Elder K. Human preimplantation embryo selection

Подождите немного. Документ загружается.

HUMAN PREIMPLANTATION EMBRYO SELECTION

the term to describe gene expression phenomena that

seem to be ‘outside’ normal inheritance is a some-

what confusing distinction.

Epigenetic phenomena can be thought of as an

evolutionary ‘fine-tuning’ of gene expression. In terms

of mechanism, epigenetics concerns the silencing/

activation of genes and chromosomal regions, and is

related to ancient and ubiquitous processes of chro-

matin silencing and genetic control. As life evolved,

selection pressure resulted in the requirement for

silencing or activation of specific genetic informa-

tion above and beyond baseline developmental and

housekeeping control. The form of expression con-

trol we now term epigenetic is particularly involved

in both the silencing/activation of genes based on

external (i.e. environmental) stimuli, and in the

control of gene dosage.

2,6

The delineation of genetic

phenomena that are classified as epigenetic is some-

what variable: the term is also used to describe

processes such as X-inactivation, the silencing of for-

eign (i.e. viral DNA), and aberrant gene activation/

silencing in cancer.

2,7,8

The best description of what

constitutes an epigenetic control mechanism, begin-

ning with environmental effects, can be provided by

specific examples.

In flowering plants, the time of flowering is a

critical developmental decision which must be

correlated with environmental cues. Cold-response

vernalization affects this process by promoting

flowering following a prolonged period of cold tem-

perature. It is now recognized that vernalization

involves epigenetic changes to the expression of cer-

tain genes that control flowering.

Expression of the

FLOWERING LOCUS C (FLC) gene in Arabidopsis

is proportionally repressed in adult plants in response

to cold treatment perceived by the germinating seed.

This repression is stable, but high FLC expression is

re-established in progeny plants. The exact mecha-

nism of this repression remains to be elucidated, but

it involves specific chromatin modification mediated

by gene products like VERNALIZATION2,a protein

related to the polycomb group proteins in Drosophila

known to be involved with gene silencing.

In this

example of epigenetic control, the perception of an

environmental cue during early development is trans-

formed into chromatin modifications that specifically

alter downstream gene expression in a stable fash-

ion and bring about an appropriate response in the

mature organism. Furthermore, these modifications

are ‘reset’ in the next generation to allow for a new

cycle of perception and response. These are hallmark

components of an epigenetic process.

In mammals, the best understood example of

epigenetic programming involves the allele-specific

expression of certain genes depending on the parent

of origin.

In the 1980s, nuclear transfer and uni-

parental disomy experiments in mice definitively

demonstrated for the first time the absolute require-

ment for both a maternally and a paternally derived

genome in directing normal mammalian develop-

ment.

11–13

Androgenetic or gynogenetic embryos

(derived from diploid paternal or maternal genomes)

exhibited aberrant and incomplete development,

indicating that the two parental genomes were not

equivalent.

11,12

Mice that inherit uniparental dis-

omies (UPD) (i.e. chromosome complements where

a component is entirely derived from one parent) also

exhibit a variety of developmental problems which are

usually lethal.

It is now understood that certain genes

are only expressed when inherited from the maternal

or paternal germline. These genes are described as

being ‘imprinted’ by the germline of origin, with a

stable and defining status that determines expres-

sion during development. For instance, androge-

netic embryos (or those with an androgenetic UPD)

lack expression of genes that are only active when

derived from the maternal germline (and exhibit

overexpression of paternally derived genes) and are

thus developmentally incompetent.

Current theory proposes that the origin of this

epigenetic control lies in the conflict between paternal

and maternal evolutionary interests during genome

transmission.

In mammals, this process necessarily

involves the maternal contribution to the developing

offspring during gestation. Paternal interests, appar-

ently accelerated by polyandry (multiple male part-

ners for each female), seek to maximize this maternal

contribution to the development of each offspring.

Paternal imprinting seems to promote the expression

of genes with activities related to maximizing fetal

growth and development.

The differential maternal

imprinting seeks to control fetal growth to protect



HPE_Chapter19.qxp 7/14/2007 5:51 PM Page 236

GENOMIC IMPRINTING AND CONSEQUENCES FOR EMBRYONIC DEVELOPMENT

the mother’s own interests and future gestational

activity. For example, the maternally derived copy

of a gene product involved with the promotion of

fetal growth, insulin-like growth factor 2 (Igf2), is

imprinted to an inactive state, effectively reducing

the dosage of this gene.

Conversely, the gene for

the Igf2 receptor (which binds to and inactivates

Igf2) is inactivated on the paternal chromosome.

This is a perfect example of a genomic ‘tug-of-war’

between the promotion and suppression of fetal

growth. The result is that a complement of imprinted

mammalian genes is only active in a single copy –

the one bearing the active parent-of-origin imprint.

As is discussed, dosage control for these genes is

critical to normal development and perturbation to

the process can have critical consequences.

EPIGENETICS: MECHANISTIC CONCERNS

A thorough description of epigenetic mechanisms

from a molecular standpoint is well beyond the

scope of this review. Epigenetic control mechanisms

are among the most complex in molecular genetics,

and are still only partly understood.

1,6,17

For the

most part, epigenetic silencing or activation involves

stable covalent modification to the chromatin, creat-

ing specific states that have an over reaching effect on

other mechanisms of expression control. Two hall-

mark epigenetic chromatin modifications are cytosine

methylation (at CpG sequences) and methylation/

acetylation of histone proteins.

18,19

Generally, the

genome is divided into heavily methylated regions

that are silent, and hypomethylated regions that are

actively transcribed. However, as discussed below, in

some cases uniquely methylated control regions are

associated with gene activation. Cytosine methylation

seems to be the principal protection mechanism for

silencing foreign DNA (viral sequences, retrotrans-

posons, etc.) in the genome, and such sequences are

heavily methylated. In epigenetic regulation, methyl-

ation and other modifications come under allele-

specific control schemes that ensure the maintenance

of stable expression patterns. As the genome transits

through the developing germline and in the devel-

oping embryo, waves of methylation, demethylation,

and other types of chromatin modification occur.

These global chromatin modification processes are

modified at imprinted loci to bring about specific

active and silenced chromatin states. Once again,

specific examples provide the best explanation.

Returning to the Igf2 locus in mice, this gene is

located on chromosome 7 in a region that is subject

to imprinting effects.

17,20

On chromosomes derived

from a paternal germline, Igf2 is expressed; the gene

is silenced on maternally derived chromosomes. A

directly adjacent locus, H19, exhibits the reverse epige-

netic expression pattern, with only maternal chromo-

some expression.

These two loci and their complex

interactions with germline-specific transregulatory

proteins form an epigenetic molecular switch that

controls Igf2 expression. Both genes require and com-

pete for expression promoting enhancer proteins.

region adjacent to H19 exhibits differential methyla-

tion on the maternal and paternal chromosomes.

This imprinting control region (ICR) acts as a binding

site for regulatory proteins, such as the CCCTC-

binding factor protein (CTCF), that create a boundary

or insulator function between the two loci that blocks

enhancer activity.

On the maternal chromosome,

the ICR is unmethylated, allowing CTCF binding and

downstream silencing effects that prevent Igf2 expres-

sion.

However, on the paternal chromosome, the

ICR is methylated, CTCF binding is prevented, and

enhancer mediated Igf2 expression results. Thus

methylation-sensitive differential protein–chromatin

interaction acts as a tight bipolar switch, initiating

the creation of unique maternal and paternal chro-

matin conformation loops with the Igf2 locus in an

active or silent state. CTCF ‘reads’ the imprinting

marks and facilitates differential expression.

25,26

CTCF is known to bind to all currently known

imprinting-related control regions, including those

involved with X-inactivation.

Interestingly, a

homolog of CTCF has recently been discovered in

Drosophila.

The Drosophila genome is highly com-

pact, and basic expression control involves an almost

ubiquitous use of ‘insulator’ and blocking ele-

ments similar to mammalian ICRs between genes.

Apparently CTCF was one of many such proteins in

Drosophila that has been co-opted to a similar but

specific function in mammals.



HPE_Chapter19.qxp 7/14/2007 5:51 PM Page 237

HUMAN PREIMPLANTATION EMBRYO SELECTION

An obvious question remains concerning the

mechanism whereby the germline-unique imprints

are created in the first place. Germline specific

mechanisms must exist that allow existing imprints

(from parental chromosomes) to be erased, and

new imprints in the transmitted germ cell chromo-

somes to be established. This process is still only partly

understood on a molecular level. A male germline-

specific protein, BORIS, has recently been identified

that exhibits exquisitely identical DNA binding regions

to CTCF, but a divergent functional sequence.

The

two proteins exhibit a mutually exclusive expression

pattern during male germ cell development. In fact,

BORIS is uniquely expressed in primary spermato-

cytes – the only cell type known where CTCF expres-

sion is absent. BORIS expression occurs at the time

when existing parental methylation imprints are

erased, and its down regulation coincides with the

wave of methylation occurring at the round sper-

matid stage, when new paternal imprints may be

established. BORIS may be involved with binding to

ICR sequences and targeting them for methylation –

in this case only in the male germline. CTCF bind-

ing has been shown to protect its target sequence from

downstream methylation, so a reciprocal expression

pattern with BORIS makes sense in this regard.

A model can be envisioned in which the parent-

specific germline expression pattern of ICR binding

proteins such as CTCF and BORIS control methyla-

tion and other modifications at critical ICR sites dur-

ing genome transmission. Therefore, as the maternal

and paternal genomes transit through the early

germline, gametogenesis, fertilization, and early devel-

opment, an increasing number of global processes

involving methylation and chromatin modification

are modulated to produce the unique parentally

imprinted loci observed.

This model is most certainly an over simplifica-

tion, and is clearly only one mechanism among sev-

eral that describe the generation and maintenance

of imprints and downstream expression control. In

general, maternal germline-specific methylation seems

to be a more common characteristic of imprinted

genes.

18,30

A wave of paternal genome demethylation

occurs in the oocyte following fertilization, and this

is speculated to be a component of the evolutionary

conflict for control of imprinted loci.

Other expres-

sion control mechanisms include direct repression

by methylated DNA binding proteins, and the inter-

action of anti-sense non-coding transcripts in down

regulating imprinted targets.

1,17,19

In fact, the H19

locus codes for one such transcript, adding another

level of complexity to that story. It is worth noting

that in the absence of methylation (in methyltrans-

ferase knockout experiments) some imprinted genes

are active from both alleles and some are silenced at

both alleles.

In general, epigenetic expression con-

trol should be thought of as arising from a complex

synergy of target sequence-specific control regions

and transcriptional effects, along with germline/

developmental specific protein expression patterns.

Furthermore, the majority of normal epigenetic

processing seems to take place during gamete devel-

opment and early embryonic development. As dis-

cussed in the next section, this aspect clearly has

implications for potential perturbation by ART

procedures.

ART AND EPIGENETICS: EVIDENCE

FOR PROBLEMS

The evidence for perturbations to epigenetic pro-

cessing associated with assisted reproduction proce-

dures comes from two main areas: research and large

animal experiments related to assisted reproduction/

cloning, and the incidence of imprinting related

disorders following human ART application.

EVIDENCE FROM ANIMAL EXPERIMENTS

Assisted reproduction protocols in large animals

such as the sheep and cow developed concurrently

with, or followed the first successful use of assisted

reproduction in the human. However, unlike in the

human, such large animal protocols have been

plagued by a variety of distinct developmental prob-

lems. It has become clear that some of these prob-

lems are directly related to abnormal imprinting and

epigenetic processing associated with in vitro culture

and gamete/embryo manipulation. Particularly

during nuclear transfer cloning protocols (but also



HPE_Chapter19.qxp 7/14/2007 5:51 PM Page 238

GENOMIC IMPRINTING AND CONSEQUENCES FOR EMBRYONIC DEVELOPMENT

observed following basic in vitro embryo culture),

aberrations of gestation length and in utero fetal

growth control have been so frequent that a specific

terminology, ‘large offspring syndrome’ (LOS), was

coined to describe the situation in ruminants.

this syndrome, animals produced via ART are born

after extended gestation periods with greatly aug-

mented body mass and a variety of physical defects.

It is now known that such animals manifest abnor-

malities in the epigenetic status and expression of

imprinted genes. This particular relationship has

not been observed in human ART, although reduced

birth weight has been correlated with ART in single-

tons. In perhaps the best study using sheep, in vivo

fertilized eggs were simply cultured in vitro for 5

days and then transferred to establish pregnancies.

One-quarter of fetuses recovered just prior to term

exhibited body weights 22–82% increased above

the largest non-cultured control fetuses. In the LOS

fetuses, IGF2 expression was normal. However, expres-

sion of the IGF2R gene was decreased 30–60% relative

to controls, and this reduction was not observed in

normal weight offspring also derived from in vitro

culture. Reduced IGF2R expression was correlated

with an aberrant loss of methylation at a key imprint-

ing control region for the gene.

A similar scenario

has been observed in other large animal ART and

cloning applications.

Based on the large animal results, studies in the

mouse have pursued a clear and specific connection

between in vitro culture/manipulations and epige-

netic abnormalities. The gist of these experiments

is that transient preimplantation in vitro culture

does seem to result in changes in gene expression

in mice.

34,35

Both imprinted and apparently non-

imprinted genes can be affected. Furthermore, the

type of culture media and serum supplementation

used can affect this process. For example, one study

showed that imprinting of the murine H19 gene

was lost, with abnormal hypomethylation of the

paternal allele following in vitro culture in a simple

media (Whitten’s media). However, this abnormal-

ity was absent when culture took place in a complex

amino acid-augmented media (KSOM).

It is

interesting to note that the abnormal bi-allelic H19

expression pattern observed in the Whitten’s media

embryos was lost following implantation, but

remained in the extraembryonic tissues. Recently, mice

derived from in vitro culture were blindly evaluated and

compared with non-cultured littermates, delineated

after analysis via a transgene maker.

After exhaustive

evaluation examining numerous developmental and

behavioral markers, in vitro and in vivo derived

mice were found to be essentially identical. No sig-

nificant differences were found, after assessment of

more than a dozen individual developmental mile-

stones and complex behavioral/learning patterns.

However, a slight but significant difference was

observed in spatial learning retention, and in one com-

ponent of an anxiety assessment maze, particularly

in male animals.

The authors of this study have

recently reported that there was no distinguishable

difference in lifespan between in vitro and in vivo

derived animals.

The actual origin of imprinting

disturbance following in vitro manipulations in

mice and ruminants is still undefined. It can be spec-

ulated that either basic epigenetic mechanisms are

simply disturbed by ‘suboptimal’ culture conditions,

or that normally functioning epigenetic mechanisms

are responding to the altered environment of in vitro

culture, resulting in gene expression changes.

Large-scale cloning experiments in the mouse

seem to indicate that gene expression abnormalities

are exacerbated by the nuclear transfer procedure.

Considering the importance of stage- and germline-

specific events in correct epigenetic processing –

certainly challenged by the introduction of a somatic

cell nucleus into an oocyte – this is not surprising. In

fact, it is surprising that such protocols can produce

developmentally functional embryos and offspring

at all. As mentioned above, several recent studies

have indicated widespread dysfunction of gene

expression in cloned embryos, and many cloned

animals exhibit developmental, late onset problems,

and in some cases reduced lifespan.

However, in

other cases, nuclear transfer offspring, even those

known to harbor clear gene expression abnormalities,

have exhibited normal development and unremark-

able life histories.

Other experiments in the mouse have examined

this issue from a different perspective by attempting

to circumvent parental-specific imprinting controls.



HPE_Chapter19.qxp 7/14/2007 5:51 PM Page 239

HUMAN PREIMPLANTATION EMBRYO SELECTION

Gynogenetic embryos, derived from the genome of

early ‘non-growing’stage oocytes (lacking later-stage

epigenetic modifications), exhibit extended devel-

opment well beyond that previously observed fol-

lowing parthenogenesis.

These experiments have

recently come to fruition, and via the direct manip-

ulation of the expression pattern of only the Igf2 and

H19 genes, resulted in the birth of apparently nor-

mal parthenogenetic offspring with a gynogenetic,

i.e. diploid, maternal genome source.

Despite their

genetic makeup with supposedly only maternally

imprinted chromosomes, the offspring from these

experiments manifested apparently normalized epi-

genetic expression patterns at other imprinted genes.

The expression of Igf2 and H19 was artificially

manipulated via genetic modification to these loci.

However, a second set of non-manipulated pater-

nally imprinted genes, Dlk1 and Dtl2, exhibited nor-

mal methylation patterns and expression in surviving

term offspring.

Future experiments will hopefully

provide further insight into how this epigenetic ‘nor-

malization’ occurs. This has important relevance to

ART and other protocols that involve early embryo

development.

Taken together, the data from large animal and

mouse experiments suggests that in vitro embryo

culture and perturbations can definitely result in

epigenetic disturbance and gene expression alter-

ations in these species. Furthermore, such epigenetic

expression issues can result in both subtle and pro-

found phenotypic consequences. However, consid-

erable variation and flexibility seems to exist in both

the origin and the downstream manifestations of

such disturbance. It is also worth noting at this point

that differences are known to exist between rodents,

ruminants, and primates, including very specific

epigenetic processing issues. For example, the IGF2R

gene, which is known to exhibit epigenetic misregu-

lation in the ruminant LOS is not subject to epi-

genetic control in primates. Despite the presence

of an almost identical imprinting control region

between the human and ruminant genes, IGF2R is

not imprinted in the human – this regulation was

apparently lost during the evolution of the primate

lineage.

Other genes continue to be identified that

display differential epigenetic behavior between mice

and humans.

It is important to point out that the

experimental findings are stage- as well species-

specific, and that environmental factors in vitro

may play a role. Embryologists experimenting with

animal embryos tend to be less concerned about

environmental criteria than human embryology

specialists, and such differences may account for

some of the observations.

EVIDENCE FROM HUMAN POPULATION STUDIES

In 2002, two ICSI-conceived children were reported

to be affected by an imprinting-related disorder;

this was followed by a flood of reports that showed

a slight but significant increase in the incidence

of human imprinting-related disorders in children

born following the application of ART.

45–48

Beckwith-

Wiedemann (BW) and Angelman (AS) syndromes

are serious human developmental disorders caused

by loss, uniparental disomies, or perturbations in

the epigenetic expression of imprinted genes. Recent

reports examining databases of affected individuals

indicate that there is a significant increase in the

incidence of both BW and AS when ART techniques

were used.

In most cases, molecular genetic analy-

sis of affected individuals conceived by ART has

confirmed an epigenetic origin for their syndrome

phenotypes.

49,50

Another study showed an increased

ART-related incidence of retinoblastoma (another

disorder with a known epigenetic component).

However no molecular data linking the disorder with

epigenetic perturbations were presented.

In these studies, the overall incidence of these

imprinting disorders was still very small. However,

the incidence in ART offspring compared with base-

line population levels was in every case significantly

increased. The apparent incidence (based on a com-

parison using the overall numbers of ART births)

seems to be approximately one per 10 000 births.

However, one study of BW in Australia calculated an

expected risk of BW following ART as approxi-

mately one per 4000.

Of course, it is not possible for such comparisons

to demonstrate a direct connection between ART tech-

niques themselves and epigenetic disturbance, since

entirely different populations are being compared.



HPE_Chapter19.qxp 7/14/2007 5:51 PM Page 240

GENOMIC IMPRINTING AND CONSEQUENCES FOR EMBRYONIC DEVELOPMENT

ART offspring are conceived from a population of

individuals manifesting a variety of serious medical

problems associated with their infertility. In some

cases, the proximal cause of such infertility is cryptic

and as yet poorly understood. Obviously, a predis-

position for epigenetic aberrations could be an under-

lying cause of infertility in some couples, and other

individuals may exhibit specific ‘sensitivity’ to the

potential for perturbations caused by ovarian stim-

ulation, in vitro culture, or other components of

ART protocols. The increased incidence of apparent

imprinting disturbance has been observed following

a variety of ART interventions, including simple

stimulation/insemination protocols without any in

vitro culture or manipulation.

In the majority of

the studies published so far, detailed parental histo-

ries are lacking. However, in at least one case, there

was a clear family history of unexplained infertility

over multiple generations on the maternal side, poten-

tially indicating a genetic predisposition for develop-

mental (perhaps imprinting related) problems.

is of course also entirely possible that a certain level

of epigenetic disturbance does directly result from

ART protocols. Obviously, this must occur (or at least

be manifested phenotypically) at a greatly reduced

level compared with what has been observed in

large animal and research experiments, and as men-

tioned, significant differences in imprinting are evi-

dent between the human and these animal model

systems, and perhaps others are yet to be defined. As

mentioned earlier, it is still unclear from animal exper-

iments what the origin of in vitro culture-associated

epigenetic disturbance might be. Also, the true

correspondence between animal model systems and

the human in terms of both in vitro manipulations

and basic epigenetic processes also remains an open

question.

CONCLUSIONS, CONSEQUENCES,

AND FUTURE PROSPECTS

Taken as a whole, current knowledge about the

molecular and developmental biology of epigenetic

phenomena, the aberrations (as well as flexibility)

observed in model systems of in vitro manipulation,

and the increased incidence of serious imprinting-

related disorders in human ART-associated offspring

suggests that there is some cause for concern about a

relationship between human ART and abnormal

epigenetic regulation. Without question this issue

deserves serious further attention in terms of basic

research, population studies, follow-up studies on

ART offspring, and in the assessment, counseling, and

treatment of ART patients. As alluded to earlier,

epigenetic gene expression phenomena can theoreti-

cally manifest themselves throughout an individual’s

life history, including association with late onset dis-

eases such as cancer. Furthermore, such phenomena

can also be manifest across multiple generations.

The first human ART offspring are now reaching

adulthood and having children of their own, and

therefore data collection on potential epigenetic

abnormalities following ART is still very much ‘in

progress’. However, drawing premature conclusions

based on the data available to date is simply bad sci-

ence, with the potential to manifest itself in com-

promised treatment of ART patients. For example,

the suggestion that ICSI poses a danger and is specif-

ically associated with imprinting aberration is based

on very little evidence, and none of it is remotely con-

clusive as to a causative relationship. The fact that

imprinting disorders have been observed following

a range of ART treatments argues against a specific

relationship.

One direct study of the chromosomal

region associated with Angelman’s syndrome showed

no evidence of epigenetic misregulation in 92 ICSI-

conceived offspring.

A study of this magnitude

must be considered preliminary, since the disease

frequency is observed 50–100 times more rarely.

Although further population and continued fol-

low-up studies are most definitely needed to provide

a true assessment of the incidence of imprinting-

related consequences in ART offspring, this is only

part of the picture. As discussed above, simple pop-

ulation incidence studies cannot determine whether

a direct connection exists between ART procedures

and abnormal imprinting, or whether ART patients

represent a group with increased incidence or sensi-

tivity to epigenetic abnormalities. One future chal-

lenge will be to address this issue in terms of both

potential realities. In cases of completely unexplained



HPE_Chapter19.qxp 7/14/2007 5:51 PM Page 241

HUMAN PREIMPLANTATION EMBRYO SELECTION

infertility in particular, it will be prudent to develop

procedures for screening patients for potential

imprinting issues and perhaps embryos via PGD. In

addition, regardless of the current ambiguity, ART

patients should be informed and counseled regard-

ing the potential for an increased incidence of

imprinting-related disorders.

While the general success of human ART would

suggest that more general epigenetic problems are

not widespread, it is possible that subtle alterations

in epigenetic processing related to ART may have

some effect on development. For example, it is pos-

sible that such subtle alterations could be involved

in the slightly decreased birth weight observed in

ART singleton offspring in some studies.

Our

basic knowledge about imprinting status and more

general epigenetic status during human gamete and

embryo development needs to be greatly increased,

particularly in relation to ART issues and protocols.

Information about the timing, establishment, and

maintenance of epigenetic modification needs to

be gained in the relevant human material. Clearly,

model systems do not provide an accurate or com-

plete picture in this regard. For example, a recent

study revealed significant differences between the

human and the mouse in expression patterns of

genes related to epigenetic processing such as methyl-

transferase and methylated DNA binding proteins.

The imprinting-related methyltransferase DNMT3L

was highly expressed in developing mouse oocytes,

but did not appear in human material until after

fertilization in developing embryos, indicating that

the timing of imprint establishment may differ

between the two species.

Due to a variety of technical and ethical con-

cerns, little direct analysis of epigenetic status has

been conducted in human gametes and embryos.

However, the studies that have been published indi-

cate that, at least in the case of some genes that are

known to be imprinted, correct epigenetic status

appears to be undisturbed. An examination of the

methylation-imprint status of the imprinted SNRPN

gene (associated with Angelman syndrome) in human

spermatozoa, oocytes, and preimplantation embryos

revealed the expected pattern in all material ana-

lyzed.

Furthermore, these results indicated that

the correct maternal imprint was already stably

established in human oocytes by the germinal vesicle

stage, and not during later stages or following fertil-

ization as had been previously thought.

Another

meticulous study examined the status of the imprinted

H19 gene (associated with Beckwith-Wiedemann

syndrome) in human male germ cells obtained by

laser dissection of samples from the seminiferous

tubules of individuals who exhibited impaired sper-

matogenesis, a situation of particular importance

to the treatment of severe male factor patients. No

evidence of abnormal imprinting at H19 was evi-

dent, even in cells from tubules that exhibited arrest

at the spermatogonial stage.

Studies such as these

need to be extended in both number and scope.

As indicated above, the timing and nature of

normal epigenetic modifications in the human is a

basic issue. Some global epigenetic events, such as

paternal genome demethylation in the zygote, are

known to occur during the in vitro culture period of

ART protocols, and this process has been studied in

a preliminary fashion in human zygotes using methyl-

ated DNA-specific antibodies.

However, other epi-

genetic processing occurs prior to or following the

period of ART manipulation, and a full and accurate

picture of such processing in the human is currently

lacking. In addition, based on the results from ani-

mal culture studies (with the caveat of potential

cross-species differences), brief periods of transient

postfertilization in vitro culture can theoretically

cause perturbations to prior or subsequent epigenetic

modifications.

A variety of molecular methods are available for

determining the imprint status at particular loci,

based on an assessment of site-specific DNA methyl-

ation. Microarray techniques have recently been

adapted to such protocols, allowing multiple targets

to be assessed simultaneously.

If such microarray

protocols could be modified to allow for the greatly

reduced target concentration that is available from

single oocytes or early embryos, this would seem to

provide an ideal methodology for pursuing epigenetic

analysis in human ART-related material. A more

daunting question concerns assessing the possibility

of more general epigenetic perturbations, perhaps

to cryptically imprinted or non-imprinted genes.



HPE_Chapter19.qxp 7/14/2007 5:51 PM Page 242

GENOMIC IMPRINTING AND CONSEQUENCES FOR EMBRYONIC DEVELOPMENT

Single-cell expression analysis has been accomplished

in human gametes and embryos, and data from such

studies on both general variation as well as the spe-

cific expression of epigenetic-related gene products

will need to be obtained and examined with these

issues in mind.

This type of data will be crucial in

determining the real status of epigenetic disturbance

during human ART protocols. It could also assist in

the development of alternative protocols (culture

media/conditions, etc.) that, if necessary, might be

used to avoid epigenetic perturbation.

Examining the situation in preimplantation

embryos will provide only a partial analysis of poten-

tial epigenetic problems in ART, and it should perhaps

be admitted that, due to the complexity of this issue, a

truly complete assessment is simply impossible. As

mentioned in the introduction, the basic worldwide

status of human ART results to date, both anecdotally

and in terms of formal follow-up studies, does not

indicate the presence of any widespread or significant

negative effect that could result from epigenetic aber-

ration. Nevertheless, ART scientists and practitioners

owe their patients an ongoing effort to ensure that

the methodology in ART protocols ‘does no harm’.

Understanding the true nature and ramifications of

potential epigenetic problems related to ART proto-

cols or in the ART patient population should now be

seen as a critical component of this ongoing effort.

REFERENCES

1. Jaenisch R, Bird A. Epigenetic regulation of gene expression: how the

genome integrates intrinsic and environmental signals. Nat Genet

2003; 33: 245–53.

2. Surani MA. Imprinting and the initiation of gene silencing in the germ

line. Cell 1998; 93: 309–12.

3. Maher ER. Imprinting and assisted reproductive technology. Hum

Mol Genet 2005; 14: R133–8.

4. Lederberg J. The meaning of epigenetics. The Scientist 2001; 15: 6.

5. Waddington C. The genetic control of wing development in Drosophila.

J Genet 1940; 41: 75–80.

6. Latham KE. Epigenetic modification and imprinting of the mam-

malian genome during development. Curr Top Devel Biol 1999; 43:

1–49.

7. Lee JT, Jaenisch R. The (epi)genetic control of mammalian X-

chromosome inactivation. Curr Opin Genet Dev 1997; 7: 274–80.

8. Feinberg AP, Tycko B. The history of cancer epigenetics. Nat Rev Cancer

2004; 4: 143–53.

9. Finnegan EJ, Sheldon CC, Jardinaud F et al. A cluster of Arabidopsis

genes with a coordinate response to an environmental stimulus.

Curr Biol 2004; 14: 911–16.

10. Finnegan EJ, Kovac KA, Jaligot E et al. The downregulation of FLOW-

ERING LOCUS C (FLC) expression in plants with low levels of DNA

methylation and by vernalization occurs by distinct mechanisms.

Plant J 2005; 44: 420–32.

11. McGrath J, Solter D. Completion of mouse embryogenesis requires

both maternal and paternal genomes. Cell 1984; 37: 179–83.

12. Surani MAH, Barton SC, Norris ML. Development of reconstituted

mouse eggs suggests imprinting of the genome during gametogenesis.

Nature 1984; 308: 548–50.

13. Cattanach BM, Kirk M. Differential activity of maternally and paternally

derived chromosome regions in mice. Nature 1985; 315: 496–8.

14. Moore T. Genetic conflict, genomic imprinting and establishment

of the epigenotype in relation to growth. Reproduction 2001; 122:

185–93.

15. Moore T, Haig D. Genomic imprinting in mammalian development: a

parental tug-of-war. Trends Genet 1991; 7: 45–9.

16. Haig D, Graham C. Genomic imprinting and the strange case of the

insulin-like growth factor II receptor. Cell 1991; 64: 1045–6.

17. Pfeifer K. Mechanisms of genomic imprinting. Am J Hum Genet 2000;

67: 777–87.

18. Bestor T. Cytosine methylation and the unequal developmental

potentials of the oocyte and sperm genomes. Am J Hum Genet 1998;

62: 1269–73.

19. Schübeler D, Elgin SCR. Defining epigenetic states through chromatin

and RNA. Nat Genet 2005; 37: 917–18.

20. DeChiara TM, Robertson EJ, Efstratiadis A. Parental imprinting of the

mouse insulin-like growth factor II gene. Cell 1991; 64: 849–59.

21. Tremblay KD, Saam JR, Ingram RS et al. A paternal-specific methyla-

tion imprint marks the alleles of the mouse H19 gene. Nat Genet 1995;

9: 407–13.

22. Leighton PA, Saam JR, Ingram RS et al. An enhancer deletion affects

both H19 and Igf2 expression. Genes Dev 1995; 9: 2079–89.

23. Bell AC, West AG, Felsenfeld G. Insulators and boundaries: versatile

regulatory elements in the eukaryotic genome. Science 2001; 291:

447–50.

24. Holmgren C, Kanduri K, Dell G et al. CpG methylation regulates the

Igf2/H19 insulator. Curr Biol 2001; 11: 1128–30.

25. Kanduri C, Pant V, Loukinov D et al. Functional association of CTCF

with the insulator upstream of the H19 gene is parent-of-origin

specific and methylation-sensitive. Curr Biol 2000; 10: 853–6.

26. Murrell A, Heeson S, Reik W. Interaction between differentially

methylated regions partitions the imprinted genes Igf2 and H19 into

parent-specific chromatin loops. Nat Genet 2004; 36: 889–93.

27. Lee JT. Molecular links between X-inactivation and autosomal imprint-

ing: X-inactivation as a driving force for the evolution of imprinting?

Curr Biol 2003; 13: R242–54.

28. Moon H, Filippova G, Loukinov D et al. CTCF is conserved from

Drosophila to humans and confers enhancer blocking of the Fab-8

insulator. EMBO Rep 2005; 6: 165–70.

29. Loukinov D, Pugacheva E, Vatolin S et al. BORIS, a novel male germ-

line-specific protein associated with epigenetic reprogramming events,

shares the same 11-zinc-finger domain with CTCF, the insulator pro-

tein involved in reading imprinting marks in the soma. Proc Natl Acad

Sci USA 2002; 99: 6806–11.

30. Li E, Beard C, Forster AC et al.DNA methylation,genomic imprinting,

and mammalian development. Cold Spring Harb Symp Quant Biol

1993; 58: 297–305.

31. Oswald J, Engeman S, Lane N et al. Active demethylation of the paternal

genome in the mouse zygote. Curr Biol 2000; 10: 475–8.

32. Young LE, Wilmut I. Large offspring syndrome in cattle and sheep.

Rev Reprod 1998; 3: 155–63.



HPE_Chapter19.qxp 7/14/2007 5:51 PM Page 243

HUMAN PREIMPLANTATION EMBRYO SELECTION

33. Young LE, Fernandes K, McEvoy TG et al. Epigenetic change in

IGF2R is associated with fetal overgrowth after sheep embryo culture.

Nat Genet 2001; 27: 153–4.

34. Erbach GT, Lawitts JA, Papaioannou VE et al. Differential growth of

mouse preimplantation embryos in chemically defined media. Biol

Reprod 1994; 50: 1027–33.

35. Doherty TL, Mann MR,Tremblay KD et al. Differential effects of culture

on imprinted H19 expression in the preimplantation mouse embryo.

Biol Reprod 2000; 62: 1526–35.

36. Ecker DJ, Stein P, Xu Z et al. Long-term effects of culture of preim-

plantation mouse embryos on behavior. Proc Natl Acad Sci USA 2004;

101: 1595–600.

37. Sommovilla J, Bilker WB, Abel T et al. Embryo culture does not

affect the longevity of offspring in mice. Reproduction 2005; 130:

599–601.

38. Humpherys D, Eggan K, Akutsu H et al. Epigenetic instability in ES

cells and cloned mice. Science 2001; 293: 95–7.

39. Rideout WM 3rd, Eggan K, Jaenisch R. Nuclear cloning and epigenetic

reprogramming of the genome. Science 2001; 293: 1093–8.

40. Kono T, Obata Y, Yoshimzu T et al. Epigenetic modifications during

oocyte growth correlates with extended parthenogenetic development

in the mouse. Nat Genet 1996; 13: 91–4.

41. Wu Q, Kumagai T, Kwahara W et al. Regulated expression of two sets

of paternally imprinted genes is necessary for mouse parthenogenetic

development to term. Reproduction 2006; 131: 481–8.

42. Killian JK, Nolan CM, Wylie AA et al. Divergent evolution in

M6P/IGF2R imprinting from the Jurassic to the Quaternary. Hum Mol

Genet 2001; 10: 1721–8.

43. Arnaud P, Monk D, Hitchins M et al. Conserved methylation imprints

in the human and mouse GRB10 genes with divergent allelic expres-

sion suggests differential reading of the same mark. Hum Mol Genet

2003; 12: 1005–19.

44. Cox GF, Burger J, Lip V et al. Intracytoplasmic sperm injection may

increase the risk of imprinting defects. Am J Hum Genet 2002; 71:

162–4.

45. Neimitz EL, Feinberg AP. Epigenetics and assisted reproductive

technology: a call for investigation. Am J Hum Genet 2004; 74:

599–609.

46. Jacob S, Moley KH. Gametes and embryo epigenetic reprogramming

affect developmental outcome: implication for assisted reproductive

technologies. Ped Res 2005; 58: 437–6.

47. Maher ER. Imprinting and assisted reproductive technology. Hum

Mol Genet 2005; 14: R133–8.

48. Sutcliffe AG, Peters CJ, Bowdin S et al. Assisted reproductive technolo-

gies and imprinting disorders – a preliminary British survey. Hum

Reprod 2006; 21: 1009–11.

49. DeBraun MR, Neimitz El, Feinberg AP. Association of in vitro fertilization

with Beckman-Weidemann syndrome and epigenetic alterations of

LIT1 and H19. Am J Hum Genet 2003; 72: 156–60.

50. Gicquel C, Gaston V, Mandelbaum J et al. In vitro fertilization may

increase the risk of Beckwith-Wiedemann syndrome related to the

abnormal imprinting of the KCN10T gene. Am J Hum Genet 2003; 72:

1338–41.

51. Moll AC, Imhof SM, Cruysberg JR et al. Incidence of retinoblastoma

in children born after in-vitro fertilisation. Lancet 2003; 361: 309–10.

52. Halliday J, Oke K, Breheny S et al. Beckwith-Wiedemann syndrome

and IVF: a case-controlled study. Am J Hum Genet 2004; 75: 526–8.

53. Chang AS, Moley KH, Wangler M et al. The association between

Beckwith-Wiedemann syndrome and assisted reproductive technology:

a series of 19 patients. Fertil Steril 2005; 83: 349–54.

54. Örstavik KH, Eiklid K, van der Hagen CB et al. Another case of

imprinting defect in a girl with Angelman syndrome who was conceived

by intracytoplasmic sperm injection. Am J Hum Genet 2003; 72: 218–9.

55. Manning M, Lissens W, Bonduelle M et al. Study of DNA-methylation

patterns at chromosome 15q11–q13 in children born after ICSI reveals

no imprinting defects. Mol Hum Reprod 2000; 6: 1049–53.

56. Schieve LA, Meikle SF, Ferre C et al. Low and very low birth weight in

infants conceived with the use of assisted reproductive technology.

N Engl J Med 2002; 346: 731–7.

57. Huntriss J, Hinkins M, Oliver B et al. Expression of mRNAs for DNA

methyltransferases and methyl-CpG-binding proteins in the human

female germline preimplantation embryos and embryonic stem cells.

Mol Reprod Dev 2004; 67: 323–6.

58. Geuns E, De Ryke M, Van Steirteghem et al. Methylation imprints of

the imprint control region of the SNRPN-gene in human gametes and

preimplantation embryos. Hum Mol Genet 2003; 12: 2873–9.

59. El-Marrii O, Buiting K, Peery EG et al. Maternal methylation imprints

on human chromosome 15 are established during or after fertilization.

Nat Genet 2001; 27: 341–4.

60. Hartmann S, Bermann M, Bohle RM et al. Genetic imprinting during

impaired spermatogenesis. Mol Hum Reprod 2006; 12: 407–11.

61. Fulka H, Mrazek M, Tepla O et al. DNA methylation pattern in human

zygotes and developing embryos. Reproduction 2004; 128: 703–8.

62. Schumacher A, Kapranov P, Kaminsky Z et al. Micrroarray-based

DNA methylation profiling: technology and applications. Nucleic

Acids Res 2006; 34: 528–42.

63. Nikaido I, Saito C, Mizuno Y et al. Discovery of imprinted transcripts

in the mouse transcriptosome using large-scale expression profiling.

Genome Res 2003; 13: 1402–9.

64. Bermudez MG, Wells D, Malter H et al. Expression profiles of individ-

ual human oocytes using microarray technology. Reprod Biomed

Online 2004; 8: 325–7.



HPE_Chapter19.qxp 7/14/2007 5:51 PM Page 244

BACKGROUND

FRAMING THE PROBLEM

In vitro fertilization (IVF) is often criticized for its

generally low success rates and high treatment costs.

The procedure is also confounded by high multiple

birth rates that contribute to preterm delivery, small

for gestational age babies, and increased neonatal

mortality with its associated healthcare costs. These

problems could be significantly ameliorated if sin-

gle embryos of known viability and high develop-

mental potential could be identified within a cohort

group and selected for transfer. Moreover, the selec-

tion of such embryos by non-invasive techniques is

necessary.

Early embryonic wastage, which is usually caused

by abnormalities in the number of chromosomes

(aneuploidy), normally occurs in nature and is

endemic in the IVF treatment procedure. Mounting

scientific evidence suggests that as many as 60–75%

of all oocytes generated in IVF are abnormal. Likewise,

in nature the number of abnormal eggs increases

dramatically with age and this is the primary cause

of reduced fecundity in older women. Abnormalities

in the genetic makeup of the egg are usually mani-

fested as embryo wastage and/or congenital defects

in the newborn. Since embryo aneuploidy is trace-

able back to oocyte aneuploidy, it follows that

oocyte competence plays a central role in reproduc-

tive success. However, although a large number

of strategies have been employed, there are no

well-defined biological measurements or analytical

procedures available today that allow embryologists

to absolutely distinguish between viable and non-

viable oocytes and embryos.

A major limitation in IVF is the inability to

predict embryo viability. The ‘holy grail’ of IVF is

embryo selection; more specifically, how to deter-

mine which embryos in a cohort group are viable,

competent embryos that will implant and produce a

pregnancy versus those that will not. Therefore, it

follows that IVF treatments that (unknowingly) use

non-viable, defective embryos will likely result in

poor pregnancy outcomes and increased medical

risk. Although there are a number of non-invasive

methods to assess embryo quality, microscopic

examination of embryo morphology remains the

primary parameter of embryo selection.

Genetic

testing of the embryo by preimplantation genetic

diagnosis is an invasive, controversial procedure

that provides information about genetic traits of an

embryo, but, with few exceptions (e.g. trisomy 22),

not the embryo’s intrinsic biological competence.

These factors contribute to the finding that only

approximately 15% of embryos transferred in IVF

result in a pregnancy.

In order to overcome this

limitation, the practice of multiple embryo transfer

has been adopted as the standard of clinical prac-

tice. This, in turn, has led to the high incidence

of multiple births that is typically observed in IVF.

This situation has given rise to the ‘IVF dilemma’:

how can IVF programs maintain high preg-

nancy rates while reducing the number of embryos

transferred, to limit the incidence of multiple

births?

20. Selection of viable embryos and gametes

by rapid, non-invasive metabolomic

profiling of oxidative stress biomarkers

James T Posillico and The Metabolomics Study Group

for Assisted Reproductive Technologies



HPE_Chapter20.qxp 7/17/2007 3:38 PM Page 245