Elder K. Human preimplantation embryo selection

Подождите немного. Документ загружается.

HUMAN PREIMPLANTATION EMBRYO SELECTION

with the sequential Bonferroni procedure. The out-

come of these analyses is indicated in the tables.

The highest implantation and pregnancy rates

were indeed obtained following transfer of embryos

with 0–5% fragmentation. The embryos in this

group had significantly more cells on average than

the embryos in all other groups, demonstrating an

association between fragmentation and total cell

number (Table 6.2).

The original study

had also shown a general

decline in implantation and pregnancy rates when

the majority of the transferred embryos had type I,

type II, type III, or type IV fragmentation, but

the largest decrease in implantation rate occurred

when embryos with fragmentation type IV were

transferred. Embryos with type IV fragmentation

also showed the highest degree of fragmentation,

on average about 25%, which is likely to have

contributed to their low viability.

An updated analysis of transfers in which all

embryos were in the same category with respect to

the pattern of fragmentation (1263 transfers and

2703 transferred embryos) (Tables 6.3 and 6.4; same

statistical analyses as described above) showed that

the highest implantation rate was achieved follow-

ing exclusive transfer of embryos with type I frag-

mentation, while the poorest implantation and

pregnancy outcomes were obtained with embryos

with type IV fragmentation. Embryos with type II

fragmentation implanted as frequently as those with

type III fragmentation. On average, these two pat-

terns showed similar degrees of fragmentation, but

type II embryos had significantly fewer cells than

type III embryos (5.4 ⫾ 1.5 vs 6.1 ⫾ 1.3 cells,



Table 6.3 Pregnancy and implantation outcome in homogeneous transfer groups, including oocyte recipients

Fragmentation

pattern transfer No. of embryos Implantation

group No. of procedures No. of pregnancies Pregnancy rate

(%) transferred No. of FHB rate

(%)

(1) Type I 130 72 55.4 242 101 41.7

(2) Type II 76 25 32.9 133 37 27.8

(3) Type III 840 412 49.0 1957 613 31.3

(4) Type IV 217 50 23.0 371 61 16.4

FHB, fetal heart beat.

Pair-wise contrasts of pregnancy results in groups 1–4 show statistically significant differences between all except group 1 vs group 3; and group 2 vs

group 4.

Pair-wise contrasts of implantation results in groups 1–4 show statistically significant differences in all except group 2 vs 3.

Table 6.4 Other characteristics of embryos in the homogeneous fragmentation pattern transfer groups

Average day-3

Fragmentation Average day-3 Average day-3 Average day-3 fragmentation

pattern transfer cell number

fragmentation

cell number

(transferred

group (all embryos) ⫾SD (all embryos) (%) ⫾SD (transferred embryos) ⫾SD embryos) (%) ⫾SD

(1) Type I 6.5 1.4 13.1 9.3 7.5 1.3 5.3 2.7

(2) Type II 5.4 1.5 21.7 13.6 6.6 1.7 13.5 6.2

(3) Type III 6.1 1.3 20.5 10.7 7.3 1.3 13.3 5.7

(4) Type IV 5.2 1.4 30.1 15.0 6.3 1.3 24.6 10.9

a) Differences among the groups are significant at the 0.05 significance level (Newman–Kuels multiple comparisons test).

b) Difference in average day 3 fragmentation in transferred embryos in group 2 vs group 3 is not significant.

HPE_Chapter06.qxp 7/13/2007 4:46 PM Page 56

respectively, overall; 6.6 ⫾ 1.7 vs 7.3 ⫾ 1.3 cells,

respectively, in transferred embryos). Pregnancy rate,

however, was significantly lower following transfer

of type II embryos than type III embryos (32.9 vs

49%, respectively).

Compared to embryos with all other patterns of

fragmentation, those with type IV fragmentation

had fewer cells (5.2 ⫾ 1.4) on day 3 and led to far

fewer implantations and pregnancies. Fluorescence

in situ hybridization (FISH) analysis has shown a

significantly higher incidence of chromosomal

abnormality in embryos with type IV fragmenta-

tion than those with types I, II, or III fragmentation

(M Alikani, G Tomkin and S Munné, unpublished

data).

The same fragmentation pattern classification

system

was adopted by others

who studied the

incidence of different patterns among transferred

embryos in pregnant and non-pregnant cycles.

Prior to embryo transfer, the zona pellucida in all

embryos was partially opened with acidified

Tyrode’s solution but fragments were not removed.

The degree of fragmentation was not noted in these

embryos. The results showed that in pregnancy

cycles, 48% of transferred embryos had type I frag-

mentation, while in non-pregnant cycles, a gener-

ally lower incidence of type I and a generally higher

incidence of type II fragmentation had occurred.

Implantation failed in seven cases in which embryos

with fragmentation types III and IV were transferred

exclusively. The latter outcome is not consistent

with our own experience with these fragmenta-

tion patterns, but the discrepancy may be partly

explained by the very low number of observations

in the study of Desai et al.

Moreover, since the

classification of the patterns still has an element of

subjectivity, observer variation may also account for

these differences.

In another interesting study,

embryos with more

than 25% fragmentation showed very low viability

(0.8%). When cell number, degree of fragmenta-

tion, and cell asymmetry (the latter defined as ‘none,

some, and severe’) were considered together, there

was a measurable negative impact of asymmetry on

viability of 8-cell embryos, regardless of the degree

of fragmentation (⬍10% or 10–25%). Transfer of

7-cell and ⬍7-cell embryos with severe asymmetry

led to complete failure of implantation.

Collectively, these data point to a definition of

‘top quality’ embryos as those with less than 20%

fragmentation, at least seven blastomeres on day 3

of development

30,31

, little asymmetry (attributable

to asynchronous division of cells), and no major

size discrepancy attributable to fragmentation or

uneven division of cells.

21,27,32

The impact of fragmentation on neonatal out-

come is less clear. One study

has suggested that

transfer of embryos with 25% to ⬎50% fragmenta-

tion leads to significantly higher rates of fetal abnor-

malities than transfer of embryos with ⬍25%

fragmentation. Four minor malformations occurred

among the 180 children born following 309 transfers

in the ⬍25% fragmentation group. In another group

consisting of 75 transfers that included embryos

with 25–50% fragmentation, one case of trisomy 21

and one of fibroma were seen among 13 newborns.

Among 19 children born following 76 transfers that

included embryos with ⬎50% fragmentation, two

cases of trisomy 18, one of hydrocephalus with anal

atresia, and one of hydrocele were found.

Several aspects of this study are puzzling. For

example, the large proportion of transfers that

reportedly included embryos with ⬎50% fragmen-

tation (75/460 or 16%) brings into question the

accuracy of the fragmentation estimates. In our data-

base of 3322 transfers homogeneous with respect to

the degree of fragmentation, only 1.3% (43/3322)

had embryos with ⬎35% fragmentation. The total

number of embryos with 50% or more fragmenta-

tion within this group was 25. When the entire

database was considered, 96 of 25 372 (0.38%)

transferred embryos had 50% or more fragmenta-

tion. These figures reflect the deliberate exclusion of

such embryos from transfer and the relatively low

frequency with which such extensive fragmentation

occurs in the first 2–3 days of culture (5893/79 936

or 7.4%).

Another point of debate is the implied associa-

tion of aneuploidy – in this case, trisomies – with

fragmentation. So far, such an association has not

been established by chromosomal analysis of

large numbers of IVF embryos. The incidence of

ORIGINS AND CONSEQUENCES OF FRAGMENTATION



HPE_Chapter06.qxp 7/13/2007 4:46 PM Page 57

HUMAN PREIMPLANTATION EMBRYO SELECTION

aneuploidy (single chromosome loss or gain) does

not seem to be related to embryo morphology, but

to maternal age; the predominant abnormalities

among fragmented embryos are polyploidy and

extensive mosaicism which are often incompatible

with development to term.

34–37

The question of a potential relationship between

early embryo morphology and structural abnor-

malities in fetuses should perhaps be more carefully

examined in larger series of cases where several

morphological parameters have been well defined

and thoroughly examined. In the study of Ebner

et al.,

the embryos were evaluated by one static

observation on day 2 of development, at a magnifi-

cation of ⫻200.

We have not systematically studied neonatal

outcome following transfer of fragmented embryos.

Although it has been suggested that embryo mor-

phology on the day of transfer does not predict first

trimester pregnancy loss,

our data indicate a high

incidence of early pregnancy loss in cases where

extensively fragmented embryos were transferred.

Among 43 cases in which all embryos had ⬎35%

fragmentation at the time of evaluation (before

fragment removal), 13 pregnancies resulted, but

only three babies were born (two male and one

female). Seven pregnancies were classified as bio-

chemical, having shown at least three consecutive

rises in ␤hCG levels but having failed to show a

gestational sac with or without fetal heart activity.

Three other pregnancies were lost between 7 and 9

weeks of gestation, after detection of a gestational

sac. This is an early loss rate of 77% in an IVF pro-

gram where the overall loss rates after a positive

␤hCG pregnancy test and detection of cardiac

activity are approximately 25% and 11%, respec-

tively. Such an extraordinarily high loss rate is an

argument against transfer of embryos with exten-

sive fragmentation (as also argued in ref. 33).

Cytogenetic analysis was not performed on the

aborted fetuses, so it is not known for certain

whether chromosomal or other abnormalities were

the cause of spontaneous abortion. However, such

abnormalities have been found in 50–65% of very

early pregnancy losses following IVF.

39,40

Moreover,

transcervical embryoscopy (direct visualization of

the fetus) combined with cytogenetic analysis of

missed abortions has shown chromosomal abnor-

malities in 75%, and normal karyotype but structural

defects in 18% of the cases.

The observed incidence of implantation failure

of embryos with extensive fragmentation reflects a

persistent negative effect of fragmentation on

embryo development in utero associated with frag-

mentation. The question is whether these effects are

manifested during prolonged culture in vitro?

THE IMPACT OF FRAGMENTATION ON

PREIMPLANTATION DEVELOPMENT

The relationship between embryo morphology

and viability has become more evident following

assessment of embryos in elective single embryo

transfers.

It is also of interest that a large study

including a data set of 10 000 embryo transfers

recently concluded that embryo quality (assessed on

day 2) is the best predictor of pregnancy (following

day 2 transfer) even when considered together with

16 other possible treatment cycle variables that

included maternal age, duration and type of infer-

tility, ovarian stimulation protocol, number of IVF

attempts, progesterone level at hCG administration,

sperm count, motility and morphology, number of

retrieved and mature oocytes, number of embryos,

and number of transferred embryos. The highest

embryo quality score was assigned to 4-cell embryos

with no fragmentation or fragmentation ⬍20% and

evenly sized cells. Thus the suggestion that embryo

morphology, including fragmentation, has no cor-

relation with blastocyst formation or quality during

extended culture

43,44

seems both counter intuitive

and generally unsupported.

Our assessment of the impact of fragmentation

on blastulation in over 1200 surplus embryos

showed that with increasing fragmentation (0–15%

vs ⬎15%), fewer embryos compacted, cavitated,

and formed normal blastocysts. When fragmenta-

tion exceeded 35%, all processes were severely

compromised, even though in some cases, fragments



HPE_Chapter06.qxp 7/13/2007 4:46 PM Page 58

ORIGINS AND CONSEQUENCES OF FRAGMENTATION

were excluded in the perivitelline space by the

embryo (see Figure 6.4). When the origin of all the

blastocysts that were selected for transfer (i.e. the

best available embryos) was considered, we found

that nearly 90% had developed from embryos

with ⱕ15% fragmentation on day 3.

Interestingly, fragmentation has been found to

influence allocation of cells during differentiation.

According to this study, increasing fragmentation

resulted not only in reduced blastocyst formation

but also in lower cell numbers when blastocysts did

form. Moreover, in the case of minimal to moderate

fragmentation, the reduction in cell number was

apparently confined to the trophectoderm, while a

steady number of inner cell mass cells was main-

tained. However, when fragmentation exceeded

25%, cell numbers in both lineages were reduced.

These results are difficult to reconcile with the

suggestion that blastocyst viability on day 5 is not

affected by the degree of fragmentation on day 3 of

development.

Perhaps this can be partly attributed

to the relatively small groups of patients and

embryos examined.

In the case of embryos with

⬎25% fragmentation, for instance, only five blasto-

cysts (two of which implanted) and possibly two

patients were included in the study.

Thus the

study’s conclusion cannot be considered definitive.

Observational studies also suggest that the pat-

tern of fragmentation has an influence on blastula-

tion. The presence of large scattered fragments in

embryos with uneven cell size is associated with a

significant reduction in normal blastocyst forma-

tion compared with the other patterns.

The

reduced ability to form morphologically normal

blastocysts often becomes obvious at compaction,

which normally occurs on day 4 of development in

the human.

‘Regional’ or partial compaction,

with exclusion of a number of cells and fragments

from the morula, occurs frequently in fragmented

embryos (Figure 6.4).

In our study,

nearly half of normally compact-

ing embryos formed blastocysts, but this incidence

was reduced to only one third in regionally com-

pacted embryos and to 10% in embryos that did

not show compaction on day 4 in culture. These

data demonstrate that evidence of compaction on



ABCDE

FGHI

Figure 6.4 Serial 4 micron thick optical sections through a day 5 human embryo, obtained on the laser scanning confocal

microscope, showing ‘regional’ compaction. The excluded cells and fragments (A)–(E) as well as a small morula (M in (I)) are visible.

The embryo has been stained with antibodies against E-cadherin (green). The nuclei are stained with propidium iodide (red).

HPE_Chapter06.qxp 7/13/2007 4:46 PM Page 59

HUMAN PREIMPLANTATION EMBRYO SELECTION

day 4 is highly prognostic for normal blastocyst

formation.

The relationship between fragmentation pat-

tern and blastocyst formation was assessed in

another study,

in which the progress of 1566

surplus embryos was monitored until day 5 of

development. Embryos with no fragmentation, or

with type I fragmentation

had a higher blastocyst

formation rate than embryos with type II or III

fragmentation, while embryos that were assessed

to have type IV fragmentation failed to form any

normal blastocysts.

Based on the fragmentation patterns described

by Antczak and Van Blerkom,

the majority of

cleavage stage embryos with types II (multiple layers

of fragments with cell size reduction) or III (com-

plete fragmentation of one blastomere) fragmenta-

tion continued to divide beyond the 4-cell stage,

and about half reached the blastocyst stage, albeit in

some cases with a delay of 12–24 hours. It was also

observed that fragmentation at the 8-cell stage did

not preclude development to the blastocyst stage,

suggesting that the occurrence of fragmentation

early in development was more likely to be detri-

mental to development. How does fragmentation

interfere with normal embryonic development in

vitro or in vivo?

THE IMPACT OF FRAGMENTS ON THE

FRAGMENTING BLASTOMERE AND

ITS SISTER BLASTOMERES

Cytoplasmic blebs that emerge often during the

process of division can become ‘reincorporated’

into blastomeres.

19,49

Furthermore, evidence sug-

gests that small apical fragments may lyze, while

larger fragments may swell and burst, possibly due

to depletion of mitochondria and ATP stores in

these structures.

However, most fragments, once

they are clearly formed, obviously persist through-

out the early cleavage stages and it would be incor-

rect to imply that all fragments are transitory

structures.

Fragments may be excluded during compaction

or even found in the blastocoel cavity of the

blastocyst. According to one study, they are not

static and can ‘move in concert’ with the underlying

blastomeres.

Clearly, fragmentation can affect the size of the

fragmenting blastomeres, albeit not always appre-

ciably. It has been shown that mean blastomere size

decreases significantly with increasing degree of

fragmentation.

Highly fragmented embryos show

a 43–67% reduction in blastomere volume. It is

possible that the reduction in size, if substantial

can lead to arrest of the affected blastomeres.

Examination of individual ‘cells’ from fragmented

embryos suggests that ‘cells’ ⬍45 ␮m in diameter

in day 2 embryos and those ⬍40 ␮m in day 3

embryos should be considered fragments since they

never contain a nucleus and only seldom contain

chromosomal DNA.

It has been suggested that certain patterns of

fragmentation can result in ‘partial or near total

loss’ of several regulatory proteins from specific

blastomeres, with developmental consequences for

both the affected blastomere and the embryo as a

whole.

The products of cell lysis or degeneration might

cause deterioration of neighboring blastomeres or

interfere otherwise in normal embryo development.

This appears to be the case in the mouse, since the

presence of deliberately lyzed cells among other

viable cells reduces the incidence of blastocyst

hatching, and removal of the lyzed cells restores

hatching ability.

At the same time, preliminary

evidence suggested that removal of cryo-damaged

cells from frozen-thawed human embryos was not

only feasible but potentially beneficial.

This tech-

nique has since been shown in clinical trials to

improve pregnancy outcome.

53,54

We investigated the question of fragment ‘toxic-

ity’ in a mouse model,

taking advantage of mouse

blastomere totipotency at first cleavage.

Two-cell

embryos were dissociated (generating two blas-

tomeres), and each blastomere (or half-embryo)

was placed in a host zona pellucida (ZP; mouse,

bovine, or human) either alone or aggregated with

mouse or human egg/embryo fragments. The

development of the half embryo-fragment aggre-

gates and the control half embryos was monitored



HPE_Chapter06.qxp 7/13/2007 4:46 PM Page 60

ORIGINS AND CONSEQUENCES OF FRAGMENTATION

over the course of the following 3–4 days in

culture. The fragments were not spontaneously

generated by the half embryos, and therefore the

experiments measured the impact of fragments or

the microenvironment created by their presence

on division of the original single blastomere (or

half embryo) and its descendent cells. The pres-

ence of large numbers of fragments (in bovine or

human ZP) appeared to be disruptive to develop-

ment in some cases, but the overall results suggest

that at least in the absence of an intact ZP, mouse

blastomeres are neither overtly sensitive to close

physical association with heterologous (often

degenerative) cytoplasmic fragments nor respon-

sive to the intrazonal microenvironment created

by them (Figure 6.5).

The loss of fragments through large slits in the

zona pellucida presented a problem in these experi-

ments and may have affected the outcome, and

therefore the results should be interpreted with

caution. However, mouse embryos are not nega-

tively affected by exposure to cytoplasts, either

removed from the zygote and reinserted in the

perivitelline space, generated by enucleation of two

blastomeres of an 8-cell embryo,

or by enucleation

of one or two blastomeres of a 2-cell embryo.

Thus,

in the mouse, there is no clear evidence of fragment

‘toxicity’ except under extreme circumstances.

In the human, the picture is somewhat different.

Clinical data suggest that removal of fragments

from some fragmented embryos prior to intrauter-

ine transfer leads to higher frequency of implanta-

tion. In the past 10 years, we have routinely used

microsurgery to remove the majority of cytoplasmic

fragments when fragmented embryos were to be

transferred. This was initially based on finding a 4%

overall increase in implantation rate when zona

drilling (assisted hatching) and fragment removal

were applied simultaneously, in comparison with

embryos which were zona-drilled only.

The analy-

sis of a larger series of cases provided more support

for the suggestion that the removal of the fragments

contributes to the survival of fragmented embryos

following intrauterine transfer: embryos with

10–35% fragmentation showed similar implanta-

tion rates, and a very low implantation rate (6%)

occurred only when fragmentation exceeded ⬎35%

before fragment removal.

In a more recent study,

327 non-donor cycles

in which embryos with ⬎10% fragmentation were

subjected to assisted hatching and microsurgical

fragment removal prior to transfer were retrospec-

tively evaluated. Three groups were identified:

74 cycles in which at least one embryo required

fragment removal, 39 cycles in which all embryos

were subjected to this procedure, and a control

group of 234 cycles that included embryos with

⬍10% fragmentation subjected to assisted hatching

only. The data showed that the rates of implanta-

tion, live birth, spontaneous abortion, and fetal

abnormalities in the first two groups were equiva-

lent to those in the control group, suggesting a

beneficial effect of fragment removal on pregnancy

outcome.

Thus it appears that the reduced viability of frag-

mented human embryos is at least partly attributa-

ble to the presence of the fragments per se. However,

the beneficial effects of fragment removal are obvi-

ously limited. A recent analysis of a larger set of

homogeneous transfers (described above) suggests

that even after removal of fragments, the implanta-

tion rate of embryos with ⬎15% but ⬍35% frag-

mentation is still lower than that of embryos with

0–15% fragmentation (32% vs 18%, respectively).

Moreover, removal of fragments from embryos with

minimal fragmentation (0–15%) or from those in

which fragmentation exceeds 35%, has little or no

influence on the transfer outcome in a majority of

cases. In the case of 0–15% fragmentation, viability is

not substantially reduced in the first place, and in

the case of ⬎35% fragmentation even though removal

of fragments may occasionally lead to survival of the

affected embryo (as mentioned above), it is unlikely

to address the underlying cause of the abnormality

or its associated developmental problems.

Further insight into the possible impact of frag-

ments on the neighboring cells was provided by

examining the developmental capacity of blasto-

meres isolated from fragmented embryos. This was

done either by individual culture of these

blastomeres

61,62

or their artificial aggregation in a

host zona pellucida.



HPE_Chapter06.qxp 7/13/2007 4:46 PM Page 61

HUMAN PREIMPLANTATION EMBRYO SELECTION



Day 2

Aa3a4a5

Bb3b4b5

Cc3c4c5

Dd3d4d5

Ee3e4e5

Day 3 Day 4 Day 5

Figure 6.5 (Continued)

HPE_Chapter06.qxp 7/13/2007 4:46 PM Page 62

In the first series of experiments, nucleated

blastomeres and fragments were biopsied from

discarded embryos with 1–10 cells and 20–75%

(primarily type IV) fragmentation on day 3 of

development. Roughly 40% of these arrested when

placed in culture in isolation but the remaining 60%

divided during the course of culture. Of those that

divided, a substantial proportion went on to form a

‘cavity’ and blastulate, albeit with fewer than four

cells. These observations suggest that the mixing of

potentially normal and abnormal cells and cell frag-

ments may reduce the development potential of the

normal cells by reducing the likelihood of normal

blastulation.

The second set of experiments lends support

to this proposal: non-viable embryos with exten-

sive fragmentation and other abnormalities were

disaggregated and blastomeres of normal appear-

ance from two or more embryos were combined

into ‘chimeric’ aggregates.

Roughly one-third of

these aggregates formed blastocysts with distinct

inner cell masses and relatively high cell numbers,

ranging from 31 to 56 cells (Figure 6.6). Of these

cells, 52–90% were diploid. Chaotic mosaicism was

the most common abnormality found in these

embryos. This outcome demonstrates the develop-

mental potential and regulatory capacity of a pro-

portion of cells derived from non-viable embryos,

again suggesting that poor development of the latter

is at least partly attributable to the presence of the

fragments and/or abnormal nucleated cells. Does

the presence of fragments influence other aspects of

early development?

THE IMPACT OF FRAGMENTS ON CELL–CELL

INTERACTIONS/EMBRYO ORGANIZATION

It has been speculated that disruption of the spatial

arrangement of cells through fragmentation may

be a cause for reduced cell–cell contact and com-

munication.

Transmission electron microscopy

does not support this proposal. According to Van

Blerkom et al.

(2001), although ‘lysed, ‘prelytic’

and intact fragments interposed between blasto-

meres seemed to prevent close apposition of adja-

cent plasma membranes’.

Serial section analysis of

images ‘demonstrated that these separated zones

were focal and that close contact between opposed

plasma membranes existed in other regions of these

blastomeres.’ Whether this is a general phenomenon

or one that is restricted to certain cases of fragmen-

tation remains to be seen.

It is nonetheless clear that during the course of

culture in vitro, fragmented human embryos

often fail to compact or they undergo abnormal

compaction, excluding a number of cells and frag-

ments (Figures 6.3–6.5).

We investigated this phenomenon further by

examining the localization of E-cadherin, a vital cell

adhesion protein, in a large number of non-viable

human embryos with normal and abnormal mor-

phology.

In all other mammalian species studied

so far, E-cadherin is actively relocated in the course

of embryogenesis.

64,65

Relocation first occurs at the

time of compaction, and involves the cells that

form the outer layer of the embryo. In these cells,

ORIGINS AND CONSEQUENCES OF FRAGMENTATION



Figure 6.5 The development of experimental and control half embryos from day 2 through day 5 in culture.In panel A, Cavitation

and blastulation appear to be abnormal in both control and the experimental half embryos. In panel B (second row from top), on

day 2 of development, five large fragments (arrowheads) aggregated with one blastomere are visible, while the sister blastomere from

the same 2-cell embryo is seen in a host zona pellucida alone. One day later, on day 3 of development (b3), the five fragments can be

seen surrounding three blastomeres (1⫻1/4 and 2⫻1/8). The control half embryo has exactly the same number of cells (1⫻1/4 and

2⫻1/8). On day 4 (b4), the fragments in the experimental half embryo have moved to one side of a compacted mass that has begun

to cavitate. The control half embryo has also compacted and cavitated but appears better organized. By day 5, both half embryos have

formed blastocysts. The fragments (arrows), now degenerated, are loosely attached to the experimental half embryo. Development of

control and experimental blastomeres depicted in panels C, D, and E is comparable. Scale bar is 50 ␮m.

HPE_Chapter06.qxp 7/13/2007 4:46 PM Page 63

HUMAN PREIMPLANTATION EMBRYO SELECTION

E-cadherin is transported from the cytoplasm to the

cell membrane in areas of contact between cells.

This relocation is a preliminary to the formation of

junctional complexes between the trophectodermal

cells, which are responsible for the integrity of the

blastocyst.

It is therefore reasonable to expect that frag-

mentation may be associated with the failure of

proper expression, localization, or distribution of

E-cadherin, all of which are required for normal

compaction. Laser scanning microscopy (LSM)

images suggest that the characteristic distribution

pattern of E-cadherin is perturbed and erratic in

abnormally cleaving human embryos (Figure 6.7).

Although it is not clear whether the erratic distribu-

tion is a cause or an effect of abnormal develop-

ment, including fragmentation, these disturbances

can nonetheless lead to failure of compaction,

which in turn leads to failed or abortive blastula-

tion. Moreover, the presence of non-interacting cells

and fragments and interactive cells in the same

embryo may contribute to the problem by disrupt-

ing cell signaling processes that are mediated by E-

cadherin.

It is therefore plausible that fragmentation,

regardless of its underlying cause, disrupts early

development via a mechanism that involves cell–cell

interaction or communication.



A⬘ B⬘ C⬘

Figure 6.6 Development of a human aggregate. (A–C) Three discarded non-viable embryos on day 3 of development. (A⬘–C⬘)

Dissociated cells of the respective non-viable day 3 human embryos in A–C, comprising mononucleated, multinucleated, and

anucleate blastomeres/fragments. (D) Eleven of the mononucleated cells from the three embryos were used on day 3 to construct an

aggregate. (E) Following 2 days of culture, the aggregate formed a blastocyst.

HPE_Chapter06.qxp 7/13/2007 4:46 PM Page 64

ORIGINS AND CONSEQUENCES OF FRAGMENTATION



Figure 6.7 The distribution of E-cadherin (green) in human embryos at different stages of development. (A) Hatched day 7 blasto-

cyst from a normally fertilized egg. The cells of the trophectoderm (both polar and mural) show an intense ‘belt’ of fluorescence

indicating localization of E-cadherin in the membranes. Cells within the inner cell mass (arrows in panels a and b) show diffuse cyto-

plasmic E-cadherin. (B) An embryo showing abnormal blastulation and ‘belt’ staining of trophoectoderm cell membranes; many

excluded cells and fragments which do not show any staining are visible (arrows in panels B, a, and b). (C) A day 5 human embryo

with excluded cells and fragments (arrows) and a small morula (M) consisting of about 6 cells. (D) Compacting 8-cell embryo show-

ing weak staining in areas of cell–cell contact (arrowheads). The images in A–D are projections of multiple 4–5 ␮m thick optical

sections obtained on the laser scanning confocal microscope. Images in a–d are single optical sections.

HPE_Chapter06.qxp 7/13/2007 4:47 PM Page 65