Cell death activation in human blastocysts was accompanied by changes in Bcl-x splicing, favoring production of Bcl-xS, an activator of cell death.. Altering the Bcl-xL⁄ Bcl-xS ratio in
Trang 1during preimplantation embryo development
Alagammal Perumalsamy1, Roxanne Fernandes1,2,3, Ingrid Lai1,2,3, Jacqui Detmar1,2, Sue Varmuza4, Robert F Casper1,2,3and Andrea Jurisicova1,2,3
1 Samuel Lunenfeld Research Institute, Mount Sinai Hospital, Toronto, Canada
2 Division of Reproductive Endocrinology and Infertility, Department of Obstetrics & Gynecology, University of Toronto, Canada
3 Department of Physiology, University of Toronto, Canada
4 Department of Cell and System Biology, University of Toronto, Canada
Introduction
For one in 10 couples, failure to achieve a pregnancy
remains a major problem Clinical results of assisted
reproductive technology procedures in Canada suggest
that clinical pregnancy is achieved in only 34% of
patients under the age of 39 years [1] The cellular and
molecular reasons behind such a high rate of failure
are unclear; however, low developmental competence
of conceived embryos is the most significant factor affecting pregnancy outcome
Successful development of mammalian preimplanta-tion embryos requires the integrapreimplanta-tion of a series of molecular and cellular events, culminating in the for-mation of two distinct cell types – the trophectoderm (TE) and the inner cell mass (ICM) These two cellular
Keywords
apoptosis; Bcl-x; blastocyst; embryo; in vitro
culture
Correspondence
A Jurisicova, Samuel Lunenfeld Research
Institute, Mount Sinai Hospital, 25 Orde
Street, Room 6-1001-AJ, Toronto, ON,
Canada
Fax: +1 416 586 8588
Tel: +1 416 586 4800 ext 2052
E-mail: jurisicova@lunenfeld.ca
(Received 23 June 2009, revised 10
November 2009, accepted 18 December
2009)
doi:10.1111/j.1742-4658.2010.07554.x
Elevated cell death in human preimplantation embryos is one of the cellu-lar events compromising pregnancy rates after assisted reproductive tech-nology treatments We therefore explored the molecular pathways regulating cell death at the blastocyst stage in human embryos cultured
in vitro Owing to limited availability of human embryos, these pathways were further characterized in mouse blastocysts Gene expression studies revealed a positive correlation between the cell death index and the expres-sion of Bcl-x transcript Cell death activation in human blastocysts was accompanied by changes in Bcl-x splicing, favoring production of Bcl-xS,
an activator of cell death Expression of Bcl-xS was detected in a subset of human blastocysts that show particular clustering in dying and⁄ or dead cells Altering the Bcl-xL⁄ Bcl-xS ratio in mouse embryos, in antisense experiments, confirmed that upregulation of Bcl-xS, with concomitant downregulation of Bcl-xL, compromised developmental potential and com-mitted a subset of cells to undergoing cell death This was accompanied by increased accumulation of reactive oxygen species levels without any impact on mtDNA content In addition, altered Bcl-x splicing in favor of Bcl-xSwas stimulated by culture in HTF medium or by addition of exces-sive glucose, leading to compromised embryo development Thus, we con-clude that inappropriate culture conditions affect Bcl-x isoform expression, contributing to compromised preimplantation embryo development
Abbreviations
AS, antisense; CDI, cell death index; DAPI, 4¢,6-diamidino-2-phenylindole; DCHFDA, 2¢,7¢-dichlorodihydrofluorescein diacetate; E, embryonic day; ICM, inner cell mass; IGF-1, insulin-like growth factor-1; MM, mismatched; PCD, programmed cell death; qPCR, quantitative PCR; ROS, reactive oxygen species; SC, scrambled; SEM, standard error of the mean; TE, trophectoderm; TUNEL, terminal deoxynucleotidyl transferase dUTP nick end labeling; DWm, mitochondrial membrane potential.
Trang 2lineages are committed to differentiation by activation
of a set of distinctive transcription factors, which
engage diverse downstream molecular pathways [2] In
addition, the TE and ICM not only differ in their
con-tribution to separate cellular compartments (e.g
pla-centa and epiblast respectively), but also exhibit
different susceptibilities to cell death
The presence of dead and dying cells in blastocysts
has been confirmed in a number of mammalian species
by the use of several techniques, implicating apoptosis
as a mode of cellular demise during blastocyst
forma-tion [3,4] In vitro culture of mammalian embryos from
the zygote to blastocyst increases the rate of cell death,
particularly in the ICM lineage [4,5] This becomes even
more evident in embryos that have been conceived,
manipulated and cultured in vitro [6–8] However,
under these conditions, death occurs in both ICM and
TE cells [4] Several growth factors, cytokines and
nutrients have been implicated in the regulation of cell
death of mammalian blastocysts [9,10] In particular,
imbalances in glucose metabolism have been linked to
increased activation of several death pathways in
rodent preimplantation embryos Cell death induced by
exposure to high glucose levels in vitro downregulates
glucose transporters, resulting in a drop in
intraembry-onic free glucose [11], triggering massive apoptosis in
the ICM lineage [12], causing malformations,
congeni-tal abnormalities, and cranial and neural tube defects
[13,14] The molecular pathways mediating cell death in
these animal models have been partially dissected, and
involve p53, Bax and caspase-3 activation [14–17]
At the present time, even with improved
formula-tions for culture medium, only approximately 50% of
human zygotes develop to the blastocyst stage, and
only a subset of these blastocysts ( 40%) are able to
implant after transfer [18,19] As excessive cell death is
one of the logical candidates for early developmental
failure, we explored cell death activation in human
blastocysts under suboptimal culture conditions, and
correlated this with cell death expression of several cell
death regulatory genes The data presented here
indi-cate that alternative processing of Bcl-x, which may be
mediated in part by culture conditions, is a key
deter-minant regulating the developmental fate of human
and murine embryos in vitro
Results
Cell death rates in human blastocysts cultured
in vitro
We have previously shown that the development of
human preimplantation embryos in vitro is influenced
by the culture medium used [20] In order to under-stand the effect of culture medium on embryo quality,
we have reanalyzed this cohort of normally fertilized human embryos (n = 96) for the extent and cellular location of cell death A low level of cell death was detected in almost all blastocysts, and this rate increased with prolonged culture in vitro, as the cell death index (CDI) increased from embryonic day (E)6 to E7 In addition, we confirmed that cells show-ing condensed chromatin also contained fragmented DNA, as assessed by terminal deoxynucleotidyl trans-ferase dUTP nick end labeling (TUNEL) assays (Fig 1A–D) Occasionally, we noticed a few cells that displayed features of necrosis (0.8 ± 0.6 per blasto-cyst, data not shown) These cells were most likely arrested blastomeres from earlier cleavage divisions [21]
In the majority of embryos, cell death activation appeared to be random, affecting both the ICM and TE; however, in 13.5% of embryos (13 ⁄ 96), the majority of cells in the ICM exhibited extensive chromatin condensation and⁄ or fragmented DNA (Fig 1D), indicating elimination of the whole ICM Furthermore, some blastocysts (15⁄ 96) had an increased overall CDI (up to 56%), which, on aver-age, did not exceed 14% on E6 and 21% on E7 (Fig 1E) Finally, there was a trend towards a higher rate of ICM depletion in blastocysts cultured in HTF medium To determine whether caspase-3⁄ 7 activity is associated with induction of cell death, we employed cleavage assays of rhodamine-conjugated DEVD pep-tide This approach revealed a high level of
caspase-3⁄ 7 activity in cells exhibiting chromatin condensation
in human blastocysts that were cultured in HTF med-ium (Fig 1F,G), whereas limited caspase activity was detected in blastocysts with low death rates Taken together, our results suggest that cell death rate and CDI are elevated upon extended in vitro culture of human embryos Furthermore, the increased cell death might be correlated with high caspase-3⁄ 7 activity
Expression of cell death regulatory genes
in human blastocysts
In order to correlate cell death rates with the gene expression profile in individual preimplantation embryos, we exposed live embryos to a nuclear fluoro-chrome [4¢,6-diamidino-2-phenylindole (DAPI)], and determined the total number of cells with normal or condensed chromatin Embryos were immediately lysed and subjected to quantitative RT-PCR dot blot analy-sis We first performed preliminary experiments on
Trang 3mouse blastocysts to determine whether this approach
would result in altered gene expression None of the
assessed genes (Bax, Mcl-1, Bag-1, Bcl-x, or b-actin)
showed significant changes induced by exposure to
DAPI (data not shown)
We next analyzed the expression levels of genes
impli-cated in regulating cell survival (Mcl-1, Bcl-x, Bcl-w, and
Bag-1), cell death inducers (Bax and Hrk), and cell death
executioners (Casp-2 and Casp-3) After hybridization of
cDNA amplified from 37 single blastocysts against a probe for EF-1a (used as a housekeeping gene), only those embryos giving a robust signal (n = 32) were used for further analysis
Levels of both cell death suppressors and activators were variable among the embryos, often ranging from undetectable to highly expressed Using multiple regression analysis, we determined that only Bcl-x transcript level correlated significantly with CDI
Day 7 Day 6
n = 48
n = 42
20 18 16 14 12 10 8 6 4 2 0
25 20 15 10 5 0
Fig 1 Cell death in human blastocysts Representative photomicrographs of expanded human blastocysts analyzed for cell death by assess-ing chromatin status and fragmentation of DNA (A) Embryo stained with DAPI Very few cells show signs of apoptosis (B) Blastocyst with extensive activation of cell death in many cells, as shown by condensed chromatin with DAPI staining (C) Blastocyst with preferential elimi-nation of ICM (arrow), as assessed by DNA condensation using DAPI staining (D) The same embryo depicted in (C) analyzed by TUNEL using biotinylated nucleotides and streptavidin ⁄ Texas Red conjugate The same cells that show intense DAPI fluorescence are also TUNEL-positive, as indicated by the bright red fluorescence (E) CDI ± SEM (proportion of cells exhibiting hallmarks of apoptosis); incidence of either ICM depletion or elevated rates of cell death in human blastocysts at day 6 or day 7 of in vitro culture (F, G) Differential interference con-trast (DIC) light microscopy (F) and cleavage of the DEVD substrate (red) merged with nuclear DAPI staining (blue) (G) in a day 7 fully expanded human blastocyst (containing over 100 cells and an ICM of normal appearance), with many cells (arrow) of the ICM showing higher activity for caspase-3.
Trang 4(P = 0.005) In addition, after plotting Bcl-x
transcript levels versus the number of dead cells per
blastocyst, it became clear that two distinct
popula-tions of embryos could be detected (Fig 2A), those
with high (n = 5) and those with low (n = 27) Bcl-x
expression
Alternative splicing of the Bcl-x transcript and subcellular distribution of Bcl-x in human blastocyst
It was previously shown that Bcl-x is alternatively spliced into at least two different isoforms: Bcl-xL
DAPI
A
B
Bcl-xL
1 2
Number of dead cells per blastocyst
0
0
500
1000
1500
5 10 15 20 25
3 4 5 6
Bcl-xS
Actin
Fig 2 Bcl-xL ⁄ Bcl-xS expression in human blastocysts (A) Correlation between Bcl-x expression and the number of dead cells in human blastocysts The levels of Bcl-x mRNA are expressed as units of specifically bound c.p.m Two populations of blastocysts with either high levels (squares, n = 5) or low levels (circles, n = 27) of Bcl-x expression are present (B) Analysis of Bcl-x splice variants in human blast-ocysts RT-PCR southern blot gel hybridized with a radiolabeled human Bcl-x cDNA probe that recognizes both splice variants (see arrows) Differences in the long ⁄ short ratio were observed among individual embryos Shown below is ethidium bromide staining of gel-separated products following RT-PCR amplification of b-actin Each lane represents a single human expanded blastocyst (C) Expression of Bcl-xS pro-tein in three E7 fully expanded human blastocysts All embryos were obtained from different patients, had good morphology, with distinct ICM cells, contained > 80 cells, and had low (4–8%) CDI Magnification · 400 Left panles: nuclear DAPI staining; right panels: Bcl-xS Whereas some embryos did not express detectable levels of Bcl-xS, others exhibited faint diffuse nuclear Bcl-xS immunostaining in a subset
of healthy cells but strong punctuate immunostaining in some, but not all, cells showing evidence of DNA condensation (arrowheads) Inset: control blastocyst exposed to the primary antibody preabsorbed with Bcl-xS peptide.
Trang 5(long or death-suppressing) and Bcl-xS (short or
death-inducing) Thus, we next determined whether
this alternative splicing of the Bcl-x transcript could
explain the cell death profile observed in human
blast-ocysts Using RT-PCR and Southern blot analysis, we
simultaneously detected variable levels of Bcl-xL and
Bcl-xSin individual human blastocysts (representative
results are shown in Fig 2B)
Whereas some embryos (Fig 2B, lanes 1, 3, 5 and 6)
predominantly expressed the long isoform, others
(lane 4) expressed either low levels of the Bcl-xL
iso-form, or approximately equal levels of the long and
short Bcl-x splice variants (lane 2) Minimal differences
were observed in the levels of b-actin expression We
further investigated the presence and subcellular
locali-zation of Bcl-xS Confirming our data obtained by
RT-PCR, we observed variable levels of Bcl-xS
immu-nostaining, ranging from undetectable to high (Fig 2C)
A moderate level of immunostaining was observed in
both the cytoplasmic and nuclear compartments of
healthy cells, whereas elevated levels of nuclear staining
were observed in some cells undergoing apoptosis In
many cases, these distinct staining patterns of Bcl-xS
were observed in the same blastocyst Moreover, in
some embryos, staining in the mural TE lineage was
lar-gely restricted to the nucleus, whereas in the ICM and
overlaying polar TE, staining was mostly cytoplasmic
(Fig 2C) As Bcl-xS is the logical candidate for the cell
death inducer, our results suggest that elevated levels of
Bcl-xSmay be responsible for activation of cell death in
human embryos cultured in vitro under suboptimal
conditions
To confirm our hypothesis that Bcl-xS expression is
associated with induction of cell death, we designed
antisense (AS) experiments targeting the splice junction
boundaries of Bcl-x [22] This approach is based on
the premise that a complementary
oligodeoxynucleo-tide, binding to the 5¢-splice site of the Bcl-xL template
strand in exon II, will downregulate expression of this
isoform via suppression of Bcl-xL splicing Moreover,
this approach would simultaneously induce Bcl-xS
transcription, owing to redirection of the splicing
machinery to the 5¢ Bcl-xS site First, we tested a wide
range of oligodeoxynucleotide concentrations from 1
to 50 lm on mouse morulae Doses of 1, 2.5 and 5 lm
had no effect on Bcl-x expression, or on embryo
devel-opment, whereas a dose of 50 lm decreased the rate of
blastocyst formation to below 50%, even in the control
[scrambled (SC)] group, and thus was deemed
unsuit-able for further studies An oligodeoxynucleotide
con-centration of 25 lm both downregulated Bcl-xL
expression and induced Bcl-xS expression in mouse
blast-ocysts in the AS treatment group only Correspondingly,
mismatched (MM) oligodeoxynucleotide completely downregulated Bcl-xL, but failed to induce Bcl-xS, at this dose, whereas SC oligodeoxynucleotide had no effect (Fig 3A,B) It has been shown previously that
MM oligodeoxynucleotide with a 5 bp mismatch as compared with AS decreased the ability to induce Bcl-xS transcript, implying that the induction is sequence-dependent [22]
Thus, having established the optimal concentration for experiments as 25 lm, we performed further experi-ments to assess the impact of altered isoform expres-sion Embryonic development was not affected by simultaneous downregulation of Bcl-xL and Bcl-xS (MM group), as 81% of embryos (n = 113) formed fully expanded or hatching blastocysts 48 h after initia-tion of treatment However, developmental competence was dramatically affected by upregulation of Bcl-xS,
as only 54% of embryos (n = 118) reached blastocyst stage in the AS group (P < 0.0001) Control (untreated; n = 111) and SC (n = 93) groups were not significantly different from the MM group (86% and 79%, respectively) Furthermore, upon evaluation
of the cell number and death rates, it was evident that Bcl-xS induction was accompanied by decreased cell numbers and elevated cell death rates (P < 0.0001; Fig 3C) Embryos in the AS group appeared to be developmentally delayed, and frequently degenerated
in culture (Fig 3D)
To determine whether caspase-3⁄ 7 activity is associ-ated with the induction of cell death mediassoci-ated by altered Bcl-x splicing, we again employed the caspase cleavage assay As was observed in human blastocysts, caspase-3⁄ 7 activity was detected in cells undergoing programmed cell death (PCD) in murine blastocysts However, DEVD cleavage was also noted in cells prior
to evidence of nuclear collapse (DNA condensation)
As expected, treatment of murine embryos with AS oligodeoxynucleotides resulted in dramatically upregu-lated caspase-3⁄ 7 activity, which was not observed in the control or MM groups (Fig 3E) The results sug-gest that, as seen with human blastocysts, induction of Bcl-xS expression is positively correlated with cell death and caspase-3⁄ 7 activity in mouse embryos
Altered Bcl-x splicing affects reactive oxygen species (ROS) production and mtDNA copy number Bcl-xL has been previously implicated in inhibiting ROS production [23] To determine whether decreasing Bcl-xL and increasing Bcl-xS expression increases ROS production in murine blastocysts, individual murine embryos were treated with AS, MM and SC oligodeoxynucleotides and subsequently exposed to
Trang 62¢,7¢-dichlorodihydrofluorescein diacetate (DCHFDA),
a fluorescent ROS indicator that reflects accumulation
of H2O2 (Fig 4A–D) In the AS group, decreasing
Bcl-xL expression and increasing Bcl-xS expression
significantly increased the level of H2O2 as compared
with control or SC embryos (P < 0.05; Fig 4E) In
addition, significant increases in ROS levels were also
seen in the MM group (P < 0.05) (Fig 4E), indicating
that depletion of Bcl-xL is sufficient to cause elevated
oxidative stress
As there is a close relationship between ROS genera-tion and mitochondrial activity, we also measured changes in mitochondrial membrane potential (DWm); however, no significant differences were observed due
to either induction of Bcl-xS or downregulation of Bcl-xL (data not shown) In order to determine whether the difference in ROS production was due to the difference in mtDNA content, quantitative PCR (qPCR) was performed to analyze the amount of mtDNA present in embryos of the SC, MM and AS
Bcl-xL
A
B
C
D
E Bcl-xS
Con SC MM AS
Con SC MM AS
*
*
80
2
20
40
60
4
Fig 3 Altered Bcl-xS expression causes induction of cell death in murine embryos (A) Representative results of analysis of Bcl-x splice vari-ants and b-actin levels in murine blastocysts upon treatment with antisense oligodeoxynucleotides Each lane represents five embryos sub-jected to RT-PCR analysis Con, untreated; SC; 25 l M scrambled oligodeoxynucleotides; MM, 10 or 25 l M mismatched oligodeoxynucleotides; AS, 10 l M AS or 25 l M antisense oligodeoxynucleotides Amplification of alternative splicing products (long versus short isoforms) of the Bcl-x gene hybridized with a radiolabeled murine Bcl-x cDNA probe that recognizes both splice variants Note that increased accumulation of Bcl-xS in the SC group represents variability in the abundance of this isoform in murine blastocysts, as similar results were observed in untreated embryos in other experiments (B) Ethidium bromide staining of gel-separated products following RT-PCR amplification of b-actin (C) Effect of oligodeoxynucleotide treatment on embryo quality, as evidenced by total cell number and cell death Data shown are means ± SEMs Statistical significance was observed in the AS group (significantly different from all three other groups, as denoted with asterisk) in all categories assessed, with P < 0.0001 (D) Nuclear staining (DAPI in blue) of murine embryos exposed to 25 l M phosphothiolated oligodeoxynucleotides (C, SC, MM, and AS), as described above Magnification: · 250 Only AS treat-ment, associated with induction of Bcl-xS, resulted in embryos with poor developmental competence, caused by reduced cell numbers and induction of cell death, as observed by DNA condensation (see arrows) (E) Caspase-3-like activity in mouse blastocysts Images were obtained by serial sectioning on a deconvolution microscope (· 200) SC, MM or AS embryos (dose: 25 l M ) Only AS treatment associated with induction of the Bcl-xS splicing isoform was accompanied by upregulation of caspase-3-like activity.
Trang 7groups As we did not observe any difference in ROS
levels between the control and SC groups, a control
group was not included in the mtDNA content assay
Surprisingly, only Bcl-xL downregulation resulted in
increases in mtDNA in the MM group (Fig 4F)
In summary, the data presented here show that
induction of Bcl-xS with a concomitant decrease in
Bcl-xL transcript level increases ROS levels without
any change in DWmand mtDNA content
Culture in HTF medium and exposure to high
glucose levels alters Bcl-x splicing in mouse
embryos
As we observed variable ratios of Bcl-xL⁄ Bcl-xS
splic-ing in human embryos, we investigated whether HTF
medium could be responsible for induction of Bcl-xS
expression We first determined that HTF medium
sig-nificantly decreased the rate of blastocyst formation
(P = 0.0303) (Fig 5A) and caused an increased rate
of cell death by E4.5 (Fig 5B) This was accompanied
by a change in Bcl-x splicing, favoring formation of
Bcl-xS(Fig 5C) and accumulation of ROS (Fig 5D),
thus recapitulating many features of the AS experi-ments
KSOM and HTF medium differ in the presence or absence of several compounds; one of the most obvi-ous differences is a 10-fold higher glucose level in HTF medium Previously, in vitro exposure to high glucose levels has been linked to poor embryo development, causing increased rates of cell death and embryonic malformations [12] Therefore, we explored whether exposure to glucose could alter Bcl-x splicing In order
to confirm the biological effect, we assessed the develop-mental progression of embryos exposed to in vitro hyperglycemia The rate of embryonic progression was reduced, particularly at E4.5, when a significantly lower proportion of embryos in high-glucose (KSOM + 50
mm d-glucose) conditions developed to the appropriate stage (P = 0.0265) (Fig 6A)
We next determined whether an alteration in Bcl-x splicing could be associated with the lower rates of pro-gression in high-glucose conditions In vitro-cultured embryos reaching the morula stage or later at E4.5
in KSOM + 50 mm d-glucose were assessed for alter-ation in Bcl-x splicing Indeed, embryos cultured in
Control
MM
SC
F
Con
a
a
a
b
a
0
1.3 1.28 1.26 1.24 1.22 1.2 1.18 1.16 1.14
20 40 60 80 100 120 140 160 180
AS
Fig 4 Altered Bcl-xL ⁄ Bcl-xS expression results in increased ROS production in murine embryos (A–D) Representative photomicrographs of mouse blastocysts exposed to 25 l M phosphothiolated oligodeoxynucleotides [control (Con) (n = 15), SC (n = 18), MM (n = 22), and AS (n = 23)], treated with 10 l M DCHFDA, and analyzed for ROS (E) Analysis of ROS in oligodeoxynucleotide-treated mouse blastocysts Values represent relative fluorescence units ± SEM Increased ROS production was observed in MM and AS oligodeoxynucleotide-treated murine blastocysts Different letters indicate that the treated groups are significantly different (P < 0.05) (F) Quantitation of mitochondrial DNA copy number in treated murine blastocysts Elevated mitochondrial copy number was observed with MM oligodeoxynucleotide-treated murine blastocysts Different letters indicate that they are significantly different (P < 0.05).
Trang 8glucose-enriched KSOM medium had a significantly
lowered Bcl-xL⁄ Bcl-xS ratio as compared with the
KSOM-only group (P = 0.031) (Fig 6B)
Further-more, culture of embryos in the presence of high
glu-cose concentrations caused elevated production of
ROS (Fig 6C)
In summary, our results indicate that developmental
rate is compromised by culture in HTF medium or in
the presence of high glucose levels, and that this can
be attributed to an increase in Bcl-xS transcript levels
Discussion Our results on cell death rates in the human embryos indicate that almost 15% of human blastocysts cultured in vitro completely eliminate the embryonic
7
0
E1.5 E2.5
HTF
*
*
*
KSOM
E3.5 E4.5 20
40
60
80
100
120
D B
4.50 4.00 3.50 3.00 2.50 2.00 1.50 1.00 0.50 0.00
450 400 350 300 250 200 150 100 50 0
6
5
4
3
2
1
0
HTF
CDI MI
KSOM
a a
a
b
Fig 5 Developmental progression, cell death and altered Bcl-xL ⁄ Bcl-xS expression in murine embryos (A) Zygotes cultured either in HTF medium (n = 102) or in KSOM (n = 101) were assessed daily for developmental progression from E1.5 to E4.5; the proportion of embryos reaching the appropriate developmental stage is expressed as a percentage A significant decrease in blastocyst formation in HTF medium-cultured embryos (black line) as compared with KSOM-medium-cultured embryos (gray line) was observed by E4.5 (B) Cell death and mitotic index (MI) were calculated by staining the embryos with DAPI, and are expressed as percentages based on total cell number A significantly increased CDI (black bar) was seen in the embryos cultured in HTF medium (n = 20) as compared with those cultured in KSOM (n = 19) (P = 0.006) Different letters indicate that values are statistically significant (C) Pools of five embryos at the morula ⁄ blastocyst stage cultured in HTF medium (n = 8) and KSOM (n = 9) were assessed for changes in Bcl-xL ⁄ Bcl-xS ratio Embryos cultured in HTF medium had
a significantly decreased ratio as compared with those cultured in KSOM (P = 0.027), owing to the decreased expression of Bcl-xS (D) Anal-ysis of fluorescence generated by DCHFDA probe, reflecting ROS generation, was performed in embryos cultured in either HTF medium (n = 13) or KSOM (n = 10) at E3.5 Values represent relative fluorescence units ± SEM Significantly increased ROS generation was observed in the HTF medium group (P = 0.0009), as indicated by the asterisk Representative images for ROS are shown.
Trang 9lineage destined to form the fetus (i.e the ICM),
leav-ing only precursor cells for the placenta Such
blast-ocysts may initiate implantation, but are capable of
creating only extra embryonic tissues with an empty
embryonic sac, a condition clinically recognized as a
‘blighted ovum’ [24] The absence of ICM cells has
been previously described in human embryos cultured
in vitro[25,26], but this was attributed to an abnormal
allocation of cells into the ICM and TE lineages [26]
Our results suggest that, in some embryos, cells of the
ICM are established, but are subsequently eliminated
An elevated rate of cell death in human ICM cells has
been previously described [27]; however, the molecular
pathways responsible for this phenomena are unclear
Of the genes examined in this study, Bcl-x was the
only one whose expression showed a statistically
signif-icant correlation with CDI in human blastocysts
Importantly, alternative splicing within the coding region of Bcl-x results in the production of several dis-tinct protein products with opposing functions [28] Bcl-xL has been reported to function as a prominent death-suppressing molecule during postimplantation embryonic development [29], as targeted mutation
of the murine Bcl-x gene that disrupts both isoforms triggers midgestational embryonic lethality [30] By comparison, very little is known about the mode of action and physiological importance of Bcl-xS, which lacks the BH1 and BH2 domains Bcl-xS renders cells more susceptible to apoptotic signals [21], even when these cells simultaneously express Bcl-xL [31,32] In addition, it has been observed that Bcl-xS is able to form homodimers as well as heterodimers with Bcl-xL [33], but this ability may depend on cell type More-over, cell death induction is associated with Bcl-xS
B
C 80
70 60 50 40 30 20
10 0
KSOM + 50 m M
Glucose
*
*
KSOM
KSOM + 50 m M
Glucose KSOM
100
A
90
80
70
60
50
KSOM + 50 m M
glucose
30
20
10
0
4.50
4.00
3.50
3.00
2.50
2.00
1.50
1.00
0.50
0.00
Fig 6 Exposure to glucose is associated with decreased embryo development and induction of Bcl-xS (A) Embryos were assessed daily from E1.5 to E4.5, and the proportion of cleaved embryos reaching the correct developmental stage at the correct time was calculated Embryos in high-glucose conditions (KSOM + 50 m M D -glucose; gray line, n = 170) had compromised developmental competence, as a sig-nificantly lower proportion of these embryos than of those cultured in KSOM only (black line, n = 60) reached the blastocyst stage at E4.5 (P = 0.0265) (B) At E4.5, pools of five embryos at the morula stage or later were assessed for changes to Bcl-x splicing KSOM-cultured embryos had an average Bcl-xL ⁄ Bcl-xS ratio of 3.49 ± 0.74 (same as in Fig 5C), which was significantly greater than the average Bcl-xL ⁄ Bcl-xS ratio of 2.03 ± 0.27 in embryos cultured in KSOM + 50 m M D -glucose (P = 0.031) (C) Representative images for analysis
of ROS are shown ROS analysis was performed in embryos cultured in KSOM and KSOM + 50 m M glucose at E3.5 A significant increase
in ROS was seen in embryos cultured with 50 m M glucose (P < 0.0001).
Trang 10localization to mitochondria, as deletion of the
trans-membrane region prevented mitochondrial localization
and diminished the death-inducing properties of this
molecule [33] Our experiments on native Bcl-xS
showed either diffused cytoplasmic or nuclear
accumu-lation of Bcl-xS in early embryos, with colocalization
to condensed DNA in pyknotic cell corpses Although
a role of Bcl-2 family members in regulating survival
has been extensively established, new evidence exists
suggesting a role of these proteins in regulation of cell
cycle progression Both Bcl-x and Bcl-2 delay cell cycle
entry and enhance the differentiation potential of
immature cells [34,35] Furthermore, Bcl-xS not only
increases susceptibility to apoptosis [22], but also alters
the kinetics of the cell cycle progression [36] Thus,
nuclear subcellular localization in healthy cells in some
embryos may reflect the ability of Bcl-xS to interfere
with cell cycle progression, as we also observed
decreased cell numbers with upregulated Bcl-xS
tran-script
Alternative splicing of the Bcl-x gene appears to play
an important role in human blastocyst development
Human blastocysts have variable implantation
poten-tial and exhibit variable Bcl-x long⁄ short form ratios
Alternative splicing favoring increased levels of Bcl-xS
may be one mechanism involved in upsetting the
balance of PCD regulatory molecules in the developing
embryo Indeed, AS experiments performed on murine
embryos confirmed this hypothesis Abnormal
embry-onic development, associated with delayed cleavage,
and activation of cell death accompanied by activation
of caspase-3⁄ 7, causing embryonic degeneration, was
only observed in the group of embryos with
upregula-tion of Bcl-xS and was not affected by downregulaupregula-tion
of Bcl-xL These data, which fully agree with previous
reports on the action of Bcl-xS in some somatic cell
types [22,36], implicate alternative splicing of the Bcl-x
gene as a major determinant of successful embryonic
development
In conjunction with its antiapoptotic activity, Bcl-xL
has been implicated in the regulation of ROS
produc-tion [37] Both of these cellular roles have been linked
to the BH4 domain [38] The results of our study
indicate that altering the Bcl-x isoform ratio (e.g
decreasing the amount of Bcl-xL and increasing the
amount of Bcl-xS) results in an elevated ROS level,
without any obvious effect on DWm Concomitantly,
altered mtDNA content was detected only in the
Bcl-xLknockdown, suggesting the initiation of
prema-ture mtDNA replication, which otherwise occurs only
after the blastocyst stage [39] It is thus possible that
the increase in ROS production in the Bcl-xL
knock-down group is due to increased mitochondrial copy
number, as previously observed in yeast [40] As mtDNA is in close proximity to the site of H2O2 gen-eration, the effect on mtDNA could be immediate, and this might result in premature replication of mitochon-dria [40] However, mtDNA number remained unchanged in the AS group, implying that the increased level of ROS caused by downregulation of Bcl-xL and induction of Bcl-xS was not due to the increased mitochondrial number It has been previ-ously reported that increased ROS levels in mitochon-dria signals the nucleus and, as a result, telomere shortening occurs, leading to the loss of proliferative capacity and replication senescence [41–43] In addi-tion, elevated ROS levels have been linked to perma-nent early embryo arrest, and this may be due to mitochondrial dysfunction and activation of p66Shc [44] Thus, we propose that altering the Bcl-xL⁄ Bcl-xS ratio in blastocysts results in mitochondrial stress, and increased ROS production, causing diminished cell numbers and cell senescence, resulting in compromised embryo development
In addition, Bcl-xL has been postulated to have a role in cellular metabolism, although this role is not well defined Bcl-xL overexpression affects several meta-bolic parameters, most notably causing a shift from oxi-dative phosphorylation to aerobic glycolysis and promoting glucose consumption [45] In conditions of nutrient deprivation, Bcl-xL facilitates the metabolism
of endogenous substrates such as phosphocreatine [45] The ability to maintain metabolic function in conditions
of limited nutrient availability is thought to contribute
to the antiapoptotic function of Bcl-xL [46] However, transgenic mice with Bcl-xL overexpression showed sig-nificant disturbances of mitochondrial metabolism asso-ciated with defective insulin secretion and induction of diabetes [47] Thus, these data suggest that fine regula-tion of Bcl-xL expression, perhaps via alternative splic-ing, regulates metabolism Hyperglycemic conditions have been previously linked to embryopathies [48] that result in changes in mitochondrial morphology along with elevated ROS production [49,50] Furthermore, in mice, hyperglycemia-induced cell death and embryo death were correlated with the induction of the pro-apoptotic members bax, p53 and caspase-3 [14,17,51]
As hyperglycemia has been previously established as a potent inducer of cell death in murine embryos, we assessed whether glucose can alter the Bcl-xL⁄ Bcl-xS ratio Recent findings suggest that insulin and insulin-like growth factor-1 (IGF-1) are capable of altering cell death via downregulation of Bcl-xS without affecting Bclx-L in a caspase-dependent manner [52] This is very exciting, as IGF-1 has been shown to improve rates of blastocyst formation and decrease rates of PCD in