Elder K. Human preimplantation embryo selection

Подождите немного. Документ загружается.

HUMAN PREIMPLANTATION EMBRYO SELECTION

discuss all aspects of the field of immunology.

2,3

For

a more in-depth discussion of the immune system

with references to the primary literature, the reader

is directed to Paul’s Fundamental Immunology.

Briefly, the innate immune system is non-specific,

present before exposure to any pathogen, and pro-

vides the first line of defense against disease. Some

aspects of the innate immune system are maintaining

a mechanical barrier (e.g. the skin and mucous

membranes), a physiological barrier (e.g. temperature,

pH, and certain chemicals, such as lysozyme and

complement), a cellular barrier (e.g. macrophages,

neutrophils, and natural killer (NK) cells), and an

inflammatory barrier (e.g. acute-phase proteins). On

the other hand, the adaptive immune system is not

triggered until the organism receives an antigenic

challenge. Unlike innate immunity, adaptive immu-

nity displays four special characteristics: specificity for

antigen, diversity, memory, and self/non-self recogni-

tion. The adaptive immune system involves T and B

lymphocytes and four major classes of membrane-

bound proteins: membrane-bound antibodies on

B cells, T cell receptors, and class I MHC proteins,

class II MHC proteins. (Class I MHC proteins are also

involved in the innate immune system.) In addition,

antigen presenting cells such as macrophages and

dendritic cells are part of the adaptive immune sys-

tem. We have recently proposed an interesting theory

which suggests that the mammalian immune system

may have developed as an accidental consequence

of a mechanism whereby semi-allogeneic fetuses can

escape destruction by the mother. Thus, a mechanism

to distinguish self from non-self and to tolerate the

allogeneic fetus may have resulted in the ability to

recognize foreign pathogens. However, regardless of

whether this theory is correct, both the innate and the

adaptive immune systems are intimately involved in

the protection of the preimplantation embryo from

immunological destruction by the mother.

THE IMMUNE SYSTEM AND

PREIMPLANTATION DEVELOPMENT

After fertilization the resulting preimplantation

embryo undergoes a series of cleavage divisions

leading to blastocyst formation. Images showing the

stages of mouse preimplantation development are

shown in Figure 13.1. Outstanding images of human

preimplantation embryos can be found in reference

6. This cleavage process takes 4–5 days in the mouse

and 5–7 days in humans. The images shown in

Figure 13.1 are from differential interference con-

trast (DIC) microscopy and represent what the cli-

nician is used to seeing in the laboratory. However,

the DIC pictures are somewhat deceptive because

the surrounding zona pellucida is optically clear so

that the observer looks right through the zona pel-

lucida when using DIC microscopy. This point is

emphasized when one examines scanning electron



Oocyte Zygote

2-cell

8-cell

Morula Blastocyst

Figure 13.1 Images of a mouse oocyte and preimplantation

mouse embryos. The images are from differential interference

contrast microscopy.

HPE_Chapter13.qxp 7/13/2007 4:52 PM Page 156

IMMUNOLOGICAL ASPECTS OF EMBRYO DEVELOPMENT

microscope images of a mouse blastocyst with the

zona pellucida intact (Figure 13.2A) and with the

zona pellucida removed (Figure 13.2B). In many

ways the zona pellucida is a physical barrier to an

immunological attack on the embryo, in a similar

way that the skin is a barrier to invasion by pathogens

in the innate immune system.

THE ZONA PELLUCIDA

Extensive studies have been performed on the bio-

chemical composition of the zona pellucida in the

mouse,

8–10

and more recently in humans.

11,12

In the

mouse, the zona pellucida is composed of three pro-

teins, designated mZP1, mZP2, and mZP3. Recently

a fourth zona pellucida protein was discovered in

humans

and it was originally named ZPB, but has

more recently been renamed ZP4.

Thus, the four

human zona pellucida proteins are ZP1, ZP2, ZP3,

and ZP4. The existence of ZP4 in human oocytes

has been shown at both the mRNA level by using

reverse transcriptase polymerase chain reaction

(RT-PCR) and the protein level by using tandem

mass spectrometry.

There are very few data in the literature defining

the function of the four human ZP proteins. In

the mouse, it was thought for many years that the

carbohydrate moieties of mZP2 and mZP3 were

involved in sperm–egg binding and fertilization.

8,9

However, this view has recently been challenged,

analyzing data from knockout mice, and it is unclear

whether fertilization depends on carbohydrate–

carbohydrate interactions, carbohydrate–protein inter-

actions, protein–protein interactions, or some com-

bination of all three mechanisms.

The realization

that the zona pellucida is involved in fertilization has

led to the idea of developing a contraceptive vaccine

based on immunization with zona pellucida pro-

teins.

However, there is evidence that anti-zona

antibodies cause ovarian dysfunction and infertility,

so this approach needs more investigation before it

is ready for clinical trials.

Aside from a putative role in fertilization, the

zona pellucida has been used to evaluate human

embryos in ART clinics. Evidence has been presented

that surface morphology

and thickness variation,

which was originally proposed by Cohen et al

and

recently confirmed by Sun et al,

as well as amino

acid sequence of the ZP proteins

are related to

pregnancy outcome after ART procedures. In addi-

tion, a number of laboratories have engaged in par-

tial disruption of the zona pellucida, in a procedure

called ‘assisted hatching’ to increase the pregnancy

rate in ART. Disruption of the zona pellucida can

be accomplished by mechanical, chemical, or laser

methods. A recent meta-analysis of 23 randomized

trials involving 2668 women has shown that although

assisted hatching improves the odds of a clinical

pregnancy, there is no effect on live birth rates.

Thus, it seems that, in general, disruption of the

zona pellucida in vitro during the preimplantation

period does not affect pregnancy outcome after

transfer to the recipient mothers.

A further discussion of the zona pellucida as a

mechanical barrier to the destruction of the embryo

in vivo by the maternal immune system is in order.

The zona pellucida is a very loose matrix, and allows

quite large molecules to traverse it. For example, even

IgM antibody with a molecular weight of approxi-

mately 950 000 can traverse the zona pellucida. The

zona pellucida is, however, protective against attack by

immune cells, as depicted in Figure 13.3. Figure 13.3

shows that molecules can traverse the zona pellucida,



Figure 13.2 Scanning electron microscope images of a mouse

blastocyst with the zona pellucida intact (A) and the zona

pellucida removed (B). Reprinted with slight modification and

permission from reference 7.

HPE_Chapter13.qxp 7/13/2007 4:52 PM Page 157

HUMAN PREIMPLANTATION EMBRYO SELECTION

but immune cells from both the innate and adaptive

immune systems are blocked from direct interac-

tion with the membranes of the embryonic cells. We

have shown this experimentally in an elegant series

of experiments in which cytotoxic T lymphocytes

(CTL) directed to major histocompatibility com-

plex (MHC) antigens were able to kill mouse blasto-

cysts with the zona pellucida removed, but not with

the zona pellucida intact.

These experiments not

only confirmed that the zona pellucida is a barrier

to immune cells, but also showed unequivocally that

preimplantation mouse embryos express MHC anti-

gens. To the best of my knowledge, similar studies to

determine the effect of the zona pellucida on possi-

ble killing of embryos by macrophages or natural

killer (NK) cells have not yet been reported. In sup-

port of the idea that the zona pellucida protects the

embryo from destruction by maternal immune cells

is a report that shows that the implantation rate for

human embryos with a small hole in their zona

pellucida was higher when the recipient mothers

were immunosuppressed.

At the end of the preimplantation period the

embryo hatches from the zona pellucida and implants

in the uterus. Thus, there is a period of time, between

hatching and implantation, when the protection of

the zona pellucida against immunological attack by

cells is not present. Due to the difficulty of design-

ing a good experimental system, very little research

has been done on the period of development just

after hatching and before implantation.

WHAT IS UNDER THE ZONA PELLUCIDA?

The real challenge in understanding the immuno-

logical aspects of preimplantation embryos is to

generate a topological map of all proteins that are

expressed under the zona pellucida, on the embry-

onic cell surface. In spite of the fact that we now

know that humans and mice have 20 000–25 000

genes in their genomes, very few of the proteins

encoded by these genes have been identified on the

embryonic cell surface. One problem is the large size

of preimplantation embryos, which makes proteins

that are expressed at low levels difficult to detect

because they may be far apart on the embryonic cell

surface. Immunofluorescence is a relatively insensi-

tive technique, because there is no enzymatic ampli-

fication step. Enzyme linked immunosorbent assay

(ELISA)

and Immuno-PCR

23,24

are more sensitive

procedures for detection of proteins on the embry-

onic cell surface. However, all of these protein detec-

tion techniques require antibody to the specific

protein of interest.

In contrast to protein detection, mRNA expres-

sion studies need no antibody and are relatively

simple to perform; DNA sequence data allow the

design of primers, and RT-PCR allows amplification

of specific mRNAs in single embryos. In addition,

the use of microchips has allowed the identification

of thousands of genes that are transcribed during

preimplantation mouse development.

25–29

However,

whereas the absence of mRNA precludes protein

expression, the presence of mRNA does not guarantee

protein expression. Therefore, the ability to detect as

many proteins on the embryonic cell surface as

possible would be highly desirable, in order to fully

understand the mechanisms that allow embryos to

escape surveillance by the maternal immune system.

Proteins of the MHC are of particular relevance to

this review.



CTL

NK cell

Molecules

Zona pellucida

EMBRYO

Figure 13.3 Schematic of the protection of the preimplantation

embryo from attack by immune cells: NK, natural killer cell; CTL,

cytotoxic T lymphocyte; M⌽, macrophage. The zona pellucida

is porous to molecules, including large macromolecules, but

impervious to attack by cytotoxic cells.

HPE_Chapter13.qxp 7/13/2007 4:52 PM Page 158

IMMUNOLOGICAL ASPECTS OF EMBRYO DEVELOPMENT

THE MAJOR HISTOCOMPATIBILITY COMPLEX

The MHC is a set of linked genes, located on human

chromosome 6 and mouse chromosome 17, which

encode three classes of proteins, class I, class II, and

class III. An up-to-date review of the MHC can be

found in a chapter by Margulies and McCluskey.

Simplified diagrams of the human MHC (HLA)

and mouse MHC (H-2) are shown in Figure 13.4. It

is clear from these diagrams that the MHC is not

orthologous between species, because there is no

one-to-one correspondence between the human

genes and the mouse genes. The MHC class I genes

are a major focus of this review. Figure 13.4 shows

that there are two classes of MHC class I genes, des-

ignated class Ia and class Ib. The MHC class Ia genes

are the ‘classical’ genes the protein products of which

are involved in both innate and adaptive immunity.

In humans, the MHC class Ia protein products are

HLA-A, HLA-B, and HLA-C. In the mouse, the

MHC class Ia protein products are H-2K and H-2D,

found in all mouse strains, and H-2L (encoded in

the D region) found in only some mouse strains.

The location of the K region in the mouse MHC,

displaced from the rest of the MHC class Ia genes,

has no known significance. In humans,all of the MHC

class I genes, classical (Ia) and non-classical (Ib), are

clustered at the telomeric end of the HLA complex.

The MHC class Ib genes differ significantly

between mouse and man. The protein products of

the MHC class Ib genes have functions in both the

innate immune response and other non-immuno-

logical venues. In humans, there are three MHC

class Ib genes that encode the protein products

HLA-E, HLA-F, and HLA-G. In the mouse, there are

15 known MHC class Ib genes in the Q region and

24 known class Ib genes in the TL region, with the

number of genes in each region varying among

mouse strains. Because there is no direct gene-to-gene

correlation between mouse and man (orthologs), it

is difficult to decipher which, if any, of these mouse

MHC class Ib genes are functional homologs with

human MHC class Ib genes. Of especial interest to

this review are Qa-2 protein encoded in the Q region

of the mouse MHC and HLA-G encoded in the

G region of the human MHC. Qa-2 and HLA-G

MHC class Ib proteins have been shown to be

functional homologs

31,32

and are discussed in detail

later in this review.

Interestingly, one consequence of deciphering the

complete genome sequences of mouse and man was

the realization that only 80% of mouse and human

genes are orthologs and the other 20% exist as func-

tional homologs. The MHC belongs to this latter

group of genes. The MHC is so important in the

understanding of the immune response that not

one, but two Nobel prizes have been awarded to

researchers working on the role of the MHC in the

immune response. The first, awarded in 1980 to

George Snell, Jean Dausset, and Baruj Benacerraf,

was for the discovery that the MHC was involved

in the immune response to foreign antigens. The

second, awarded in 1996 to Peter Doherty and Rolf

Zinkernagel, was for the discovery of the role of

the MHC in antigen recognition by T cells. However,

we now know that the MHC is also involved in

reproduction.

The MHC class I proteins are most directly rele-

vant to immunological rejection. Because the MHC

contains dozens of genes, for simplicity the whole

group of genes within an MHC on a single chromo-

some is called a ‘haplotype.’ It should be noted that

each class I MHC gene has many polymorphic vari-

ants in the population, leading to the existence of



Human – Chromosome 6

Region

Mouse – Chromosome 17

Class of

protein

II III B C E A G F

Ia Ia Ib Ia Ib Ib

HLA-G

Region

Class of

protein

K I S D Q TL

Ia II III Ia Ib Ib

Qa-2

II III

Figure 13.4 Schematic of the major histocompatibility complex

(MHC) genes of the human and the mouse.

HPE_Chapter13.qxp 7/13/2007 4:52 PM Page 159

HUMAN PREIMPLANTATION EMBRYO SELECTION

thousands of different haplotypes. During reproduc-

tion, each embryo receives one MHC haplotype of

maternal origin and one MHC haplotype of paternal

origin, as depicted in Figure 13.5. In this diagram

each haplotype is represented by a capital letter, A,

B, C, or D. As can be seen from Figure 13.5, all off-

spring are histoincompatible (semi-allogeneic) with

the mother, as stated by Medawar and quoted in the

opening paragraph of this review. An easy answer

to the lack of immunological rejection during the

preimplantation period would be that preimplanta-

tion embryos do not express MHC class I proteins.

However, this is not true.

MHC CLASS I PROTEINS ON

PREIMPLANTATION EMBRYOS

Initially there was a great deal of controversy about

whether or not preimplantation mouse embryos

express MHC class I proteins. It now seems that

these initial studies, based on using immunofluores-

cence, were simply not sensitive enough to detect

MHC class I proteins on preimplantation embryos.

The first clear demonstration of MHC class I pro-

teins on preimplantation embryos was by Searle

et al

using electron microscopy, and was later

confirmed by our group

using a similar electron

microscopic technique. A variety of techniques other

than electron microscopy have since been used to

demonstrate that MHC class I proteins are expressed

on preimplantation mouse embryos. These studies

include

125

I-lactoperoxidase labeling,

complement-

dependent cytotoxicity,

36,37

cell-mediated cytotoxi-

city,

ELISA,

22, 38–40

and Immuno-PCR.

23,41

Early studies on human preimplantation embryos

reported the absence of MHC class I proteins.

42,43

However, although this idea has persisted,

proper,

highly sensitive techniques have never been used

to systematically evaluate preimplantation human

embryos for MHC class I expression. Moreover, at

least one MHC class I protein, HLA-G, has been

shown to be expressed by preimplantation embryos.

In addition, human embryonic stem cells created

from human blastocysts have been shown to express

MHC class I proteins.

Therefore, it seems highly

likely that with the use of appropriate techniques

and experimental protocols, expression of MHC

class I proteins on human preimplantation embryos

will be found, as they are found on mouse pre-

implantation embryos.

MEMBRANE-BOUND VERSUS SOLUBLE MHC CLASS I

PROTEINS ON PREIMPLANTATION EMBRYOS

The big question, then, is why are MHC class I pro-

teins expressed by preimplantation embryos? The

answer to this question is entangled in the fact

that preimplantation embryos not only express

membrane-bound MHC class I molecules, but also

express soluble MHC class I molecules. Of particu-

lar clinical relevance are the recent reports from

ART clinics that preimplantation human embryos

produce soluble HLA-G and that the presence of

soluble HLA-G enhances the chance of pregnancy

success.

47–52

However, at least two groups have been

unable to detect HLA-G in the supernatants of pre-

implantation human embryos.

53,54

These provoca-

tive findings have been the subject of debate, and are

addressed in a recent issue of Molecular Human

Reproduction with comments from the Editor-in-

Chief

and an excellent editorial by Sargent.

The

clinical relevance of a test for soluble HLA-G in

preimplantation embryo supernatants and its role

in preimplantation embryo development and

pregnancy outcome after ART

is reviewed in



A/C

A/D B/C B/D

Mother

A/B

Father

C/D

Figure 13.5 Inheritance of MHC haplotypes. Hypothetical

haplotypes are represented by A, B, C, and D. Each embryo

receives one haplotype of maternal origin and one haplotype

of paternal origin.

HPE_Chapter13.qxp 7/13/2007 4:52 PM Page 160

IMMUNOLOGICAL ASPECTS OF EMBRYO DEVELOPMENT

Chapter 12 of this book, and will not be pursued

further here. In addition, a thoughtful and complete

review of the role of both membrane-bound and

soluble HLA-G in human reproduction has recently

been published.

Whether or not soluble HLA-G turns out to be a

clinically relevant marker for the prediction of preg-

nancy success, the existence of soluble isoforms of

MHC class I proteins is of direct relevance to the

embryo’s escape from destruction by the maternal

immune system. MHC class I proteins other than

HLA-G are known to exist in soluble form, but to

my knowledge no attempt has been made to assay

for the presence of soluble HLA-A, HLA-B, or HLA-C

in the supernatants of preimplantation embryos.

This is consistent with the previous discussion

about the lack of rigorous experiments to test for

the presence of membrane-bound MHC class I

molecules other than HLA-G on preimplantation

human embryos. As pointed out previously, the zona

pellucida protects the preimplantation embryo from

direct contact and killing by cytotoxic cells. Moreover,

the zona pellucida is porous and allows the exchange

of macromolecules between the embryo and its

environment, whether that environment is the cul-

ture medium in the ART clinic or the environment

in the uterus. Several reports from animal models

indicate that soluble MHC class I molecules sup-

press the immune response and prolong the survival

of tissue allografts.

58,59

(The embryo is an allograft,

as shown in Figure 13.5.) Several reports in human

systems have also shown that binding of soluble

MHC class I proteins to cytotoxic T lymphocytes

(CTL) induces them to undergo apoptosis.

60,61

addition, it has been shown that soluble HLA-G

inhibits the activity of both CTLs (CD8⫹ cells) and

natural killer (NK) cells, prevents proliferation of

helper T cells (CD4⫹ cells), and produces immuno-

logical tolerance in dendritic cells.

62–66

Therefore,

cells of both the innate and adaptive immune systems

are inhibited by soluble MHC class I molecules. This

may be an important mechanism in protecting the

embryo from attack by the maternal immune system

during the window of time between hatching and

implantation, when the zona pellucida no longer

presents a physical barrier to attack by immune cells.

THE PED GENE

One of the major criteria used to determine which

embryos should be transferred to the mother after

ART procedures is the embryo’s rate of develop-

ment.

The transfer of faster developing embryos

generally leads to a greater chance of pregnancy

success than does transfer of slower developing

embryos. The rate at which preimplantation embryos

cleave is dependent on both environmental and

genetic parameters. The Ped (preimplantation embryo

development) gene, which was discovered in my

laboratory,

has a profound influence on the rate of

development of preimplantation mouse embryos.

Of particular relevance to the topic of this review is

that the Ped gene product is an MHC class Ib pro-

tein, Qa-2 in the mouse and HLA-G in humans, and

these proteins are encoded in the Q region of the

mouse MHC and the G region of the human MHC

(Figure 13.4). Embryos that express Qa-2/HLA-G

on their cell surface cleave at a faster rate than

embryos that do not express these proteins, and

therefore are more likely to lead to a successful preg-

nancy. As discussed previously, it is difficult to

image MHC class I proteins on preimplantation

embryos because of their low concentration per sur-

face area. However, we have recently succeeded in

visualizing the Ped gene product, Qa-2 protein, on

the mouse blastocyst cell surface by immunogold

labeling, as shown in Figure 13.6 (Newmark and

Warner, unpublished). Based on the analysis of

representative electron microscopic samples, we

estimate that there are approximately 2000 Qa-2

protein molecules on the surface of a typical mouse

blastocyst. This is the first direct demonstration of

Qa-2 protein on the embryonic cell surface and

emphasizes that even a small amount of cell-surface

protein can have a very big effect on development

and reproduction.

MECHANISM OF ACTION OF THE PED GENE

The major challenge in elucidating the role of

the Ped gene in reproduction is to understand how the

presence of a cell surface molecule signals that the



HPE_Chapter13.qxp 7/13/2007 4:52 PM Page 161

HUMAN PREIMPLANTATION EMBRYO SELECTION

preimplantation embryo should increase its cleavage

rate. We have shown that Qa-2 must be expressed

on the cell surface in order for the protein to medi-

ate a rapid rate of preimplantation development.

This is based on the fact that enzymatic cleavage of

Qa-2 from the embryonic cell surface slows the rate

of preimplantation development.

Neither Qa-2 nor

HLA-G reach far enough into the inside of the cell

to act independently as signaling molecules. Qa-2 is

attached to the outer leaflet of the cell membrane by

a glycosylphosphatidylinositol (GPI) linkage, whereas

HLA-G traverses the cell membrane but has a trun-

cated cytoplasmic tail of only six amino acid residues.

Therefore, Qa-2 and HLA-G must bind to accessory

molecules(s) to transmit a signal from the cell sur-

face to the inside of the cell. We also know that both

Qa-2 and HLA-G are located in lipid rafts,

imply-

ing a signaling role for these molecules.

Although we do not yet know the nature of the

Qa-2/HLA-G transmembrane binding partner(s),

we have recent evidence that the pathway to activa-

tion is via phosphatidylinositol-3 (PI-3) kinase (De

Fazio and Warner, unpublished). Figure 13.7 shows

a hypothetical scheme for Qa-2/HLA-G activation

of cells, based upon using cross-linking of Qa-2 on

T cells to induce proliferation of the cells (unpub-

lished results). In these experiments we showed that

cross-linking of Qa-2 with anti-Qa-2 antibody, in

the presence of a second signal provided by phor-

phol myristate acetate (PMA), induces activation of

PI-3 kinase and phosphorylation of Akt. In addi-

tion, we have shown that Fyn co-immunoprecipi-

tates with anti-Qa-2 antibody, implying that Qa-2

and Fyn are in close physical proximity. This sug-

gests, but not yet proves, that Fyn is involved in the

activation scheme. Thus, we are well on the way

towards understanding the pathways by which Qa-2

enhances cell cycle progression.

PRESENCE VERSUS ABSENCE OF QA-2/HLA-G

AND EMBRYO SURVIVAL

One curious aspect of the Ped gene product, Qa-2 in

the mouse, and HLA-G in humans, is that although

the absence of these proteins is compatible with

embryo survival, their presence enhances pregnancy

success.

57,69,70

The majority of mouse strains and

humans have genes that can potentially be trans-

cribed and translated to produce Qa-2/HLA-G pro-

teins. In humans about 3% of the population has a

‘null allele’ at the DNA level, which results in the

lack of HLA-G protein expression.

We have

recently shown that the genes encoding Qa-2 in the

wild mouse population are absent at a similar fre-

quency (Byrne, Jones, and Warner, unpublished).

We speculate that the non-expression of Qa-2/HLA-G

has been retained through evolution because it is

advantageous to have embryos that develop at a

slower rate under some circumstances. In order for

implantation to occur, the timing of cleavage

divisions of the preimplantation embryo and the

preparation of the uterus for implantation must be

coordinated. Embryos that develop at a slow rate

can wait for the uterus to catch up, but embryos that

have missed the opportunity to implant because they

are developing too rapidly will die. Therefore, it seems



200 nm

Figure 13.6 Image of Qa-2 on the surface of a mouse blastocyst.

The image is from transmission electron microscopy with

immunogold labeling of Qa-2 protein. The arrow points to the

gold particle that identifies the location of Qa-2.

HPE_Chapter13.qxp 7/13/2007 4:52 PM Page 162

IMMUNOLOGICAL ASPECTS OF EMBRYO DEVELOPMENT

that a range of developmental rates would be advan-

tageous for species with litters, so that at least some

embryos could implant. In humans the same argu-

ment can be made for successive reproductive cycles.

RECENT FINDINGS USING A MOUSE MODEL

TO STUDY THE PED GENE

Animal models provide an excellent opportunity for

studying the immunological aspects of early preg-

nancy. Many experiments can be carried out in ani-

mal embryos that cannot be performed on human

embryos.As mentioned previously, of the many pos-

sible animal models, the mouse is particularly valu-

able because of the availability of inbred, congenic,

transgenic, and knockout strains. Preimplantation

mouse embryos can be cultured in chemically defined

media, and research can thus be performed that is

directly comparable with human embryos that are

grown in vitro in the ART clinic. In the mouse, the

congenic strains B6.K1 and B6.K2 have been a valu-

able resource for studies on the Ped gene.

69,70

These

two strains differ genetically only in that B6.K1 lacks

the Qa-2 encoding genes, and these are present in

B6.K2. We recently tested whether the sex of pre-

implantation embryos is a confounding factor in

mediating the rate of preimplantation development

of B6.K1 and B6.K2 embryos, and found that sex

has no effect on the rate of development of embryos

from both B6.K1 and B6.K2 strains of mice.

Therefore, Qa-2 protein, and only Qa-2 protein (the

Ped gene product), is responsible for regulating

the rate of preimplantation development in the

B6.K1/B6.K2 mouse model.

Another interesting recent finding has an impact

on the mechanism that enables B6.K1 embryos to

survive without the presence of Qa-2 protein. We

cultured embryos from B6.K1 and B6.K2 mice from

the 2-cell to the blastocyst stage, and measured

the amount of platelet activating factor (PAF) in the



Thr

Ser

PMA

Akt

Qa-2

Hypothetical

transmembrane

signaling partner

Fyn?

p85

p110

PI-3 kinase

Cell cycle progression

Inhibition of apoptosis

Antibody to Qa-2

PIP

Figure 13.7 Schematic of the putative mechanism of Qa-2 signaling. After antibody-mediated cross-linking of two Qa-2 protein

molecules, Fyn may be activated, possibly via a transmembrane protein partner that interacts with Qa-2 on the cell exterior and with

Fyn inside the cell. Fyn can directly interact with the regulatory domain (p85) of phosphatidylinositol-3 (PI-3) kinase. PI-3 kinase

catalyzes the formation of phosphatidylinositol 3,4,5 triphosphate (PIP

), which results in phosphorylation (P) of a specific

threonine residue (Thr) in Akt. Phorbol myristic acetate (PMA) completes the activation of Akt by inducing phosphorylation of a

specific serine residue (Ser). Once activated, Akt promotes a number of events that support cell proliferation and inhibit apoptosis.

HPE_Chapter13.qxp 7/13/2007 4:52 PM Page 163

HUMAN PREIMPLANTATION EMBRYO SELECTION

culture supernatant. PAF is known to have a role in

many reproductive events, including ovulation, fer-

tilization, implantation, and parturition. We found

that B6.K1 embryos produce twice as much PAF as

B6.K2 embryos.

The conclusion is that B6.K1 pre-

implantation embryos may enhance their chance of

survival in order to compensate for the lack of Qa-2

expression by increasing PAF production. It seems

logical to hypothesize that embryonic expression of

molecules other than PAF may also be affected by the

presence or absence of Qa-2 protein. Such molecules

may be proteins, but may also be other molecules.

Studies to identify these other putative molecules are

underway in my laboratory, using the B6.K1/B6.K2

mouse model system. It remains to be determined

which, if any, of these hypothesized molecules are

involved in mediating embryo survival and pro-

tecting the embryo from immune destruction by

the mother.

CLINICALLY RELEVANT PROPERTIES

OF THE MOUSE PED GENE

Our basic research on the mouse Ped gene is ripe for

transition from laboratory to clinic. A large body of

literature on the properties of the Ped gene and its

protein product, Qa-2 in the mouse and HLA-G in

humans, has recently been reviewed.

5,57,69,70

The

original definition of the Ped gene phenotype was

its relation to rapid or slow development during the

preimplantation period. Thus, it is likely that selec-

tion of rapidly developing embryos in the ART

clinic will select for the ‘fast’ allele of the human Ped

gene, as well as for other as yet unidentified genes

that affect the rate of preimplantation development.

As shown in Table 13.1, there are many parameters

other than rate of preimplantation development

that are affected by the presence of Qa-2 protein in

mouse embryos. In Table 13.1, clinically relevant

parameters have been separated into the effects of

Qa-2 protein on preimplantation embryos and

effects of Qa-2 protein later in life. Some of the data

in Table 13.1 are based on our studies in the

B6.K1/B6.K2 mouse model system described above.

It is fascinating to note that the presence of the

Ped gene product, Qa-2 protein, has effects beyond

the preimplantation period into adult life. These

results are consistent with the ‘Barker Hypothesis’,

which states that growth of the embryo during the

preimplantation period affects health during adult-

hood, including effects on body weight, heart

disease, and blood pressure. Our recent study

has

indeed shown that both blood pressure and levels of

angiotensin converting enzyme (ACE) are lower in

B6.K2 (Qa-2 positive) mice compared with B6.K1

(Qa-2 negative) mice. It has been speculated that

perhaps the lifespan of B6.K2 mice is shorter than

the lifespan of B6.K1 mice, but this has not been

tested experimentally.

However, this speculation is

at odds with the Barker Hypothesis, so the potential

outcome of a longevity study on the B6.K1 and

B6.K2 mice is not clear. The overall conclusion is

that selection of fast-developing embryos for trans-

fer to the mother after ART may have health impli-

cations that are far beyond the preimplantation

period of development.

CONCLUSIONS AND FUTURE DIRECTIONS

I now return to the quote by Medawar at the begin-

ning of this article, in which he wonders why the



Table 13.1 Clinically relevant properties of the mouse

Ped

gene product, Qa-2 protein. Data are based

on the B6.K1 (Qa-2 negative)/B6.K2 (Qa-2 positive)

mouse model system

Effect Reference

Preimplantation effects of the presence of Qa-2 protein

Earlier ovulation Comiskey and Warner, unpublished

Earlier 1st cleavage Comiskey and Warner, unpublished

Faster rate of development Reviewed in 69 and 70

Lower levels of platelet 73

activating factor

Earlier implantation Comiskey and Warner, unpublished

Later effects of the presence of Qa-2 protein

Enhanced survival to birth Reviewed in 69 and 70

Higher birth weight Reviewed in 69 and 70

Enhanced survival to weaning Reviewed in 69 and 70

Higher weaning weight Reviewed in 69 and 70

Lower adult blood pressure 75

Lower adult levels of 75

angiotensin converting enzyme

HPE_Chapter13.qxp 7/13/2007 4:52 PM Page 164

IMMUNOLOGICAL ASPECTS OF EMBRYO DEVELOPMENT

allogeneic embryo is not rejected by the maternal

immune system. In this chapter I have shown that

not only do preimplantation embryos express allo-

geneic MHC proteins, but that the expression of at

least one of them, Qa-2 in the mouse and its homolog,

HLA-G in humans, enhances pregnancy outcome.

An interesting recent article

suggests that HLA-G

protein must be expressed at just the right level to

enhance reproduction, too much HLA-G or too lit-

tle HLA-G are detrimental to reproductive success.

The validity of this hypothesis awaits testing in a

clinical setting. It also remains to be determined if

all human embryos that express cell surface HLA-G

also produce soluble HLA-G. It has been shown that

embryos that express cell surface HLA-G cleave

faster than embryos that lack cell surface HLA-G,

but whether the rate of development is correlated

with the amount of soluble HLA-G found in the

supernatant of the cultured embryos has not yet

been rigorously tested.

The proteins encoded by the MHC are only one of

many classes of proteins involved in the immune

response. It has been estimated that perhaps 10% of

the 20 000–25 000 human genes are involved in

immune function. Therefore, the challenge is not

only to identify these proteins on the surface of the

embryo and in culture supernatants from preimplan-

tation embryos, but also to ascertain the amount of

protein that is indicative of a healthy embryo. In

addition, diagnostic methods could also be applied to

the parents donating eggs and sperm in the ART

clinic. For example, it has been shown that there is an

association between the mother’s HLA-G genotype

and success of IVF and pregnancy outcome.

In the

future it may be possible to determine the genotype

of both parents and to predict which couples will

have the greatest chance of a successful outcome after

IVF. This approach should apply to all genes and not

only to those involved in the immune system.

SINGLE EMBRYO TRANSFER

In the United States about 1% of all births result

from ART procedures, with the percentage reaching

2–3% in many European countries and in Australia.

At the present time there is a worldwide effort towards

single embryo transfer in order to reduce the fre-

quency of multiple births, which are detrimental to

both the mothers and the babies. In the United

States the multiple birth rate is 35% after ART com-

pared with 1% after natural conception. A recent

paper

presents data on the first prospective study

in which single embryo transfer on day 3 (cleavage

stage embryos) is compared with embryo transfer

on day 5 (blastocyst stage embryos): the delivery

rate was 22% for day 3 transfer, and 32% for the day 5

transfer. The embryos in this study were all pro-

duced by women under 36 years of age, thereby

excluding half of the women who visit the ART

clinic. Clearly future prospective studies need to be

performed on women in the older age group. The

fact that even the best procedure gave only a 32%

success rate emphasizes the issue of identifying

which embryos should be transferred to the mother.

In this study, only morphological methods (DIC

microscopy) were used for embryo evaluation on

both day 3 and day 5, and the number of blastomeres

(cells) and the degree of fragmentation were evalu-

ated. Day 5 embryos (blastocysts) were also evalu-

ated on the basis of size of the inner cell mass (ICM),

and the degree of expansion of the blastocoel cavity.

It seems clear that if additional data were available,

such as the nature of the molecules secreted by these

embryos, the methods by which embryos could be

selected for single embryo transfer could be greatly

enhanced.

BETTER REPRODUCTION THROUGH CHEMISTRY

How can we identify the proteins on the surface of

preimplantation embryos and those secreted into

the culture medium? After identification, how can

we quantitate the levels of these proteins and corre-

late the levels with embryo health? How can we

determine the genotype of the embryos and their

parents? Clearly new methodologies for genomics

and proteomics are needed. There is growing inter-

est among chemists in the application of nanotech-

nologies to biomolecular detection and medical

diagnostics.

The long range goal is to provide

personalized medicine based on the screening of

biological fluids such as blood and urine. It seems



HPE_Chapter13.qxp 7/13/2007 4:52 PM Page 165