Elder K. Human preimplantation embryo selection

Подождите немного. Документ загружается.

HPE_Chapter22.qxp 7/13/2007 4:53 PM Page 286

OVARIAN STIMULATION AND HIGH-ORDER

MULTIPLE PREGNANCY

The use of ovarian stimulation has provided a means

of obtaining multiple oocytes per assisted reproduc-

tive technology (ART) treatment cycle. In most

cases the majority of oocytes retrieved will be suc-

cessfully fertilized, and as a result, several embryos

are typically available for transfer. Because the prob-

ability of an individual embryo forming a viable

pregnancy is relatively low (approximately 15%) it

has long been standard practice to transfer more

than one embryo per cycle. This approach has been

highly successful at increasing take-home-baby rates,

thus reducing the likelihood that a patient will need

to face the emotional, physical, and financial burden

of multiple treatment cycles.

Unfortunately, the transfer of multiple embryos

has also led to an explosion in the incidence of high-

order multiple gestations. Since the introduction of

in vitro fertilization (IVF) more than 20 years ago

the incidence of such pregnancies has increased

400% for women in their thirties and 1000% for

women in their forties (National Center for Health

Statistics). Although some patients initially welcome

the prospect of a ‘ready made family’, the medical

reality is that pregnancies of this type carry a sig-

nificantly elevated risk of serious complications

for both the mother and children. For the mother

there is increased risk of problems such as pre-

eclampsia, gestational diabetes, and vaginal and

uterine hemorrhaging. Fetal complications include

a high-risk of prematurity and very low birth weight,

as well as significantly elevated risks of miscarriage,

infant mortality, and cerebral palsy. High-order

multiple pregnancies also have financial implications

for both the parents and the health care provider.

1–3

For example, a triplet pregnancy will typically cost a

health care provider in the United States approxi-

mately $300 000.

EMBRYO VIABILITY ASSESSMENT –

MORPHOLOGY

The problem of multiple gestations in IVF stems

from the fact that current methods for predicting

embryo viability and implantation potential are

relatively poor. Multiple embryos are transferred

because this is the only way to maximize the proba-

bility that at least one viable embryo has been

selected. It is inevitable that in some cases all of

the embryos transferred will have the potential to

implant.

There is an increasing desire within the field of

reproductive medicine to overcome the problem of

multiple gestations. Additionally, in some countries,

legislation to limit the number of embryos trans-

ferred has come into force.

Although it may be

generally accepted that high-order multiple preg-

nancies are not a desirable outcome of IVF treatment,

many laboratories remain concerned about the

impact of a reduction in the number of embryos

transferred on pregnancy rates. This is particularly

true in countries where IVF success rates are pub-

lished and influence the number of patients seeking

treatment at a particular center.

If pregnancy rates are to be maintained while

the number of embryos transferred is reduced,

ultimately culminating in the routine use of single

embryo transfer at some point in the future, it will

be necessary to dramatically improve our ability to

assess embryo viability. One of the most important

challenges encountered each day in IVF laboratories

23. Future genetic and other technologies for

assessing embryos

Dagan Wells



HPE_Chapter23.qxp 7/13/2007 5:31 PM Page 287

HUMAN PREIMPLANTATION EMBRYO SELECTION

is the determination of which embryos to select for

transfer to the uterus. Which embryos have the

greatest probability of providing the patient with a

successful pregnancy?

In the majority of laboratories, microscopic

assessment of developmental stage and morphology

of the embryo is the most important guide to future

viability. Many different systems for scoring zygotes

and embryos based on their morphology have been

described, and in most cases these schemes have

been demonstrated to assist in distinguishing viable

embryos from non-viable to some extent.

5–11

However, morphological analysis remains a rela-

tively inefficient means of assessing embryos. All

physicians and embryologists working in IVF can

recount stories of patients who received two

embryos of extremely poor morphology and became

pregnant with twins. Similarly, we are all familiar

with cases in which embryo morphology was

perfect and yet no pregnancy ensued.

EMBRYO VIABILITY ASSESSMENT –

PREIMPLANTATION GENETIC

SCREENING

One reason why morphologically normal, appar-

ently healthy, IVF embryos may fail to form a preg-

nancy is the presence of aneuploidy – an incorrect

number of chromosomes. During the early 1990s,

research using fluorescence in situ hybridization

(FISH) revealed that chromosome anomalies in

human preimplantation embryos are startlingly

common. Analysis of between three and six of the

24 types of chromosome (22 pairs of autosomes

plus the X and Y), revealed that more than 50%

of cleavage stage embryos contained at least one

abnormal cell.

12–15

Later investigations using

comparative genomic hybridization (CGH), a more

comprehensive cytogenetic method that permits the

entire chromosomal set to be screened in a single

cell, suggested that between two-thirds and three-

quarters of human embryos contain abnormal cells

on day 3 postfertilization.

16,17

Most of the anom-

alies detected have been shown to be lethal,

and

their high prevalence in embryos produced during

IVF treatment is thought to explain many cases of

embryonic arrest, implantation failure, and early

spontaneous abortion. A critical point of note is

that most chromosomally abnormal embryos are

morphologically indistinguishable from their

euploid counterparts.

A major breakthrough in embryo assessment

came with the development of preimplantation

genetic diagnosis (PGD) for the detection of chro-

mosome abnormalities, known as PGD-AS (PGD-

aneuploidy screening) or PGS (preimplantation

genetic screening). The most widely applied PGS

strategy involves the biopsy of a single blastomere

on day 3 postfertilization. The blastomere can be

subjected to a wide range of genetic analyses. In the

case of chromosomal screening, the cell is spread

onto a microscope slide, fixed, and then analyzed

using FISH. At present, the most experienced labo-

ratories assess up to ten chromosomes per cell. This

is achieved by employing two sequential rounds

of hybridization using chromosome-specific probes

labeled with distinct fluorescent dyes.

The use of PGS as a guide to selecting embryos

to be considered for transfer, supplementing (not

replacing) morphological analysis, has been shown

to significantly improve several IVF outcomes. Not

surprisingly, the screening of chromosomes 13, 18,

21, X, and Y has led to a reduction in the number of

pregnancies affected by aneuploid syndromes (i.e.

Patau, Edward, Down, Turner and Klinefelter).

This is particularly evident in IVF patients where

the female partner is over 35 years of age, an age

after which the risk of producing a chromosomally

abnormal oocyte begins to increase rapidly. It is also

clear that certain groups of IVF patient display a

decline in the incidence of spontaneous abortion

following PGS. This is not surprising, given that

chromosome imbalance is known to be the major

cause of early miscarriage in natural cycles. Of first

trimester pregnancy losses, 60–70% display chromo-

some anomalies.

20–22

Of most significance to the issue of multiple

pregnancies is the impact of PGS on embryo implan-

tation rates. Although some controversy remains,

there is mounting evidence that PGS provides spe-

cific groups of IVF patient with a significant increase



HPE_Chapter23.qxp 7/13/2007 5:31 PM Page 288

FUTURE GENETIC AND OTHER TECHNOLOGIES FOR ASSESSING EMBRYOS

in implantation rate per embryo transferred.

23,24

Apparent contradictions in the literature regarding

the impact of PGS on implantation rate seem to be

a consequence of methodological differences between

studies.

Data suggest that increased implantation

is only conveyed by PGS strategies employing biopsy

of a single blastomere (rather than two) and analy-

sis of at least seven chromosomes. If PGS truly has

the ability to identify embryos with high implanta-

tion potential, then it is likely that this type of

screening can help to reduce the incidence of multiple

pregnancy, while maintaining an acceptable preg-

nancy rate. A recent study performed by Munné

et al has provided preliminary data in support of

this suggestion.

The patients taking part in the study were divided

into two groups: (1) embryos assessed using PGS

and transfer limited to one to two embryos per cycle;

and (2) a control group in which embryos were

transferred without any genetic screening and no

restriction was placed on the number of embryos

transferred to the patient. Data from over 100 cycles

were documented in each group, with patients

matched for a variety of factors including maternal

age. As with earlier studies, embryos transferred

after PGS displayed an approximate doubling of

implantation rate compared with the control group.

However, despite the increased implantation rate

for the PGS embryos, pregnancy rates for the two

groups were essentially identical (30% for the PGS

group, 33% for the control). This is because fewer

embryos were transferred per cycle in the PGS group

(average of 1.5 per cycle for PGS versus 3.6 for con-

trol). Importantly, no high-order multiple pregnan-

cies were seen in the PGS group, whereas two triplet

pregnancies and more twins were present in the con-

trol group (Munne et al, personal communication).

PREIMPLANTATION GENETIC SCREENING –

FUTURE DIRECTIONS

Although PGS has provided significant benefits in

terms of embryo screening and IVF outcome, exist-

ing protocols assess less than half of the chromoso-

mal complement. Examination of the chromosomes

not usually tested during PGS in a research context,

has shown that any chromosome can display aneu-

ploidy during preimplantation development. Thus,

it is inevitable that some aneuploid embryos are

incorrectly classified ‘normal’ using current PGS

methods, and may be inadvertently transferred.

Such embryos are believed to have little potential for

forming a viable pregnancy and most likely fail to

implant or spontaneously abort at an early stage of

gestation.

It seems logical that an expanded chromosomal

screen, permitting assessment of every chromo-

some, would improve the ability of PGS to identify

viable embryos. However, a comprehensive chro-

mosomal screen cannot be achieved using FISH due

to the restricted number of spectrally distinct fluo-

rochromes available for probe labeling. Although

this problem can be partially overcome by perform-

ing sequential FISH experiments on the same cell,

accuracy declines with each additional hybridiza-

tion and consequently most PGS laboratories limit

the number of rounds of FISH to two.

Although methods such as G-banding, multi-

plex-FISH (M-FISH), and spectral karyotyping

(SKY) provide information on every chromosome,

they are dependent on the presence of cells in

metaphase. Unfortunately the vast majority of blas-

tomeres sampled from day 3 embryos are found to

be in interphase. During this phase of the cell cycle

chromosomes are contained within the nucleus and

cannot be distinguished from one another. On the

rare occasions when a blastomere in metaphase is

biopsied, artifacts of the method used for fixation

and spreading, most notably loss of chromosomes,

are often observed.

Comparative genomic hybridization (CGH) is a

method related to FISH that permits analysis of the

entire chromosome complement. Aneuploidy detec-

tion is achieved via a competitive hybridization of

differentially labeled DNA samples to normal meta-

phase chromosomes

(Figure 23.1). In the case of

PGS, DNA from the sample (blastomere) is labeled

with a green fluorochrome, while chromosomally

normal reference DNA (usually from 46 XY lym-

phocytes) is labeled with a red fluorochrome. The

two labeled samples are simultaneously hybridized



HPE_Chapter23.qxp 7/13/2007 5:31 PM Page 289

HUMAN PREIMPLANTATION EMBRYO SELECTION

to normal metaphase chromosomes on a microscope

slide. Red and green DNA fragments compete for

hybridization to their complementary sequences

on the chromosomes, causing each chromosome to

adopt a coloration related to its copy number in

each of the samples. For example, if the green (blas-

tomere) DNA was from an embryo with an extra

copy of chromosome 16, there would be relatively

more green copies of chromosome 16 DNA frag-

ments than red. This will cause green DNA to out-

compete red DNA for hybridization to chromosome

16. The overall effect is that chromosome 16 appears

more green than the other chromosomes. Chromo-

some imbalances of this type are most clearly

indicated by analyzing the ratio of green : red fluo-

rescence along the length of each chromosome.

CGH represents an excellent method for the simul-

taneous assessment of all chromosomes, and can

even reveal errors involving fragments of chromo-

somes, provided they are at least 5 Mb in size.

Unfortunately, CGH requires ~1 ␮g of DNA

whereas a single blastomere contains only 5–10 pg.

For this reason it is necessary to perform whole

genome amplification (WGA) prior to CGH analy-

sis of individual cells. The method of choice for this

purpose is known as degenerate oligonucleotide

primed (DOP) PCR.

27,28

DOP-PCR utilizes a semi-

degenerate oligonucleotide primer that has the abil-

ity to anneal at many sites throughout the genome.

With the addition of a thermostable DNA poly-

merase, DNA synthesis is initiated at each site of

primer annealing. Thermal cycling using the princi-

ples of the polymerase chain reaction (denaturation,

annealing, synthesis) permits a dramatic amplifica-

tion of DNA from the original sample and provides

enough material for subsequent CGH analysis.



Normal

1 : 1

Trisomy

3 : 2

Monosomy

1 : 2

Normal DNA

46 XY

Test DNA

Ratio red : green

Figure 23.1 Comparative genomic hybridization (CGH) involves the simultaneous hybridization of differentially labeled DNA

samples – a chromosomally normal DNA (labeled with a red fluorochrome) and sample DNA (labeled with a green fluorochrome).

The relative number of red and green DNA fragments hybridizing to a given chromosome is related to the number of copies of the

chromosome in question in the sample compared with the control. An excess of chromosomal material in the sample causes a green

coloration, while a deficiency leads to a red coloration. Identical copy numbers of sample (green) and control (red) chromosomes

results in equal hybridization of red and green fragments and a yellow coloration. In most cases a computer is used to assist in the

calculation of red : green fluorescence intensity along the length of each chromosome.

HPE_Chapter23.qxp 7/13/2007 5:32 PM Page 290

FUTURE GENETIC AND OTHER TECHNOLOGIES FOR ASSESSING EMBRYOS

Studies applying CGH to the analysis of cleavage

stage embryos of good morphology confirmed that

aneuploidy could affect any chromosome during

human preimplantation development. Abnormalities

included chromosome breakage and varieties of

aneuploidy that are not seen in prenatal samples or

in material from spontaneous abortions (e.g. imbal-

ance affecting the larger chromosomes of groups A

and B).

16,17

These unfamiliar forms of abnormality

are presumed to be lethal during early stages of

development.

Analysis of the published CGH data reveals that

25–30% of embryos carry chromosome abnormali-

ties that would not be detectable using the nine-

chromosome PGS examination employed by most

PGD laboratories.

29,30

The fact that current PGS

protocols are not 100% successful in preventing the

transfer of aneuploid embryos argues for the devel-

opment of methods that permit a comprehensive

analysis of blastomere chromosomes. However, it

may be possible to significantly improve embryo

selection simply by adjusting the combination of

chromosomes screened.

In most cases the chromosomes selected for

screening are included because they are frequently

found to be abnormal in prenatal samples and mis-

carriages. However, these chromosomes are not

necessarily the most relevant in terms of implanta-

tion failure.

Unfortunately, there are insufficient

published CGH data to allow the incidence of

aneuploidy to be precisely calculated for individual

chromosomes, and consequently it is difficult to

determine precisely which chromosomes should be

prioritized for screening. However, pooling all exist-

ing CGH data can provide a rough indication of the

most common aneuploidies at the cleavage stage. In

most cases this analysis confirms the importance

of the chromosomes currently assessed by PGS.

However, it also suggests that screening might be

improved if some chromosomes were substituted by

others with higher preimplantation aneuploidy rates.

Chromosomes that have not been extensively

studied in human embryos, but appear to show above

average aneuploidy rates include chromosomes 2, 4,

7, 9, 12, and 20. This observation is based on a small

number of samples and requires confirmation using

CGH or FISH on a much larger series of embryos

before adjustment of clinical PGS strategies. In

addition to the CGH data, some FISH studies have

also indicated that a revision of the chromosomes

assessed by PGS may be necessary.

From the pooled CGH data, one can predict that

the proportion of aneuploid embryos detected could

be increased from 70–75% to ⬎95% if the number

of chromosomes screened by FISH was expanded

from nine to 15. However, although it may be possi-

ble to achieve high detection rates for aneuploid

embryos without screening the entire chromosome

complement, it is inevitable that maximum accu-

racy will only be achieved if all the chromosomes

are evaluated. To date CGH has been the most

promising method for this purpose. However, as

with FISH there are limitations to this technology.

The principal difficulties in applying CGH clini-

cally are the complexity of the technique and the

length of time required (approximately 5 days),

which is incompatible with the restricted timeframe

available for PGS. One strategy for overcoming the

problem of timeframe involves cryopreservation of

embryos after biopsy, with embryo transfer occur-

ring in a subsequent cycle.

29,30,33

The main drawback

of this approach is that freezing and thawing can

lead to a reduction in embryo implantation poten-

tial, a problem that is exacerbated by embryo biopsy.

An alternative strategy for the clinical applica-

tion of CGH is based upon assessment of the first

polar body (PB).

The first PB is available for analy-

sis 3 days earlier than biopsied blastomeres, provid-

ing sufficient time to perform CGH without the

need for cryopreservation. However, confirmatory

FISH analysis of blastomeres is advisable since chro-

matid anomalies detected in meiosis I have only a

50% chance of leading to an aneuploid embryo.

There is also the potential for misdiagnosis due to a

meiosis II error or a chromosome abnormality of

paternal origin.

Whatever the CGH strategy employed for PGS,

the complexity of the technique remains a major

problem. The protocol is labor intensive and neces-

sitates expertise in both molecular genetic and cyto-

genetic methods that are not generally available to

fertility clinics. A less complex approach will be



HPE_Chapter23.qxp 7/13/2007 5:32 PM Page 291

HUMAN PREIMPLANTATION EMBRYO SELECTION

necessary if CGH is to be widely applied. At the

present time, the best hope for a simplified method-

ology is microarray CGH. As with conventional CGH,

this method involves the competitive hybridization

of differentially labeled test and reference DNA

samples. However, in this case the labeled DNAs are

hybridized to DNA probes fixed to a microscope

slide rather than metaphase chromosomes. Each

probe is specific to a different chromosomal region

and occupies a discrete spot on the slide. Chromo-

somal loss or gain is revealed by the color adopted

by each spot after hybridization (ratio of red : green

fluorescence). Importantly, the evaluation of red :

green fluorescence is easily automated, circumvent-

ing the need for an experienced cytogeneticist.

Microarray CGH has been successfully applied

for the detection of aneuploidies in single cells after

whole genome amplification using DOP-PCR or an

alternative method known as multiple displacement

amplification (MDA).

35–37

Using this approach, com-

prehensive chromosome analysis has been achieved

in less than 48 hours (i.e. within the timeframe nec-

essary for PGS) and consequently the future appli-

cation of microarray CGH for PGS appears to be

extremely encouraging.

Although comprehensive chromosomal analy-

sis would seem to be advantageous for PGS, some

physicians have expressed concern that a full-

chromosome screening will increase the number of

embryos excluded due to aneuploidy, leading to

more IVF cycles in which there are no embryos

eligible for transfer. Even though current protocols

only assess nine or ten chromosomes, a small but

significant number of PGS cycles end without

embryo transfer due to all embryos receiving an

unfavorable diagnosis.

Fortunately, the small amount of data currently

available indicate that the proportion of embryos

diagnosed as normal may actually increase follow-

ing CGH analysis. In a small study comparing FISH

and CGH for the purposes of PGS, 40% of embryos

analyzed by CGH were found to be chromosomally

normal, while the proportion of embryos diag-

nosed normal by FISH was 33%.

The most likely

explanation for this counterintuitive observation is

inaccuracy in the FISH methodology caused by the

presence of micronuclei. Micronuclei, which are a

relatively common phenomenon in human blas-

tomeres, contain chromosomal material and are

prone to loss during fixation and spreading prior to

FISH. This may explain why an excess of mono-

somies (relative to trisomies) have been detected

following FISH analysis of human embryos. CGH

does not involve the spreading of the sample and

consequently avoids this source of error.

RELATIONSHIP BETWEEN GENE EXPRESSION

AND EMBRYO VIABILITY

The application of PGS has been shown to benefit

several groups of IVF patients and proves that chro-

mosome abnormality is an important factor affect-

ing embryo implantation potential. However, when

considering implantation failure it is also clear that

chromosome anomaly is only part of the story.

Many IVF-PGS cycles fail to provide a pregnancy,

despite the transfer of embryos classified as chro-

mosomally and morphologically normal. Further-

more, there are significant subgroups of IVF patients

that show little or no improvement in embryo

implantation following PGS (e.g. patients under 35

years of age). Although some of these failures may

be attributed to uterine or endocrine problems, it is

likely that many (perhaps most) are the conse-

quence of nonchromosomal embryological issues.

A host of vital, yet largely invisible, processes

occur during the first few days of life. These include

the transition away from a reliance on maternal

protein and mRNA that accompanies activation of

the embryo genome, and the first types of cellular

differentiation. Given the complex and fundamen-

tal nature of events the embryo must successfully

undertake at this time, it is perhaps unsurprising

that developmental arrest during the preimplanta-

tion phase is common. One of the great challenges

now facing scientists involved in translational IVF

research is to characterize the cellular pathways

perturbed in embryos destined for implantation

failure. An improved understanding in this area

may lead to new viability markers, revealing embryos

that are under pressure and heading for arrest, and



HPE_Chapter23.qxp 7/13/2007 5:32 PM Page 292

FUTURE GENETIC AND OTHER TECHNOLOGIES FOR ASSESSING EMBRYOS

assist in the optimization of ovarian stimulation

and embryo culture methods.

An extremely useful approach for investigating

the processes occurring in a biological sample (e.g. a

cell, a tissue, an embryo, etc.) is to examine gene

expression. The activity of individual genes is con-

stantly changing, fluctuating in response to the

changing needs of the cell. Variation in environ-

mental conditions will induce changes in gene

expression, as will factors such as altered metabolic

requirements, the presence of aneuploidy or DNA

damage, progress through the cell cycle, and processes

such as differentiation and apoptosis.

Quantification of the number of mRNA tran-

scripts derived from a given gene provides an indi-

cation of how actively expressed it is. This in turn

provides information on the activity of the cellular

pathways that require the product of that gene.

Embryonic gene expression at different preimplan-

tation stages has been assessed using real-time poly-

merase chain reaction (PCR). This method involves

extraction of RNA from the sample, reverse tran-

scription in order to produce cDNA and PCR

amplification of a cDNA fragment from the gene(s)

of interest. The accumulation of amplified DNA is

measured in each sample tube in real-time (i.e. once

per PCR cycle) by monitoring the fluorescence

emitted by sequence specific probes or generic DNA

stains. All samples are assessed relative to standards

containing known numbers of template molecules.

Measurements taken while amplification is still

proceeding in an exponential fashion, allow the

number of templates in each sample at the begin-

ning of the PCR to be calculated by reference to the

standards.

A recent gene expression study focused on genes

with roles in cell cycle regulation, DNA repair, sig-

naling pathways, and apoptosis, functions of great

importance during early development.

Although

transcripts from every gene assessed were detectable

at every stage, from mature oocyte to hatched blas-

tocyst, the number of transcripts at different stages

varied considerably. The quantity of mRNA tran-

scripts was generally high in oocytes, but decreased

dramatically after fertilization, in agreement with

earlier observations.

40–42

The data obtained clearly

indicate that transcripts from many genes reach extre-

mely low levels in 2- and 3-cell embryos, in some

cases scarcely above the threshold of detection.

Depletion of maternal mRNA after fertilization

appears to be a normal occurrence and yet may

have great significance for the embryo. Most cellular

pathways are controlled to some degree by the tran-

scriptional activation or repression of specific genes.

In other words, gene expression allows the subtle

control of cellular mechanisms in response to changes

in the intracellular and extracellular environment. It

is likely that prior to the activation of the embryonic

genome many important cellular pathways exist in a

rigid form, controlled by a reservoir of proteins

inherited from the oocyte. Consequently, the embryo

may have limited ability to respond to environ-

mental challenges during this period. This may have

important implications for the practice of in vitro

fertilization and embryo culture methods.

No embryonic gene expression was detected

until the 4-cell stage, consistent with previous data

demonstrating the onset of gene activation.

43,44

most cases the increase in transcript number upon

genome activation was modest, however, in a subset

of embryos dramatic increases in expression were

seen for the BRCA1 gene (several hundred fold).

BRCA1 is a multifunctional protein with roles in

cell cycle regulation and DNA repair. The increased

expression of BRCA1 in cleavage stage embryos may

indicate the presence of DNA damage. Such genetic

damage could be derived from the sperm (as indi-

cated by studies employing the sperm chromatin

structure assay)

or the oocyte. It is possible that

DNA repair is relatively inefficient until the embry-

onic genome is activated and fresh gene expression

permits the stimulation of DNA repair pathways.

From the 10-cell to morula stages the activity of

most genes examined stabilized or even underwent

a small decline. Thereafter, expression levels rose

proportionately with increasing cell number, and as

a result, most genes displayed greater transcript

numbers in blastocysts than at any other stage. The

high gene activity in blastocysts is not surprising, as

most of the genes assessed produce proteins that

interact with DNA or chromosomes (i.e. the quan-

tity of protein required is likely to be closely related



HPE_Chapter23.qxp 7/13/2007 5:32 PM Page 293

HUMAN PREIMPLANTATION EMBRYO SELECTION

to the number of nuclei). Gene expression was

clearly linked to developmental stage, with embryos

at similar stages having very similar patterns of gene

expression.

The gene expression data were also analyzed

using a hierarchical cluster algorithm, a mathemati-

cal method for identifying samples with similar

characteristics. The results are displayed graphically

as a dendogram, embryos with similar gene expres-

sion being placed on closely associated branches of a

tree diagram. Cluster analysis grouped embryos of

specific stages together, confirming statistically that

patterns of gene expression and developmental

stage are closely associated. Only 20% of embryos

failed to cluster with counterparts of similar stage.

Interestingly, more than half of these embryos dis-

played morphological abnormalities, suggesting

that such aberrations may cause (or be caused by)

perturbations of gene expression

(Figure 23.2).

As well as separating embryos by developmental

stage, cluster analysis was able to identify subgroups

of identically staged embryos that had distinct

expression profiles. Analysis of the morphological

notes for the embryos within these subgroups

revealed that one group was associated with the

presence of granular cytoplasm, condensed orga-

nelles, and multiple nuclei – features that are nega-

tively correlated with embryo implantation. A

second group of 4–10-cell stage embryos was found

to contain a preponderance of embryos with uneven

cleavage divisions.

Interestingly, not all the groups of embryos

displaying a distinct pattern of gene activity were

associated with a characteristic morphology. This

demonstrates that differences in expression are

frequently related to non-morphological factors,

which could include factors such as the cell cycle,

cellular stress, chromosomal status (e.g. aneuploid



ABCDEFGH I J ABCDEFGH I J

ABCDEFGH I J

Figure 23.2 Quantification of gene expression reveals that most embryos show an expression profile characteristic of their develop-

mental stage. However, abnormalities can lead to disturbances in the normal pattern of gene expression. This figure shows the results

of gene expression analysis for ten genes (A–J) in four blastocysts (1–4). Blastocysts 1, 2, and 3 displayed very similar patterns of gene

expression and appeared to be morphologically identical (good morphology). However, analysis of blastocyst 4 revealed altered

expression of several genes. This blastocyst had a poorly defined inner cell mass and minor growth retardation.

HPE_Chapter23.qxp 7/13/2007 5:32 PM Page 294

FUTURE GENETIC AND OTHER TECHNOLOGIES FOR ASSESSING EMBRYOS

versus euploid), or microenvironment of individual

embryos. The potential of gene expression analysis

to reveal ‘hidden’ information concerning processes

occurring within the embryo may be of clinical

value, as some of these processes are likely to have a

bearing on viability.

Further studies will continue to improve our

understanding of the cellular processes occurring

during the preimplantation phase. A recent advance

has been the development of in vitro transcription

techniques that permit the mRNA content of single

oocytes to be amplified to levels sufficient for

microarray analysis.

Microarrays allow the simul-

taneous analysis of ⬎29 000 genes in a single exper-

iment and have allowed us to catalogue over 9000

genes that are expressed in human oocytes (Wells

et al, unpublished data). It is anticipated that the

application of such methods to the study of human

preimplantation development will greatly accelerate

the progress of discovery in this area. Our analysis of

human oocytes has already revealed several key cel-

lular pathways that are differentially expressed when

a chromosome abnormality is present. Not only

does this open up new possibilities for the detection

of aneuploidy, but also it reveals the cellular mecha-

nisms underlying this problem. The knowledge

gained from such studies may lead to the develop-

ment of preventative interventions, such as improved

ovarian stimulation regimens or optimized culture

media.

PROTEOMIC ANALYSIS OF HUMAN EMBRYOS

Investigation of proteins may prove to be even more

useful for assessing embryo viability than either

cytogenetic or gene expression analysis. Although

the gene expression data obtained provide an inter-

esting insight into the genetic activity of the early

embryo, a change in the number of mRNA transcripts

derived from a given gene does not necessarily indi-

cate altered utilization of the pathway in which it

functions. Most genes experience some degree of

regulation at the post-translational level, through

protein modification, degradation, or sequestra-

tion. Consequently, there may be occasions when a

change in the concentration of active protein is not

mirrored by an alteration in gene activity. Further-

more, individual genes usually produce more than

one type of protein, accomplished by utilizing

mechanisms such as alternative splicing and post-

translational modification. It is not possible to detect

post-translational modifications by looking at gene

expression.

Advances in mass spectrometry have led to the

development of techniques with sufficient sensitiv-

ity to permit the analysis of individual embryos. A

recent study by Katz-Jaffe and colleagues utilized

surface-enhanced laser desorption and ionization

time-of-flight mass spectrometry (SELDI-TOF MS)

to produce proteomic profiles for individual human

blastocysts.

The method involved the lysis of whole

blastocysts and binding their proteins to chips.

Bound proteins were released from the surface of

the chip and ionized by laser activation and the mass :

charge ratio of the liberated ions determined by

time-of-flight mass spectrometry. The mass : charge

ratios of the many different protein fragments pro-

duced during this process produce a unique pro-

teomic fingerprint for each sample assessed.

Analysis of individual blastocysts revealed differ-

ences in proteomic profiles related to morphology

and developmental stage. For example, degenerat-

ing embryos displayed numerous alterations in

protein expression when compared with developing

blastocysts.

Proteomic analysis also identified dif-

ferences between embryos that were not correlated

with either developmental stage or morphology. In

some cases such changes may be related to altered

utilization of specific cellular pathways that influ-

ence or reflect embryo viability. Analysis of the pro-

teins involved may allow morphologically identical

embryos to be distinguished in terms of their

implantation potential.

ASSESSMENT OF OOCYTES AND EMBRYOS

WITHOUT BLASTOMERE BIOPSY

Analyses of genes and proteins have demonstrated

that dynamic fluctuations in expression occur

throughout preimplantation development and that



HPE_Chapter23.qxp 7/13/2007 5:32 PM Page 295