRYOS
stage, but not at the blastocyst stage. The initial overexpression of the Bex1 mRNA by all 4 classes of embryos
may reflect a developmental delay relative to fertilized control embryos, because the Bex1 mRNA displays a transient
peak in expression at the 8-cell stage, with a second increase in expression at the blastocyst stage when its expression becomes trophectoderm-specific [33]. The modest
reduction in expression of the EF1a mRNA in parthenogenones may also reflect such a delay, as this mRNA increases steadily in abundance over the course of preimplantation development [28]. The continued overexpression
of the Bex1 mRNA in haploid parthenogenones, however,
distinguishes them from diploid parthenogenones, raising
the possibility that this mRNA may be overexpressed because of a lack of dosage compensation in the haploid parthenogenones. This effect may be more readily apparent for
the Bex1 gene than for some other genes, because the Bex1
mRNA undergoes a strong increase in abundance at the
blastocyst stage, in contrast to other X-linked transcripts
(e.g., Hprt, Prps1, Pdha1) that actually decrease in relative
abundance between morula and blastocyst stages [29]. Bex1
is preferentially expressed in the trophectoderm and may
play a regulatory role in commitment of cells to that lineage. The increased expression of Bex1 in haploid parthenogenetic embryos, therefore, may negatively affect development of the inner cell mass in these embryos, which
could in turn inhibit development via an insufficiency of
trophic factors for the trophectoderm.
Our observations reveal that expression of X-linked genes
is altered in haploid embryos. These alterations appear subtle
among the surviving embryos that can be collected for analysis, but may be more severe among those embryos that
arrest development and degenerate. It is not clear whether
such effects at the X chromosome are sufficient by themselves to account for the limited developmental potential in
these embryos or merely contribute to their demise. The difference between haploid androgenones and haploid parthenogenones suggests a possible role for autosomal imprinted
genes. The striking differences between haploid and diploid
androgenones, and between haploid and diploid parthenogenones, however, indicate that any other imprinting effects
must be superimposed on other effects of haploidy. Some
genes (imprinted or nonimprinted) may be subject to stochastic inactivating events, such as heterochromatization like
that observed for position effect variegation [36]. For such
loci, the effects of gene silencing would be expected to be
more evident in haploid uniparental embryos, as they would
be 100% penetrant, whereas a much smaller fraction of diploid uniparental embryos would be affected, as this would
require silencing of both gene copies. It would be informative to devise a strategy for producing haploid embryos that
possess a combination of both maternal and paternal chromosomes. This would permit a definitive test of to what
degree uniparental chromosomal composition contributes to
haploid embryo developmental arrest.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
REFERENCES
1. Rasch EM, Cassidy JD, King RC. Evidence for dosage compensation
in parthenogenetic Hymenoptera. Chromosoma 1977; 59:323–340.
2. McGrath J, Solter D. Nucleocytoplasmic interactions in the moue embryo. J Embryol Exp Morphol 1986; 97(suppl):277–289
3. Witkowska A. Parthenogenetic development of mouse embryos in
vivo. J Embryol Exp Morphol 1973; 30:519–545.
4. Kaufman MH, Gardner RL. Diploid and haploid mouse parthenogenetic development following in vitro activation and embryo transfer.
J Embryol Exp Morphol 1974; 31:635–642.
5. Kaufman MH, Sachs L. Complete preimplantation development in
28.
29.
30.
391
culture of parthenogenetic mouse embryos. J Embryol Exp Morphol
1976; 35:179–180.
Modlinski JA Haploid mouse embryos obtained by microsurgical removal of one pronucleus. J Embryol Exp Morphol 1975; 33:897–905.
Tarkowski AK. In vitro development of haploid mouse embryos produced by bisection of one-cell fertilized eggs. J Embryol Exp Morphol
1977; 38:187–202.
Tarkowski AK, Rossant J. Haploid mouse blastocysts developed from
bisected zygotes. Nature (Lond) 1976; 259:663–665.
Modlinski JA. Preimplantation development of microsurgically obtained haploid and homozygous mouse embryo and effects of pretreatment with cytochalasin B on enucleated eggs. J Embryol Exp
Morphol 1980; 60:153–161.
Ito M, Kaneko-Ishino T, Ishino F, Matsuahshi M, Yokoyama M, Katsuki M. Fate of haploid parthenogenetic cells in mouse chimeras during development. J Exp Zool 1991; 257:178–183.
Kaufman MH, Robertson EJ, Handyside AH, Evans MJ. Establishment of pluripotential cell lines from haploid mouse embryos. J
Embryol Exp Morphol 1983; 73:249–61.
Escriba MJ, Garcia-Ximinez F. Influence of sequence duration and
number of electric pulses upon rabbit oocyte activation and parthenogenetic in vitro development. Anim Reprod Sci 2000; 59:99–107.
Kaufman MH. The chromosome complement of single-pronuclear
haploid mouse embryos following activation by ethanol treatment. J
Embryol Exp Morphol 1982; 71:139–154.
Henery CC, Kaufman MH. The incidence of aneuploidy after single
pulse electroactivation of mouse oocytes. Mol Reprod Dev 1993; 34:
299–307.
Latham KE, Solter D. Effect of egg composition on the developmental
capacity of androgenetic mouse embryos. Development 1991; 113:
561–568.
McGrath J, Solter D. Completion of mouse embryogenesis requires
both maternal and paternal genomes. Cell 1984; 37:179–183.
Latham KE, Patel B, Bautista FDM, Hawes SM. Effects of X chromosome number and parental origin on X-linked gene expression in
preimplantation mouse embryos. Biol Reprod 2000; 63:64–73.
Chatot CL, Tasca RJ, Ziomek CA. Development of 1-cell embryos
from different strains of mice in CZB medium. Biol Reprod 1990; 42:
432–440.
Kimura Y, Yanagimachi R. Intracytoplasmic sperm injection in the
mouse. Biol Reprod 1995; 52:709–720.
Wakayama T, Perry ACF, Zuccotti M, Johnson KR, Yanagimachi R.
Full-term development of mice from enucleated oocytes injected with
cumulus cell nuclei. Nature 1998; 394:369–374.
Kuretake S, Kimura Y, Hoshi K, Yanagimachi R. Fertilization and
development of mouse oocytes injected with isolated sperm heads.
Biol Reprod 1996; 55:789–795.
Cummins JM, Wakayama T, Yanagimachi R. Fate of microinjected
spermatid mitochondria in the mouse oocyte and embryo. Zygote
1998; 6:213–222.
Wakayama T, Yanagimachi R. The first polar body can be used for
the production of normal offspring in mice. Biol Reprod 1998; 59:
100–104.
Mikamo K, Kamiguchi Y. A new assessment system for chromosomal
mutagenesis using oocytes and early zygotes of the Chinese hamster.
In: Ishihara T, Sasaki MS (eds.), Radiation-Induced Chromosome
Damage in Man. New York: Alan R. Liss; 1983: 411–432.
Summer AT. A simple technique for demonstrating centromeric heterochromatin. Exp Cell Res 1972; 15:467–470.
Rambhatla L, Patel B, Dhanasekaran N, Latham KE. Analysis of G
protein a subunit mRNA abundance in preimplantation mouse embryos using a rapid, quantitative RT-PCR approach. Mol Reprod Dev
1995; 41:314–324.
Latham KE, De La Casa E, Schultz R. Analysis of mRNA expression
during preimplantation development. In: Tuan RS, Lo CW (eds.),
Methods in Molecular Biology: Developmental Biology Protocols,
vol. II. Totowa, NJ: Humana Press; 1999: 315–331.
Wang Q, Latham KE. Translation of maternal mRNAs encoding transcription factors during genome activation in early mouse embryos.
Biol Reprod 2000; 6:969–978.
Latham KE, Rambhatla L. Expression of X linked genes in androgenetic, gynogenetic, and normal mouse preimplantation embryos.
Dev Genet 1995; 17:212–222.
Bos-Mikich A, Whittingham DG, Jones KT. Meiotic and mitotic Ca21
oscillations affect cell composition in resulting blastocysts. Dev Biol
1997; 182:172–179.
392
LATHAM ET AL.
31. Kay GF, Barton SC, Surani MA, Rastan S. Imprinting and X chromosome counting mechanisms determine Xist expression in early
mouse development. Cell 1994; 77:639–650.
32. Kaneko KJ, Cullinan EB, Latham KE, DePamphilis ML. Transcription
factor mTEAD2 is selectively expressed at the beginning of zygotic
gene expression in the mouse. Development 1997; 124:1963–1973.
33. Williams J, Hawes SM, Patel B, Latham KE. Trophectoderm-specific
expression of the X-linked Bex1/Rex3 gene in preimplantation stage
mouse embryos. Mol Reprod Dev 2002; 61:281–287.
34. Kono T, Sotomaru Y, Sato Y, Nakahara T. Development of androgenetic mouse embryos produced by in vitro fertilization of enucleated
oocytes. Mol Reprod Dev 1993; 34:43–46.
35. Kishikawa H, Wakayama T, Yanagimachi R. Comparison of oocyteactivating agents for mouse cloning. Cloning 1999; 1:153–159.
36. Festenstein R, Tolaini M, Corbella P, Mamalaki C, Parrington J, Fox
M, Miliou A, Jones M, Kioussis D. Locus control region function and
heterochromatin-induced position effect variegation. Science 1996;
271:1123–1125.