TenCritical
CriticalTopics
Topicsin
in
Ten
ReproductiveMedicine
Medicine
Reproductive
Produced
the
Science
/AAAS
Produced
byby
the
Science
/AAAS
Custom
Publishing
Office
Custom Publishing Office
Sponsored
Sponsored
by by
Sponsored
Supplement
Science
AA
Sponsored
Supplement
toto
Science
Produced by the Science/AAAS Custom Publishing Office
2
3
4
8
The Science of ART
Sean Sanders
From the Chief Executive of the Serono Symposia
International Foundation Rachel Clark
“Top Ten” Conference Promotes Good Science
in Human Reproduction
HISTORY AND PERSPECTIVE
Reproductive Medicine Before and After the Nobel
J.L.H.EversandAntonioPellicer
10
Germline Stem Cells in Adult Mammalian Ovaries
13
Bidirectional Communication Between the Oocyte and
Surrounding Somatic Cells is Required for Successful
Follicular Development and Oocyte Maturation
DoriC.WoodsandJonathanL.Tilly
JiaPeng,JohnJ.Eppig,andMartinM.Matzuk
15
An Introduction to Human Embryo Development
18
Environmentally Induced Epigenetic Transgenerational
Inheritance of Disease
21
23
CONTENTS
ReneeA.Reijo-Pera
MichaelK.Skinner
Aberrant Endocannabinoid Signaling is a Risk Factor
for Ectopic Pregnancy
XiaofeiSunandSudhansuK.Dey
Single Embryo Transfer in IVF: Live Birth Rates and
Obstetrical Outcomes
ChristinaBergh
AAAS is here – promoting universal science literacy.
In 1985, AAAS founded Project 2061 with the goal of helping all Americans become literate in science, mathematics, and
technology. With its landmark publications Science for All Americans and Benchmarks for Science Literacy, Project 2061 set out
recommendations for what all students should know and be able to do in science, mathematics, and technology by the time they
graduate from high school. Today, many of the state standards in the United States have drawn their content from Project 2061.
Every day Project 2061 staff use their expertise as teachers, researchers, and scientists to evaluate textbooks and assessments,
create conceptual strand maps for educators, produce groundbreaking research and innovative books, CD-ROMs, and professional development workshops for educators, all in the service of achieving our goal of universal science literacy.
As a AAAS member, your dues help support Project 2061 as it works to improve science education. If you are not yet a AAAS
member, join us. Together we can make a difference.
To learn more, visit aaas.org/plusyou/project2061
25
Oocyte Vitrification: A Major Milestone in Assisted Reproduction
27
Accurate Single Cell 24 Chromosome Aneuploidy Screening
29
What Is a “Normal” Semen Analysis?
32
Circulating Angiogenic Factors and the Risk of Preeclampsia
35
The Ovarian Factor: Cutting-Edge Basic Science Combined
with Classic Clinical Insights
AnaCobo
RichardT.Scott,Jr.andNathanR.Treff
DavidS.Guzick
S.AnanthKarumanchiandRaviThadhani
RichardS.LegroandMartinM.Matzuk
40
Infertility: The Male Factor
44
Molecular Interplay in Successful Implantation
DoloresJ.LambandLarryI.Lipshultz
JeeyeonCha,FelipeVilella,SudhansuDey,andCarlosSimón
This booklet was produced by the Science/AAAS Custom Publishing Office and sponsored by the Serono Symposia International Foundation. Materials
that appear in this booklet were commissioned, edited, and published by the Science/AAAS Custom Publishing Office and were not reviewed or assessed
by the Science Editorial staff. Articles in this booklet can be cited using the following format: [AUTHOR NAME(S)], [ARTICLE TITLE] in Ten Critical Topics in
Reproductive Medicine, S. Sanders, Ed. (Science/AAAS, Washington, DC, 2013), pp. [xx-xx]. DOI: 10.1126/science.342.6164.1393-c
The Serono Symposia International Foundation is a member of the European ‘Good CME Practice Group,’ proactively working with other medical education providers
to enhance the standards of medical education provision.
Editor: Sean Sanders, Ph.D.
Proofreader/Copyeditor: Yuse Lajiminmuhip; Designer: Amy Hardcastle
© 2013 by The American Association for the Advancement of Science.
All rights reserved. 13 December 2013
Cover design by the Serono Symposia International Foundation.
All Images © istockphoto.com.
1
Produced by the Science/AAAS Custom Publishing Office
The Science of ART
From the Chief Executive of the Serono
Symposia International Foundation
This publication
In 1978, Robert Edwards and Patrick
Improved genetic screening of embryos was also on the agenda,
At the Serono Symposia International Foundation (SSIF), we be-
ity andSterility, respectively, discuss
Our aim with the
provides a brief
thought impossible: the birth of a
tial penetrance and the fear expressed by some of the specter
best independent continuing medical education (CME) is avail-
believe reproductive medicine and
Top Ten project
Steptoe achieved what many likely
overview of the
baby conceived outside of a woman’s
conference, a
first test tube baby. This was not only
first of its kind
as far as format
body, commonly referred to as the
a breakthrough in biological research
and a significant advance in the
field’s understanding of human repro-
duction, but also a cultural and soci-
and intent is
etal leap that released woman (and
concerned.
reproductive vagaries of their bodies.
men) from being at the mercy of the
A solution for many forms of infertility
was now available, affording couples
previously unable to bear biological
children that opportunity. Although
controversial at the time, assisted re-
productive technology (ART) is now
seen as a standard instrument in the
toolbox at a doctor’s disposal to as-
of so-called designer babies. This publication provides a brief
overview of the conference, a first of its kind as far as format and
intent is concerned. We also present a review article from each
of the authors of the 10 original manuscripts, providing context
to their work in the current research landscape. The 10 articles
advances is this area was the mo-
tivation for a conference that took
place in September 2013 in Florence,
Italy, and ultimately this publication.
Ten previously published journal
articles, five on basic research and
five on clinical work, were chosen
sessment of ART following Edwards and Steptoe’s long-awaited
Nobel prize in Medicine in 2010. We close out the booklet with
three excellent reviews from leaders in reproductive medicine:
ing ART and how this might be ad-
dressed through new techniques.
of internationally renowned experts with audiences worldwide
through live educational activities and, increasingly, online.
TheTopTeninReproductiveMedicine project has been both
summer of 2012, we brought together an expert panel to select
la, Dey, and Simón, who together review the complex topic of
the molecular mechanisms involved in embryo implantation in
the uterus.
It is our hope that this booklet, made possible by the sponsor-
tion, will provide a broad review of the important and impactchallenging dogmas, and advancing new paradigms,
as seen from the perspective of conference
attendees and presenting authors. The
intention is not to stop here, but to
use this conference as a model to
continue the discussion in this
field as well as many others
that impact our lives and
well-being.
Custom Publishing Office
incidence of multiple births follow-
proach that SSIF takes to all of the areas we cover and the many
search has focused on the male role in infertility, and Cha, Vilel-
function and related treatments, Lamb and Lipshultz whose re-
in the field at this unique two-day
Topics covered included the high
This booklet focuses on one very important strand of our work
exciting and challenging. With so much groundbreaking research
Sean Sanders, Ph.D.
meeting.
for patients.
Legro and Matzuk who discuss disorders that affect ovarian
by a panel of experts to be presented, and challenged, by peers
sis and treatment mean new opportunities to improve outcomes
events we hold each year, sharing the knowledge and insights
ful subject of reproductive medicine, summarizing advances,
implications of new research and
the world, new research findings and improvements in diagno-
medicine journals and experts in this field, who provide their as-
tonio Pellicer, both editors in chief of prestigious reproductive
out its challenges, difficulties, and
sion of these hurdles as well as the
able to clinicians, scientists, and researchers. Every day across
in CME: reproductive medicine. It clearly exemplifies the ap-
ship and support of the Serono Symposia International Founda-
failings. An honest and open discus-
lieve that it is more important than ever to ensure that the very
are prefaced by a short perspective from J.L.H. Evers and An-
sist infertile patients.
However, ART has not been with-
2
including the challenges regarding testing for diseases with par-
Editor
Science/AAAS
worldwide, how could we identify just 10 papers? During the
the best papers from a shortlist of 50 produced by a special-
the exciting directions in which they
biology are heading. There also are
is above all
productive medicine. It all makes for a
to highlight
will enjoy. We sincerely appreciate
research that
featured in this booklet. It would not
has taken
reviews of key research areas in restimulating read, which we hope you
the contributions of all of the authors
have been possible without their
reproductive
their dedication and passion for
medicine
Carlos Simón, who was instrumental
forward to
concerted writing efforts or indeed
their work. In particular, Professor
in the conception and execution of
the conference and this booklet.
SSIF is committed to sharing medi-
ist independent bibliographer. The shortlist featured 25 papers
cal knowledge and debate beyond
based. The papers were chosen by the number of downloads
Ten event we wanted to reach out to
related to basic research and 25 that were clinical and fieldthey had received and the number of citations in the past 10
years. Our expert panel of eight leading figures in reproductive
medicine then spent many hours debating which would form the
final Top Ten, asking which papers had really challenged tra-
ditional views or which had actually changed diagnosis
and treatment?
Our aim with the Top Ten project is to highlight research that has taken reproductive
the benefit of
patients.
our educational events. With our Top
key opinion formers who might not be
aware of our work, and that is why
we are pleased to be able to provide
financial support for the publication
of this important educational book-
let with the assistance of Science/
AAAS.
medicine forward in benefiting patients.
Following a year of preparation, in September 2013, we welcomed an inter-
national audience to Florence, Italy to
hear from the top 10 authors them-
selves and challenge their findings.
Out of that two day event grew
Rachel Clark
the top 10 papers and the subse-
Serono Symposia International
this booklet, featuring synopses of
quent debates as well as contain-
ing so much more. Hans Evers and
Chief Executive Officer
Foundation
Antonio Pellicer, the editors in chief
of Human Reproduction and Fertil-
3
Produced by the Science/AAAS Custom Publishing Office
“Top Ten” Conference Promotes Good
Science in Human Reproduction
The “Top Ten in
Reproductive
Medicine”
conference
highlighted the
best papers as
an example of
how research
on human
reproduction
and infertility
should be done.
4
In September 2013, scientists and clinicians in the field of human reproduction participated in a groundbreaking
conference held in Florence, Italy. Over
the course of two days, authors of the
10 most cited, viewed, or downloaded
research articles in the field presented
their papers. Each author was confronted by a “challenger,” after which participants discussed the paper’s findings
and significance. The goal: to advance
rigorous science in a field dominated
by small-scale studies and trial and
error. “Everyone is aware of the clinical relevance of assisted reproduction
technology [ART],” said Carlos Simón,
professor of obstetrics and gynecology
at the University of Valencia, Spain,
and a conference scientific organizer.
“We need to increase our scientific relevance.”
In 2009, the World Health Organization officially recognized infertility as a
disease. Not all countries, however,
have followed suit. Consequently, largescale funding of fertility treatments has
been scarce. Regulation differs among
countries, making adoption of standard
practices difficult. And on the basic research front, strict embryo protection
laws in some countries restrict studies
on human embryos. “If you look at all
the things we do in the clinic, I would
bet half of them have never been proven efficient in multicenter trials,” said
Hans Evers, a professor in the department of obstetrics and gynecology at
Maastricht University Medical Centre in
Maastricht, The Netherlands. Intracytoplasmic Sperm Injection (ICSI) whereby
sperm is injected directly into an egg,
for example, is now being used to treat unexplained infertility in
about half of all treatment cycles, said Evers. “But no one has ever
shown that ICSI is necessary in unexplained infertility.”
ICSI was an advance initially developed to help men with poor
sperm quality become fathers when traditional in vitro fertilization
failed. The past 10 years have seen additional advances in technology aimed at helping greater numbers of people become parents,
too. Several of the most promising of these were discussed at the
conference. Embryo cryopreservation, for example, may make in vitro fertilization (IVF) safer and less stressful by enabling a woman to
undergo ovarian stimulation only once to retrieve eggs (See Cobo,
page 25). And embryo selection technology that can distinguish
healthy embryos from unhealthy ones promises to raise success
rates. Further in the future it will be possible to diagnose any number
of gene mutations in an embryo and harvest eggs from ovarian stem
cells so that older couples can have biological children (see Legro,
page 35 and Lamb, page 40).
However, the best application of these technologies should be debated before they come into widespread use, said Evers. Moreover,
some technologies pose ethical questions that require societal input.
For example, while diagnostic tests can detect the presence of a
gene, the tests may be meaningless since genes don’t predict with
100% certainty whether a person will actually develop a disease.
Also, if women beyond normal reproductive age can indeed one day
have biological children, then how old is too old? “You have to think
about these things before the possibility arises,” said Evers.
To start addressing these issues, The “Top Ten in Reproductive
Medicine” conference highlighted the best papers as an example
of how research into human reproduction and infertility should be
done. The format, inspired by Simón and developed by the Serono
Symposia International Foundation (SSIF), was the first of its kind
in the field. By discussing each paper—five clinical and five basic
research—at length, conference organizers hope to promote good
clinical practice that comes from adopting technologies and practices that are backed by strong scientific evidence and that have
been validated by ethical debate.
ASSiSTeD RePRODuCTiOn TeChnOlOgieS:
We CAn DO BeTTeR
IVF has changed little since Robert Edwards and Patrick Steptoe
introduced the technique over 35 years ago. While there have been
advances in the field, such as the
ability to freeze sperm or eggs
and test embryos for specific genetic abnormalities, the process
remains much the same: Fertilize egg with sperm, transfer the
resulting embryo into a woman’s
uterus, and hope for a full-term
pregnancy. The technique has
brought five million babies into the world and fulfilled the longing of
parenthood for couples struggling with infertility.
But IVF has a downside. The drugs given to stimulate the ovaries to
produce multiple eggs can overstimulate the ovaries and land women in the hospital, when the clinical condition is particularly risky. The
procedure is stressful, primarily because it’s hard to predict whether
it will work. Depending on the age of the eggs, the clinic, and the
procedure used, any embryo transfer cycle has between a 25% and
40% chance of resulting in a baby. With the odds in favor of failure,
doctors have typically transferred more than one embryo to boost the
chance of success. While the rate of triplets and higher order births
has come down drastically, the number of twins born from reproductive technologies remains at about 30% of all ART births in the U.S.
A twin or multiple pregnancy has long been known to raise the risk
of harm to both mother and babies. Maternal complications include
increased rates of pre-eclampsia, gestational diabetes, and preterm
labor. Risks for babies include low birth weight and prematurity as
well as a higher risk of long-term disability and death as compared
with singleton pregnancies.
WiTh neW TeChnOlOgieS, SuCCeSS iS MORe likely
At the conference, Christina Bergh from the Sahlgrenska University
Hospital in Gothenburg, Sweden presented her 2004 paper that for
the first time provided evidence that transferring a single embryo
yields almost the same birth rate as transferring two in women under age 36 (see page 23). Bergh’s paper was the first clinical paper
presented and set the backdrop for the presentations that followed.
It began with the question: How can we treat infertility and inflict the
least harm? The paper was among the most highly viewed over the
last 10 years, primarily because it was published when rates of multiple births using ART were much higher than they are today. Since
the paper was published, single embryo transfer (SET) has been
widely adopted, particularly in
Northern Europe. In Sweden,
doctors use SET in 75% to 80%
of all IVF cycles in women of any
age, said Bergh. But there remain several countries in which
it hasn’t caught on, she added,
and this is likely because of differences in reimbursement prac-
tices for reproductive procedures.
Bergh presented additional data showing that SET works well in
women up to age 42. Additionally, she studied children born from IVF
procedures and showed that at the same time, the rate of complications had also decreased drastically. “Her work has really pushed the
SET idea worldwide,” said her challenger Robert Fischer, medical
director of Fertility Center in Hamburg, Germany and SSIF Scientific Committee Member. However, Swedish statistics may not hold
for other populations, said a clinician from the audience. Patients in
Scandinavia opt for IVF treatments much more quickly than patients
in southern Europe, he added. Consequently, IVF patients in southern Europe tend to be older and have poorer outcomes. In such a
setting, SET may not be as successful. Another participant pointed
out that for SET to really take off, success rates would need to be
much higher.
eMBRyO SeleCTiOn AnD FReezing
One way to raise success rates would be to distinguish good embryos from bad ones. Enter Richard Scott’s 2010 paper on detecting aneuploidy in an embryo using whole genome amplification and
single nucleotide polymorphism arrays (see page 27). Scott showed
that the technology can distinguish abnormal from normal embryos
with a 4% error rate. Since publication, additional data shows that
the error rate is actually only around 1%, said Scott, director of reproductive science at Robert Wood Johnson Medical School in Basking
Ridge New Jersey. The paper was the first of a series of four that
together show that the technology gives clinically meaningful results,
Scott said.
For example, in one study, Scott and his colleagues biopsied embryos and transferred them back into women as part of their infertility
care, before they knew the biopsy results. Once they determined the
success or failure of the procedure, they then compared the biopsy
5
Produced by the Science/AAAS Custom Publishing Office
analysis data with birth outcomes and showed that the DNA array
technology could detect, with 99% accuracy, an abnormal embryo
that was unlikely to develop into a baby. Another study showed that
selecting embryos using this technology does indeed yield higher
birth rates. The predictive power of the technology is so great that
the Reproductive Medicine Associates of New Jersey fertility clinic
of which Scott is the director now performs SET using the diagnostic
technology in 70% of all IVF cycles, he said. The clinic also offers the
diagnostic service to other clinics.
The technology isn’t easy to implement, however, and is expensive compared with other technologies on the market, challenger
Carlos Simón pointed out, and no one has yet demonstrated that
the array technology yields better clinical outcomes as compared
with those alternatives that are cheaper and easier to use. Discussion also raised the point that array technologies may be rendered
obsolete in the near future by next generation technologies that can
sequence genes faster and cheaper. In fact, there are already companies selling tests that purport to predict the chance of a child inheriting certain traits based on sequencing specific parental genes.
Adapting such tests to screen embryos for specific traits would be
a simple task—the prospect of which has raised more than a few
eyebrows. Nevertheless, technologies to parse normal embryos
from abnormal ones are a boon to the field because they raise success rates. Scott’s clinic is seeing pregnancy rates of greater than
50%, he said, and preliminary data suggests that embryo selection is bringing preterm births rates down to a level on par with the
general population.
Microarrays aren’t the only promising embryo selection technology available. During the meeting Renee Reijo Pera, a professor
in the department of obstetrics and gynecology at Stanford University School of Medicine, presented her 2010 paper on noninvasive
imaging of human embryos (see page 15). The paper showed that
time-lapse image analysis of two-day old zygotes together with gene
expression profiling could with 93% accuracy predict which embryos
would develop to the blastocyst stage. Three parameters captured by imaging appeared to be the
most predictive: duration of the first cell
division, the time interval between the
end of first mitosis and initiation of
the second (mitosis is the replication of chromosomes that precedes cell division), and the
time interval between the
second and third mitoses.
“This is an example of
out of the box thinking,”
said challenger Simón.
However, he pointed
out there is a lack
of randomized clinical
trial data showing that
embryo selection using
this technology results in
higher birth rates. Discussion also raised the issue
of a lack of standards, with
6
groups trying to replicate the results using different measurement
techniques.
In yet another advance, embryo freezing would make IVF safer
and finally make it possible for young women suffering from ovarian
cancer and premature ovarian failure to have families later in life.
Ana Cobo, an embryologist from the Infertility Institute in Valencia,
presented her 2008 paper comparing fresh eggs to those cryopreserved by Cryotop—a commercially available method to vitrify human tissues (see page 25). Vitrification is the conversion of a water
containing solution to a glass-like solid that is free from the ice crystals normally formed during freezing. She and her colleagues found
that cryopreserved eggs were just as readily fertilized and able to
develop into blastocysts as fresh ones. Cobo also presented recent
data showing that the pregnancy rate from cryopreserved embryos
is similar to that of fresh embryos.
“The paper got a lot of interest because oocytes have been extremely difficult to freeze,” said challenger Evers. “The Cobo study
showed for the first time in a large number of patients that you can
do it and have good outcomes.” Moreover, embryo cryopreservation if broadly implemented might reduce the need to put back more
than one embryo, Evers added, because you can freeze several
and put them back in subsequent cycles if a woman doesn’t become pregnant. It may even enhance pregnancy rates because the
endometrial environment appears to be “more friendly” during normal cycles as compared to drug stimulated cycles, he added.
exPAnDing The BOunDARieS OF TReATMenT
In the future, there may be no limit to the age at which a woman
is able to bear children, according to Jonathan Tilly, professor and
chair of biology at Northeastern University. His 2012 paper provided
evidence that human ovarian tissue contains pools of ovarian stem
cells that develop into human oocytes—confirming what researchers
have found in other animals and overturning the long-held dogma
positing that women are born with a fixed number of eggs that are
never replenished (see page 10). Tilly first shocked the field in 2004
when he and his colleagues extracted ovarian stem cells from female mice for the first time.
In the recent paper, Tilly and his colleagues developed a more sensitive method to identify and collect mouse ovarian stem cells. Once
the team confirmed that it had isolated the mouse ovarian stem cells
by this new method, it repeated the technique with frozen ovaries donated from sex-reassignment patients. When cultured, the retrieved
oogonial stem cells (OSCs) turned into immature oocytes. The team
then tagged the cells with green fluorescent protein to trace them
and injected them into pieces of human ovarian tissue, which were
then transplanted into mice. After about two weeks, the OSCs had
differentiated into green-glowing cells that by all measures appeared
to be human oocytes.
“This was one of the most astonishing papers published in 2012,”
said challenger Felipe Vilella from the Valencia Institute of Infertility
in Spain. However, no one has yet shown that the human-derived
OSCs can be fertilized to form an embryo, Vilella said, and no one
knows whether any offspring would be healthy. Although studies on
animal derived OSCs have produced offspring, no data on the health
of the offspring has been provided, said Vilella. Nevertheless, research on monkeys and ultimately humans is in the planning stages,
Tilly said. If the research pans out, it would give women with ovarian
cancer the chance to have children later in life. It would also open the
possibility that women well into their 50’s and even older could bear
biological children—the ethics of which “are not for me to decide,”
concluded Tilly.
Vilella’s questions about the health of offspring raised another important point of debate: whether any of the procedures used in IVF,
currently or in the future, might unwittingly be harming the resulting offspring. At the meeting, Michael Skinner, director of the center
for reproductive biology at Washington State University in Pullman
Washington, presented his serendipitous finding revealed in his
2005 paper on the epigenetic transgenerational effects of endocrine
disruptor chemicals on rodents (see page 18). If a pregnant female
rodent was exposed to endocrine disrupting chemicals during a
window when the fetus’s sex cells were developing, the chemical
exposure caused epigenetic alterations—changes in methylation
patterns in the genome—in sperm. These alterations were passed
down through at least four generations and reduced male fertility.
Skinner has since tested several additional chemicals and found the
same or similar effects on sperm, and that these epigenetic changes
can cause disease in subsequent generations. In one study [Reprod.
Toxicol.34:708 (2012)], for example, Skinner showed that pregnant
female rats exposed to the pesticide mixture DEET—commonly
found in insect repellants—gave birth to pups with higher rates of
disease that included pubertal abnormalities, testis disease, and
ovarian disease. Disease susceptibility was predominantly passed
via males and persisted for three generations.
A similar issue has been raised before in regard to IVF techniques. Could the egg or culture media be exerting an effect on the
epigenetics of the developing embryo? To date, rigorous studies
have not yet been done to answer this
question. But the debate certainly generated enough interest and ideas for
future studies.
WRAP uP
After the meeting, participants lauded
the challenge and debate format. “From
my perspective this type of format
should be broadly applied,” said Tilly.
“It’s an awesome idea,” Scott added,
“and I’m not prone to hyperbole.” Some,
however expressed a desire for more
probing questions. “We were all being
a little too nice,” Reijo Pera said. But
everyone agreed that in contrast to
large conferences where there is little
time for discussion and questions, this
one enabled much more opportunity
to learn. Such feedback will be important in organizing the second Top Ten
in reproductive medicine conference,
which is already in the works. According to Simón, the next one will likely
take place in four years—enough time
to publish new papers and to measure
whether the first conference had an impact on spreading ideas and improving
the way research in reproductive medicine is done.
everyone
agreed that in
contrast to large
conferences
where there is
little time for
discussion and
questions, this
one enabled
much more
opportunity to
learn.
7
Produced by the Science/AAAS Custom Publishing Office
HISTORY AND PERSPECTIVE
Reproductive Medicine Before and After the Nobel
J.L.H. Evers and Antonio Pellicer
Several methods
for performing
genetic screens
on embryos, both
invasive and
non-invasive, are
in the research
pipeline right
now.
J.L.H. Evers
Antonio Pellicer
8
The 1960’s through the 1970’s was a
period of great change in reproductive
medicine. Knowledge about reproduction increased dramatically. Before in
vitro fertilization (IVF), gynecology had
little more to offer infertile couples than
making a diagnosis and offering tender,
loving care (1). When the efforts of the
two pioneers, Robert G. Edwards and
Patrick C. Steptoe, resulted in the birth
of the first IVF baby in 1978, the world
slowly started to understand the insurgency that these two icons had caused
in the medical field. Not without open
antagonism from some, Edwards became acclaimed as the father of IVF
and his discoveries were belatedly recognized by the awarding of the Nobel
Prize in Physiology or Medicine in 2010.
By then, several millions of babies had
been conceived through IVF and the
term assisted reproductive technology
(ART) was coined to include other procedures that improved infertility treatment. Edwards’ work laid the foundation for further exciting developments in
reproductive medicine, such as preimplantation genetic diagnosis (PGD).
ART is focused on bringing better
gametes into close proximity in order
to generate a viable embryo, which can
then be placed in a receptive endometrium to generate a healthy, full-term pregnancy. The greatest progress has been
made in improving embryo viability and
increasing implantation rates which, unintentionally, has led to the main drawback—the unacceptably high incidence
of multiple births. The rate of multiples
has been substantially reduced by the
introduction of elective single embryo
transfer (eSET), one of the few ART developments that has been
introduced based on robust clinical data (2). Regulatory initiatives for
eSET approval have followed around the world. However, there is
still much to be done before eSET is universally accepted.
An ART process begins with controlled ovarian stimulation (COS),
the manipulation of the reproductive physiology to generate several mature oocytes simultaneously in a single cycle. In the early
days of ART, this was a necessary functional step because the
techniques for fertilization and culture were poor, so several eggs
needed to be fertilized to obtain just a few viable embryos. Today,
the procedures used in ART labs are much more sophisticated, so
there is a diminished need for aggressive COS. The field is now
moving towards more gentle and patient-friendly stimulation protocols which will produce only an adequate number of embryos, and
no more (3).
in ViTRO SCReening
Several methods for performing genetic screens on embryos, both
invasive and noninvasive, are in the research pipeline right now. After some initial controversy regarding the biopsy and screening of
embryos for a limited number of chromosomes, the introduction of
more accurate technologies that provide more detailed information
on all chromosomes—such as single nucleotide polymorphism arrays—is looking promising, especially for use in older patients (4). In
vitro embryo observation utilizing time-lapse technology (5), and the
analysis of proteins and metabolites consumed and secreted by the
developing embryo in culture, are other noninvasive methods that
hold promise.
The same molecular techniques are applied during PGD in order to
detect those embryos carrying particular disease-associated mutations. The list of diseases is steadily increasing and in the future it
will become possible to diagnose multiple disorders simultaneously
in a single embryo, shifting the frontiers in human health care. These
advances will, however, also raise important ethical questions that
will need to be addressed (6), such as how to contend with the partial
expression or limited penetrance of certain diseases that might only
manifest in later life.
CRyOPReSeRVATiOn
Although the failure of embryo implantation was a concern in the
early days of IVF and a valid justification for the use of multiple
embryos, the problem was compounded by the
poor performance of embryo cryopreservation
techniques. However, vitrification—introduced
as a method of cryopreservation in the 1980’s—
completely changed the field for both embryo
and oocyte freezing (7). Today, survival rates
after freezing/thawing of human oocytes and
embryos are greater than 90%, bringing about
new possibilities for ART, while also reducing
the need for ethically questionable selective
abortions.
Vitrification has also been applied as a
means to preserve fertility in young women
with cancer, allowing for the freezing of oocytes
prior to chemotherapy or radiation treatment.
Meanwhile, fertility preservation has more
recently been introduced to avoid the deleterious effects of aging on the oocytes (‘social
freezing’). Since age correlates positively with
aneuploidies, an increasing number of women
are requesting oocyte freezing before age 38
in an attempt to abrogate this effect (8).
Two other complications of COS deserve our
Robert Edwards holds Louise Brown, the first baby conceived through in vivo fertilization, as Patrick Steptoe
attention: an altered endometrial receptivity
looks on, July 25, 1978.
that decreases the chance of implantation, and
the development of ovarian hyperstimulation
syndrome (OHSS), a rare but potentially life-threatening condition. A cytes by exploring the cellular and molecular hallmarks of aging
new two-step ART approach that involves the implantation of previ- and attempting to at least partially reverse them (12). Similarly,
ously frozen embryos in subsequent natural cycles will potentially the creation of gametes by cell reprogramming could be another
lead to a safer and more patient-friendly ART approach with fewer source of healthy oocytes (and spermatozoa), an advance that
will certainly change our perspective on infertility in the coming
complications (9,10).
years (13,14).
The enDOMeTRiAl enViROnMenT
Most of these potential advancements in ART are still based on
Most attention in ART is focused on the embryo, but endometrial small, observational clinical “proof-of-concept” studies. They dereceptivity is an issue that is often neglected, predominantly because mand to be followed up with more comprehensive, multicenter
insufficient methods exist to ascertain the functional status of the randomized clinical trials. Subfertility couples belong to a vulneruterine lining. In today’s ˈomics era, however, new tools are coming able group and should not be exploited (15). We should provide
to the fore that will allow the best embryo(s) to be placed in a fully them with dedicated care and evidence-based treatment. During
receptive endometrium (11).
the next decade, the medical establishment needs to provide more
Despite numerous advances in reproductive technologies, robust evidence for the value and integrity of the treatments curwe still have very limited possibilities for addressing the main rently offered to patients. As Robert Edwards, our very own Nocause of infertility, namely the woman’s age. The next two de- bel Laureate stated: “the legacy to future generations is a matter
cades will see research in developing methods to rejuvenate oo- of life” (16).
REFERENCES
1. J. L. H. Evers. Lancet 360,151 (2002).
2. A. Thurin et al., N. Engl. J. Med. 351, 2392 (2004).
3. R. G. Edwards, R. Lobo, P. Bouchard, Hum. Reprod. 11, 91
(1996).
4. N. R. Treff et al., Fertil. Steril. 94, 2017 (2010).
5. C. C. Wong et al., Nat. Biotechnol. 28, 1115 (2010).
6. J.C. Harper et al., Hum. Reprod. Update 18, 234 (2012).
7. A. Cobo et al., Fertil. Steril. 89, 1657 (2008).
8. J.A. Garcia-Velasco et al., Fertil. Steril. 99, 1467 (2013).
9.
10.
11.
12.
13.
14.
15.
16.
A. Maheshwari, S. Bhattacharya, Hum. Reprod. 28, 6 (2013).
P. Devroey, N.P. Polyzos, C. Blockeel, Hum. Reprod. 26, 2593 (2011).
P. Díaz-Gimeno et al., Fertil. Steril. 95, 50 (2011).
C. López-Otín, M. A. Blasco, L. Partridge, M. Serrano, G. Kroemer,
Cell 153, 1194 (2013).
Y.A. White, et al., Nat. Med. 18, 413 (2012).
K. Hayashi, et al., Science 338, 971 (2012).
A. W. Nap, J. L. H. Evers., Hum. Reprod. 22, 3262 (2007).
R. G. Edwards, P. C. Steptoe, A Matter of Life. The Story of IVF - a
Medical Breakthrough (Hutchinson & Co., London, 1980).
9
Produced by the Science/AAAS Custom Publishing Office
Germline Stem Cells in Adult Mammalian Ovaries
Dori C. Woods and Jonathan L. Tilly
inTRODuCTiOn
Until recently, a decades-old belief in the field of mammalian reproductive biology was that females are born with a finite endowment of
oocytes, termed the ‘ovarian reserve,’ which is progressively depleted throughout life by either ovulation or atresia (1). Once exhausted,
the ovaries cease to function. In women, this event heralds the onset
of menopause (2). Although the traditional explanation for the cessation of ovarian function—namely that a woman simply runs out of
oocytes, fits the model of the ovarian reserve being set forth as a
nonrenewable resource at birth, a number of recent studies indicate
that this paradigm is probably incorrect. In its place has emerged a
model that aligns gametogenesis in mammals more closely with that
observed in non-mammalian species, in that adult ovaries actively
support formation of new oocytes and follicles (3–14). The underlying foundation of this major paradigm shift is rooted in the development of detailed methods for the isolation and characterization of
a rare population of mitotically active germ cells from adult ovaries
termed female germline stem cells (fGSCs) or oogonial stem cells
(OSCs) (4,6,8,12). Once isolated, these cells can be stably propagated in vitrofor months without loss of their ability to differentiate
into oocytes following either defined culture in vitro or transplantation
into ovarian tissue in vivo (4,5,7,8,14). In mice, oocytes formed
from transplanted OSCs complete maturation to the metaphase-II
stage of development, and these eggs can be fertilized to produce
viable embryos and offspring (4,7,8,12).
While intraovarian transplantation models clearly show that mammalian OSCs are capable of generating fully functional eggs, the role
of these cells, if any, in adult ovaries under normal physiological conditions has not yet been reported. In non-mammalian vertebrates,
such as the teleost Medaka (Oryziaslatipes), genetic-lineage–tracing approaches have been successfully employed to show that fGSCs actively contribute to the adult oocyte pool used for reproduction
(15). Development and analysis of similar genetic models in mice,
which are currently under investigation in the Tilly laboratory, should
provide a means to map the rates of new oocyte formation during
postnatal life, and the contribution of these oocytes to offspring production. In addition, the ability to track a ‘marked’ pool of oocytes
formed at a specific point in life will facilitate studies of in vivo oocyte
lifespan to determine how long a given oocyte, once generated, persists in adult ovaries. This information will help elucidate if the quality
of eggs deteriorates in women with age simply because all of the oocytes have been sitting around for decades accumulating damage,
or if the decline in egg quality is instead tied to a progressive impairment in the ability of OSCs to generate competent oocytes with age,
as is the case in the aging Drosophila ovary (16).
Thisarticlebasedontheoriginalpublication:
Y. A. R. White et al., Nature Med. 18, 413 (2012).
Department of Biology, Northeastern University, Boston, MA
Corresponding Authors: [email protected] or [email protected]
10
ChARACTeRizATiOn OF MOuSe AnD huMAn OSCS
Although the beginnings of contemporary evidence for the existence
of OSCs and postnatal oogenesis date back nearly a decade (3),
the recent development of protocols that allow for the enrichment
or purification of OSCs from adult ovaries has greatly accelerated
research in this area (4,6,8,12). Building on earlier observations
that mouse OSCs expose the C-terminus of the germ cell-specific
protein, Ddx4 [DEAD box polypeptide 4; also commonly referred
to as Vasa or Mouse vasa homolog (Mvh)] on the outer cell surface (4), we developed and validated an antibody-based fluorescence-activated cell sorting (FACS) approach that purifies OSCs
based on detection of this extracellular epitope of Ddx4 (extracellular Ddx4- or ecDdx4-positive cells) (8, 12, 13). The reason why
Ddx4 is exposed on the surface of OSCs but is completely internalized in oocytes is unknown, but may be related to movement of
the protein from a potentially sequestered transmembrane location
to a more functionally active position in the cytoplasm as primitive
germ cells undergo meiotic differentiation (13). Of note, similar OSC
isolation strategies have been reported using antibodies against
the externalized domain of the primitive germ cell marker, Ifitm3
(interferon-induced transmembrane protein 3; also referred to as
Fragilis) (6,12).
Following isolation and expansion of OSCs in vitro, the cells can
be genetically modified to express a traceable gene (such as green
fluorescent protein, GFP) and returned to the ovaries of recipient
wild type mice where they actively undergo differentiation into oocytes that mature, ovulate, and fertilize to produce viable embryos
and offspring (4,7,8,12). Although this type of cell fate-tracking experiment is not feasible in humans, we have successfully employed
a mouse xenograft model to establish the oocyte-forming capabilities of human OSCs, expressing GFP, in adult human ovarian tissue in an in vivo environment (8,12). The ensuing observations that
human OSCs form what appear to be meiotically arrested oocytes
[positive for expression of Y-Box protein 2 (YBX2), which is specifically expressed in germ cells at the diplotene stage of meiosis
(17, 18)] contained in follicles within one to two weeks of grafting
not only provides definitive evidence that adult human ovaries are
fully capable of supporting oogenesis and folliculogenesis, but also
that oocyte and follicle formation from purified OSCs returned to their
natural environment are events that can occur in a fairly rapid timeframe (8,12).
COnTROVeRSiAl neW FinDingS?
Although independent verification of the existence of OSCs in adult
mouse ovaries is now available from several labs (4–8,12–14), two
new studies—both based on circumstantial negative findings with
mice—claim otherwise despite the fact that neither study actually
attempted to reproduce what we and others have done or to isolate OSCs from mouse ovaries using published protocols employed
successfully by us and others. The first of these, from Liu and col-
Figure 1. Assessment of human oogenesis using adult ovary-derived oogonial stem cells (OSCs). Ovarian cortical tissue from women is dissociated into single cell
suspension, and OSCs are isolated by FACS using an antibody targeting the extracellular domain of DDX4 (ecDDX4) (8, 12). Purified OSCs are established in culture and
expanded ex vivo to assess egg maturation potential or the rate of oogenesis. To evaluate egg formation, OSCs are transformed to express a reporter (i.e., GFP) and
directly injected into human ovarian cortical strips, which are then cultured in vitro to monitor formation of GFP-expressing oocytes contained in follicles. Growing follicles
with GFP-positive oocytes are dissected and cultured further to obtain oocytes that can be in vitro-matured to the metaphase-II (egg) stage of development (28–30). To
evaluate oogenesis rates, human OSCs are transformed to express DsRed under control of the promoter of the Stra8 gene, which is activated during meiotic commitment
in germ cells. The cells are then screened for factors that activate DsRed expression. Positive hits, representing candidate meiotic (oogenic) activators, are then assessed
with a secondary screen in which the number of oocytes formed in vitro by a fixed number of OSCs per well exposed to the candidate factor is determined (12, 14). If a
given factor is identified as a positive hit in both assays (increased DsRed expression and oogenesis in vitro), it is then tested for its ability to increase oocyte-containing
primordial follicle numbers in adult mouse ovaries in vivo.
leagues, relied entirely on a transgenic reporter mouse generated
by crossing Ddx4-Cre mice with Rosa26 rbw/+ mice, on the assumption that OSCs could be specifically tracked due to Ddx4-driven Cre
activation in these cells (as would be the case with all germ cells,
irrespective of their differentiation status) (19). Unfortunately, an absence of many key control experiments, discussed in detail recently
(12), precludes any clear interpretation of the findings. In fact, in
as-yet unpublished studies we have since evaluated ovaries from
the same genetic mouse line, and combined this approach with our
FACS-based protocol for OSC isolation, to highlight the erroneous
nature of their primary conclusion that, in contrast to substantial evidence from us and others (3–14,20), mitotically active germ cells do
not exist in postnatal mouse ovaries.
The second paper relies heavily on two key assumptions (21).
The first is that oogenesis in embryonic and adult ovaries is accomplished through identical processes. In fetal ovaries, primordial
germ cells proliferate and form cyst-like clusters prior to meiosis (22).
Upon meiotic commitment, the cyst-like clusters associate with somatic cells, forming the ovigerous cords that will break down shortly
after birth to produce the initial primordial follicle pool (23,24). In this
new study, Lei and Spradling propose a model, with no supporting
data, that relies on formation of germ cell cysts as an indispensable
step in postnatal oogenesis as well. When they failed to observe
evidence of cyst-like germ cell clusters in adult mouse ovaries, the
authors concluded that oogenesis is not occurring (21). Another assumption relates to the “lineage tracing model” used, in which both
Cre-recombinase and estrogen receptor are expressed in essentially
all cells of the mice under study. Short-term induction with Tamoxifen
excises a stop-cassette located within a Rosa26-yellowfluorescent
protein (YFP) transgene, yielding indiscriminate labeling in which
all cell types are ‘marked’ with YFP at very low frequency (1%–
2%). Since all germ cells, including OSCs and oocytes, express
YFP at equivalent frequencies, it is impossible to discern if any
labeled oocytes originated from labeled OSCs or if any unlabeled
oocytes originated from unlabeled OSCs. In fact, the authors acknowledge identification of “a few germ cells at a prefollicular state
of development,” but do not discuss the meaning of such findings
(21). Since oocytes are incapable of surviving in ovaries unless
surrounded by granulosa cells as follicles, the rare “prefollicular”
germ cells identified by Lei and Spradling are likely OSCs or their
immediate progeny.
Whatever the case, assuming 2% of OSCs are YFP labeled in their
11
Produced by the Science/AAAS Custom Publishing Office
model, one would expect the ratio of labeled and unlabeled oocytes
comprising the primordial follicle pool to remain constant over time
since a fixed proportion of labeled and unlabeled oocytes would be
constantly entering the pool through de novo oogenesis from a chimeric population of OSCs and subsequently exiting the pool through
follicle growth activation. This is exactly what Lei and Spradling observe (21). It bears mention that a toxicity model was also employed
to assess if OSC activation and follicle renewal in postnatal ovaries occurs in response to damage. When such evidence was not
produced, these negative findings were offered as further proof that
OSCs do not exist. Surprisingly, however, the cytotoxic drug used
was busulfan, which is a widely known GSC toxicant (3,25–27). Accordingly, it is unclear why one would expect to find damage-induced
activation of a population of cells that are killed by the agent used to
‘activate’ them.
nexT STePS AnD huMAn heAlTh iMPliCATiOnS
As work with laboratory animal models continues to fill in key gaps
in our understanding of mammalian OSCs, the discovery of oocyte
progenitor cells in human ovaries opens up many opportunities for
their utilization, some from the perspective of the clinical management of infertility and others from that of basic biomedical discovery
(8,9). Given the remarkable similarities between mouse and human
OSCs, and in particular the ability of OSCs from both species to
participate in new oocyte and follicle formation when delivered back
into adult ovarian tissue (8, 12), it would stand to reason that, as
observed in mice (4,7,8,12), human OSCs also have the potential
to generate developmentally competent eggs. However, because
such functional testing of human OSCs in vivo is not possible, strategies to mature human eggs derived from OSCs entirely ex vivo will
be needed. In this regard, Telfer and McLaughlin have reported a
multistep culture system in which microthin human ovarian cortical
strips are maintained under defined serum-free conditions to initiate primordial follicle growth activation (28,29). Once the preantral
stage of development is reached, the follicles are dissected out and
subsequently cultured individually until the enclosed oocytes can be
aspirated for in vitro maturation to metaphase-II eggs. Should this
methodology prove successful (30), it may provide an opportunity to
test the developmental potential of human OSCs to form functional
eggs (Fig. 1), given that human OSCs have already been shown
to generate primordial oocytes contained in follicles in adult human
ovarian cortical strips (8,12).
In addition to the ability to test the competency of oocytes derived
from OSCs recombined with somatic cells and other support necessary for ensuring ex vivo development, the innate oocyte-forming
capability of these cells is valuable for other reasons. During serial
passage of pure OSC cultures (i.e., with no somatic or feeder cells)
under defined conditions, a small subset of OSCs spontaneously initiate a differentiation process that results in the formation of what
appear to oocytes (8,12). Although these oocytes are not functionally competent (having never been associated with nor instructed by
granulosa cells), they can be used as a powerful bioassay to rapidly
decipher the factors and signaling pathways that guide oogenesis
(Fig. 1). This can be achieved by assessing the rate of formation of
in vitro-derived oocytes from a fixed number of OSCs exposed in
different wells to various agents (12,14). In addition to the scientific
12
value of such knowledge, large-scale screening of cultured OSCs
may facilitate identification and development of therapeutics to sustain or restore the ovarian reserve in women of advancing maternal
age or after exposure to noxious insults such as chemotherapy. Although more work is clearly needed to test these ideas, at this stage
we believe it is appropriate to, at minimum, end further debate over
whether mammalian OSCs exist, and instead constructively focus
on what additional experiments are needed to more clearly define
the regulation and function of these newly discovered cells in female
reproductive biology.
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
29.
30.
REFERENCES
S. Zuckerman, Rec. Prog. Horm. Res. 6, 63 (1951).
H. Buckler, J. Br. Menopause Soc. 11, 61 (2005).
J. Johnson et al., Nature. 428, 145 (2004).
K. Zou et al., Nat. Cell Biol. 11, 631 (2009).
J. Pacchiarotti et al., Differentiation 79, 159 (2010).
K. Zou et al., Stem Cells Dev. 20, 2197 (2011).
Y. Zhang et al., J. Mol. Cell Biol. 3, 132 (2011).
Y. A. R. White et al., Nat. Med. 18, 413 (2012).
D. C. Woods, J. L. Tilly, Fertil. Steril. 98, 3 (2012).
D. C. Woods, Y. A. R. White, J. L. Tilly, Reprod. Sci. 20, 7 (2013).
D. C. Woods, J. L. Tilly, Semin. Reprod. Med. 31, 24 (2013).
D. C. Woods, J. L. Tilly, Nat. Protoc. 8, 966 (2013).
A. N. Imudia et al., Fertil. Steril. 100, 1451 (2013).
E. S. Park, D. C. Woods, J. L. Tilly, Fertil. Steril. 100, 1468 (2013).
S. Nakamura et al., Science 328, 1561 (2010).
R. Zhao et al., Aging Cell. 7, 344 (2008).
W. Gu et al., Biol. Reprod. 59, 1266 (1998).
J. Yang et al., Proc. Natl. Acad. Sci. U.S.A. 102, 5755 (2005).
H. Zhang et al., Proc. Natl. Acad. Sci. U.S.A. 109, 12580 (2012).
Y. Hu et al., Cell Prolif. 45, 287 (2012).
L. Lei, A. C. Spradling, Proc. Natl. Acad. Sci. U.S.A. 110, 8585
(2013).
M. Pepling, A. Spradling, Development 125, 3323 (1998).
M. Pepling, A. Spradling, Dev. Biol. 234, 339 (2001).
C. Nicholas, K. Haston, R. Reijo-Pera, BMC Dev. Biol. 10, 2 (2010).
L. R. Bucci, M. L. Meistrich, Mutat. Res. 176, 259 (1987).
M. C. Pelloux et al., Acta Endocrinol. 118, 218 (1988).
C. J. Brinster et al., Biol. Reprod. 69, 412 (2003).
E. E. Telfer et al., Hum. Reprod. 23, 1151 (2008).
E. E. Telfer, M. McLaughlin, Semin. Reprod. Med. 29, 15 (2011).
C. E. Dunlop, E. E. Telfer, R. A. Anderson, Maturitas doi:10.1016/j.
maturitas.2013.04.017 (2013).
Acknowledgments: We thank Cooper Graphics (www.cooper247.
com) for preparation of the final figure. Work conducted by the authors
discussed herein was supported by grants from the United States National
Institutes of Health (R37-AG012279, R21-HD072280, F32-AG034809),
the Henry and Vivian Rosenberg Philanthropic Fund, and the Glenn Foundation for Medical Research.
Competing Interests Statement: D.C.W. is a scientific consultant for
OvaScience, Inc. (Cambridge, MA); J.L.T. discloses interest in intellectual
property described in U.S. Patent 7,195,775, U.S. Patent 7,850,984 and
U.S. Patent 7,955,846, and is a co-founder of OvaScience.
Bidirectional Communication Between the Oocyte and Surrounding
Somatic Cells is Required for Successful Follicular Development
and Oocyte Maturation
Jia Peng1,2, John J. Eppig3, Martin M. Matzuk1,2,4,5,6,7*
Follicular development and oocyte maturation are essential processes for successful ovulation to produce competent eggs ready
for fertilization. Historically, it has been unclear whether the oocyte
or the surrounding granulosa cells orchestrate the rate of follicular
growth. Elegant work from the last decade shed considerable light
on this process. Studies utilizing chimeric reaggregated mouse
ovaries consisting of half-grown 12-day-old oocytes and newborn
granulosa cells revealed that these stage-mismatched ovaries could
undergo follicular development from primary to antral follicle stage at
twice the normal rate. However, recent studies also uncovered the
importance of ovarian somatic cells in the regulation of folliculogenesis and oogenesis. Here, we discuss bidirectionalcommunication
between oocytes and their companion somatic cells during follicular
development and oocyte maturation, which ensures normal female
fertility.
inTRODuCTiOn
In mammals, primordial germ cells (PGCs) divide and migrate to
the germinal ridge to become oogonia, which begin meiotic division during prenatal life. Around the time of birth, a cohort of oocytes recruits a single layer of flattened pre-granulosa cells to form
the primordial follicles (1). Folliculogenesis starts with activation of
a primordial follicle, which progresses through primary, secondary,
and antral follicle stages, and finishes by either generating a fertilizable oocyte or follicular atresia. It was not clear previously that the
rate of ovarian follicular development was dominantly controlled by
either oocytes or neighboring somatic cells until a key study was
done in the last decade (2). In this study, mid-sized oocytes from the
secondary follicles of 12-day-old mice were isolated and combined
with somatic cells from primordial follicles of newborns to produce
reaggregated ovaries. As a result, the 12-day-old oocyte prompted the proliferation and differentiation of both cumulus and mural
granulosa cells and doubled the rate of follicular development to
generate competent oocytes ready for fertilization. Thus, this study
demonstrated that a developmental program intrinsic to the oocyte
is crucial for orchestrating the rate of follicular growth and fundamentally changed our understanding of the regulation of ovarian
follicular development.
Thisarticlebasedontheoriginalpublication:
J. J. Eppig et al., Proc. Natl. Acad. Sci. U.S.A. 99, 2890 (2002).
Departments of 1Pathology & Immunology, 2Molecular &
Human Genetics, 4Molecular and Cellular Biology, and
5
Pharmacology, 6Center for Drug Discovery and
7
Center for Reproductive Medicine, Baylor College of Medicine,
Houston, TX; 3The Jackson Laboratory, Bar Harbor, ME
*Corresponding Author: [email protected]
Figure 1. Bidirectional communication between oocyte and surrounding
somatic cells. There are three major ways of oocyte-somatic cells communication
in follicular development and oocyte maturation: (i) oocyte-secreted factors GDF9,
BMP15, and GDF9:BMP15 heterodimer; (ii) granulosa cell-derived kit ligand and its
receptor on the oocyte; and (iii) secondary messengers cAMP and cGMP.
OOCyTe-SeCReTeD FACTORS in The TRAnSFORMing
gROWTh FACTOR β (TgF-β) PAThWAy
In follicular development, the mammalian oocyte modulates granulosa (directly) and thecal cell (indirectly) proliferation and differentiation rates through multiple oocyte-secreted factors (3). Among
those factors, growth differentiation factor 9 (GDF9) and bone morphogenetic protein 15 (BMP15) are well known as two of the most
important paracrine factors to prompt primary follicle growth as well
as subsequent participation in the various stages of folliculogenesis
(4). GDF9 and BMP15 are close paralogs in the TGF-β superfamily
and signal via the downstream P-SMAD2/3 or P-SMAD1/5/8 pathway, respectively, to stimulate proliferation and differentiation of
granulosa cells. Animal models deficient in these two ligands have
been used to study their in vivo functions at the early stages of follicular development between poly- and mono-ovulatory mammals
(Table 1). In mice, Gdf9 null females are sterile due to an arrest of
follicular growth at the primary follicle stage (5). Bmp15 null mice
exhibit later defects in ovulation and fertilization rates, which lead to
reduced litter size (6). In sheep, homozygous BMP15 mutants exhibit a block in folliculogenesis at the primary follicle stage, resulting
13
Produced by the Science/AAAS Custom Publishing Office
Species
Gene
Gdf9
Mouse
Bmp15
GDF9
Sheep
BMP15
GDF9
Human
BMP15
Mutant
Phenotype
Heterozygous
Normal fertility (5)
Homozygous
Sterility due to block at primary follicle stage (5)
Heterozygous
Normal fertility (6)
Homozygous
Subfertility due to defects in cumulus expansion (6)
Heterozygous
Increased ovulation rate resulting in increased offspring (e.g., dizygotic twins) (8)
Homozygous
Sterility due to block at primary follicle stage (8)
Heterozygous
Increased ovulation rate resulting in increased offspring (e.g., dizygotic twins) (7)
Homozygous
Sterility due to block at primary follicle stage (7)
Heterozygous
Associated with premature ovarian failure (9) or spontaneous dizygotic twinning (12)
Homozygous
Not reported
Heterozygous
Associated with premature ovarian failure (10,11)
Homozygous
Not reported
have identified three major pathways that mediate the communication between oocyte and somatic cells (Fig. 1): (i) oocyte-secreted
GDF9, BMP15, and GDF9:BMP15 heterodimer which stimulate
TGFβ signaling in granulosa cells; (ii) granulosa cell-derived kit
ligand that binds to kit receptor on the oocyte and triggers downstream PI3K signaling; and (iii) secondary messengers, which are
transferred via oocyte-cumulus cell gap junctions. Although the oocyte is the dominant factor orchestrating the onset of folliculogenesis and regulating somatic cell proliferation and differentiation during follicular development, other regulators in the oocyte-somatic
cell regulatory loop are also indispensable for normal female fertility (22). Thus, progression through successive stages of follicular development and oocyte maturation requires bidirectional
communication between the oocyte and surrounding somatic
cells.
Table 1. GDF9 and BMP15 mutants in mouse, sheep, and human.
in sterility similar to that of Gdf9 null mice (7). A homozygous point
mutation in sheep GDF9 also results in abnormalities at early stages
of follicular development (8). In humans, although there is no direct
evidence to prove that BMP15 and GDF9 are major causes of ovarian insufficiency, multiple studies have found that these two genes
are associated with premature ovarian failure (9–11) or spontaneous dizygotic twinning (12). Therefore, these genetic data indicate
that both GDF9 and BMP15 play important roles in the transition
between primary and secondary follicles as well as in ovulation.
However, which of the two proteins is the dominant regulator might
differ due to varying protein activities and expression patterns among
species.
To further resolve the functions of these two growth factors and
their potential synergistic functions in later follicle developmental
stages, recombinant GDF9 or/and BMP15 were used to treat granulosa cells in vitro. In a recent study, our group purified human (h) and
mouse (m) recombinant GDF9 and BMP15 homodimers and heterodimers to define their activities in pre-ovulatory follicles (13). We
showed that mGDF9 homodimer was a potent trigger of the TGF-β
signaling cascade and upregulated downstream cumulus expansionrelated gene expression to induce the proliferation and differentiation of granulosa cells, while mBMP15 homodimer was inactive. By
contrast, hGDF9 homodimer was inactive, and hBMP15 homodimer
had a relatively low activity compared to mGDF9. In our studies, m/
hGDF9:BMP15 heterodimers were the most biopotent ligands in the
regulation of ovarian function.
negATiVe OOCyTe RegulATORS in The
PhOSPhATiDylinOSiTOl 3 kinASe (Pi3k) PAThWAy
Granulosa cell-derived kit ligand is a positive regulator, stimulating oocyte growth and somatic cell proliferation. After binding with
kit ligand, the kit receptor mediates PI3K signaling cascades in the
oocyte via many downstream regulators. Among these regulators,
several key components of the PI3K pathway function as suppressors of follicular activation at the early stages of follicular development. Phosphatase and tensin homolog deleted on chromosome
10 (PTEN) is a major negative regulator. Mice lacking Pten in the
oocyte exhibit premature ovarian failure due to depletion of primor-
14
dial follicles in early adulthood (14). FOXO3A, a forkhead transcription factor, is another important downstream negative effector of
the PI3K pathway. Phosphorylation of FOXO3A leads to its nuclear
export and the regulation of genes involved in primordial follicle activation. Foxo3a knockout female mice are phenotypically similar to
Pten oocyte-specific knockout mice (15). These animal models could
help unravel the etiology of infertility and premature ovarian failure in
women and enable clinical intervention.
MeiOTiC ARReST AnD ReSuMPTiOn RegulATiOn
ViA gAP JunCTiOnS
Synchrony between oocyte meiosis and follicular ovulation is required
for female fertility. Granulosa cells maintain oocyte meiotic arrest via
the natriuretic peptide C/natriuretic peptide receptor 2 (NPPC/NPR2)
system (16). Mural granulosa cells produce the NPPC ligand, which
binds to the receptor NPR2, a guanylyl cyclase in cumulus cells,
resulting in cyclic guanosine monophosphate (cGMP) generation.
cGMP diffuses from cumulus cells to the oocyte via gap junctions
and then inhibits PDE3A, an oocyte-specific phosphodiesterase, to
maintain elevated cyclic adenosine monophosphate (cAMP) in the
oocyte (17,18). High levels of cAMP require the orphan Gs-linked
receptor GPR3 to prevent precocious gonadotropin-independent
resumption of meiosis (19). After the LH surge, PDE3A becomes
activated, decreasing the oocyte cAMP concentration and thus triggering the downstream pathways to resume meiosis during ovulation
(20). As a dominant factor, the oocyte controls meiotic arrest not only
by maintaining high cAMP levels but also by secreting certain paracrine growth factors (including GDF9 and BMP15) to elevate NPR2
expression in cumulus cells (16). Furthermore, a novel oocyte protein, named meiosis arrest female 1 (MARF1), has been identified as
an oocyte maturation regulator by recent ENU mutagenesis screening (21). Mutations of Marf1 leads to infertility characterized by the
failure of meiotic resumption, upregulation of protein phosphatase 2
catalytic subunit beta (PPP2CB), and an increase in DNA damage in
the ovulated oocytes.
COnCluSiOnS AnD iMPliCATiOnS
In summary, genetic data generated over the past few decades
1.
2.
3.
4.
5.
6.
REFERENCES
H. Peters, Acta Endocrinol. (Copenh) 62, 98 (1969).
J. J. Eppig, K. Wigglesworth, F. L. Pendola, Proc. Natl. Acad. Sci.
U.S.A. 99, 2890 (2002).
P. G. Knight, C. Glister, Reproduction 132, 191 (2006).
J. A. Elvin, C. Yan, M. M. Matzuk, Mol. Cell. Endocrinol. 159, 1
(2000).
J. Dong et al., Nature 383, 531 (1996).
C. Yan et al., Mol. Endocrinol. 15, 854 (2001).
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
S. M. Galloway et al., Nat. Genet. 25, 279 (2000).
J. P. Hanrahan et al., Biol. Reprod. 70, 900 (2004).
T. T. Wang et al., Hum. Reprod. 28, 2473 (2013).
E. Di Pasquale, P. Beck-Peccoz, L. Persani, Am. J. Hum. Genet. 75,
106 (2004).
H. Dixit et al., Hum. Genet. 119, 408 (2006).
G. W. Montgomery et al., Twin Res. 7, 548 (2004).
J. Peng et al., Proc. Natl. Acad. Sci. U.S.A. 110, E776 (2013).
P. Reddy et al., Science 319, 611 (2008).
D. H. Castrillon, L. Miao, R. Kollipara, J. W. Horner, R. A. DePinho,
Science 301, 215 (2003).
M. Zhang, Y. Q. Su, K. Sugiura, G. Xia, J. J. Eppig, Science 330, 366
(2010).
S. Vaccari, J. L. Weeks, 2nd, M. Hsieh, F. S. Menniti, M. Conti, Biol.
Reprod. 81, 595 (2009).
R. P. Norris et al., Development 136, 1869 (2009).
L. M. Mehlmann et al., Science 306, 1947 (2004).
F. J. Richard, A. Tsafriri, M. Conti, Biol. Reprod. 65, 1444 (2001).
Y. Q. Su et al., Science 335, 1496 (2012).
M. M. Matzuk, K. H. Burns, M. M. Viveiros, J. J. Eppig, Science 296,
2178 (2002).
Acknowledgments: Studies on oocyte-somatic cell bidirectional
communication have been supported by the Eunice Kennedy Shriver
National Institute of Child Health and Human Development through
R01 grants HD33438 (M.M.M.) and HD23839 (J.J.E.).
An Introduction to Human Embryo Development
Renee A. Reijo-Pera
Human embryo development begins with an oocyte-to-embryo transition, a process that includes the fusion of the egg and sperm, migration of the germ cell pronuclei into the uterus, pronuclear membrane
breakdown, and a series of cleavage divisions. This culminates in
the activation of the unique human embryonic genome, compaction
of the blastomeres to form a morula, and subsequent differentiation
of the first cell lineages—the trophectoderm and inner cell mass (1).
In humans, there is a minor wave of transcriptional activation early in
development and a major wave that constitutes embryonic genome
activation (EGA) on day three of embryogenesis (2). Human embryo
development is remarkable in that the oocyte provides the majority of the required resources to direct the complex developmental
pathways during the first three days of development, in the absence
of large-scale transcriptional activity (2). Consider further that native
human reprogramming from gametic pronuclear programs to embryonic programs involves the epigenetic remodeling of one of the most
challenging sets of chromosomes to reprogram, those inherited from
Figure 1. LAMIN-B1
(green) and CENP-A
(orange) expression
in a DAPI-stained
(blue) cleavage-stage
human embryo by
confocal microscopy
reveals chromosomecontaining micronuclei in the blastomeres
and cellular fragments
of human embryos.
Image from (6).
Thisarticlebasedontheoriginalpublication:
C. C. Wong et al., Nature Biotech. 28, 1115 (2010).
Institute for Stem Cell Biology & Regenerative Medicine,
Department of Obstetrics and Gynecology, Stanford University
School of Medicine, Stanford, CA
[email protected]
15
Produced by the Science/AAAS Custom Publishing Office
Figure 3. Proposed model for the origin
of human embryonic aneuploidy based
on fragmentation timing, fragment resorption, and underlying chromosomal
abnormalities. Human embryos with meiotic errors (monosomies and trisomies) and
those that appear to be triploid typically
exhibit fragmentation at the one-cell stage,
while fragmentation is most often detected
at the two-cell stage in embryos with mitotic
errors. We also demonstrated that missing
chromosome(s) are contained within fragments termed “embryonic micronuclei” and
for those embryos with mitotic errors, propose that the embryo likely divided before
these chromosomes properly aligned on the
mitotic spindle. The correlation between the
timing of fragmentation and the type of embryonic aneuploidy suggests that the embryo can sense chromosomal abnormalities
and undergoes fragmentation as a survival
mechanism. As development proceeds,
these fragments either remain or are reabsorbed by the blastomere from which they
originated, or a neighboring blastomere, to
generate the complex human aneuploidies
observed. Image from (6).
Figure 2. Automated tracking of cellular fragmentation for embryo assessment. Time sequence of cumulative segment lengths
in pixels for each frame of the time-lapse image analysis for three embryos was observed: embryo C3 with high fragmentation, embryo
B2 with low fragmentation and embryo C1 with no fragmentation. Note that the embryo with high fragmentation exhibits much larger
cumulative length of segments than that of embryos with low or no fragmentation. Image from (6).
the sperm (1). These chromosomes are highly condensed, highly
methylated, and organized around a protamine scaffold rather than
a histone scaffold. Yet, it is clear from several studies that native reprogramming in the human oocyte-to-embryo transition succeeds in
over 75 percent of embryonic blastomeres (3); moreover, advances
in somatic cell nuclear transfer (SCNT) procedures have revealed
the robust ability of human oocytes to reprogram somatic DNA (4).
uSe OF nOninVASiVe TiMe-lAPSe iMAging TO PReDiCT
DeVelOPMenTAl FATe
Over the last decade, key studies in human embryo development
have used time-lapse imaging of embryogenesis to elucidate the role
of various cell cycle parameters and factors that may predict success
or failure in the embryo reaching the blastocyst stage and beyond.
In 2010, Wong etal. reported the use of time-lapse microscopy and
gene expression profiling at the single embryo and blastomere level
to document the growth of human embryos from zygote to blastocyst (3). Key developmental features were extracted and algorithms
generated to predict success and failure in development leading up
to the blastocyst stage; morphological assessment was augmented
by a comparison of gene expression in embryos that succeeded or
failed to develop. These studies indicated that the characteristics
of the first cytokinesis and the first two cleavage divisions could be
used as markers for a positive outcome. Successful development
was shown for the first time to be highly regulated, following strict
timing in pre-implantation development even prior to EGA. In normal
development, degradation of maternal mRNA was shown to precede
16
cell autonomous EGA. In contrast, arrested embryos most often displayed aberrant cytokinesis during the first cleavage divisions prior
to EGA, accompanied by equally aberrant gene expression in the
embryo or individual blastomeres. Since the studies by Wong and
colleagues suggested that human embryo fate could be predicted
accurately prior to EGA, it was proposed that success and failure is
often inherited. Embryos that begin life with defective maternal programs, or inherit defective paternal components, are likely to display
aberrant cytokinesis, embryo fragmentation, and arrest of individual
blastomeres or the entire embryo.
The methods and algorithms described by Wong et al. provided
a platform for improved and earlier diagnosis of embryo potential to
hopefully allow for the transfer of fewer embryos earlier in development during assisted reproduction procedures. Subsequent reports
have documented that the parameters identified are able to provide
predictive power beyond the blastocyst stage, to implantation, and
that other parameters may be added for ease of use or to adapt the
technology to clinics that differ in terms of cell culture media, patient
base, or other characteristics (5).
exTenSiOn OF STuDieS TO unDeRSTAnDing
geneSiS OF AneuPlOiDy
Based on the results of Wong et al., we hypothesized that measurement and analysis of specific parameters in preimplantation human
embryos could provide insight into fundamental characteristics of the
embryo, including ploidy. We first sought to correlate normal and abnormal development by correlating continual imaging with genetic
that a combination of time-lapse parameter analysis, along with assessment of blastomere fragmentation dynamics (Fig. 2), may reduce the inadvertent transfer of aneuploid embryos, with implications
for decreasing miscarriage risk and improving in vitro fertilization
success. Based on this work, we have offered a model of aneuploidy
development during human embryogenesis that is currently being
further examined (Fig. 3).
analysis of single embryos and embryonic blastomeres (6). Results
indicated that euploid embryos could not be accurately distinguished
from those with aneuploidies when only cell cycle parameters were
analyzed. By adding additional confocal microscopy analysis (which
included reconstruction of all blastomere karyotypes to the four-cell
stage), however, we concluded that the complexity in human embryonic aneuploidy is not simply due to errors on the meiotic/mitotic
spindle. Rather, results suggested that generation of aneuploidy may
also occur during interphase and can be explained by the containment of a subset of missing chromosomes within cellular fragments,
which bud off from embryonic blastomeres and may persist or become reabsorbed. We also noted that chromosome-containing fragments may arise from micronuclei, or embryonic structures distinct
from primary nuclei, with encapsulated chromosome(s) detected
only in the blastomeres of cleavage-stage human embryos (Fig. 1).
These findings suggest that individual human blastomere behavior is
diagnostic of chromosomal status and provides the first example of
the use of non-invasive imaging and automated fragmentation tracking to distinguish euploid and aneuploid cells. These studies predict
SuMMARy
Numerous basic and clinical laboratories are now investigating the use of noninvasive time-lapse imaging to provide reliable information regarding the development of human embryos
to the blastocyst stage, implantation, and beyond. It is hoped
that advances will enable the separation of those embryos
that are most likely to develop successfully from those that likely would result in miscarriage or still-birth due to severe birth
defects.
REFERENCES
1. K. Niakan, J. Han, R. Pedersen, C. Simon, R. R. Pera, Development
139, 829 (2012).
2. Z. Xue et al., Nature 500, 593 (2013).
3. C. Wong et al., Nat. Biotechnol. 28, 1115 (2010).
4. M. Tachibana et al., Cell 153, 1228 (2013).
5. M. Meseguer et al., Hum. Reprod. 26, 2658 (2011).
6. S. Chavez et al., Nat. Commun. 3, 1251 (2012).
17
Produced by the Science/AAAS Custom Publishing Office
Environmentally Induced Epigenetic Transgenerational
Inheritance of Disease
Figure 2. Role of germline
in epigenetic transgenerational inheritance. An environmental factor acts on the
F0 generation gestating female (left) to influence the developing F1 generation fetus,
resulting in reprogramming of
the methylation state of the
primordial germ cell DNA.
This altered DNA methylation pattern becomes fixed,
similar to an imprinted gene,
and is transferred through the
germline to subsequent generations. Somatic cells in the
offspring in later generations
also carry the altered epigenome, generating an aberrant transcriptome, and promoting adult-onset disease.
Modified from (4).
Michael K. Skinner
inTRODuCTiOn
We have investigated the effects
of two environmental toxicants
(vinclozolin and methoxychlor)
following exposure of rats during
fetal gonadal sex determination,
and the impact on gonadal development and function (1,2). A serendipitous observation was made
when F1 generation animals were
mistakenly bred to generate the
F2 generation offspring: the vast
majority of the testes in the F2
generation carried a spermatogenic cell defect that induced
apoptosis. This prompted a multiple year study that demonstrated
the phenomenon of environmentally induced epigenetic transgenerational inheritance of disease
for the first time (3). This study
showed an increase in apoptosis
in testis spermatogenic cells in the
F1, F2, F3, and F4 generations,
as well as in outcrossed offspring,
through non-Mendelian genetic
inheritance, affecting 90% of the male population. The causative epigenetic effects were traced to specific DNA methylation sites. The
impact of this study on our understanding of the molecular control of
inheritance, disease etiology, and environmental toxicology is significant. Over the past ten years a series of studies have further investigated this phenomenon, including environmental factor specificity,
germline epigenetics, and disease etiology (4,5).
18
Figure 1. Environmenally induced epigenetic
transgenerational inheritance of disease. The
environmental factors such as toxicants, nutrition,
and stress influence a gestating female during the
period of fetal gonadal sex determination to then
affect disease incidence (percentage) in the F1,
F2, F3, and F4 generations. The disease incidence
for vinclozoiin lineage males compared to control
lineage animals is shown for each generation. The
incidence of tumor development, prostate disease, kidney disease, testis disease, and immune
abnormalities are presented. Modified from (20).
ePigeneTiC TRAnSgeneRATiOnAl inheRiTAnCe
The definition of epigenetic transgenerational inheritance is “germline mediated inheritance of epigenetic information between generations that leads to phenotypic variation in the absence of direct
environmental influences” (6). This is in contrast to epigenetic inheritance that involves direct environmental germline or somatic cell
exposure, and epigenetic responses during early development that
influence phenotypes in later life. An example is the Agouti mouse
model, which responds to an environmental signal in utero through a
change in an epiallele, thus shifting the coat color of the offspring (5,
7). Environmental epigenetic inheritance responses may be corrected in subsequent generations during epigenetic reprogramming of
the germline or early embryo, such that the phenotype is lost (8,9).
In order for environmentally induced epigenetic transgenerational
inheritance to occur, the external insult must act on a gestating female during fetal gonadal sex determination, influencing the epigenetic programming through altered DNA methylation patterns in the
germline (Fig. 1). The epigenetic information is then carried through
the epigenome of the germline, into the embryonic stem cell, and
thus to all somatic cells of the offspring. Susceptible tissues in offspring will potentially develop disease and the defect will continue to
be transgenerationally inherited (Figs. 1 and 2) (3–5). Several studies, described below, support this hypothesis of the molecular etiology of epigenetic transgenerational inheritance.
Thisarticlebasedontheoriginalpublication:
M. D. Anway et al., Science 308, 1466 (2005).
Center for Reproductive Biology, School of Biological Sciences,
Washington State University, Pullman, WA
[email protected]
geRMline ePiMuTATiOnS AnD exPOSuRe
(TOxiCAnT) SPeCiFiCiTy
The environmental compound (toxicant) administered in the initial
study was vinclozolin, one of the most widely used agricultural fungicides (3). The outcross information from this work indicated that
the transgenerational phenotype was transmitted through the male
sperm (3). A genome-wide promoter analysis identified approximately 50 differentially methylated regions (DMRs) in the sperm of
the F3 generation from vinclozolin-treated animals when compared
with sperm from matched control (vehicle exposed) F3 rats (10). The
DMRs carry epimutations that can transmit this epigenetic information transgenerationally (4).
A critical question was whether this phenomenon was unique to
vinclozolin. A series of studies investigated the actions of dioxin (11),
a pesticide and an insect repellent (permethrin and DEET) (12),
plastics (BPH and phthalates) (13), and hydrocarbons (JP8 jet fuel)
(14). All were found to promote the transgenerational inheritance of
sperm epimutations and of disease (15). Interestingly, each toxicant
promoted a unique set of epimutations with negligible overlap (Fig.
3) (15). These studies did not measure true environmental risk, but
provide a good foundation for future analysis to determine the environmental hazards of exposure. Others have now shown transgenerational inheritance of disease as a result of exposure to various
environmental factors including nutrition (16), stress (17), and other
toxicants (18). Combined observations suggest that epigenetic biomarkers for ancestral toxicant exposure and adult onset disease
may exist.
TRAnSgeneRATiOnAl inheRiTAnCe
OF DiSeASe AnD PhenOTyPiC VARiATiOn
The first transgenerational abnormality observed was an increase
in spermatogenic cell apoptosis (3, 19). Other abnormal pathologies seen (Fig. 1) included prostate disease (11–15, 20, 21), kidney disease (11–14, 20), mammary tumors (20), immune system
abnormalities (14, 20), and behavioral effects such as anxiety
(22). Other laboratories have shown transgenerational effects on
reproduction (23, 24), stress response (25), and obesity (14, 26).
Some transgenerational diseases were induced by a variety of
different environmental exposures (15), including polycystic ovaries and reduction of primordial ovarian follicle pool size, which
were found in the majority of females from all exposure groups
examined (27).
Figure 3. Venn diagram of transgenerational epimutations associated with different exposure groups showing a number of common epimutations in the F3 generation of rats due to ancestral exposure of F0 generation gestating females to dioxin,
pesticide, plastics, or hydrocarbons. Modified from (15).
Epigenetic transgenerational inheritance may also impact other areas of biology. One study found that the F3 generation of vinclozolintreated rats had significant altered mate preference behavior (28).
Since sexual selection is a major determinant in evolutionary biology, this work suggests that transgenerational epigenetics may play
a critical role in evolution (4).
TRAnSgeneRATiOnAl SOMATiC TRAnSCRiPTOMeS
AnD The eTiOlOgy OF RePRODuCTiVe DiSeASe
Transgenerational germline transmission of epimutations results in
all somatic cells in offspring carrying an altered methylation pattern
and transcriptome (4, 6). Building upon earlier transgenerational
transcriptome studies in fetal rat testis (29), we examined the
19
Produced by the Science/AAAS Custom Publishing Office
Figure 4. Schematic showing a 2–5 Mbp
epigenetic control region containing multiple distal
gene regulatory units (black boxes) under the
control of a differentially methylated region (white
triangle, DMR) and a long noncoding RNA (white
box, lncRNA). Modified from (30).
transgenerational transcriptomes of 11 different tissues in male
and female vinclozolin-treated versus control lineage rats (30).
All tissues had a transgenerational transcriptome that was unique
to the specific tissue, with negligible overlap between tissues. It is
intriguing that a relatively small number of epimutations can produce
such a large number of specific transcriptome changes (30). The
identification of epigenetic control regions—areas of 2–5 megabases
with an overrepresentation of regulated genes close to DMRs and
long noncoding RNA regions—may provide a clue (Fig. 4) (30),
suggesting unique molecular regulatory mechanisms that will require
further investigation.
To further elucidate the role of epimutations in adult onset disease,
the molecular etiology of ovarian diseases were studied (27). Follicular somatic granulosa cells were found to have an altered epigenome
and transcriptome that suggested specific signaling pathways were
affected. Similarly, somatic Sertoli cells in the testis were found in a
separate study to have transgenerational alterations in their epigenomes and transcriptomes, affecting genes previously shown to be
involved in male infertility (31).
iMPACT AnD FuTuRe STuDieS
Our original publication (3) described the phenomenon of environmentally induced epigenetic transgenerational inheritance of a
disease. Subsequent publications confirmed and clarified the molecular and physiological parameters involved. The research shows
the existence of a non-genetic (i.e., epigenetic) form of transgenerational inheritance that impacts the etiology of certain diseases,
as well as a potential molecular mechanism describing how environmental factors could directly influence gene expression and
therefore disease (4, 5). Further studies clearly are needed to
clarify the role of epigenetics and transgenerational inheritance in
disease etiology, evolutionary biology, and other areas of cell and
developmental biology. The specific next steps needed include investigation into why specific sites are more susceptible to becoming transgenerationally programmed and the mechanism by which
this occurs, as well as the translation of this work from animals to
humans. These and other studies will undoubtedly have significant impact on our understanding of normal biology and disease
etiology.
REFERENCES
1. A. S. Cupp et al., J. Androl. 24, 736 (2003).
2. M. Uzumcu, H. Suzuki, M. K. Skinner, Reprod. Toxicol. 18, 765
(2004).
3. M. D. Anway, A. S. Cupp, M. Uzumcu, M. K. Skinner, Science 308,
1466 (2005).
20
4. M. K. Skinner, M. Manikkam, C. Guerrero-Bosagna, Trends
Endocrinol. Metab. 21, 214 (2010).
5. R. L. Jirtle, M. K. Skinner, Nat. Rev. Genet. 8, 253 (2007).
6. M. K. Skinner, Epigenetics 6, 838 (2011).
7. M. E. Blewitt, N. K. Vickaryous, A. Paldi, H. Koseki, E. Whitelaw,
PLoS Genet. 2, e49 (2006).
8. R. R. Newbold, Toxicol. Appl. Pharmacol. 199, 142 (2004).
9. R. A. Waterland, M. Travisano, K. G. Tahiliani, FASEB J. 21, 3380
(2007).
10. C. Guerrero-Bosagna, M. Settles, B. Lucker, M. Skinner, PLoS ONE
5, e13100 (2010).
11. M. Manikkam, R. Tracey, C. Guerrero-Bosagna, M. K. Skinner, PLoS
ONE 7, e46249 (2012).
12. M. Manikkam, R. Tracey, C. Guerrero-Bosagna, M. Skinner, Reprod.
Toxicol. 34, 708 (2012).
13. M. Manikkam, R. Tracey, C. Guerrero-Bosagna, M. Skinner, PLoS
ONE 8, e55387 (2013).
14. R. Tracey, M. Manikkam, C. Guerrero-Bosagna, M. Skinner, Reprod.
Toxicol. 36, 104 (2013).
15. M. Manikkam, C. Guerrero-Bosagna, R. Tracey, M. M. Haque, M. K.
Skinner, PLoS ONE 7, e31901 (2012).
16. R. A. Waterland, Horm. Res. 71 Suppl. 1, 13 (2009).
17. P. Jensen, Prog. Biophys. Mol. Biol. (Epub ahead of print) (2013). doi:
10.1016/j.pbiomolbio.2013.01.001.
18. V. Bollati, A. Baccarelli, Heredity (Edinb.) 105, 105 (2010).
19. M. D. Anway, M. A. Memon, M. Uzumcu, M. K. Skinner, J. Androl.
27, 868 (2006).
20. M. D. Anway, C. Leathers, M. K. Skinner, Endocrinol. 147, 5515
(2006).
21. M. D. Anway, M. K. Skinner, Prostate 68, 517 (2008).
22. M. K. Skinner, M. D. Anway, M. I. Savenkova, A. C. Gore, D. Crews,
PLoS ONE 3, e3745 (2008).
23. E. E. Nilsson, M. D. Anway, J. Stanfield, M. K. Skinner, Reproduction
135, 713 (2008).
24. T. J. Doyle, J. L. Bowman, V. L. Windell, D. J. McLean, K. H. Kim,
Biol. Reprod. 88, 112 (2013).
25. D. Crews et al., Proc. Natl. Acad. Sci. U.S.A. 109, 9143 (2012).
26. R. Chamorro-Garcia et al., Environ. Health Perspect. 121, 359
(2013).
27. E. Nilsson et al., PLoS ONE 7, e36129 (2012).
28. D. Crews et al., Proc. Natl. Acad. Sci. U.S.A. 104, 5942 (2007).
29. M. D. Anway, S. S. Rekow, M. K. Skinner, Genomics 91, 30 (2008).
30. M. K. Skinner, M. Manikkam, M. M. Haque, B. Zhang, M.
Savenkova, Genome Biol. 13, R91 (2012).
31. C. Guerrero-Bosagna, M. Savenkova, M. M. Haque, I. SadlerRiggleman, M. K. Skinner, PLoS ONE 8, e59922 (2013).
Aberrant Endocannabinoid Signaling is a Risk Factor for Ectopic Pregnancy
Xiaofei Sun and Sudhansu K. Dey*
According to the U.S. Centers for Disease Control and
Prevention, ectopic pregnancy is the leading cause of
pregnancy-related death during the first trimester (1). In the
United States, around 100,000 women encounter ectopic
pregnancies each year, and 6% of maternal deaths are
attributable to ectopic pregnancies (2). Although ectopic
pregnancy adversely affects female reproductive health, its
etiology remains elusive. It was discovered that cannabinoid
receptor 1 (CB1)-deficient mice demonstrated spontaneous
oviductal retention of embryos at the blastocyst stage (3),
making these mice a good model to study certain aspects of
ectopic pregnancy.
In normal pregnancy, eggs are fertilized and undergo further development to embryos within the oviduct in mice or the
loss in Cb1-/- mice. (A)
Fallopian tubes in humans. Mouse embryos develop within Figure 1. Impaired oviductal embryo transport causes pregnancy
Percentage of embryos recovered from oviducts or uteri in Cb1-/- and wild-type (WT) fethe oviduct for about 72 hours, and then transit through the males mated with WT males, on day 4 of pregnancy. Oviductal retention of embryos is
utero-tubal junction to enter the uterus. The normal passage observed in Cb1-/- females. (B) A representative histological section of a day 7 pregnant
-/of embryos through the oviduct into the uterus requires coor- Cb1 oviduct showing a trapped blastocyst (Bl, arrow) at the isthmus. Bar, 100 µm. Mus,
muscularis; S, serosa; Mu, mucosa. The figure is adapted from (3).
dinated oscillation of the cilia on the oviductal epithelium, as
well as muscle contractions. Upon arrival in the uterine cavity, embryos develop to the blastocyst stage and become activated played in the mouse oviduct. Nonetheless, mouse models have been
(implantation competent); they can then implant in the uterus if the used to study certain aspects of oviductal embryo retention (8). One
uterine environment is conducive to implantation.
such rodent model, CB1-deficient mice, was the first to demonstrate
In ectopic pregnancies, embryos implant outside of the uterine spontaneous oviductal retention of embryos, providing a useful tool
cavity; in humans, 98% of ectopic pregnancies occur in the Fallo- to study certain mechanisms of ectopic pregnancy (9).
pian tubes and are known as tubal pregnancies. Two prerequisites
CB1 is a G protein-coupled cannabinoid receptor that is targeted
of tubal pregnancy are Fallopian tube retention of embryos and an by cannabis and endocannabinoids, a group of endogenous cannaabnormal tubal environment that is conducive to implantation. A di- bis-like compounds that act as lipid mediators. The two most extenagnosis of tubal pregnancy often requires multiple visits to the doc- sively studied endocannabinoids are anandamide (AEA) and 2-arator and there is the danger that the implanted embryo will rupture chidonoylglycerol (2-AG) (9). Levels of endocannabinoids in vivo are
within the Fallopian tube potentially resulting in maternal death (4). tightly regulated by enzymes controlling their synthesis and degradaAlthough some of the risk factors for tubal pregnancy are known, tion. The magnitude and duration of AEA signaling is regulated by a
such as tubal damage from surgery or infection, cigarette smoking, membrane-bound fatty acid amide hydrolase (FAAH), which is also
and in vitro fertilization procedures, the precise mechanism by which capable of degrading 2-AG to arachidonic acid and glycerol (9). In
embryo implantation in the Fallopian tube occurs is not completely mice, endocannabinoids and CB1 are present in the oviduct. FAAH
understood (5).
protein levels in the isthmus are lower than those in the ampullary
Tubal pregnancy occurs rarely in other mammals and the lack of region (10,11), suggesting a gradient of AEA in the oviduct.
animal models has made it difficult to study the underlying mechaIn mice, the movement of embryos within the oviducts is aided
nisms. Extra-uterine implantation usually occurs in the abdominal mainly by the beating of cilia located on the oviductal epithelium
cavity in animals, and just a few cases of tubal pregnancy in primates (12, 13); meanwhile, the transport of embryos from the oviduct to
have been reported (6,7). There are no known cases of full ectopic the uterus is facilitated by a wave of oviductal muscle movement
pregnancy in mice, possibly because the walls of mouse oviducts (14). Muscular movement is controlled by the peripheral sympathetic
are thin compared with human Fallopian tubes. It is also possible nervous system (15). Stimulation of the β2 adrenergic receptor (β2that implantation-specific genes expressed in the uterus are not dis- AR) causes sphincter muscle relaxation, whereas stimulation of the
α1-AR produces muscle contraction. It is believed that reciprocal
stimulation of these two receptors causes a wave of contraction and
relaxation, which is conducive to the passage of embryos from the
Thisarticlebasedontheoriginalpublication:
oviduct into the uterine lumen (14,15).
H. Wang et al., Nature Med. 10, 1074 (2004).
Division of Reproductive Sciences, Perinatal Institute, Cincinnati
Our group has shown that oviductal retention of embryos occurred
Children’s Research Foundation, Cincinnati, OH
inCB1-deficient mice (3). Reciprocal embryo transfer experiments
*Corresponding Author: [email protected]
21
Produced by the Science/AAAS Custom Publishing Office
further confirmed that maternal CB1 is critical for appropriate
oviductal transport of embryos, since only CB1-deficient recipients
displayed oviductal retention irrespective of embryo genotype
(Fig. 1) (3). Our laboratory also found that higher cannabinoid
signaling affects oviductal embryo transport. Wild-type (WT) mice
exposed to Δ9-tetrahydrocannabinol (THC, the active psychoactive
component of marijuana) or methanandamide (a stable AEA analog)
showed similar oviductal embryo retention, which was rescued
by treatment with a CB1 antagonist (11). Studies using FAAHdeficient mice further confirmed that higher oviductal AEA levels
disrupt normal oviductal embryo transport. Taken together, the
results demonstrated that either higher or lower endocannabinoid
signaling impairs normal wave movement through the oviduct,
and consequently derailed normal oviductal transportation of
embryos.
Our studies in mice stimulated research on ectopic pregnancy in
women and many of the findings in humans were consistent with
the results of the mouse studies. In women with ectopic pregnancies, the Fallopian tubes and endometria showed lower levels of
CB1 compared with those in women with normal pregnancies (16).
AEA levels in ectopic Fallopian tubes were significantly higher
than those in normal pregnancies (17), and increased AEA levels were associated with low FAAH activity in ectopic Fallopian
tubes. In addition, treatment with oleoylethanolamide, an endocannabinoid, significantly decreased cilia oscillation frequency
displayed on the Fallopian tube epithelium (18). These results
suggest that disturbances in cannabinoid signaling during early
pregnancy could result in ectopic pregnancy in humans. It would
be interesting to explore whether polymorphisms in the gene encoding the CB1 receptor are associated with ectopic pregnancy
in humans.
A national survey of marijuana use from 1991 to 2011 in students
grades 9–12 showed a rise in usage of this Schedule I drug (19).
Since synthetic cannabinoids appeared in 2008, the popularity
among high school seniors and persons under 25 years of age has
increased sharply (20,21). Synthetic cannabinoids are found in illicit
“fake marijuana” with street names such as “Spice” or “K2.” Many
synthetic cannabinoids have much higher affinities for cannabinoid
receptors and are more potent compared to natural cannabis. This
increase in marijuana and synthetic cannabinoid use is potentially
troubling when combined with the aforementioned statistic that
ectopic pregnancy accounts for about 6% of all pregnancy-related
deaths in the United States (2). Therefore, our studies not only
provide a useful animal model to study ectopic pregnancy, but also
draw more public attention to the adverse effects of maternal usage
of marijuana and synthetic cannabinoids.
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21
REFERENCES
CDC, MMWR Morb. Mortal Wkly Rep. 44, 46 (1995).
K. W. Hoover, G. Tao, C. K. Kent, Obstet. Gynecol. 115,
495 (2010).
H. Wang et al., Nat. Med. 10, 1074 (2004).
S. Senapati, K. T. Barnhart, Fertil. Steril. 99, 1107 (2013).
J. L. Shaw, S. K. Dey, H. O. Critchley, A. W. Horne, Hum.
Reprod. Update 16, 432 (2010).
C. P. Jerome, A. G. Hendrickx, Vet. Pathol. 19, 239 (1982).
J. K. Brown, A. W. Horne, Curr. Opin. Obstet. Gyn. 23, 221
(2011).
J. Zenteno, C. Silva, H. Cardenas, H. B. Croxatto, J. Reprod. Fertil.
86, 545 (1989).
H. Wang, S. K. Dey, M. Maccarrone, Endocr. Rev. 27, 427 (2006).
Y. Guo et al., J. Biol. Chem. 280, 23429 (2005).
H. Wang et al., J. Clin. Invest. 116, 2122 (2006).
S. A. Halbert, P. Y. Tam, R. J. Blandau, Science 191, 1052 (1976).
S. A. Halbert, D. R. Becker, S. E. Szal, Biol. Reprod. 40, 1131
(1989).
G. R. Howe, D. L. Black, J. Reprod. Fertil. 33, 425 (1973).
R. D. Heilman, R. R. Reo, D. W. Hahn, Fertil. Steril. 27, 426
(1976).
A. W. Horne et al., PLoS ONE 3, e3969 (2008).
A. K. Gebeh, J. M. Willets, E. L. Marczylo, A. H. Taylor, J. C.
Konje, J. Clin. Endocr. Metab. 97, 2827 (2012).
A. K. Gebeh et al., J. Clin. Endocr. Metab. 98, 1226 (2013).
CDC. (The National Youth Risk Behavior Survey, 2011). http://
www.cdc.gov/healthyyouth/yrbs/pdf/us_drug_trend_yrbs.pdf
ONDCP. (Office of National Drug Control Policy Fact Sheets synthetic drugs, 2012). http://www.whitehouse.gov/ondcp/ondcpfact-sheets/synthetic-drugs-k2-spice-bath-salts
http://www.acep.org/uploadedFiles/ACEP/News_Room/NewsMedi
aResources/Media_Fact_Sheets/Synthetic%20Drugs%20
Fact%20 Sheet-Final.pdf
Acknowledgments: We thank Serenity Curtis for her careful editing of
the manuscript. Work in the Dey laboratory was funded in part by the
National Institute on Drug Abuse and National Institute of Child Health
and Human Development at the U.S. National Institutes of Health. XS was
supported by a Lalor Foundation postdoctoral fellowship in reproductive
biology.
Single Embryo Transfer in IVF: Live Birth Rates and Obstetrical Outcomes
Christina Bergh
The wide application of in vitro fertilization has caused a dramatic increase in multiple births, associated with adverse neonatal outcome,
mainly as a result of the transfer of several embryos per cycle. Randomized controlled trials, observational studies, and meta-analyses
were carried out to investigate pregnancy and live-birth rates after
single (SET) or double embryo transfer (DET) as well as obstetrical
outcomes. Results showed that the live-birth rate was significantly
higher after DET than SET if only fresh cycles were compared. If one
additional frozen/thawed SET was added to the SET group, live-birth
rates did not differ substantially. Obstetrical outcomes were considerably improved with the SET strategy. We therefore conclude that
SET is the more desirable option for a majority of patients.
inTRODuCTiOn
The rapid development of different in vitro fertilization (IVF) techniques has resulted in increasing pregnancy and live-birth rates per
cycle. In parallel, a dramatic increase in multiple births has occurred.
Numerous publications have demonstrated higher risks for an adverse outcome, including preterm birth (PTB, <37 weeks), low birth
weight (LBW, <2,500 g), and perinatal mortality for children born
after IVF, compared with children born after spontaneous conception, mainly due to the high rate of multiple births following IVF (1–3).
Also, for IVF singletons, increased rates of PTB and LBW were noted
(3–5), likely caused by inherent characteristics of the infertile couple
such as the mother’s age and nulliparity. An increase in the frequency of neurological sequele has also been found, strongly associated
with PTB and multiple births (6). An increased rate of physical malformations has been reported for children born following IVF (7).
The most important factor affecting the rate of multiple births is the
number of embryos transferred during IVF. Reducing the number of
transferred embryos from three to two had little effect on live births,
but decreased the rate of multiple births; the observed twin rate was,
however, still high (8). Data from large registry studies—many performed in Sweden—on obstetrical outcomes after IVF showed an increased rate of severe medical problems in IVF children, generating
an intensive debate in Sweden among obstetricians, paediatricians,
doctors in reproductive medicine, and politicians. There was a call
for transferring only one embryo during IVF, a strategy supported
by some reports that single embryo transfer results in satisfactory
pregnancy rates (9).
The aim in the trial discussed here (10) was to investigate whether
a similar live-birth rate could be achieved if two embryos were transferred one at a time—one fresh embryo, and if no live birth was de-
Thisarticlebasedontheoriginalpublication:
A. Thurin et al., N. Engl. J. Med. 351, 2392 (2004).
Department of Obstetrics and Gynaecology, Institute of Clinical
Sciences, Sahlgrenska Academy, Gothenburg University,
Reproductive Medicine, Sahlgrenska University Hospital,
Gothenburg, Sweden
[email protected]
22
tected, one additional frozen/thawed embryo—rather than the standard practice of transferring two embryos on a single occasion, and
whether this would decrease the multiple birth rate.
MeThODS
Between 2000 and 2003, eleven clinics in Sweden, Denmark, and
Norway participated in the trial in which 661 women under 36 years
of age, performing their first or second IVF cycle and having at least
two good quality embryos available for transfer, were randomly assigned to two groups, SET and DET. The study was double blind,
so neither the physician nor the patient knew whether one or two
embryos had been transferred. The number of patients needed in
each group (330) was based on the assumptions that the true livebirth rate in both groups would be 0.30 and the probability is 0.80 that
the upper limit of the 95% confidence interval (CI) for the difference
in live-birth rates between the groups did not exceed 0.10 (a=0.05).
MAin ReSulTS
The results showed that the live-birth rate was 128/330 (38.8%) in
the SET group and 142/331 (42.9%) in the DET group (95% CI for
the difference: 0.3-11.6). Thus equivalence between the two groups
could not be demonstrated, but the live birth rate did not differ substantially between SET and DET. Moreover, the multiple birth rate
was dramatically reduced in the SET compared to the DET group:
0.8% (1) vs 33.3% (47) (p<0.001). Most important, the neonatal outcome was considerably better for children born to the SET group,
with significantly lower rates of PTB (11.6% vs. 29.1%, p=0.002)
and LBW (7.8% vs. 27.5%, p<0.0001). The SET group showed less
severe neonatal complications demanding neonatal care in hospital
(17.8% vs. 33.9%, p=0.003) and also a lower rate of complex morbidity (7.8% vs. 24.3%, p=0.0001) (11). A financial analysis indicated
that the SET strategy was cost-effective; the incremental cost-effectiveness-ratio (ICER) was €73,307 ($99,089) per additional delivery
in the DET alternative.
FuRTheR DeVelOPMenT
Several randomized trials and meta-analyses (12, 13) have compared the delivery rates in SET and DET groups. The studies have
shown significantly higher pregnancy and delivery rates after DET
compared to SET when only fresh cycles were compared. Performing an additional frozen/thawed SET in the SET group resulted in a
delivery rate comparable with that in the DET group (10). The Scandinavian countries, particularly Sweden, have played a pioneering
role in reducing multiple births by introducing SET on a large scale.
In Sweden, this strategy has resulted in no change in delivery rates
for fresh or frozen/thawed cycles, while the rate of multiple births
has decreased dramatically, from 25% to around 5%. The overall
SET rate is 70%–80% (Fig. 1). Delivery-and multiple birth rates
after SET and DET according to age are illustrated in Figure 2. A
gradual increase in SET rates has been seen worldwide, including
23
Produced by the Science/AAAS Custom Publishing Office
PeRCenT
ated with multiple births,
there is still an ongoing
debate whether, in particular, twins are a desired outcome for IVF. It
has been claimed that
the risks associated with
IVF twins are dramatically exaggerated and that
twins should be encouraged (17).
There are also financial
variables for patients involved to be considered,
yeAR
and large differences exist in reimbursement systems for IVF in different
countries, which influences utilization of SET.
Figure 1. Delivery rate, multiple-birth rate, and single embryo transfer rate in Sweden 2000–2011 (fresh cycles).
In countries where IVF is
Scandinavia and other northern European countries such as Bel- completely or partially covered by public funding, SET has been acgium and the Netherlands, as well as in Australia, New Zealand, and cepted more readily. If patients are self-funded, they might choose
Japan. The United States has lagged behind in applying SET (14). more embryos to be transferred in order to maximize the chance of
Although the rate of multiple births has decreased in recent years, becoming pregnant.
it is still high in many countries. The latest report indicates that the
Other factors impacting SET acceptance include a lack of knowlmultiple birth rate in Europe is 22% (2008) and in the USA is 31% edge concerning the risks of multiple births, failure to consider cumu(2009) (15,16).
lative live birth rates that include frozen/thawed cycles, differences
in legislation between countries, and impediments stemming from
RATiOnAle FOR nOT APPlying SeT
cultural beliefs.
Fears among both patients and clinicians that SET will lower pregnancy and live-birth rates is probably the most common reason given COnCluDing ReMARkS
for not applying SET more broadly. A focus on short-term higher suc- The most important risk associated with IVF is the higher rate of
cess rates (pregnancy rate per cycle), rather than optimal outcomes, multiple births, resulting in increased child morbidity and mortality.
i.e., the birth and health of a singleton, has slowed progress in this In addition to problems in the neonatal period, preterm birth and low
birth weight may have long-term consequences for future health. Acarea.
Despite the overwhelming data indicating increased risks associ- cording to the Barker hypothesis (18), these adverse outcomes may
Figure 2. Percent delivery per SET and DET, percent multiple birth per SET and DET, and the overall SET rate according to age of the mother. National data
from the Swedish Quality Registry for Assisted Reproduction (www.ucr.uu.se/qivf/) for 2011 (n=9317 fresh IVF cycles).
24
lead to increased risk of type 2 diabetes and cardiovascular diseases
in adulthood.
The association between IVF and a poorer obstetric outcome for
singletons has also been widely documented, but is quantitatively a
much smaller problem. Maternal characteristics seem to be the largest contributors to these adverse outcomes (19), while the contribution of IVF-related factors is less clear.
Recent data from Sweden indicates that it is possible to
have two consecutive singleton births in the same or similar number of fresh IVF cycles using the SET strategy as with
the DET strategy by simply adding a small number of frozen/
thawed cycles, while concomitantly avoiding the risk of multiple
births (20).
Current knowledge strongly supports SET as an IVF strategy that
is safe and effective for mothers and children.
REFERENCES
1. T. Bergh , A. Ericsson , T. Hillensjö , KG. Nygren , U. B. Wenner
holm, Lancet 354, 1579 (1999).
2. L. A. Schieve et al., N. Engl. J. Med. 346, 731 (2002).
3. F. M. Helmerhorst, D. A. Perquin, D. Donker, M. J. Keirse, BMJ 328,
261 (2004).
4. R. A. Jackson, K. A. Gibson, Y. W. Wu, M. S. Croughan, Obstet.
Gynecol. 103, 551 (2004).
5. S. D. McDonald et al., Eur. J. Obstet. Gynecol. Reprod. Biol.
146,138 (2009).
6. B. Strömberg et al., Lancet 359, 461 (2002).
7. C. Bergh, U. B. Wennerholm, Best Pract. Res. Clin. Obstet. Gynecol.
26, 841 (2012).
8. A. Templeton, J. K. Morris, N. Engl. J. Med. 339, 573 (1998).
9. S. Vilska, A. Tiitinen, C. Hyden-Granskog, O. Hovatta, Hum. Reprod.
14, 2392 (1999).
10. A. Thurin et al., N. Engl. J. Med. 351, 2392 (2004).
11. A. Thurin-Kjellberg, P. Carlsson , C. Bergh , Hum. Reprod. 21,
210 (2006).
12. Z. Pandian, S. Bhattacharya, O. Ozturk, G. Serour, A. Templeton,
Cochrane Database Syst. Rev. 15, CD003416 (2009).
13. D. J. McLernon et al., BMJ, 341: epub c6945 doi:10.1136/bmj.c6945
(2010).
14. A. Maheshwari, S. Griffiths, S. Bhattacharya, Hum. Reprod. Update
17, 107 (2011).
15. A.P. Ferraretti et al., Hum. Reprod. 27, 2571 (2012).
16. S. Sunderam et al., MMWR Surveill. Summ. 61, 1 (2012).
17. N. Gleicher, D. Barad. Fertil. Steril. 91, 2426 (2009).
18. D. J. Barker, BMJ 311, 171 (1995).
19. A. Pinborg et al., Hum. Reprod. Update 19, 87 (2013).
20. A. Sazonova, K. Källen, A. Thurin-Kjellberg, U. B.Wennerholm, C.
Bergh, Fertil. Steril. 99, 731 (2013).
Oocyte Vitrification: A Major Milestone in Assisted Reproduction
Ana Cobo
The cryopreservation of female gametes has been pursued since
the onset of assisted reproductive technology (ART). After a lengthy
period of failures and sporadic successes, the introduction of refined
vitrification methods has meant a huge step forward in infertility practice as it allows efficient egg cryostorage. Today, the clinical use of
oocytes that have been stored in cryogenic tanks is an accepted
practice. The very broad scope of this technology encompasses various new areas such as fertility preservation for social or oncological
reasons, the possibility of creating egg banks for ovum donation, and
the opportunity to avoid ovarian hyperstimulation syndrome, or to accumulate oocytes in low-responder patients. Even segmented infertility treatments can be offered by stimulating ovaries, vitrifying embryos, and then transferring them during a natural cycle. All of these
options were not available just a few years ago, but are now being
routinely applied in increasingly more infertility centers worldwide.
Thisarticlebasedontheoriginalpublication:
A. Cobo et al., Fertil. Steril. 89, 1657 (2008).
Instituto Valenciano de Infertilidad, University of Valencia,
Valencia, Spain
[email protected]
inTRODuCTiOn
The development of efficient cryopreservation techniques for female
gametes has been a real challenge as both freezing and thawing
expose oocytes to severe stress that can result in cell death. The
introduction of improved vitrification methods has meant increasingly successful outcomes, where the most commonly applied methodology since the onset of ART—slow freezing—has led to limited
success (1, 2). Among the reasons for failure resulting from slow
freezing, chilling injury and ice formation are recognized as being
the most deleterious (3). Chilling injury is avoided with vitrification
because cooling occurs extremely rapidly and ice formation is circumvented very efficiently by using high concentrations of cryoprotectants (CPAs). Application of vitrification has been limited since it
was first employed in embryology in the early 1980s (4) because the
concentration of CPAs required for the earliest vitrification protocols
can have dramatic toxic effects. Significant changes made to these
protocols have reduced the concentration of CPAs to circumvent deleterious consequences to cells (5). The efficiency of modified protocols has been successfully tested clinically, which is an essential
requisite for fertility preservation, ovum donations, or other clinical
applications, and also to overcome certain clinical obstacles that
ART posed, such as the absence of a partner’s semen sample on
25
Produced by the Science/AAAS Custom Publishing Office
the day of ovum collection.
A comparison of embryo development using vitrified and fresh oocytes was performed (6) to provide valuable information about the
further potential of vitrified oocytes, and to serve as a foundation for
additional studies.
OOCyTe ViTRiFiCATiOn AnD eMBRyO QuAliTy
A prospective cohort study was used to perform a first of its kind
analysis of the quality of embryos emerging from simultaneously
generated vitrified and fresh donor oocytes (6). All of the metaphase
II oocytes assigned to a single recipient were randomly allocated
one of two groups: vitrified (oocytes were vitrified and then warmed
for one hour before implantation) or fresh (maintained in culture at
37°C). Intracytoplasmic sperm injection (ICSI) using the husband’s
semen sample was performed simultaneously on both vitrified and
fresh oocytes. A high overall survival rate was achieved (97%; N=224
of 231 oocytes). No significant differences in fertilization rates for day
one (76.3% vs. 82.2%), day two (94.2% vs. 97.8%), or day three
(77.6% vs. 84.6%) embryo cleavage or blastocyst formation rates
(48.7% vs. 47.5%) were detected for the vitrified and fresh oocytes,
respectively. The ratios of good quality embryos on day three (80.8%
vs. 80.5%) and in the blastocyst stage (81.1% vs. 70.0%) were similar in both groups. Additionally, promising clinical outcomes were
obtained following the transfer of embryos developed from vitrified
oocytes (40.8% for the implantation rate and 47.8% for the ongoing
pregnancy rate). These outcomes are in contrast to those previously
reported by other authors using a slow freezing methodology (1).
The widespread introduction of oocyte vitrification into clinical practice has been greatly strengthened by these findings, which have
demonstrated that both embryo developmental capability and implantation potential are essentially unaltered by oocyte vitrification.
COnTRiBuTiOn OF OOCyTe ViTRiFiCATiOn
TO CuRRenT CliniCAl PRACTiCe
Our findings were further replicated by other authors with both donor and own oocytes [see (7) for review]. In a follow up controlled,
randomized clinical trial we used a refined methodology to compare
clinical outcomes utilizing both cryopreserved and fresh oocytes in
an ovum donation program (8). In this study, vitrified oocytes could
not be shown to be inferior to fresh oocytes in terms of the ongoing
pregnancy rate (odds ratio = 0.921, 95% CI 0.667 to 1.274). This
finding definitively confirms our previous observations regarding the
unimpaired potential of vitrified oocytes to develop into competent
embryos capable of generating ongoing pregnancies in a proportion
comparable to that of fresh oocytes.
Most of the difficulties posed by fresh donations, such as long waiting lists, the need for donor/recipient synchronization, and the lack of
quarantine, can be overcome by oocyte cryobanking. Stored oocytes
are available anytime they are needed, significantly alleviating donation logistics.
The benefits of oocyte cryostorage have also been demonstrated in autologous in vitro fertilization (IVF) cycles (8–10) leading to
26
high cumulative pregnancy rates (11). Rienzi and coworkers have
mirrored our findings on embryo quality and clinical outcomes in
a prospective, randomized sibling-oocyte study conducted with
infertile patients undergoing own-oocyte IVF cycles (9). The consistency and reliability of oocyte vitrification for this population
was further demonstrated in a multicentric study (12). Additionally, important findings regarding the number of oocytes vitrified,
the patient’s age, and the developmental stage of the embryo at
the time of transfer have been published and have proven useful
for patient counseling (12). Oocyte vitrification has also been suggested to be an interesting therapeutic alternative for low responders (13), improving outcomes for a group that has been very difficult
to treat successfully, and may even save patients the disappointment of failure after IVF cycles with a poor response using fresh
oocytes (14).
The extensive research carried out to date has laid the groundwork for the establishment of successful protocols for extremely efficient oocyte cryopreservation. These preservation programs have
provided a viable alternative that avoids the downside of an unfavorable uterine environment resulting from administration of exogenous
gonadotrophins during controlled ovarian hyperstimultation (15–18).
The research described here has fundamentally changed the
clinical landscape and strongly supports the position that oocyte vitrification offers advantageous clinical results in diverse
populations, opening up a wide range of possibilities in the field
of ART.
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
REFERENCES
D. A. Gook, D. H. Edgar, Hum. Reprod. Update 13, 591 (2007).
A. Cobo, C. Diaz, Fertil. Steril. 96, 277 (2011).
G. Vajta, M. Kuwayama, Theriogenology 65, 236 (2006).
W. F. Rall, G. M. Fahy, Nature 313, 573 (1985).
M. Kuwayama, G. Vajta, O. Kato, S. P. Leibo, Reprod. Biomed.
Online 11, 300 (2005).
A. Cobo et al., Fertil. Steril. 89, 1657 (2008).
A. Cobo, J. A. Garcia-Velasco, J. Domingo, J. Remohi, A. Pellicer,
Fertil. Steril. 99, 1485 (2013).
A. Cobo, M. Meseguer, J. Remohi, A. Pellicer, Hum. Reprod. 25,
2239 (2010).
L. Rienzi et al., Hum. Reprod. 25, 66 (2010).
C. C. Chang et al., Fertil. Steril. 99, 1891 (2013).
F. Ubaldi et al., Hum. Reprod. 25, 1199 (2010).
L. Rienzi et al., Hum. Reprod. 27, 1606 (2012).
A. Cobo, N. Garrido, J. Crespo, R. Jose, A. Pellicer, Reprod. Biomed.
Online 24, 424 (2012).
J. Cohen, G. Grudzinskas, M. Johnson, Reprod. Biomed. Online 24,
375 (2012).
B. S. Shapiro et al., Fertil. Steril. 96, 344 (2011).
B. S. Shapiro et al., Fertil. Steril. 96, 516 (2011).
A. Maheshwari, S. Bhattacharya, Hum. Reprod. 28, 6 (2013).
A. Aflatoonian, H. Oskouian, S. Ahmadi, L. Oskouian, J. Assist.
Reprod. Genet. 27, 357 (2010).
Accurate Single Cell 24 Chromosome Aneuploidy Screening
Richard T. Scott, Jr.1,2* and Nathan R. Treff1,2
inTRODuCTiOn
Given a prevalence of one in six couples, infertility is one of the
largest contemporary public health issues in Western countries. In
vitro fertilization (IVF) is now widely available and is the most effective treatment for infertile couples. While clinical outcomes have
improved since IVF was first introduced, the efficiency of the process
remains low. In the most recent U.S. Centers for Disease Control
and Prevention summary of IVF outcomes in the United States, fewer than 17% of embryos deemed of sufficient quality to merit transfer
actually implanted and progressed to delivery. This fact reflects that
morphologic and temporal markers of in vitro embryonic development have only modest predictive value when trying to assess the
true reproductive potential of any single embryo. As a result of this
uncertainly, patients and IVF providers elect to transfer multiple embryos in the hope that one with true reproductive potential will be
included. While improving delivery rates, unacceptable rates of multiple gestations with significantly increased maternal and neonatal
morbidity ensue.
It might seem logical that if test results are consistent among all the
cells in an embryo, then the test is valid. However, embryonic mosaicism is a real phenomenon impacting approximately one in five
embryos and therefore when testing embryonic cells from the same
embryo, some disagreement is expected. When that happens, does
it represent an error in the assay or mosaicism? A number of investigators have attributed these differences to mosaicism, but there is no
scientific rationale to preferentially attribute them to either mosaicism
or laboratory.
Given the critical impact that screening has on management of
individual embryos, the challenges of mosaicism, the fact that the assay may be required to work with only a single cell, and the complexity in demonstrating clinical benefit, validation of embryonic aneuploidy screening is a complex process requiring multiple studies at the
basic laboratory and translational/clinical level. The study reviewed
herein describes the first step in the complex process of validating a
tool—single nucleotide polymorphism (SNP) arrays—for embryonic
aneuploidy screening, namely standardizing the assay.
ACCuRATe ASSeSSMenT OF eMBRyOniC
RePRODuCTiVe POTenTiAl
Assessing embryonic competence brings special challenges not encountered elsewhere in clinical science. The reality is that embryos
deemed to lack reproductive potential are discarded as biologic
waste. Thus a misdiagnosis leads embryologists to throw out an
embryo which might have been capable of becoming a baby. Some
couples only rarely produce a competent embryo or may have limited access to care and thus such a misdiagnosis may lead them to
discard their only meaningful opportunity for delivery. There is no
other area of medicine where a single abnormal laboratory result
causes clinicians or scientists to discontinue care with such irrefutable finality.
TARgeT DnA AMPliFiCATiOn
SNP array technology provides an opportunity to simultaneously investigate the copy number of all 24 chromosomes. Unlike fluorescence in situ hybridization (FISH)-based testing, arrays require the
incorporation of DNA amplification for which a variety of methodologies could be employed. Methods for whole genome amplification
(WGA) were compared for performance on SNP arrays (2). Results
demonstrated superior performance of a PCR-based strategy when
evaluating copy number accuracy (most relevant to aneuploidy
screening).
VAliDATing ASSAyS FOR eMBRyOniC
AneuPlOiDy SCReening
Screening embryos for the correct number of chromosomes (euploidy) represents a well-founded concept given the direct correlation
between increasing maternal age, decreased fecundity, and oocyte
aneuploidy (1). Developing and validating a single cell embryonic
aneuploidy assay is more complex than for other cell types, in part
because access to biologic material for validation is difficult and limiting. There are no cell lines available of embryonic cells at this stage
of development, but discarded embryos can be used for validation.
Thisarticlebasedontheoriginalpublication:
N. R. Treff et al., Fertil. Steril. 94, 2017 (2010).
1
Reproductive Medicine Associates of New Jersey, Basking Ridge, NJ
2
Division of Reproductive Endocrinology, Department of Obstetrics
Gynecology and Reproductive Sciences, Robert Wood Johnson
Medical School, Rutgers University, New Brunswick, NJ
*
Corresponding Author: [email protected]
A FOunDATiOnAl STuDy OF SnP ARRAy SCReening
This study was conducted in three phases (3). Single cells from a
variety of cell lines with known chromosomal abnormalities were
evaluated and data analysis settings were established to give the
highest level of consistency. This calibration was necessary to define
the methodology. Next, the same cell lines were used to prepare
and blind single cell samples for analysis. Blinded analysis provided
a rigorous test of the accuracy of the method on positive controls.
Finally, multiple single cells from human embryos available for research were evaluated to demonstrate consistency of diagnoses in
the relevant tissue type.
AnalysesofCellLines
The SNP array approach analyses the relative intensity of binding for
a large number of SNPs spread across the genome. Average intensities are considered to have a copy number of two. Those with lower
intensity are assigned copy numbers of one (monosomy) and those
with higher intensities are assigned copy numbers of three (trisomy).
Given that embryonic aneuploidy typically involves the gain or loss of
an entire chromosome, the results on a single chromosome should
27
Produced by the Science/AAAS Custom Publishing Office
all be the same. The reality is that there is not uniform intensity for
the SNPs on a single chromosome and individual SNPs may be
assigned copy numbers of one, two, or three. This is overcome by
application of a Gaussian smoothing algorithm that evaluates the
intensities amongst adjacent SNPs and averages them to enhance
accuracy. However, the smoothing algorithms used for conventional
karyotyping of the cell lines or clinical samples, where hundreds of
millions of cells are available, do not function well following whole
genome amplification of a single cell.
The optimal smoothing distance was determined on cells with
known karyotypes. It was important to validate smoothing on chromosomes that were monosomic and trisomic, and not just disomic.
Tolerances were established for the level of discordance amongst
individual SNPs on a single chromosome. The upper limit of discordance, meaning the percentage of SNPs on a given chromosome
that produce varying results, was established for each chromosome.
If that level was exceeded, the assay was deemed “non-concurrent”
and the result considered unreliable. The non-concurrent rate per
chromosome should be <0.1% and per cell should be <2%. These
data were all generated on samples where the karyotype of the cell
being tested was known. Functionally, this is starting with the answer
and working backwards; a critical step in the process, but far from
sufficient to validate the assay.
In the second phase, the prospective blinded evaluation, the
testing algorithm was applied to blinded samples. The accuracy
per chromosome was >99.9%. More importantly, the overall accuracy of single cell aneuploidy screening using this methodology
was 98.6%.
AnalysesofSingleCellsfromArrestedEmbryos
In the third phase, individual cells from arrested, cleavage-stage embryos were evaluated. Given the underlying mosaicism, there was
no expectation that the results would be uniform. However, when discrepancies occurred, it was logical that they would typically involve
reciprocal errors of the same chromosomes. For example, a trisomy
X in one cell might be accompanied by a monosomy X in another cell
from the same embryo. Additionally, the level of non-concurrence of
the SNP assignments on each chromosome should be similar to that
found with the cell line samples.
This prospective blinded assessment indeed found non-concurrence rates similar to those in single cells taken from cell lines, indicating similar accuracy levels. Additionally, the majority of the errors
were reciprocal, which was highly reassuring.
This study provided the foundation for validating clinical embryonic aneuploidy screening, but represents only the first of several
critical steps. The predictive values of both euploid and aneuploid
results cannot be determined solely in the laboratory. Clinical studies
assessing those predictive values as well as the overall impact on
implantation, delivery rates, and clinical decision-making were now
possible.
CliniCAl STuDieS eMPOWeReD By SnP ASSAy
DNA Fingerprinting
SNP arrays provide data on genetic polymorphisms that may be
used for DNA fingerprinting (4). This provides a unique opportunity
for a paired, randomized controlled trial (RCT) of sibling embryos
28
since two embryos could be simultaneously transferred and resulting fetuses or newborns could be precisely linked to the embryo
from which they developed. This has enabled powerful studies on
the impact of oocyte vitrification (5), cleavage, and blastocyst stage
embryo biopsy (6), and other ongoing clinical trials.
BenchmarkforAlternativeMethodsofCCS
A validated method for comprehensive chromosome screening
(CCS) provides an opportunity to test other screening methodologies. For example, SNP arrays were used to demonstrate that FISHbased blastomere analysis is inconsistent (7) and poorly predictive
of aneuploidy in the developing blastocyst (8), indicating that FISH
is limited by more than just the inability to test all 24 chromosomes.
In addition, it served as a standard when developing quantitative
real-time (q)PCR (9) and next generation sequencing based methodologies (10). Finally, SNP arrays were used to demonstrate significantly reduced concurrence of CCS by array comparative genomic
hybridization compared to qPCR, based on blinded analysis across
multiple laboratories (11).
ClinicalTrials
Validation of the assay and DNA fingerprinting was followed by a
prospective trial to determine the predictive value of euploid and aneuploid screening results for actual clinical outcome (12). Embryos
selected by standard morphological criteria were biopsied and immediately transferred prior to any analysis of the sample. SNP array predictions of aneuploidy and euploidy were compared to the
actual clinical outcomes and used to directly calculate the predictive
values of the assay. Approximately 4% of embryos designated as
aneuploid actually implanted and developed euploid fetuses. Subsequent improvements have reduced this number to <2%. Embryos
that screened euploid were more than twice as likely to implant and
deliver. This study (12) met a critical requirement for validating embryonic aneuploidy screening—to directly measure the predictive
values of an aneuploid result so that the risk for discarding a euploid
embryo may be specifically determined.
Given the demonstrated equivalence of qPCR- and SNP-array–
based CCS described above, SNP arrays indirectly provided an opportunity to test qPCR clinical efficacy in subsequent RCTs. The first
RCT demonstrated that, in women under the age of 42 with no more
than one prior failed IVF cycle and a normal ovarian reserve, CCS
improves the success of IVF (13). A second trial further established
the utility of CCS by demonstrating equivalent success rates with the
transfer of a single euploid blastocyst compared to the transfer of
two untested blastocysts, while eliminating twins (14) and improving
obstetrical outcomes (15).
ADDiTiOnAl APPliCATiOnS
SNP array CCS has also been demonstrated effective at identifying
sub-chromosomal imbalances associated with reciprocal translocations (16,17). These studies showed the importance of evaluating
aneuploidy of all 24 chromosomes in parallel with analysis of the
derivative chromosomes from the translocation, since many otherwise balanced/normal embryos were aneuploid for chromosomes
not involved in the translocation.
SNP arrays have also provided an opportunity to use loss of
heterozygosity to address the clinical relevance of uniparental disomy [found to be extremely rare in the blastocyst (0.06%)],
to confirm the parthenogenetic status of embryos derived after swapping nuclei between oocytes to eliminate mitochondrial
DNA variants (18), to investigate the parental origins of aneuploidy using genotype comparisons (19), and to confirm the genetic status of human embryos derived from somatic cell nuclear
transfer (20).
REFERENCES
1. T. Hassold, P. Hunt, Nat. Rev. Genet. 2, 280 (2001).
2. N. R. Treff, J. Su, X. Tao, L. E. Northrop, R. T. Scott, Jr., Mol. Hum.
Reprod. 17, 335 (2011).
3. N. R. Treff, J. Su, X. Tao, B. Levy, R. T. Scott, Jr., Fertil. Steril. 94,
2017 (2010).
4. N. R. Treff et al., Fertil. Steril. 94, 477 (2010).
5. E. J. Forman et al., Fertil. Steril. 98, 644 (2012).
6. R. T. Scott, Jr., K. M. Upham, E. J. Forman, T. Zhao, N. R. Treff,
Fertil. Steril 100, 624 (2013).
7. N. R. Treff et al., Mol. Hum. Reprod. 16, 583 (2010).
8. L. E. Northrop, N. R. Treff, B. Levy, R. T. Scott, Jr., Mol. Hum.
Reprod. 16, 590 (2010).
9. N. R. Treff et al., Fertil. Steril. 97, 819 (2012).
10. N. R. Treff et al., Fertil. Steril. 99, 1377 (2013).
11. A. Capalbo et al., 69th Annual Meeting of the American Society for
Reproductive Medicine (2013).
12. R. T. Scott, Jr. et al., Fertil. Steril. 97, 870 (2012).
13. R. T. Scott, Jr. et al., Fertil. Steril. 100, 697 (2013).
14. E. J. Forman et al., Fertil. Steril. 100, 100 (2013).
15. E. J. Forman, Hong, K.H., Scott, R.T. Jr., 61st American College of
Obstetricians and Gynecologists Annual Clinical Meeting (2013).
16. N. R. Treff et al., Fertil. Steril. 96, e58 (2011).
17. N. R. Treff et al., Fertil. Steril. 95, 1606 (2010).
18. D. Paull et al., Nature 493, 632 (2013).
19. N. R. Treff et al., Fertil. Steril. 90, S37 (2008).
20. Y. Chung et al., Cloning Stem Cells 11, 213 (2009).
What Is a “Normal” Semen Analysis?
David S. Guzick
Reference values for semen characteristics have been based on
the distribution of values in fertile men. In an effort to provide more
meaningful reference values, in 2001 members of the National Institutes of Health’s (NIH) Reproductive Medicine Network (RMN)
reported values for semen measurements based on a comparison
of fertile and infertile men. Instead of a single value for each semen
measurement that would distinguish between “normal” and “abnormal,” data analysis yielded ranges of values that delineated three
groups—fertile, indeterminate, and subfertile. These results can be
more meaningfully applied in clinical practice and research than previous measurements.
inTRODuCTiOn
During a standard infertility evaluation, both male and female partners undergo a series of diagnostic tests in an effort to shed light on
etiology (1). In the male partner, a semen specimen is typically evaluated for sperm concentration, motility, and morphology. The results
are presented to the patient, usually with a statement about whether
each of these parameters is “normal.”
But what is “normal”? Most clinicians refer to values published in
the World Health Organization (WHO) Manual for Semen Analysis,
Thisarticlebasedontheoriginalpublication:
D. S. Guzick et al., N. Engl. J. Med. 345, 1388 (2001).
University of Florida, Gainesville, FL
[email protected]
the last edition of which was published in 2010 (2). Recognizing that
there is substantial overlap in sperm parameters between fertile and
infertile men (3–5), beginning with the 1999 edition of the WHO manuals, the nomenclature for semen characteristics was changed from
“normal” to “reference” values.
Two prospective studies of semen parameters and fertility among
couples who discontinued contraception, published in 1998 (6) and
2000 (7), indicated that a reexamination of the WHO criteria was
warranted. In an effort to provide more meaningful reference values,
the NIH Reproductive Medicine Network (RMN) conducted a study
to identify values for semen measurements that best discriminate
between fertile and infertile men (8).
MeThODS
In the RMN study, two semen specimens from each of the male
partners in 765 infertile couples and 696 fertile couples were evaluated using uniform and rigorous methods. The infertile couples were
drawn from a randomized clinical trial in which the female partners
all had normal results in an infertility evaluation (9). Fertile men (controls) were recruited from prenatal classes at the same hospitals in
which the infertile couples were recruited. Within each of nine RMN
sites, fertile couples were frequency matched to infertile couples according to the five-year age groups of both partners.
Technicians from each of the nine RMN sites attended training
sessions in semen analysis at a central laboratory. Semen specimens
were stained at the clinical sites and shipped to the central laboratory
for assessment of sperm morphology by a single technician, trained
29
Produced by the Science/AAAS Custom Publishing Office
by a factor of two to three
in the likelihood of infertility
when one sperm measurement was in the subfertile
range can be compared
with an increase by a factor
of five to seven in the risk of
infertility when two sperm
measurements were subfertile, and an increase by
a factor of 16 when all three
measurements are subfertile.
The frequency distributions of fertile and infertile
men with regard to the three
sperm measurements are
Figure 1. Percentage of men from infertile and fertile couples
shown in Figure 1. Overall,
with values in the subfertile, indeterminate, and fertile ranges for
sperm concentration (A), percentage of motile sperm (B), and
the degree of overlap bepercentage of sperm with normal morphologic features (C), as
tween fertile and infertile
defined by classification and regression tree (CART) analysis.
men is striking. Indeed, apArrows indicate the thresholds between subfertile and indeterminate ranges (red) and indeterminate and fertile ranges (green).
proximately equal proporAdapted from (8).
tions of fertile and infertile
men had values in the indeterminate ranges of the three sperm measurements. There was an excess of infertile men with values in the
subfertile ranges of these variables, and a corresponding excess of
fertile men with values in the fertile ranges.
The area under receiver-operating-characteristic (ROC) curves
(12), which assesses the level of discrimination between fertile and
infertile men, was significantly greater for morphology (0.66) than for
concentration (0.60; p<0.001) and motility (0.59; p<0.001).
by the developer of a strict morphology assessment (10).
All data were then sent to a central site for data coordination
and statistical analysis. Classification and regression tree (CART)
analysis (11) was used to estimate thresholds for each sperm
measurement that would discriminate between fertile and infertile
men. Two thresholds were estimated for each sperm characteristic:
one between the subfertile and indeterminate ranges, and the other
between the indeterminate and fertile ranges.
ReSulTS
Infertile men were slightly older than fertile controls (mean age:
34.7 vs. 33.5, p<0.001), and were more likely to be white (91.0% vs
85.2%, p<0.001) and to smoke (17.9% vs. 12.5%, p=0.02), but were
less likely to have completed college (55% vs. 64%, p<0.001). As
shown in Table 1, the CART-defined subfertile range for sperm measurements was a concentration of less than 13.5 million per milliliter,
less than 32% motility, and less than 9% normal morphology. Table
1 also shows CART-identified thresholds that identify the indeterminate and fertile ranges, and associated odds ratios.
Table 2 shows that the odds of infertility increase with an increasing
number of sperm measurements in the subfertile range. An increase
30
DiSCuSSiOn AnD uPDATe
A case can be made that the case-control methodology used in the
RMN study is the best of an imperfect set of options. Follow up of
couples who have discontinued contraception has the advantage of
prospectively defining couples as fertile and infertile, but such an approach requires large samples, given the low incidence of infertility,
and is complicated by the unknown extent of female infertility among
the couples so defined as infertile. To date, reference values from
such a methodology have not been proposed.
The WHO manual uses a statistical cut-point among fertile men,
defined as the 5th percentile in the last edition. Such men are at the
statistical tail of the normal distribution of fertile men, but they are
still fertile. Using a population of fertile men to predict male infertility makes little sense, biologically or methodologically. Moreover,
the databases from which cut-points were chosen in the first four
editions were constrained by imprecisely defined reference populations of fertile men, and there was variability in the standards and
protocols among andrology laboratories (13). Reference values defined in the latest edition are based on a pooled database with improved laboratory consistency and a better characterized group of
recent fathers. The 5th percentile of semen characteristics among
these fertile men were: sperm concentration <15 million per milliliter; motility <40%; and normal morphology <4%.
Interestingly, the WHO cut-point for sperm concentration is quite
Variable
Fertile range
Indeterminate range
Univariate odds ratio for
infertility (95% CI)
Subfertile range
Univariate odds ratio for
infertility (95% CI)
SemenMeasurement
Concentration
Motility
Morphology
6
(x10 /mL)
(%)
(% normal)
>48.0
>63
>12
13.5–48.0
32–63
9–12
1.5 (1.2–1.8)
1.7 (1.5–2.2)
1.8 (1.4–2.4)
<13.5
<32
<9
5.3 (3.3–8.3)
5.6 (3.5–8.3)
3.8 (3.0–5.0)
Table 1. Fertile, indeterminate, and subfertile ranges for sperm measurements using CART analysis and the corresponding odds ratios for infertility. CI denotes
confidence interval.
SpermMeasurementRange
MorphologicalFeatures
Motility
OddsRatio(95%CI)
Concentration
Fertile
Fertile
Fertile
1.0
Subfertile
Fertile
Fertile
2.9 (2.2–3.7)
Fertile
Subfertile
Fertile
2.5 (1.6–4.2)
Fertile
Fertile
Subfertile
2.2 (1.3–3.6)
Subfertile
Subfertile
Fertile
7.2 (4.3–12.2)
Subfertile
Fertile
Subfertile
6.3 (3.8–10.3)
Fertile
Subfertile
Subfertile
5.5 (3.0–10.2)
Subfertile
Subfertile
Subfertile
15.8 (8.7–29.0)
Table 2. Odds ratios for infertility for combinations of sperm measurements. The ranges of the sperm measurements were defined by the thresholds derived from
CART analysis. The reference group for the odds ratios is the mean with all three measurements in the fertile range. CI denotes confidence interval.
close to the cut-point found in the RMN study that separated subfertile from indeterminate. Comparing fertile and infertile men, the
RMN study identified a somewhat lower threshold for motility (32%
vs. 40%). This is reflected in Figure 1, which shows no difference
between fertile and infertile men in the range of 32%–42% motility.
Additionally, Figure 1 shows a clear difference between fertile and
infertile men in the range of 5%–8% normal morphology, which justifies an RMN-defined cut-point for morphology that is higher than the
WHO manual.
Semen characteristics are, biologically, continuous variables.
Whether they be sperm measurements such as concentration,
motility, and morphology, or sperm function tests such as zona
binding assays, egg penetration tests, acrosome reaction testing, or
others, no measurement has a clear cut-point that separates fertile
from infertile. Thus, clinicians cannot convey with certainty to an
infertile couple that a semen measurement below a given threshold
value is responsible for their infertility, nor that a value above such
a threshold excludes a male factor etiology. This is essentially a
probabilistic matter. In the RMN study, to capture this idea, instead of
a single value for each semen measurement that would distinguish
between “normal” and “abnormal,” we defined the two values that
best delineated three groups: fertile, indeterminate, and subfertile.
Since these values have been defined by comparing fertile and
infertile men, they provide ranges that can be meaningfully applied
in clinical practice and research.
REFERENCES
1. M. A. Fritz, Clin. Obstet. Gynecol. 55, 692 (2012).
2. WHO Laboratory Manual for the Examination of Human SpermCervical Mucus Interaction (World Health Organization, Geneva, ed.
5, 2010).
3. J. MacLeod, R. Z. Gold, J. Urol. 66, 436 (1951).
4. J. MacLeod, R. Z. Gold, Fertil. Steril. 2, 187 (1951).
5. J. MacLeod, R. Z. Gold, Fertil. Steril. 2, 394 (1951).
6. J. P. Bonde et al., Lancet 352, 1172 (1998).
7. M. J. Zinaman, C. C. Brown, S. G. Selevan, E. D. Clegg, J. Androl.
21, 145 (2000).
8. D. S. Guzick et al., N. Engl. J. Med. 345, 1388 (2001).
9. D. S. Guzick et al., N. Engl. J. Med. 340, 177 (1999).
10. T. F. Kruger et al. Fertil. Steril. 49, 112 (1988).
11. L. Breiman, J. H. Friedman, R. A. Olshen, C. J. Stone, Classification
and regression trees (Wadsworth, Belmont, CA, 1984).
12. J. A. Hanley, B. J. McNeil, Radiology 143, 29 (1982).
13. T. G. Cooper et al., Hum. Reprod. Update 16, 231 (2010).
31
Produced by the Science/AAAS Custom Publishing Office
Circulating Angiogenic Factors and the Risk of Preeclampsia
S. Ananth Karumanchi1,2* and Ravi Thadhani3
Preeclampsia remains a major cause of maternal and fetal morbidity and mortality worldwide. We have previously suggested that
aberrant vascular endothelial growth factor signaling due to excess
circulating soluble fms-like tyrosine kinase 1 plays a causal role in
the pathogenesis of preeclampsia. This article summarizes the key
pathophysiological consequences of these angiogenic abnormalities
and discusses our efforts to translate this knowledge into novel diagnostic and therapeutic strategies.
inTRODuCTiOn
As a leading cause of premature birth and maternal and infant
mortality worldwide, preeclampsia remains a tragically unmet public health challenge. Preeclampsia and related hypertensive disorders conservatively cause over 75,000 maternal and 500,000 infant
deaths globally each year (www.preeclampia.org) mostly in developing countries.
The mechanisms that initiate preeclampsia in humans have long
been elusive, leading to its description in medical textbooks as an idiopathic “toxemia” of pregnancy whose definitive treatment remains
delivery of the placenta. Nearly a decade ago, we proposed that elevated plasma soluble fms-like tyrosine kinase (sFlt1), an anti-angiogenic protein, and low levels of placental growth factor (PlGF), a
pro-angiogenic protein, were key pathophysiological characteristics
of preeclampsia (1,2) and showed robust correlations with disease
presence and severity (3). Moreover, alterations in these angiogenic
factors were noted prior to onset of clinical signs or symptoms (1,4).
In addition, we demonstrated that alterations in circulating sFlt1 may
explain a number of risk factors for preeclampsia such as multiple
gestation, trisomy 13, nulliparity, and molar pregnancies (5). These
findings led us to hypothesize that preeclampsia is an anti-angiogenic state due to excess circulating sFlt1 (6).
BiOlOgy OF AnTi-AngiOgeniC STATe in PReeClAMPSiA
In 2003, we identified upregulated sFlt1 mRNA in preeclamptic placentas (3). sFlt1 protein, a splice variant of the vascular endothelial
growth factor (VEGF) receptor Flt1 or VEGFR1, is made by the syncytiotrophoblast layer of the placenta and is secreted into maternal
circulation (7), inhibiting VEGF signaling in the vasculature (Fig. 1).
Circulating sFlt1 levels are markedly increased in women with established preeclampsia, while free VEGF levels are decreased (3)
even prior to onset of clinical signs and symptoms (8). Furthermore,
removal of sFlt1 or addition of exogenous VEGF can reverse the
anti-angiogenic phenotype noted in preeclamptic plasma (3,9). Het Thisarticlebasedontheoriginalpublication:
R. J. Levine et al., N. Engl. J. Med. 350, 672 (2004).
1
Beth Israel Deaconess Medical Center, Harvard Medical School, Boston,
MA; 2Howard Hughes Medical Institute, Boston, MA; 3Massachusetts
General Hospital, Harvard Medical School, Boston, MA
*
Corresponding Author: [email protected]
32
erologous expression in rodents produces a syndrome of hypertension, proteinuria, and glomerular endotheliosis in the kidney resembling human preeclampsia (3,10,11). Recombinant VEGF therapy
or lowering of sFlt1 ameliorates the preeclampsia phenotype in rodent models (10,12,13). Reduction of VEGF levels by 50% in the
glomeruli using genetically modified mice leads to proteinuria and
glomerular endothelial damage that resemble preeclampsia (14).
Furthermore, VEGF antagonists, used in cancer patients, can produce a preeclampsia-like phenotype including hypertension, proteinuria, and cerebral edema that is often noted in severe preeclampsia
(15–17). These data support the hypothesis that inhibition of VEGF
signaling in an adult may lead to preeclampsia-like signs/symptoms.
The studies on sFlt1 and VEGF in preeclampsia biology have also
led to new insights into basic biology of vascular and placental homeostasis in health and disease. The microvasculature of organs affected in preeclampsia such as the glomeruli and hepatic sinusoidal
vasculature are more permeable due to the presence of intracellular perforations referred to as fenestrations. While it was previously
known that VEGF induces endothelial fenestrae in culture (18), experimental data in VEGF knockout mice demonstrated that fenestral
density is regulated by constitutive expression of VEGF (14). Therefore it is not surprising that the microvascular damage in preeclampsia is largely concentrated in vasculature that constitutively express
VEGF and that loss of endothelial fenestrae in preeclampsia is due
to excess circulating sFlt1. While the placenta is the major source of
sFlt1 production, recent studies have suggested that syncytial knots
(degenerating syncytiotrophoblast tissue) in the placenta are the major site of sFlt1 production (19). These syncytial knots easily detach
from the syncytiotrophoblast, resulting in free multinucleated aggregates (50–150 mm diameter) laden with sFlt1 protein and mRNA,
which are secreted into the maternal circulation. Studies using autopsy material have confirmed that shed syncytial knots contribute to
circulating sFlt1 in preeclampsia (20).
It is likely that other synergistic anti-angiogenic proteins such as
soluble endoglin and semaphorin 3B may also contribute to the
pathogenesis of preeclampsia (21, 22). Patients presenting with
high levels of both soluble endoglin and sFlt1 had a more severe
and premature form of the disease (21, 23). Animals exposed to
high soluble endoglin and high sFlt1 developed the most severe
features of preeclampsia, including cerebral edema (21,24). More
research into the role of these synergistic factors in preeclampsia
is needed.
BiOMARkeR STuDieS in PReeClAMPSiA
Although VEGF is secreted, it acts as a juxtacrine or autocrine
growth factor locally, and circulating VEGF levels are relatively
low during pregnancy (<10 pg/mL). Furthermore, measurement of
VEGF in serum is compromised by platelet VEGF release during
clotting and hence circulating VEGF is not a useful biomarker for
clinical use. As a surrogate of VEGF impairment, placental growth factor levels
(PlGF) in the plasma has emerged as an
attractive biomarker. PlGF, a VEGF family member, is a pro-angiogenic protein
abundant during pregnancy, which binds
to the Flt1 receptor, but not other VEGF
receptors such as the kinase insert domain receptor (KDR). As sFlt1 levels rise,
free PlGF levels drop, and the sFlt1:PlGF
ratio correlates with preeclampsia phenotypes (25, 26). Working with industry
partners, we and others have developed
highly sensitive, specific, and robust high
throughput assays to quantitate levels of
sFlt1 and free PlGF in plasma and serum
(27–29).
Several groups have demonstrated
that sFlt1 and PlGF levels can be used
to differentiate preeclampsia from diseases that mimic preeclampsia such
as chronic hypertension, gestational
hypertension, kidney disease, and gestational thrombocytopenia (30). We led
a large prospective study that demFigure 1: Schematic of Impaired VEGF Signaling During Preeclampsia. During normal pregnancy, vascular
onstrated that the plasma sFlt1:PlGF homeostasis is maintained by physiological levels of VEGF signaling in the vasculature. In preeclampsia, excess
ratio at presentation predicts adverse placental secretion of sFlt1 acts as a VEGF trap by binding and preventing its interaction with cell surface VEGF
maternal and perinatal outcomes (oc- receptors (Flt1 and KDR).
curring within two weeks) in the preterm setting. This ratio alone outperformed standard clinical diag- long pregnancy duration. We have successfully extended (in a single
nostic measures including blood pressure, proteinuria, uric acid, arm unblinded study) three preeclamptic pregnancies by two to four
and others. Importantly, sFlt1:PlGF levels in the triage setting cor- weeks, all of which resulted in healthy deliveries with no neonatal or
related with preterm delivery, an observation that has now been maternal morbidity (36). During treatment, sFlt1 levels were lowered
confirmed by several groups (31–33). In addition, we demon- by ~30% on average, indicating that modestly lowering circulating
strated that women diagnosed with preeclampsia, but with a nor- sFlt1 levels may be sufficient to successfully prolong preeclamptic
mal angiogenic profile, are not at risk for adverse maternal or fetal pregnancies.
outcomes (34).
RelaxinandStatins
TheRAPeuTiC STuDieS
Experimental studies suggest that the hormone relaxin, a natuHuman and animal studies as outlined above have strongly sug- rally occurring ovarian hormone, may be used to ameliorate pregested that targeting the sFlt1 pathway may be a viable strategy to eclampsia by improving vascular compliance and upregulating
prevent or treat preeclampsia. Below are some strategies that have locally produced VEGF (38). A phase I study to test the safety of
been evaluated in either preclinical or early clinical studies.
human relaxin in women with preeclampsia is ongoing (39). Compounds that upregulate pro-angiogenic factors such as statins
TherapeuticApheresis
also have been demonstrated to reverse preeclampsia in animal
sFlt1 has a large volume of distribution, and circulating plasma lev- models (11). A clinical trial to test the safety of statins in women
els represent less than 20% of the total body sFlt1 burden (35). We with established preeclampsia has been initiated in the United
hypothesized that a selective adsorption column such as dextran Kingdom (40).
sulfate (a polyanionic derivative of dextran) could augment sFlt1 removal. Dextran sulfate binds sFlt1 efficiently (36) and these columns SuMMARy
have been used previously to bind apolipoprotein B-containing lipo- During the past decade, epidemiological, experimental, and theraprotein LDL in pregnant patients with familial homozygous hypercho- peutic studies from several laboratories have provided compelling
lesterolemia without adverse consequences to mother or fetus (37). evidence that an anti-angiogenic state due to excess circulating sFlt1
In a proof-of-concept study, we demonstrated that a ~30% reduction and related angiogenic proteins leads to preeclampsia. Further work
in circulating sFlt1 levels using dextran sulfate apheresis may be on the regulation of sFlt1 production may improve our understanding
sufficient to ameliorate preeclamptic signs and symptoms and pro- of the fundamental molecular defect in preeclampsia. We anticipate
33
Produced by the Science/AAAS Custom Publishing Office
that in the ensuing decade, these molecular discoveries will lead to
improvement of care of patients suffering from preeclampsia and its
devastating complications.
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
REFERENCES
R. J. Levine et al., N. Engl. J. Med. 350, 672 (2004).
R. Thadhani et al., J. Clin. Endocrinol. Metab. 89, 770 (2004).
S. E. Maynard et al., J. Clin. Invest. 111, 649 (2003).
R. Romero et al., J. Matern. Fetal Neonatal. Med. 21, 9 (2008).
C. E. Powe, R. J. Levine, S. A. Karumanchi, Circulation 123, 2856
(2011).
I. Agarwal, S. A. Karumanchi, Pregnancy Hypertens. 1, 17 (2011).
D. E. Clark et al., Biol. Reprod. 59, 1540 (1998).
B. M. Polliotti et al., Obstet. Gynecol. 101, 1266 (2003).
S. Ahmad, A. Ahmed, Circ. Res. 95, 884 (2004).
A. Bergmann et al., J. Cell. Mol. Med. 14, 1857 (2010).
K. Kumasawa et al., Proc. Natl. Acad. Sci. U.S.A. 108, 1451
(2011).
J. S. Gilbert et al., Hypertension 55, 380 (2010).
Z. Li et al., Hypertension 50, 686 (2007).
V. Eremina et al., J. Clin. Invest. 111, 707 (2003).
V. Eremina et al., N. Engl. J. Med. 358, 1129 (2008).
P. Glusker, L. Recht, B. Lane, N. Engl. J. Med. 354, 980 (2006).
T. V. Patel et al., J. Natl. Cancer Inst. 100, 282 (2008).
W. G. Roberts, G. E. Palade, J. Cell. Sci. 108 (Pt 6), 2369 (1995).
A. Rajakumar et al., Hypertension 59, 256 (2012).
A. J. Buurma et al., Hypertension 62, 608 (2013).
S. Venkatesha et al., Nat. Med. 12, 642 (2006).
Y. Zhou et al., J. Clin. Invest. 123, 2862 (2013).
R. J. Levine et al., N. Engl. J. Med. 355, 992 (2006).
24. A. S. Maharaj et al., J. Exp. Med. 205, 491 (2008).
25. R. J. Levine et al., JAMA 293, 77 (2005).
26. M. Noori, A. E. Donald, A. Angelakopoulou, A. D. Hingorani, D. J.
Williams, Circulation 122, 478 (2010).
27. S. J. Benton et al., Am. J. Obstet. Gynecol. 205, 469 e461 (2011).
28. S. Sunderji et al., Am. J. Obstet. Gynecol. 202, 40 e41 (2010).
29. S. Verlohren et al., Am. J. Obstet. Gynecol. 202, 161 e161 (2010).
30. H. Hagmann, R. Thadhani, T. Benzing, S. A. Karumanchi, H. Stepan,
Clin. Chem. 58, 837 (2012).
31. T. Chaiworapongsa et al., J. Matern. Fetal Neonatal. Med. Epub
ahead of print (2013) doi: 10.3109/14767058.2013.806905.
32. A. G. Moore et al., J. Matern. Fetal Neonatal. Med. 25, 2651 (2012).
33. S. Rana et al., Circulation 125, 911 (2012).
34. S. Rana et al., Hypertens Pregnancy 32, 189 (2013).
35. F. T. Wu, M. O. Stefanini, F. Mac Gabhann, A. S. Popel, PLoS ONE 4,
e5108 (2009).
36. R. Thadhani et al., Circulation 124, 940 (2011).
37. R. Klingel, B. Gohlen, A. Schwarting, F. Himmelsbach, R. Straube,
Ther. Apher. Dial. 7, 359 (2003).
38. J. T. McGuane et al., Hypertension 57, 1151 (2011).
39. E. Unemori, B. Sibai, S. L. Teichman, Ann. N.Y. Acad. Sci. 1160, 381
(2009).
40. A. Ahmed, Thromb. Res. 127 Suppl. 3, S72 (2011).
Disclosures: SAK is co-inventor of multiple commercial patents related
to diagnosis/treatment of preeclampsia that have been licensed to multiple
companies. SAK has served as a consultant to Roche, Beckman Coulter,
and Siemens Diagnostics, and has financial interest in Aggamin LLC. RT
is co-inventor on patents related to licensed biomarkers in preeclampsia.
RT also discloses financial interest in Aggamin LLC.
The Ovarian Factor: Cutting-Edge Basic Science Combined with Classic
Clinical Insights
Richard S. Legro1 and Martin M. Matzuk2
Functional ovaries are necessary in mammalian females for propagation of the species and maintenance of the hormone milieu. Throughout most of the postnatal life of a female, oocytes—the female germ
cells—are encased in somatic cells in structures called follicles. The
resting pool of follicles, called primordial follicles, either die or are
recruited into the growing pool of more developed follicles. Although
there is much debate about postnatal oocyte stem cells, it is believed
that the rate of demise of primordial follicles determines the reproductive longevity of an individual within a species. While there are
many mutations that have been identified that disrupt female fertility
in model organisms (mainly mice), few mutations in ovary-specific
genes have been identified in women with fertility problems. However, new strategies for genome sequencing may help to identify
causative fertility associated mutations. Effective treatments exist for
all types of ovarian failure, though ovulation dysfunction is more difficult to detect and empirical treatments have less certain benefits.
The BASiC SCienCe
Over the last 25 years, we have witnessed amazing and rapid advances in our understanding of the genetics of mammalian reproduction, in particular the genes and factors important for ovarian
development and physiology [reviewed in (1–5)]. In large part, this
progress has been possible because of breakthroughs in DNA sequencing and transgenic mouse technology. Knockout mouse strategies, developed in the late 1980s and early 1990s, allowed researchers to address the question, “What is the essential role of a gene
product?” Prior to this point, the reproductive genetics community
relied on spontaneous mutations or N-ethyl-N-nitrosurea (ENU) mutagenesis—a challenging process akin to finding a proverbial needle
in a haystack. Embryonic stem (ES) cell technology allowed for the
reverse genetics approach in which precise mutations could be created to disrupt a gene and subsequently elucidate its function.
This technology has permitted identification of over 100 proteins that function in the formation of female embryonic germ
cells, at every stage of ovarian folliculogenesis, in post-ovulation
events, and at various stages of meiosis and DNA repair. These
pioneering studies prompted work in other mammalian species,
including sheep, with relevant and important translational follow
up in humans. Some of the pertinent studies will be described in
depth below.
1
Department of Obstetrics and Gynecology, Pennsylvania State
University College of Medicine, Hershey, PA 2Department of Pathology
and Immunology, Baylor College of Medicine, Houston, TX
[email protected] (RSL); [email protected] (MMM)
34
geRMline STeM CellS
The pioneering transgenic mouse studies of Brinster and Palmiter
(6), and others led not only to the advances in mouse reproductive
genetics, but also to advances in oocyte and embryo culture conditions for human assisted reproduction [elegantly reviewed in (7)] and
in manipulation of the male germline (8). Spermatogonial stem cell
transplantation techniques identified factors required for the maintenance of male stem cells in an undifferentiated state and also uncovered genes that mark the differentiated state (8). More recently,
other groups have attempted to identify and define female germline
stem cells, generating enormous controversy in the literature, the lay
press, and at scientific meetings. While not wanting to join in the controversy, especially since there may be some differences between
animal models and humans, we will comment on two intriguing studies published recently. First, Liu and colleagues (9) used a genetic
trick to make DDX4-positive germ cells in the postnatal ovary, and
thereby follow the differentiation and proliferation of these cells over
time. The authors could not detect mitotically active germline precursors, in contrast to the previous studies in mice (10) or humans
(11). Second, Saitou and colleagues (12) differentiated mouse XX
ES cells and induced pluripotent stem (iPS) cells into cells resembling primordial germ cells (PGCs), reconstituted these cells with gonadal somatic cells, and showed that they could undergo meiosis. In
vivo, these cells with meiotic potential could develop to the germinal
vesicle stage and could give rise to fertile offspring when subjected
to in vitro maturation and fertilization. Thus, oocyte precursors could
be created from pluripotent stem cells and could be manipulated for
fertilization.
The above studies and other exciting research being performed in
defining sex divergent steps in the differentiation of PGCs and the
intrinsic and extrinsic factors necessary for prenatal entry into and
arrest of meiosis will continue to influence the scientific pursuit of the
elusive oocyte stem cell. These studies will also aid in the in vitro production of functional oocytes from human ES cells and/or iPS cells.
BiDiReCTiOnAl COMMuniCATiOn
DuRing FOlliCulOgeneSiS
Understanding the control of the recruitment of follicles into the growing pool has been an actively pursued area of research. The Eppig
and Nalbanov laboratories have postulated that oocyte-secreted factors could influence somatic cell function in vitro [reviewed in (1)],
and later work of Eppig showed that the oocyte dictated the phenotype of the postnatal granulosa cells (13). In 1996, knockout mouse
studies from the Matzuk laboratory uncovered for the first time the
identity of one of these oocyte-secreted germline factors, growth differentiation factor 9 (GDF9), a member of the transforming growth
factor β (TGF-β) superfamily (14). Mice lacking GDF9 showed a
35
Produced by the Science/AAAS Custom Publishing Office
Figure 1. Follicular, ovarian, and endometrial
ultrasonographic measurements at the beginning and end of the one-month baseline
period and at their maximum during r-metHuLeptin treatment. Each symbol represents one
subject. Reprinted with permission from (16).
Figure 2. Regimens of ovarian stimulation and hormone
replacement used to synchronize the development of ovarian follicles in the oocyte donor
and endometrial cycle in the
recipient. Reprinted with permission from (20).
Figure 3. The time course and quantitative change
in circulating concentrations (Mean ± SE) of luteinizing hormone (LH), follicle-stimulating hormone
(FSH), estradiol (E2), and progesterone (P) during the
menstrual cycle of four normal women treated with
a GnRH agonist for three days after menses onset
(hatched box, left). Data were centered around the
midcycle gonadotropin peak (day 0). Reprinted with
permission from (25).
block in folliculogenesis at the primary follicle stage, and later studies identified an
oocyte-secreted paralog, bone morphogenetic protein 15 (BMP15), which also influences fertility in mice, sheep, and women
(5). More recent studies demonstrated that
mouse and human GDF9:BMP15 heterodimers are far more biopotent than active
homodimers (10–30-fold higher activity in
mice; ~3,000-fold in humans) (15). However, oocyte-secreted growth factors are only
one-half of an ongoing bidirectional dialog
that regulates key metabolic synchrony of
oocytes and granulosa cells throughout folliculogenesis (see also page 13). The implications of these studies are far-reaching for animal and human
fertility, and include the development of new, defined culture media
supplemented with cocktails of oocyte and granulosa cell proteins
and metabolites that take advantage of this crosstalk.
PiTuiTARy COnTROl OF
OVARiAn FunCTiOn
Follicle-stimulating hormone (FSH) and luteinizing hormone (LH)
are key regulators of ovarian function (16). Mice lacking FSH and
women with homozygous mutations in the FSHβ or FSH receptor
gene are infertile secondary to blocks in folliculogenesis at the multilayer preantral follicle stage. Mice or women with mutations in the
LHβ or LH receptor genes have blocks prior to ovulation resulting in
infertility. The ability to predict defects at the hypothalamic or pituitary
levels based on alterations in serum FSH or LH levels has enabled
the identification of other genes that are important regulators of FSH
and LH, and that indirectly influence gonadotropin synthesis (16). An
understanding of these key ovarian regulators has been critical for
assisted reproduction.
The CliniCAl SCienCe
We will now shift focus to highly cited articles that relate to clinical
interventions that have improved (or replaced) ovarian function in
an infertile population (Table 1). The choice is highly subjective, but
the goal is to include articles that have led to a paradigm shift in
the treatment of female infertility. We will cover treatments for the
most common causes of ovulation failure, as well as lesser ovulatory problems such as those found in undiagnosed or unexplained
infertility. The World Health Organization (WHO) provides a schema
for ovulation failure with three categories (based on hypothalamic/
pituitary and ovarian function): Type I (hypogonadotropic, hypoestrogenic), Type II (normogonadotropic, normoestrogenic), and Type III
(hypergonadotropic, hypoestrogenic).
36
WHO Type I: Anovulation-Hypothalamic Amenorrhea
Hypothalamic amenorrhea can arise from genetic conditions leading
to failure of the GnRH neurons to develop or migrate to the hypothalamus or from disruption of genes relating to onset of puberty.
There are also common acquired conditions associated with stress,
excessive exercise, and loss of body fat/weight (i.e., anorexia nervosa or exercise-induced amenorrhea), which can cause hypothalamic
shutdown and downstream loss of ovarian function. While treatment
with gonadotropins effectively bypasses the hypothalamic GnRH
pulse generator and directly stimulates follicular development, the
factors leading to the hypothalamic failure remain in doubt. Cessation of activity and regain of weight should cure the condition, but no
high-quality clinical trials have yet demonstrated this.
The study of Welt etal. (17) looked at the effects of recombinant
leptin administration on ovarian function in women with hypothalamic amenorrhea. The intervention group was observed without
treatment for one month, then administered recombinant leptin for
up to three months. A control group received no treatment. Five
of eight patients in the intervention group either ovulated or had
some resumption of ovarian activity (Fig. 1). None of the control
subjects or intervention subjects during the initial observation period showed any change. This provided evidence that peripheral
fat status was critical to maintaining reproductive function and supported studies that a critical amount of fat mass is necessary to
initiate menarche.
WHO Type II: Ovulatory Dysfunction-Polycystic
Ovary Syndrome
A study of Legro etal. (18) occurred at a time of great debate as to
the best treatment for polycystic ovary syndrome (PCOS), a heterogeneous condition of hyperandrogenism and chronic anovulation,
with characteristic polycystic ovaries filled with preantral follicles.
PCOS was viewed by some as a metabolic syndrome due to dimin-
37
Produced by the Science/AAAS Custom Publishing Office
Category
Specific
Condition
WHO Type I
Anovulation
Hypothalamic
Amenorrhea due
to exercise or
excessive thinness
16
WHO Type II
Anovulation
Polycystic Ovary
Syndrome
17
Ovulation and live birth, as well as fecundity per ovulation, were better
with clomiphene than metformin despite metformin’s improved effects
on weight and insulin levels.
WHO Type III
Anovulation
Age-related
diminished ovarian
reserve
20
Oocyte donation is an effective treatment for women with diminished
ovarian reserve with live-birth rates similar to those with primary
ovarian insufficiency, suggesting uterine function is not agedependent.
Reference
KeyFinding
Recombinant leptin was able to initiate ovarian function in five of eight
women given the intervention, three of whom ovulated, implying that
fat mass is critical to reproductive function.
Ovulatory
Dysfunction
Acquired luteal
phase defect
25
A luteal phase defect, characterized by a shortened luteal phase with
inadequate circulating serum progesterone levels could be induced
by the administration of a GnRH agonist in the early follicular phase in
normally ovulating women.
Subtle Ovulatory
Dysfunction
Unexplained
infertility
26
Empiric treatment with the ovulation induction agent clomiphene or
with unstimulating intrauterine insemination is no better than expectant
management for couples with unexplained infertility.
Table 1. List of key clinical trials improving our understanding of ovulatory failure or dysfunction in humans.
ished insulin action—requiring primary treatment with insulin-sensitizing effects—and by others as a reproductive abnormality with
inappropriate gondadotropin secretion and corresponding altered
ovarian steroidal production, and lack of development of functional
follicles resulting in anovulation. The study examined the effects of
ovulation-inducing agents: clomiphene alone vs metformin alone vs
their combination in infertile women with PCOS, using a double blind
double dummy design.
Contrary to expectations, clomiphene was markedly superior to
metformin achieving a twofold higher live-birth rate, with the combination offering a further marginal but non-significant improvement.
Importantly, the quality of ovulation differed depending on ovulationinducing agent: the improved fecundity and ovulation rates on clomiphene were associated with increased body weight, waist circumference, and insulin levels from baseline, implying that improvement in
these factors were not as critical to restoration of ovulation in these
women as previously thought. This study served to justify the search
for ovulation-inducing agents that work primarily by altering ovarian
steroid feedback to the hypothalamic pituitary axis, such as aromatase inhibitors that block the conversion of excess androgens to bioactive estrogens.
WHO Type III: Anovulation Age Related
to Diminished Ovarian Reserve
Primary ovarian insufficiency can be due to genetic conditions such
as Turner’s Syndrome or single gene defects such galactosidase de-
38
ficiency or mutations acquired from exposure to radiation or alkylating agents during cancer treatment. Further, while Type III anovulation occurs relatively rarely, all women experience an age-related
decline in ovarian function, with increasing rates of embryo aneuploidy and miscarriage, culminating in the complete loss of ovulation
and menses at menopause. While preservation of fertility is a rapidly
expanding preventive treatment option, there have been no real advances in restoring ovarian function in women with age-related ovarian insufficiency. A breakthrough occurred when it was shown that
embryos created from donor oocytes could be placed in the uterus
of a woman with ovarian insufficiency whose endometrium had been
rendered receptive to embryo implantation using sex steroid treatment. Case reports describing embryo flushing in inseminated oocyte donors (19) and IVF performed using donor oocytes (20) provided evidence in support of this procedure.
The broader clinical implication, built upon this evidence, was the
extension of this therapy to women with advanced age and infertility
due to what is now described as diminished ovarian reserve (DOR).
Sauer etal. (21,22) studied women over 40 with DOR, finding that
implantation of donated oocytes yielded pregnancies and live-birth
rates markedly superior to expected fertility rates based on age.
Manipulation of hormone levels to prepare the uterus for implantation is shown in Figure 2. Rates of pregnancy after oocyte donation routinely exceeded those after standard IVF in women of all age
groups in registries throughout the world. Such high success rates
in a group that had previously experienced such dismal results reset
expectations for subsequent therapies. Further, it underscored that
the primary effects of aging impacted oocyte quality and number.
OVulATORy DySFunCTiOn—luTeAl PhASe DeFeCTS
Many women with infertility or recurrent pregnancy loss do not
present with chronic or sporadic anovulation, but are suspected to
have some form of ovulation dysfunction leading to the absence of
functional oocytes and/or to implantation failure. Several ovulatory
disorders occur within the context of what appears to be regular
ovulation, but continue to defy clear definition and diagnosis. These
include extremely rare disorders such as luteinized unruptured follicle syndrome, empty follicle syndrome, and more common disorders such as luteal phase defects (LPDs). LPDs, originally described
by Georgeanna Jones, are characterized by inadequate corpus
luteum function following ovulation, as measured by a variety of
parameters including inadequate rise in basal body temperature,
secretion of progesterone or its metabolites, an out-of-phase endometrial biopsy, or a shortened luteal phase (23). Subsequent
studies have found this disorder difficult to diagnose largely
due to the occurrence of these “abnormalities” in normal, fertile
populations (24,25).
However, the proof of concept for LPD was demonstrated in a classic study by Sheehan et al. (26) in which normally ovulating women
(verified by careful monitoring of gonadotropins and sex steroids
prior to treatment) were administered a GnRH agonist in the early
follicular phase that resulted in a temporary increase in gonadotropin
secretion, followed by a decrease in FSH levels and a shortened
luteal phase (Fig. 3). While this was seen at the time as a potential
contraceptive, it helped to justify the use of progesterone administration to supplement the luteal phase in IVF cycles where a GnRH
agonist had been administered, and was also used to treat women
with suspected LPDs.
unexPlAineD inFeRTiliTy—
eMPiRiC OVulATiOn inDuCTiOn
Unexplained infertility remains the most heterogeneous of infertility
diagnoses and contains subclinical etiologies related to ovulatory,
tubal, and male factors. Couples often receive empiric therapy which
takes a shotgun approach, trying to hit multiple targets with a single
shot. Empiric treatment with ovulation-inducing agents such as clomiphene and gonadotropin has two intended goals: to increase the
quality of ovulation (difficult to quantify, as noted above with LPDs)
and to cause superovulation, which results in multiple mature oocytes and both a greater chance for pregnancy and a higher risk of
multiple pregnancies.
The study by Bhattacharya et al. (27) randomized couples with
unexplained infertility into three treatment groups (with similar initial pregnancy rates): empiric clomiphene citrate, unstimulated intrauterine insemination (IUI), or expectant management. The trial
was unique for the inclusion of the control expectant management
group, a “watch and wait” approach not usually regarded as acceptable in an infertility clinical trial. Although subjects in the expectant
management group were less satisfied with their participation than
others, the trial did meet its enrollment goals. This trial examined
the role of subtle, subclinical ovulatory dysfunction in unexplained
infertility and, although there was no statistically significant benefit
to IUI, it trended better than clomiphene. This study raised the bar
for empiric infertility treatments, introducing the requirement that
proof of benefit of the intervention over expectant management
be shown before acceptance as a clinical treatment. It further underscores the need for better diagnosis strategies in unexplained
infertility.
COnCluSiOnS
This review summarizes groundbreaking research that offers hope
for new treatments of ovarian disorders that lead to ovulation dysfunction and anovulation. It highlights the critical role of the oocyte in
its traditional responsive role to gonadotropins and ovarian factors,
but also its newer role in guiding ovarian function. With a greater
emphasis on translation of this work to the clinic, we anticipate that
the classic treatments of the past will soon be supplanted by newer
and better evidence-based strategies.
REFERENCES
1. M. M. Matzuk, K. Burns, M. M. Viveiros, J. J. Eppig, Science 296,
2178 (2002).
2. M. M. Matzuk, D. J. Lamb, Nat. Cell Biol. 8 (S1), S41 (2002).
3. M. M. Matzuk, D. J. Lamb, Nat. Med. 14, 1197 (2008).
4. M. A. Edson, A. K. Nagaraja, M. M. Matzuk, Endocr. Rev. 30, 624
(2009).
5. M. M. Matzuk, K. H. Burns, Annu. Rev. Physiol. 74, 503 (2012).
6. R. D. Palmiter, R. L. Brinster, Annu. Rev. Genet. 20, 465 (1986).
7. A. R. Shuldiner, N. Engl. J. Med. 334, 653 (1996).
8. R. L. Brinster, Science 316, 404 (2007).
9. H. Zhang et al., Proc. Natl. Acad. Sci. U.S.A. 109, 12580 (2012).
10. K. Zou, Z. Yuan, Nat. Cell Biol. 11, 631 (2009).
11. Y. A. White et al., Nat. Med. 18, 413 (2012).
12. K. Hayashi et al., Science 338, 971 (2012).
13. J. J. Eppig, K. Wigglesworth, F. L. Pendola. Proc. Natl. Acad. Sci.
U.S.A. 99, 2890 (2002).
14. J. Dong et al., Nature 383, 531 (1996).
15. J. Peng, et al., Proc. Natl. Acad. Sci. U.S.A. 110, e776 (2013).
16. A. Roy, M. M. Matzuk, Nat. Rev. Endocrinol. 7, 517 (2011).
17. C. K. Welt et al., N. Engl. J. Med. 351, 987 (2004).
18. R. S. Legro et al., N. Engl. J. Med. 356, 551 (2007).
19. J. E. Buster et al., Lancet 2, 223 (1983).
20. P. Lutjen et al., Nature 307, 174 (1984).
21. M. V. Sauer, R. J. Paulson, R. A. Lobo, N. Engl. J. Med. 323, 1157
(1990).
22. M. V. Sauer, R. J. Paulson, R. A. Lobo, JAMA 268, 1275 (1992).
23. H. W. Jones, Jr., Fertil. Steril. 90, e5 (2008).
24. C. Coutifaris et al., Fertil. Steril. 82, 1264 (2004).
25. H. L. Hambridge, S.L. Mumford, Hum. Reprod. 28, 1687 (2013).
26. K. L. Sheehan et al., Science 215, 170 (1982).
27. S. Bhattacharya et al., BMJ 337, a716 (2008).
Acknowledgments: Studies on the female reproductive tract in the Matzuk
laboratory have been supported by the Eunice Kennedy Shriver National
Institute of Child Health and Human Development through grants RO1
HD032067, RO1 HD033438, U54 HD007495, and the Ovarian Cancer Research Fund, and in the Legro laboratory by grants R01HD056510, U54
HD034449, and U10 HD 38992.
39
Produced by the Science/AAAS Custom Publishing Office
Infertility: The Male Factor
Dolores J. Lamb1,2,3* and Larry I. Lipshultz1,2
inTRODuCTiOn
Infertility impacts the lives of 8% to 12% of couples attempting to
conceive for the first time. In roughly half of these cases a male
factor is causative yet, surprisingly, in many assisted reproductive
technology (ART) clinics the male is not clinically evaluated beyond
a routine semen analysis. While most textbooks state that primary
infertility in approximately 30% of infertile men is idiopathic, some
experts speculate that 50% to 80% of all male infertility results
from a (usually undiagnosed) genetic cause. In these patients, no
explanation can be found for their impaired semen quality. In part,
this statistic reflects our poor understanding of the basic mechanisms regulating sex determination, genital tract differentiation and
development, spermatogenesis, sperm maturation in the genital
tract, and the molecular events required for fertilization and early
embryonic development; hence, our inability to properly diagnose
the cause of the infertility. Indeed, many of the commonly used diagnostic categories for male infertility are descriptive rather than
mechanistically based. A diagnosis of non-obstructive azoospermia (NOA), testicular failure, cryptorchidism, or ductal obstruction
provides no insight into the molecular cause of the infertility. In this
review, we focus on the molecular defects that underlie the congenital genitourinary birth defects associated with infertility, endocrine
deficiencies, idiopathic spermatogenic failure, and deficiencies of
sperm function.
STRuCTuRAl AnD nuMeRiCAl ChROMOSOMe
DeFeCTS ACCOunT FOR A SigniFiCAnT
PeRCenTAge OF MAle inFeRTiliTy
Karyotype abnormalities are present in about 6% of infertile men
[reviewed in (1)]. With severe spermatogenic deficiency, the incidence of karyotype abnormalities increases. At our tertiary referral
clinic at the Baylor College of Medicine, more than 11% of NOA men
and nearly 17% of men with male genitourinary birth defects have
karyotype anomalies (2). These anomalies can be numerical and affect only the sex chromosomes—for example, Klinefelter syndrome
(46,XXY-46,XXXXY), which accounts for about 14% of all NOA—or
structural, affecting the autosomes or the sex chromosomes, including complex chromosomal rearrangements, deletions, duplications,
inversions, and translocations.
A well-recognized structural chromosome defect causing male
infertility is the occurrence of Y chromosome microdeletions encompassing the azoospermia factor (AZF) region, first described in
1995 (3). The DeletedinAzoospermia(DAZ) gene was the first AZF
spermatogenesis gene identified within a Y chromosome microdele-
Center for Reproductive Medicine, 2Scott Department of Urology, and
Department of Molecular and Cellular Biology, Baylor College
of Medicine, Houston, TX
*Corresponding Author: [email protected]
1
3
40
tion. The AZF region is now further subdivided into smaller segments
termed AZFa, AZFb, and AZFc. Sperm are unlikely to be found in
men with deletions in AZFa or b, whereas men with AZFc deletions
encompassing DAZhave a 75% chance of having rare sperm found
on a testis biopsy [reviewed in (1)]. For AZFcmicrodeletions, the rare
variant having a major effect on spermatogenesis is seen with the
b2/b4 deletion, which accounts for about 6% of severe spermatogenic failure, whereas the more commonly occurring gr/gr deletion has a
modest affect, doubling the risk of severe spermatogenic failure (4) .
Upon homologous recombination and meiotic segregation, autosomal structural defects, such as Robertsonian translocations, can
result in balanced or unbalanced translocations. Infertility can arise
due to the production of unbalanced gametes (nullisomic or disomic)
resulting in aneuploid embryos or chromosomally unbalanced offspring (5). Thus, before proceeding with ART to attempt to achieve
a pregnancy, it is important to know if chromosome anomalies are
present in the male patient.
enDOCRine PAThWAyS ARe CRiTiCAl
FOR nORMAl FeRTiliTy in MAleS
Although endocrine defects account for only about 1% of all male
infertility, the molecular abnormalities that underlie these conditions
are increasingly evident. In short, any genetic defect affecting the
hypothalamic-pituitary-gonadal axis, steroid hormone biosynthesis
and metabolism, steroid hormone receptors and their signaling
pathways, peptide hormones and their receptors and associated
signaling pathways, or paracrine factors and cytokines and their
respective signaling pathways can affect male fertility [reviewed in
(6–8)].
The best example of the relative complexity of genetic defects that
underlie a seemingly discrete infertility phenotype is reveal by the
studies of hypogonadotropic hypogonadism by Crowley and others
(9–19). Gene defects causing GnRH deficiency vary in relation to their
site of action on the ontogeny of the GnRH neurons and the clinical
phenotype observed. For patients with anosmia, neurodevelopmental gene functions that regulate GnRH neuronal migration or olfactory
bulb/tract development are affected due to gene deficiencies, such
as in KAL1andNELF. In contrast, GnRH-deficient patients with normosmia may have defects in genes that encode members of the kisspeptin, neurokinin B, and GnRH signaling pathways such asKISS,
KISS1R, TAC3, TACR3, and GnRHR. Interestingly, defects in the
prokineticin and FGF signaling pathways (FGF8, FGFR1, PROK2/R)
are variably present in patients and families with both anosmia and a
normal sense of smell within the same pedigree [reviewed in (9)]. Despite the identification of numerous monoallelic, bialellic, and digenic
mutations in the genes above, a molecular cause for slightly less than
half of isolated GnRH deficiency remains unknown and a topic of intense investigation. It is clear that even for hypogonadotropic hypogonadism, considerable genetic complexity underlies the causative
defects.
geneTiC AnD genOMiC DeFeCTS in MAle inFeRTiliTy
Using mouse models, investigators were surprised to learn that
genes never thought to be involved in male reproductive function,
when deleted, cause infertility. One of the most unexpected finds
was that loss or pharmacological blockage of testis-expressed taste
genes caused male infertility in the mouse (20). The targeted deletion of TAS1R3, a component of taste receptors, and the gustducin
α-subunit GNAT3, lead to male sterility due to immotile sperm with
poor morphology (detached or amorphous heads, tails flipped over
heads, and multiple kinks and loops in the tails), indicating a potential role in spermiogenesis and sperm maturation (20).
In 2008, Matzuk and Lamb summarized the gene defects in mouse
models and human patients associated with male infertility [see supplement to (7)], and a large number of additional gene defects have
since been defined affecting genitourinary birth defects, disorders of
sexual determination and differentiation, puberty, spermatogenesis,
sperm function, and other aspects of male reproductive health. For
example, cationic channel of sperm (CatSper)—the main flagellar
Ca2+ channel in sperm—was shown to play a role in nongenomic
progresterone signaling (21). Upon activation by progesterone,
calcium influx triggers hyperactivated motility and the acrosome
reaction. While CatSper mutations occur rarely, they can result in
severe asthenozoopsermia (22). Unfortunately, mouse models are
not informative, because unlike humans, the CatSper channel in
the mouse is not sensitive to progesterone. Nevertheless, there are
literally thousands of possible gene defects identified in mice and
humans causing male infertility. In the clinic, however, just one gene
is analyzed on a routine basis in males with congenital bilateral absence of the vas deferens (CBAVD)—the CFTR gene (cystic fibrosis
transmembrane regulatory protein) (23). CBAVD is a genital form of
cystic fibrosis. For patients of North African ethnicity, patients may
also be tested for mutations of the Aurora Kinase C gene, specifically
the c.144delC mutation (24). This mutation results in large-headed,
polyploid, multi-flagellar sperm, because this protein is necessary for
male meiotic cytokinesis. As research continues, additional genetic
testing will no doubt become standard of care in the evaluation of the
infertile male.
FeRTiliTy PReSeRVATiOn FOR PReVenTiOn
OF SeCOnDARy inFeRTiliTy
In the testis, long-cycling isolated spermatogonia, identified by morphological criteria, have been proposed to be testicular stem cells
(25–27). The pioneering work of Brinster and colleagues demonstrated, with certainty, that these testicular stem cells exhibit the
unique properties of prolonged proliferation, self-renewal, generation
of differentiated progeny, maintenance of developmental potential,
and proliferation in response to injury (28). Although work in this area
is predominantly in rodent models and more recently primates, this
represents a research area of significant promise for patients facing
secondary infertility.
Spermatogenesis is damaged by exposure to chemotherapeutic
agents, radiation, toxic agents, and occupational exposures to gonadotoxins, resulting in secondary infertility. Prolonged periods of
azoospermia or severe oligospermia may result. While sterility appears permanent, spermatogenesis may spontaneously recover
years after treatment (29,30) when the surviving reserved stem cells
(type A spermatogonia) reinitiate the proliferative and differentiative
events required for spermatogenesis. Today, cancer patients should
be advised to bank cryopreserved semen prior to potentially harmful
treatment. Children who undergo cancer treatment cannot produce
an ejaculate for cryopreservation, and even for adults, most couples would prefer a natural conception. Accordingly, application of
spermatogonial stem cell isolation and cryopreservation, followed by
germ cell transplantation to restore spermatogenesis after recovery
from cancer treatment, may provide a very useful approach to preservation of fertility for these cancer patients.
A significant roadblock to translation of these studies to clinical
practice has been the concern that the testis may serve as a reservoir for cancer cells (hematological cancers, in particular) and transplantation of spermatogonial stem cells obtained prior to chemotherapy could transmit the malignant cells back into the now healthy
recipient. Recently, a novel cell sorting strategy was devised (21) to
remove malignant contamination while simultaneously enriching the
population of human spermatogonial stem cells. A human to mouse
xenotransplantation biological assay was used to define both the
spermatogonial stem cell activity and malignant contamination of
the enriched cell populations. The development of these separation
technologies will be a major advance towards eventual application
of spermatogonial stem cell transplantation in the clinic. However,
human studies demonstrating safety and efficacy will be needed, as
current purities of 98.8% to 99.8% are still insufficient; even small
numbers of malignant cells present during hematopoietic stem cell
transplantation will prove problematic. Importantly, hematopoietic
stem cell transplantation is required to treat a life-threatening condition whereas fertility preservation is an elective procedure [reviewed
in (31)].
Human spermatogonial stem cells can be cultured under conditions that allow them to exhibit pluripotent potential (32, 33). The
cells exhibit properties similar to embryonic stem cells and can produce teratomas when transplanted into immunodeficient mice. The
human adult germline stem cells can differentiate into the full range
of cell types, including germline cells (32,33).
In the future, spermatogonial stem cells will not only offer the
promise of seminiferous tubule rejuvenation after exposure to
gonadotoxins, but perhaps also the potential to regenerate organs and tissues. Much further in the future, these cells may be
used for gene therapy to cure certain Mendalian inherited genetic
diseases.
heAlTh RiSkS FOR inFeRTile Men gO
BeyOnD TheiR RePRODuCTiVe WOeS
Although it is important to find methods to bypass or overcome severe male factor infertility, to assist severe male factor patients in
achieving a pregnancy, bypassing nature’s barriers to fertilization
by utilizing potential defective sperm for in vitro fertilization by intracytoplasmic sperm injection (ICSI) may have unrecognized consequences for the patient and his offspring. Today we know that Y
chromosome microdeletions occur through events that generate
isodicentric Y chromosomes, as well as Y chromosomes with deletions, duplications, or inversions. The first report of Y chromosome
microdeletions and their association with NOA came from David
Page’s laboratory (described above) (3), but it has been recognized
41
Produced by the Science/AAAS Custom Publishing Office
Figure 1. SHOX deletions and duplications in patients with Y chromosome microdeletions: co-existing genomic syndromes. Y chromosome microdeletions are
usually diagnosed in a clinical genetics laboratory using a multiplex PCR approach. However, when array comparative genomic hybridization is employed, in addition
to the Y chromosome microdeletions found, additional submicroscopic chromosome gains and losses in the pseudoautosomal regions were identified. Some of these
encompass the SHOX gene. Panel A depicts a region on the short arm of the Y chromosome showing a clear deletion (highlighted in red) of SHOX in a Y chromosome
microdeleted man. Panel B shows a Y chromosome microdeleted patient with a duplication (blue highlight) of the region encompassing the SHOX gene. Both of these
men had stature abnormalities as well. It is noteworthy that the plot from the array depicts these gains and losses on the X chromosome (by convention) although fluorescent in situ hybridization reveals that these variations are present on the Y chromosome.
more recently that these structural defects can be generated by a
variety of mechanisms. Indeed, intrachromosomal homologous recombination can generate pseudo-isoY chromosomes and Y chromosome pericentric inversions (34). At one time, it was thought that
about 95% of the Y chromosome (except the pseudoautosomal regions) did not undergo homologous recombination during meiosis.
However, as a result of breakpoint hotspots, as well as homologous
recombination errors due to amplicons associated with palindromic
42
structures present on the human Y chromosome, Y chromosome
microdeletions may occur (35). These rearrangements can even
occur due to amplicons on the short (Yp) and long (Yq) arms of the
Y chromosome through intrachromosomal non-allelic homologous
recombination (34).
These structural Y chromosome defects affect the male specific
Y and, although other genes not involved in spermatogenesis
were affected, the clinical consequence of these defects appeared
to predominantly affect spermatogenesis. However, in part due to
isodicentric Y chromosome formation and complex rearrangements
and deletions/duplications of the Y, about 25% of men with NOA due to
Y chromosome microdeletions have a co-existing genomic syndrome,
SHOX (short stature homeobox-containing gene) syndrome (36)
(Fig.1). SHOX syndrome is the most common cause of short stature
and results from mutations, microdeletions, or microduplications of
the SHOX gene, which is located in the pseudoautosomal region
of the sex chromosomes. SHOX is a dosage-sensitive gene that
does not undergo X-inactivation. While SHOX syndrome occurs in
Turner and Klinefelter syndrome patients (deletion and duplication,
respectively), the clinical spectrum varies. The associated Madelung
deformity of the forearm and mesomelic disproportion of the limbs
develops over time and appears during the second decade of life
(or perhaps never). There are a number of other clinical indications
(among them scoliosis, microgathia, and muscular hypertrophy of
the calves) that are found in some, but not all, SHOX syndrome
patients [reviewed in (37)]. The presence of SHOX deletion or
duplication in a subset of men with Y chromosome microdeletions
suggests that this co-existing genomic syndrome could be inherited
by the male offspring of men conceived by ICSI, although to
date there have been no reports of SHOX syndrome in ICSIconceived offspring.
There may be other associated health issues for the NOA male that
are clinically unappreciated. Our studies show a 2.9-fold increased
risk of cancer in NOA patients (38). Walsh and colleagues (39)
evaluated data on 22,562 partners of infertile couples and showed
the risk of testis cancer was nearly threefold higher in infertile men
compared with normal men (38). Jacobsen etal. similarly evaluated
more than 32,000 men who had their semen analysed over a 30year period and found a 1.6-fold increased risk of testis cancer in
men with reduced sperm counts (40). There was also an increased
incidence of cancers of the peritoneum and other digestive organs
in these men. Walsh and colleagues performed an analysis of their
patient cohort and showed that those with male factor infertility
were 2.6 times more likely to be diagnosed with high-grade prostate
cancer (39,41).
Indeed, a comprehensive male infertility evaluation identified a
number of significant (and previously undiagnosed) medical pathologies. In a landmark paper by Honig et al., significant medical pathologies were uncovered in 1.1% of 1,236 patients presenting to a
male infertility clinic (42). Ten patients (0.8%) had tumors (six testis,
three brain, and one spinal cord) and their findings point to the need
for urologic consultation when an infertile couple seeks treatment at
an ART clinic. It is believed that there are common etiologic factors
for infertility and cancer development, such as deficient DNA repair
mechanisms (39,40,43).
COnCluSiOnS
We have witnessed an exponential increase in advances in our
understanding of the molecular controls of spermatogenesis and
sperm function, as well as increasing definition of the molecular defects associated with infertility in the male. A major challenge that
remains is to enhance our abilities to translate these research advances into routine clinical practice. Additionally, it is essential to
ensure that every male factor patient has a complete history and
physical performed by a urologist or andrologist/endocrinologist,
as well as an in-depth assessment of the causes of his infertility.
Our goal is to have physicians recognize that there may be other
health risks for the infertile male that go beyond their infertility. We
thus look with hope towards the future where the research advances of today will be translated to clinical practice resulting in not
only restoration of fertility for currently infertile men after gonadotoxin exposure, but perhaps even regeneration of organs and tissues without the ethical constraints that limit embryonic stem cell
research today. Importantly, infertile couples must understand that
the cause of their infertility may remain unknown. They also need
to carefully consider the potential health risks for both the patient
and any offspring conceived by ART that go beyond possible birth
defects.
REFERENCES
1. A. W. Pastuszak, D. J. Lamb, J. Androl. 33, 1075 (2012).
2. A. N. Yatsenko et al., J. Urol. 183, 1636 (2010).
3. R. Reijo, R. K. Alagappan, P. Patrizio, D. C. Page, Lancet 347, 1290
(1996).
4. S. G. Rozen et al., Am. J. Hum. Genet. 91, 890 (2012).
5. I. Bernicot et al., Eur. J. Med. Genet. 55, 245 (2012).
6. K. Hwang et al., Ann. N.Y. Acad. Sci. 1214, E1 (2010).
7. M. M. Matzuk, D. J. Lamb, Nat. Med. 14, 1197 (2008).
8. M. M. Matzuk, D. J. Lamb, Nat. Cell Biol. 4 Suppl., s41 (2002).
9. W. F. Crowley, Jr., Mol. Endocrinol. 25, 1989 (2011).
10. B. Mosinger et al., Proc. Natl. Acad. Sci. U.S.A., Epub ahead of print
(2013) doi: 10.1073/pnas.1302827110.
11. J. F. Smith et al., Proc. Natl. Acad. Sci. U.S.A. 110, 6823 (2013).
12. M. S. Hildebrand et al., Eur. J. Hum. Genet. 18, 1178 (2010).
13. A. Anguiano et al., JAMA 267, 1794 (1992).
14. K. Dieterich et al., Hum. Mol. Genet. 18, 1301 (2009).
15. C. Huckins, Cell Tissue Kinet. 4, 335 (1971).
16. C. Huckins, Cell Tissue Kinet. 4, 313 (1971).
17. C. Huckins, Anat. Rec. 169, 533 (1971).
18. R. L. Brinster, J. W. Zimmermann, Proc. Natl. Acad. Sci. U.S.A. 91,
11298 (1994).
19. M. L. Meistrich, G. Wilson, B. W. Brown, M. F. da Cunha, L. I.
Lipshultz, Cancer 70, 2703 (1992).
20. R. M. Pryzant, M. L. Meistrich, G. Wilson, B. Brown, P.
McLaughlin, J. Clin. Oncol. 11, 239 (1993).
21. S. L. Dovey et al., J. Clin. Invest. 123, 1833 (2013).
22. S. Conrad et al., Nature 456, 344 (2008).
23. N. Kossack et al., Stem Cells 27, 138 (2009).
24. J. Lange et al., Genomics 102, 257 (2013).
25. S. Repping et al., Am. J. Hum. Genet. 71, 906 (2002).
26. C. J. Jorgez et al., J. Clin. Endocr. Metab. 96, E674 (2011).
27. G. Binder, Horm. Res. Paediatr. 75, 81 (2011).
28. M. L. Eisenberg, P. Betts, D. Herder, D. J. Lamb, L. I.
Lipshultz, Fertil. Steril., Epub ahead of print (2013) doi: 10.1016/j.
fertnstert.2013.05.022.
29. T. J. Walsh et al., Cancer 116, 2140 (2010).
30. R. Jacobsen et al., BMJ 321, 789 (2000).
31. J. M. Hotaling, T. J. Walsh, Nat. Rev. Urol. 6, 550 (2009).
32. S. C. Honig, L. I. Lipshultz, J. Jarow, Fertil. Steril. 62, 1028 (1994).
33. M. R. Maduro et al., Mol. Hum. Reprod. 9, 61 (2003).
43
Produced by the Science/AAAS Custom Publishing Office
Molecular Interplay in Successful Implantation
Jeeyeon Cha1*, Felipe Vilella2*, Sudhansu K. Dey1**, and Carlos Simón2**
ABSTRACT
Embryo implantation in the maternal womb is a fundamental event in
mammalian procreation. The phases of implantation are apposition,
attachment, and finally penetration of the blastocyst trophectoderm
through the uterine epithelium and into the stroma. Both uterine receptivity and blastocyst activation are required for successful implantation. This review briefly highlights the molecular processes responsible for endometrial receptivity, implantation, and decidualization in
mice and humans.
inTRODuCTiOn
Implantation requires an effective bidirectional communication between the receptive endometrium and an implantation-competent
blastocyst. The duration of the window of implantation (WOI) to
achieve optimal pregnancy outcome is short-lived. Blastocyst attachment to the uterine lining (luminal epithelium) initiates the implantation process, which is followed by localized stromal cell proliferation
and differentiation to generate specialized decidual cells at the site
of implantation, a process called decidualization. Appropriate decidualization directs placentation, in which the maternal circulation
is brought into contact with the embryonic vascular system, establishing a functional placenta. Maternal resources, filtered across the
placental barrier, nourish and protect the fetus. Implantation is the
first significant encounter between the mother and embryo, and is
crucial for a successful pregnancy.
MOleCulAR inSighTS AnD CliniCAl iMPliCATiOnS
OF enDOMeTRiAl ReCePTiViTy
The WOI, a transient interphase between the prereceptive and refractory phases, lasts approximately 24 hours (beginning day four
of pregnancy or pseudopregnancy) in mice and three days in humans during the mid-luteal phase of the menstrual cycle (1–3). Understanding the “molecular clock” at play during the WOI could help
to identify biomarkers of endometrial receptivity useful for increasing
pregnancy rates in in vitro fertilization (IVF) programs or to develop
novel contraceptives.
The master regulators for the acquisition of endometrial receptivity
are estrogen and progesterone (P4). In mice, the transition from the
prereceptive to the receptive phase requires priming with P4 superimposed with a small amount of estrogen. The P4 and estrogen nuclear receptors, receptor chaperones, and co-regulators all play cru-
1
Division of Reproductive Sciences, Cincinnati Children’s Hospital
Medical Center, Cincinnati, OH
2
Fundación Instituto Valenciano de Infertilidad (FIVI), Valencia University,
and Instituto Universitario IVI/INCLIVA, Valencia, Spain
*
These authors contributed equally
**
Corresponding authors: [email protected] (S.K.), Carlos.Simon@ivi.
es (C.S.)
44
cial roles in regulating the WOI. Progesterone receptors (PR-A and
PR-B) and estrogen receptors (ERα and ERβ) are expressed in the
uterus. Studies in mice have shown that these isoforms serve as the
principle modulators during specific stages of pregnancy. ERα (encoded by the Esr1 gene) is the primary mediator of estrogen activity
in the uterus during implantation, as the uteri of Esr1-/- mice are hypoplastic and unable to support implantation (4). Mice missing both
PR-A and PR-B are also infertile and have multiple ovarian and uterine defects, although PR-B–deficient mice show normal fertility (5,
6), indicating that PR-A is the key player in implantation. FKBP52, an
immunophilin co-chaperone for steroid hormone nuclear receptors,
optimizes PR activity and has profound effects during pregnancy (7,
8). In the absence of Fkbp52, P4 signaling and responsiveness are
significantly diminished, resulting in implantation failure. Depending
on the genetic background of the mice, this functional P4 resistance
can be overcome by exogenous P4 administration (9). FKBP52 is
also expressed in human endometrium (10).
ER and PR signaling during implantation are executed by juxtacrine, paracrine, and autocrine factors orchestrated by various
growth factors, cytokines, lipid mediators, homeobox transcription
factors, and morphogens. Mouse models have improved our mechanistic understanding of several factors critical for uterine receptivity and implantation (Fig. 1; also see reviews 1and2). Among the
growth factors, heparin-binding EGF-like growth factor (HB-EGF)
stands out as a major player in mediating embryo-uterine interactions during the attachment reaction in both mice and humans (Fig.
2; 11–14). Membrane-anchored HB-EGF expressed in the luminal
epithelium functions as a juxtacrine mediator for adhesion of the
blastocyst expressing EGF receptors, while soluble HB-EGF influences embryo-uterine interactions in a paracrine manner (1). Juxtacrine interactions between the luminal epithelium and blastocyst in
humans are also mediated by L-selectin ligands and their receptors
displayed on the luminal epithelium and blastocyst cell surface, respectively (15).
Leukemia inhibitory factor (LIF), a member of the interleukin (IL)6 family of cytokines, is expressed in mouse uterine glands early
on day four of pregnancy (the day of uterine receptivity) and in the
stroma surrounding the blastocyst upon attachment late on day four
(16, 17). This factor is essential for uterine receptivity and implantation via binding to its receptor, LIFR, and co-receptor, gp130, to
activate STAT3 signaling. LIF-gp130-STAT3 signaling is critical to
implantation since uterine deletion of the genes encoding gp130 or
Stat3has been shown to cause implantation failure (18,19). More
recently identified mediators critical for uterine receptivity are muscle
segment homeobox genes (Msx1/Msx2); conditional uterine deletion
of these genes results in implantation failure (Fig. 3; 20).
In some respects, implantation resembles a pro-inflammatory
reaction and involves inflammatory mediators. Cyclooxygenase-2
(Cox2), a critical enzyme for prostaglandin (PG) biosynthesis, is
Figure 1. Signaling network for uterine receptivity and implantation. Hybrid cartoon, based on mouse and human studies, portraying compartment- and cell-type
specific expression of molecules and their potential functions critical for uterine receptivity, implantation, and decidualization. Interplay of ovarian P4/estrogen-dependent
and independent factors in the pregnant uterus in specific compartments contributes to the success of implantation in a juxtacrine/paracrine/autocrine manner. During
attachment, interactions between the blastocyst and luminal epithelium (LE) involve ErbB1/4 (blue) and both transmembrane (TM) and soluble (Sol) forms of HB-EGF, as
well as L-selectin ligands expressed by the luminal epithelium to L-selectin receptors on the blastocyst. The other critical signaling pathways for uterine receptivity and
implantation are also shown. AA, arachidonic acid; COUP-TFII, chicken ovalbumin upstream promoter transcription factor-2; E, estrogens; EC, epithelial cell (luminal
and glandular epithelia); ENaC, epithelium sodium channel; ErbB1/4, epidermal growth factor receptor 1/4; GE, glandular epithelium; Hand2, Heart- and neural crest
derivatives-expressed protein 2; HB-EGF, heparin-binding EGF-like growth factor; ICM, inner cell mass; KLF5, Kruppel-like factor 5; LPA3, lysophosphatidic acid receptor
3; MSX1, muscle segment homeobox 1; PPARδ, peroxisome proliferator-activating receptor δ; Ptc, Patched; RXR, retinoid X receptor; SC, stromal cell; SGK1, serumand glucocorticoid-inducible kinase 1; Smo, Smoothened; Tr, trophectoderm. Compartment colors: blue, stroma; pink, luminal epithelium; orange, glandular epithelium;
purple, epithelium at the attachment site. Adapted from (1).
expressed in the uterus at the site of implantation (21–23). Cox2derived PGs are thought to participate in implantation since ablation
of the Ptgs2 (Cox2) gene leads to infertility (21–23). PGs also
influence vascular permeability and decidualization in the stromal
bed (24). Cox2-derived prostacyclin (PGI2) activates PPARδ, which
appears to influence implantation as Pparδ-/- mice show deferred
implantation and compromised pregnancy outcome (22). Cytosolic
phospholipase A2α (cPLA2α, encoded by Pla2g4a), which generates
arachidonic acid for Cox2-generated PG synthesis, is expressed
in the uterus similarly to Cox2. Its critical role in implantation is
evidenced by deferred implantation and related adverse effects
in Pla2g4a –/– mice, compromising pregnancy outcome (25).
Lpar3-/-mice exhibit a similar phenotype to Pla2g4a–/–mice, exhibiting
downregulated Cox2 (26; reviewed in 1and2).
Among other lipid mediators, endogenous cannabinoid-like compounds (endocannabinoids) and their G-protein coupled receptors Cnr1 (CB1) and Cnr2 (CB2) also play roles in implantation.
Endocannabinoids, N-arachidonoylethanolamine (anandamide),
and 2-arachidonoyl glycerol (2-AG), bind to CB1 and CB2. Uterine
anandamide levels are higher during the prereceptive phase, but
decrease during the receptive phase (27). Similarly increased anandamide levels resulting from deficiency of fatty acid amide hydrolase
(FAAH), which degrades anandamide, lead to multiple pregnancy
defects, including disruption of on-time implantation (28). These re-
45
Produced by the Science/AAAS Custom Publishing Office
Figure 2. HB-EGF serves as a reciprocal mediator between the luminal
epithelium and activated blastocyst during attachment reaction. (A) In
situ hybridization of HB-EGF mRNA in the mouse uterus at 1600 hours on day
four of pregnancy. Note distinct hybridization signals in luminal epithelial cells
surrounding two blastocysts in a longitudinal section. Arrows demarcate location
of blastocysts. (B) Scanning electron microscopy of blastocysts co-cultured with
32D cells. Zona-free day four (0900 hours) mouse blastocysts were co-cultured
with (a) parental 32D cells, (b) 32D cells displaying transmembrane form of HBEGFTM, or (c) 32D cells synthesizing soluble form of HB-EGF for 36 hours. After
extensive washing to remove loosely adhering cells, blastocysts were fixed
and examined by scanning electron microscopy; Magnification 800X. Arrows
point to 32D cells. (C) Bmp2 gene expression in response to beads preloaded
with HB-EGF. Beads preabsorbed either in BSA (control, a) or HB-EGF (100 ng/
mL, b) were transferred into uterine lumens of day four pseudopregnant mice.
Bmp2 expression at the sites of beads were examined on day five. Arrows
indicate the locations of the beads. Note localized stromal expression of
Bmp2 at the sites of beads preabsorbed with HB-EGF. Bar, 75 mm. Reprinted
from (12, 13, 74).
46
sults indicate that a tight regulation of endocannabinoid signaling is
critical to uterine receptivity and oviductal transport of embryos (see
page 21). Aberrant anandamide levels are also associated with recurrent pregnancy loss and ectopic pregnancy in women (29,30).
In humans, receptive status is typically defined by histological
characteristics known as the Noyes criteria (31). However, the accuracy and relevance of these criteria have been questioned (32,33).
Thus, the search is on-going for specific biomarkers that define the
WOI. Morphological changes of the endometrial luminal epithelium
include modifications of the plasma membrane (34) and cytoskeleton (35–37). The presence of pinopodes was proposed as a receptivity marker (38,39), but these ectoplasmic epithelial projections are
not specific to the receptive state (40). Biochemical markers such as
integrins (41), MUC1 (42), calcitonin (43), LIF (16–17), and HOXA10
(44) initially generated interest, but none have proven to be effective
in clinical practice.
Microarray technology has advanced the search for biomarkers
based on the transcriptomics signature during the WOI (45). Recent
evidence has helped to characterize the human endometrial cycle
by expression profiling with respect to receptive status (3,46). A diagnostic tool has been developed to identify receptive endometrium
using a specific transcriptomics signature in both natural and hormonal replacement therapy cycles (47). The endometrial receptivity array (ERA) is a customised microarray platform of 238 genes
differentially expressed during the menstrual cycle that can identify
the WOI of each patient (48). ERA appears superior to endometrial
histology, and results are reproducible in the same patient on the
same day of her menstrual cycle up to 40 months later (48). ERA
for WOI detection to schedule embryo transfer in patients with recurrent implantation failure has recently been reported as a novel
IVF therapeutic strategy (49). The proteomic profile of the human
endometrium throughout the menstrual cycle has also been studied (50–52). However, despite the large amount of information generated, a consensus profile has not been established. Analysis of
endometrial secretions (secretomics) is an emerging, noninvasive
alternative, since endometrial fluid aspiration in the same treatment
cycle does not affect pregnancy rates (53).
mans, this transformation is accompanied by the secretion of prolactin (PRL) and insulin-like growth factor binding protein-1 (IGFBP-1)
(58) with changes in extracellular matrix proteins such as laminin,
type IV collagen, fibronectin, and heparin sulphate proteoglycans
(59–61).
Proteomic analysis during human decidualization revealed 60
differentially expressed proteins including known markers (cathepsin B) and new potential biomarkers (transglutaminase 2 and peroxiredoxin 4) (59). Secretomics analysis during this process identified 13 secreted proteins including established (IGFBP-1 and PRL)
and novel (MPIF-1 and PECAM-1) markers (61).
DeCiDuAlizATiOn iS CRiTiCAl
FOR PRegnAnCy SuCCeSS
During decidualization in a normal pregnancy in mice, the implanting
blastocyst initiates changes in the surrounding stromal cells, inducing stromal proliferation and differentiation into decidual cells (1). In
humans, the initiation of this process (predecidualization) does not
require the presence of a blastocyst; however, the process becomes
robust with implantation. Predecidualization prepares the endometrium for implantation and appears akin to the extensive stromal cell
proliferation with expression of decidual markers seen prior to implantation in mice (1). Decidualization involves morphogenic, molecular, and vascular modifications that are initiated by estrogen following P4 priming. Hoxa10 and Hoxa11, critical for patterning during
development, are also expressed in the uterus and have crucial roles
in decidualization; deletion of these genes produces compromised
P4 signaling and fertility in mice (54–57). Morphologically, decidualization is characterized by elongated stromal cells transforming into
enlarged round cells with specific ultrastructural modifications. In hu-
APPROPRiATe TROPhOBlAST inVASiOn
iS CRiTiCAl TO PlACenTATiOn
Trophoblast invasion through the maternal decidua induces extracellular matrix (ECM) remodeling regulated by both the trophectoderm and the stroma (62). Metalloproteinases (MMPs) play key
roles in degrading ECM components (63). MMP-2 and MMP-9 are
expressed and secreted by the human endometrium and participate
in tissue remodelling during implantation and promote menstruation
during normal cycles (64–67). Conversely, decidual cells activate
the expression of tissue inhibitors of MMPs (TIMPs) and downregulate urokinase plasminogen activator (uPA) (65). It is thought that
TIMPs limit ECM degradation by MMPs during decidualization, restricting embryonic invasion. A balance between these factors is
crucial for a successful pregnancy outcome (1, 68–70). Aberrant
decidual display of ECM components is associated with preeclampsia or restricted intrauterine growth (71, 72). It was also recently
reported that histone acetylation derails the balance of ECM modulators, inhibiting trophoblast invasion (73).
Figure 3. Msx genes regulate uterine luminal epithelial cell polarity during receptivity. Confocal images of E-cadherin and β-catenin
colocalization. Retention of the luminal epithelium (le) in Msx1/Msx2d/d mice at the site of blastocyst apposition on day 6, as opposed to luminal
epithelium breakdown in Msx1/Msx2f/f mice with loss of E-cadherin. Arrowhead indicates the location of blastocysts. s, stroma. Scale bar,
100 µm. Reprinted from (20).
ADVAnCeS AnD ChAllengeS
Uterine transition through the prereceptive, receptive, and refractory phases is dynamic and rapid. Many genes show overlapping
expression spanning more than one phase and/or uterine tissue
compartment, making stage- and tissue-specific contributions of particular genes difficult to ascertain with current tools. For example,
Bmp2, which is expressed in the stroma during attachment and
decidualization, is considered to be critical for decidualization in
mice (74, 75), but its exact mechanism of action is still unknown.
Cre recombinase-expressing mouse lines have been used to delete or overexpress floxed genes, but expression of these genes in
multiple cell types from the uterus, ovary, and pituitary often limits
interpretation of results. Therefore, there is still a need for the development of reproductive tissue- and cell-specific Cre lines. Generation of inducible conditional knockout mouse models could be
valuable in addressing stage-specific effects of a gene with overlapping expression patterns. However, the transient and dynamic expression of some genes makes the utility of such inducible
systems questionable.
Limited numbers of systems biological approaches—embracing
genomics, transcriptomics, proteomics, lipidomics, and metabolomics—have been used to study the intricate and dynamic changes
underlying implantation. These studies have produced a wealth of
information, but effective assimilation and rigorous validation of the
results are lacking, limiting their impact.
Comparative research on implantation across species has become
rare in recent years. Studies contrasting different strategies and/or
hormonal requirements could help to identify common mechanisms
underlying implantation, better understand the causes of female
infertility, and improve pregnancy rates. In particular, the transition
47
Produced by the Science/AAAS Custom Publishing Office
of the uterus from receptive to nonreceptive states requires special
attention since the molecular interplay required remains largely
unknown and may be critical to extend the WOI during IVF. The
complexities of pregnancy necessitate continued basic and clinical
research, since aberrant implantation leads to compromised
pregnancy outcomes and may even impact the long term health of
offspring.
REFERENCES
1. J. Cha, X. Sun, S. K. Dey, Nat. Med. 18, 1754 (2012).
2. H. Wang, S. K. Dey, Nat. Rev. Genet. 7, 185 (2006).
3. M. Ruiz-Alonso, D. Blesa, C. Simon, Biochim. Biophys. Acta. 1822,
1931 (2012).
4. D. B. Lubahn et al., Proc. Natl. Acad. Sci. U.S.A. 90, 11162 (1993).
5. J. P. Lydon et al., Genes Dev. 9, 2266 (1995).
6. B. Mulac-Jericevic et al., Science 289, 1751 (2000).
7. S. Tranguch et al., Proc. Natl. Acad. Sci. U.S.A. 102, 14326 (2005).
8. Z. Yang et al., Mol. Endocrinol. 20, 2682 (2006).
9. S. Tranguch et al., J. Clin. Invest. 117, 1824 (2007).
10. H. Yang et al., Reproduction 143, 531 (2012).
11. H. Xie et al., Proc. Natl. Acad. Sci. U.S.A. 104, 18315 (2007).
12. S. K. Das et al., Development 120, 1071 (1994).
13. G. Raab et al., Development 122, 637 (1996).
14. H. J. Yoo, D. H. Barlow, H. J. Mardon, Dev. Genet. 21, 102 (1997).
15. O. D. Genbacev et al., Science 299, 405 (2003).
16. C. L. Stewart et al., Nature 359, 76 (1992).
17. H. Song et al., Mol. Endocrinol. 14, 1147 (2000).
18. J. H. Lee et al., FASEB J. 27, 2553 (2013).
19. X. Sun, Bartos, A., Whitsett, J., and Dey, S.K., Mol. Endocrinol. Epub
ahead of print (2013) doi:10.1210/me.2013.
20. T. Daikoku et al., Dev. Cell 21, 1014 (2011).
21. H. Lim et al., Cell 91, 197 (1997).
22. H. Lim et al., Genes Dev. 13, 1561 (1999).
23. H. Wang et al., J. Biol. Chem. 279, 10649 (2004).
24. H. Matsumoto et al., J. Biol. Chem. 277, 29260 (2002).
25. H. Song et al., Development 129, 2879 (2002).
26. X. Ye et al., Nature 435, 104 (2005).
27. B. C. Paria et al., J. Biol. Chem. 276, 20523 (2001).
28. H. Wang et al., J. Clin. Invest. 116, 2122 (2006).
29. M. Maccarrone et al., Lancet 355, 1326 (2000).
30. A. K. Gebeh et al., J. Clin. Endocrinol. Metab. 98, 1226 (2013).
31. R. W. Noyes, A. T. Hertig, J. Rock, Am. J. Obstet. Gynecol. 122, 262
(1975).
32. M. J. Murray et al., Fertil. Steril. 81, 1333 (2004).
33. C. Coutifaris et al., Fertil. Steril. 82, 1264 (2004).
34. C. R. Murphy, Cell Res. 14, 259 (2004).
35. J. C. Martin et al., Biol. Reprod. 63, 1370 (2000).
36. M. Thie, P. Fuchs, H. W. Denker, Int. J. Dev. Biol. 40, 389 (1996).
37. M. Thie et al., Eur. J. Cell. Biol. 66, 180 (1995).
38. G. Nikas, Hum. Reprod. 14 Suppl. 2, 37 (1999).
39. B. A. Lessey, Fertil. Steril. 96, 522 (2011).
40. C. E. Quinn, R. F. Casper, Hum. Reprod. Update 15, 229 (2009).
41. B. A. Lessey et al., J. Clin. Invest. 90, 188 (1992).
42. M. Meseguer, A. Pellicer, C. Simon, Mol. Hum. Reprod. 4, 1089
(1998).
43. S. Kumar et al., J. Clin. Endocrinol. Metab. 83, 4443 (1998).
48
44. H. S. Taylor, P. Igarashi, D. L. Olive, A. Arici, J. Clin. Endocrinol.
Metab. 84, 1129 (1999).
45. M. Schena, D. Shalon, R. W. Davis, P. O. Brown, Science 270, 467
(1995).
46. J. A. Horcajadas, A. Pellicer, C. Simon, Hum. Reprod. Update 13, 77
(2007).
47. P. Diaz-Gimeno et al., Fertil. Steril. 95, 50, e51-15 (2011).
48. P. Diaz-Gimeno et al., Fertil. Steril. 99, 508 (2013).
49. M. Ruiz-Alonso et al., Fertil. Steril. Epub ahead of print (2013) doi:
10.1016/j.fertnstert.2013.05.004.
50. T. Parmar et al., Fertil. Steril. 92, 1091 (2009).
51. F. Dominguez et al., Hum. Reprod. 24, 2607 (2009).
52. S. Altmae et al., Mol. Hum. Reprod. 16, 178 (2010).
53. M. H. van der Gaast et al., Reprod. Biomed. Online 7, 105
(2003).
54. H. Lim et al., Mol. Endocrinol. 13, 1005 (1999).
55. I. Satokata, G. Benson, R. Maas, Nature 374, 460 (1995).
56. H. M. Hsieh-Li et al., Development 121, 1373 (1995).
57. A. M. Raines et al., Development 140, 2942 (2013).
58. J. C. Irwin, L. de las Fuentes, B. A. Dsupin, L. C. Giudice, Regul.
Pept. 48, 165 (1993).
59. C. L. Dunn, R. W. Kelly, H. O. Critchley, Reprod. Biomed. Online 7,
151 (2003).
60. S. Tabanelli, B. Tang, E. Gurpide, J. Steroid. Biochem. Mol. Biol. 42,
337 (1992).
61. T. Garrido-Gomez et al., J. Clin. Endocrinol. Metab. 96, 706
(2011).
62. F. van den Brule et al., Chem. Immunol. Allergy 88, 163 (2005).
63. A. Page-McCaw, A. J. Ewald, Z. Werb, Nat. Rev. Mol. Cell. Biol. 8,
221 (2007).
64. W. H. Rodgers et al., J. Clin. Invest. 94, 946 (1994).
65. F. Schatz et al., Semin. Reprod. Endocrinol. 17, 3 (1999).
66. L. A. Salamonsen, G. Nie, Rev. Endocr. Metab. Disord. 3, 133
(2002).
67. S. K. Das et al., Dev. Genet. 21, 44 (1997).
68. P. K. Lala, C. Chakraborty, Placenta 24, 575 (2003).
69. M. Knofler, Int. J. Dev. Biol. 54, 269 (2010).
70. V. Plaks et al., Proc. Natl. Acad. Sci. U.S.A. 110, 11109
(2013).
71. J. V. Cockle et al., Reprod. Sci. 14, 629 (2007).
72. C. J. Lockwood et al., Biol. Reprod. 78, 1064 (2008).
73. C. Estella et al., PLoS ONE 7, e30508 (2012).
74. B. C. Paria et al., Proc. Natl. Acad. Sci. U.S.A. 98, 1047 (2001).
75. K. Y. Lee et al., Mol. Cell. Biol. 27, 5468 (2007).
Acknowledgments: Work of SKD and JC was funded by grants from the
National Institute of Child Health and Human Development, National Institute on Drug Abuse, and National Institute of Aging of the U.S. National
Institutes of Health. Work of CS and FV was funded by grants from the
European Union FP7 People Programme, namely the Marie Curie Actions
Marie Curie Industry-Academia Partnerships and Pathways (IAPP SARM
324509), and through the Spanish Government (CDTI-EUREKA E!6478 –
NOTED and Ministerio de Economía y Competitividad SAF2012-31017)
and Valencia Government Generalitat Valenciana (Conselleria d’Educació
PROMETEOII/2013/018).
Reproductive medicine section of the SSIF website
www.reproductive-medicine.seronosymposia.org
Take a look at the reproductive medicine section of the Serono Symposia
International Foundation (SSIF) website and register to access free e-learning
activities and receive updates on next live educational events and resources.
Are you finding it a challenge to keep up with the
latest developments in reproductive medicine?
Registering with the SSIF reproductive medicine
website could be just the solution you need.
And it is absolutely free.
All you need to do to enjoy the wealth of information
that SSIF provides about your specialist field is to
visit reproductive-medicine.seronosymposia.org and
click on the button to register.
It is quick and easy and you will have immediate
access to:
• Upcoming educational activities in reproductive
medicine
• Online video lectures by international experts
• Online courses to increase your knowledge
• Slide and video presentations from SSIF symposia
… And much more!
follow us on
SSIF_RM
http://twitter.com/SSIF_RM
Serono Symposia International Foundation | Improving the patient's life through medical education
iii
Conference Videos
Themammalianoocyte
orchestratestherateofovarian
folliculardevelopment
Speaker (L): Martin M. Matzuk
Challenger (R): Antonio Pellicer
Main Presentation: bit.ly/IInMfT
Challenger Response: bit.ly/IDt5ws
Non-invasiveimagingofhuman
embryosbeforeembryonicgenome
activationpredictsdevelopmentto
theblastocyststage
Speaker (L): Renee A. Reijo-Pera
Challenger (R): Carlos Simón
Main Presentation: bit.ly/1iuUGk1
Challenger Response: bit.ly/1g1QE0q
AberrantCannabinoid
signalingimpairsoviductal
transportifembryos
Speaker (L): Sudhansu K. Dey
Challenger (R): Hilary O.D. Critchley
Main Presentation: bit.ly/IIoMAB
Challenger Response: bit.ly/18dFTDn
Electivesingle-embryo
transferversusdouble-embryo
transferininvitrofertilization
Speaker (L): Christina Bergh
Challenger (R): Robert Fischer
Main Presentation: bit.ly/1jfJVjf
Challenger Response: bit.ly/1ciKKEI
Speaker Response: bit.ly/1cW8tJ2
iv
Link to The Top Ten in Reproductive Medicine
conference on the SSIF website here
Comparisonofconcomitant
outcomeachievedwithfreshand
cryopreserveddonoroocytesvitrified
bytheCryotopmethod
Speaker: Ana Cobo
Main Presentation: bit.ly/1bFx3S2
Accuratesinglecell24chromosome
aneuploidyscreeningusingwholegenome
amplificationandsinglenucleotide
polymorphismmicroarrays
Speaker (L): Richard T. Scott, Jr.
Challenger (R): Carlos Simón
Main Presentation: bit.ly/IBTdrk
Challenger Response: bit.ly/1bbRoN5
Spermmorphology,motility,
andconcentrationinfertile
andinfertilemen
Speaker (L): David S. Guzick
Challenger (R): Richard T. Scott, Jr.
Main Presentation: bit.ly/1iuWf1i
Challenger Response: bit.ly/1atpl7m
Circulatingangiogenicfactorsand
theriskofpreeclampsia
Speaker (L): S. Ananth Karumanchi
Challenger (R): Randy Levinson
Main Presentation: bit.ly/1c8zXJK
Challenger Response: bit.ly/18X6y88
v