¿ìè¶ÌÊÓƵ

SOON, MPs in Britain will debate the government’s bill to regulate test-tube
baby clinics and research on embryos. But the government has prevaricated
on the central issue: research on human embryos. MPs will vote either to
ban research on embryos completely, or to allow some research on embryos
less than 14 days old, under licence. Politicians will decide whether to
make illegal what is now a thriving area of progress in our understanding
of human reproduction and early development.

¿ìè¶ÌÊÓƵs and clinicians in the field are naturally worried by the
prospect of a ban. They see it as a disaster on many fronts. It would prevent
the diagnosis of severe genetic disorders in embryos, which recent research
has shown to be feasible. It would hinder the development of new forms of
contraception, such as vaccines. And it would prevent any significant improvement
in the treatment of infertility.

Arguments over whether research on human embryos is ethical tend to
rest on conflicting beliefs about when embryonic cells should be awarded
the status of a person. Here scientific findings become rather peripheral
to ideology. But it is nonetheless true that in the course of normal reproduction
between fertile individuals, most human embryos, perhaps as many as two-thirds
of those conceived in vivo, fail to develop to term. Human embryos can become
babies only if they implant into a woman’s uterus during a narrow ‘window’
of time in their development, and develop successfully thereafter. In the
natural course of events, about half fail before or during implantation,
when the embryo is a ball of about 100 cells – about the size of a full
stop – called a blastocyst.

The Warnock committee, set up by the government in 1983 to debate such
issues, struggled to accommodate those who believe that early embryos, whether
destined to develop further or not, are nonetheless endowed with a soul.
Meeting religion with rational argument, the committee tried to pinpoint
the moment when an embryo could first be said to be an ‘individual’ – that
is, one entity and not two or none. In its report, published in 1984, the
committee decided to recommend a time limit for research, set at 14 days
after conception, because that is when the ‘primitive streak’ first appears.
This event clearly demarcates those cells that might go on to form an embryo
from those that would form the supporting membranes. Before this stage,
two individuals, or none, might develop from the fertilised egg.

The Warnock committee may have hoped that this time limit might convince
even the religious; if you assign a soul to an earlier embryo you have to
be prepared to agree that the soul is not firmly established, or that it
can split, if none, or two fetuses subsequently develop. But the religious
opponents of research on embryos remained unconvinced. They maintain that
even a fertilised egg is a person.

Ironically, the Warnock committee’s attempt to placate this camp may
have inadvertently encouraged the notion that an embryo is really a tiny
person. By according a 14-day-old embryo in vitro the status of an ‘individual’,
the committee seemed to imply that ‘personhood’ begins at a precise moment
in embryonic development, whatever the embryo’s circumstances. But in fact,
an early embryo is only potentially a person if it can implant in a woman’s
uterus and be nurtured by her as it develops, until it is capable of independent
life. In the laboratory, an embryo ceases to develop soon after it reaches
the blastocyst stage.

The Warnock committee might have avoided such confusions if it had taken
the stance adopted by the Polkinghorne committee, which reported earlier
this year on the ethics of using fetal tissue. Human embryos in the laboratory
could be regarded as similar to donated human organs, and to fetal tissues
obtained from abortions. As the Polkinghorne report on fetal tissues made
clear, such material is worthy of special respect due to its human origin
and should not form part of any commercial transactions. A statutory body
should regulate the use of such material for research or therapeutic purposes,
with the informed consent of the donor.

Embryos now used in research are ‘spare’ embryos. Current techniques
of in vitro fertilisation (IVF) produce more embryos than can safely be
transferred in one cycle of treatment. The surplus could be frozen, to be
transferred at a later date, but the process has a low ‘take-home baby rate’.
Furthermore, some of these embryos look abnormal, with misshapen cells,
which makes clinicians reluctant to transfer them, although no one yet knows
exactly how to judge the health of an embryo. Women undergoing IVF or sterilisation
have donated such embryos to projects designed to tell us more about early
human development. IVF would never have developed without research on embryos
and will never improve without it.

Anti-abortionists and the Catholic church would allow experiments on
human embryos only if they are then returned to a woman’s uterus. Researchers,
in contrast, consider it unethical to do this, because one could never know
what damage might have been done as a result of the research procedure.

One way round objections to research on human embryos might be to study
animal embryos instead. Yet, as toxicologists know only too well, the findings
of studies on animals cannot always be extrapolated directly to people.
Recent research has produced examples of just how different human embryos
can be from those of mice or monkeys. These findings suggest that it is
unethical not to study human embryos, if in vitro fertilisation and embryo
transfer is to continue as a treatment for infertility. Researchers argue
that we need to study spare embryos produced through IVF first, to safeguard
and improve the chances of the ones we ‘put back’.

The dangers of extrapolating from mouse to human is demonstrated by
some recent work by Martin Johnson and his colleagues in the department
of anatomy at the University of Cambridge. He has discovered that human
eggs are much more vulnerable than mouse eggs to cooling. ‘Handling procedures
are causing avoidable problems that probably contribute to the low success
rate of IVF,’ he says.

A fall in temperature, below 37 Degree C, disrupts the microtubules
that hold the egg’s chromosomes. These microtubules, known as the spindle,
play a crucial role in the developing egg. Before they are fertilised, human
eggs are in a sort of suspended animation – arrested in a particular phase
of cell division. To complete their development, the spindle has to pull
two sets of chromosomes apart and eject one set as a so-called polar body.
Keeping eggs at room temperature (25 Degree C) for just 5 or 10 minutes
wreaks havoc with the spindle, causing it to depolymerise, unravel and eventually
break up altogether. The chromosomes then disperse throughout the giant
cell. Many mouse eggs recover if restored to body temperature, but, Johnson
finds, most human eggs do not.

Johnson thinks human eggs are so much more sensitive to lower temperatures
than mouse eggs because of differences in the way microtubules are distributed
in the two sorts of eggs. Mouse eggs have clumps of microtubules, known
as pericentriolar material, around their periphery. These clumps act as
foci for the reconstruction of the spindles damaged by low temperatures.
Human eggs have much less of this pericentriolar material, and what is there
is concentrated around the spindles themselves. When human eggs are cooled,
the material migrates away from the spindle, making it more difficult for
the human egg to rebuild a spindle.

Although infertility clinics now routinely transfer previously frozen
embryos, very few babies have been born from unfertilised eggs that have
been frozen and later fertilised. The special sensitivity of human eggs
could explain why. The damage, says Johnson, is probably done when embryologists
are preparing the eggs for freezing, and when thawing them. To make matters
worse, the chemicals added to prevent the formation of ice crystals in the
eggs also seem to damage microtubules. Johnson found that a standard cryoprotectant,
DSMO, causes a massive proliferation of microtubules – exactly the opposite
of what happens when an egg is cooled. This proliferation also disrupts
the spindle, detaching chromosomes from their orderly alignment. Removing
DSMO did not undo all the damage. ‘Simply adding cryopreservant devastated
up to 40 per cent of the oocytes irreversibly,’ said Johnson.

Johnson and his colleagues have also found that a chemical medium used
by some clinicians can damage human eggs. This medium, known as acid Tyrode’s,
softens the outer layer of the egg, the zona pellucida. The procedure, called
zona drilling, can enable sparse or feeble sperm to fertilise an egg. But
Johnson finds that as many as a third of human eggs treated with acid Tyrode’s
act as if they had been fertilised – they are parthenogenetically activated.
Such eggs look like normally fertilised ones, but are doomed to die soon
after they have implanted in the uterus. Johnson suspects that this mishap
may contribute to the loss of pregnancies after in vitro fertilisation and
embryo transfer. His research has amply demonstrated that ‘the mouse is
not necessarily an adequate model for procedures destined to be used in
humans. We have to do experiments on human eggs and embryos as well as resolve
these problems.’

Can mice stand in for people in the study of human genetic diseases?
Recent research in Cambridge and London has discovered striking differences
in the timetable for genetic events in the two species. Marilyn Monk of
the MRC’s Mammalian Development Unit in London has developed procedures
for the diagnosis of genetic diseases in mouse embryos by biochemical assays.
To use this technique to detect human genetic disorders, we have to know
precisely when a certain gene is switched on in a human embryo.

The goal of ‘preimplantation diagnosis’ is to give couples known to
be at risk of passing on a severe genetic disease the option of having their
embryos screened in the laboratory. This procedure would enable the woman
to begin a pregnancy knowing that the embryo will escape the disease. Preimplantation
diagnosis can work in one of two ways. In the first approach, researchers
could remove a cell from, say, an eight-cell embryo, and analyse it biochemically
for signs of the faulty protein produced from the defective gene. The alternative
is to analyse the DNA itself, to look directly for the defective gene.

Several years ago, Monk and her colleagues pioneered the biochemical
approach to preimplantation diagnosis. Studying mouse embryos, they showed
that they could remove a cell without damaging the embryo, and determine
whether it produced crucial enzymes such as HPRT and ADA. Deficiencies in
these enzymes cause severe genetic diseases in humans. People who do not
produce this enzyme develop a rare but horrific disorder known as Lesch
Nyhan syndrome. Those deficient in ADA usually die in the first year of
life from severe combined immune deficiency disease.

Yet this technique depends crucially on knowing when the genes in question
are ‘switched on’ and begin to produce their protein products, the enzymes.
Otherwise, the tests may simply measure the protein produced from the mother’s
genome, when the egg was formed. Monk in London, together with Peter Braude,
Johnson and their colleagues in Cambridge, have found that the biochemical
assay developed in mice cannot be simply applied to human embryos. ‘In people,
we do not know whether the HPRT gene is not switched on in the embryo, or
if it is masked by high levels of HPRT in the egg. We need human studies
to avoid the dangers of unjustified extrapolation from mice to man,’ Monk
says.

Underlining the differences between mice and people, Braude and his
colleagues at Cambridge found evidence that a human embryo first switches
on some of its genes when it is at the 4- to 8-cell stage, 2.5 days after
fertilisation. Mouse embryos seem to activate many of their genes much earlier,
at the 2-cell stage.

If we know when a gene is active in a young embryo, the biochemical
approach to preimplantation diagnosis can be useful when researchers have
already discovered the protein defect that underlies a genetic disease.
But in other cases, preimplantation diagnosis would have to rely on an analysis
of the DNA itself. Monk and her colleague Cathy Holding have recently shown
that this second approach is also feasible: they could detect whether a
mutant gene was present in a single cell of a mouse embryo. They diagnosed
a defect in a mouse carrying a mutation equivalent to that causing the human
disease beta-thalassaemia.

Monk and Holding achieved this by modifying a technique called the polymerase
chain reaction. This procedure, developed several years ago, exploits the
ability of the enzyme DNA polymerase to make more copies of a specific sequence
of DNA, provided that the bounds of the sequence copied are set by two ‘primers’.
These are short bits of synthesised DNA that bind to either end of that
DNA. The primers attach to the sequence, once the strands of DNA separate
at high temperature. Then the polymerase moves from the primers, elongating
a new strand of DNA. By repeating this cycle many times researchers can
easily produce more than a million copies of a single gene sequence, and
so obtain enough DNA to analyse accurately.

Monk and Holding increased the sensitivity and specificity of the reaction,
so that it copied just the right gene sequence in a single cell, by using
two sets of primers. By adding first the outer ones, and then the inner,
nested, ones, they ensured that the polymerase chain reaction ‘amplified’
just the fragments they wanted. The breakthrough also depended upon ‘being
absolutely rigorous about avoiding contamination’, says Monk. Success came
only after they physically segregated the preparation of the reactions from
the amplification and the analysis of the results.

Monk is now trying to adapt the test so that it could reliably diagnose
genetic disease in human embryos. Initially, she will attempt to amplify
a large fragment that encompasses most of the major mutations in the human
beta-haemoglobin gene. Such an analysis of the DNA would be the only way
to detect thalassaemia or the other disorders of haemoglobin in an embryo,
because no protein is produced from the gene until much later, when the
blood is formed in the fetus. Again, research on human embryos is crucial
before scientists and clinicians would attempt preimplantation diagnosis
on embryos destined to be transferred to a woman’s uterus.

Henry Leese of the University of Sheffield has been studying human and
animal embryos to try to find non-invasive ways of determining their quality.
In most IVF clinics, doctors give women various hormonal drugs to cause
them to ‘superovulate’, releasing far more eggs than the usual one per cycle.
The clinicians then transfer usually three fertilised eggs into the woman’s
uterus – but the problem is, which three should they select to increase
the chances that one will develop to term? And what sort of environment
should be provided for the egg, sperm and embryo while they are in the laboratory?
‘We have little idea what is the appropriate chemical or physical environment
for human embryos,’ says Leese. The oviduct, the tube through which the
embryo travels on its way to the uterus, is ‘like a coral reef, covered
with cilia beating like the tentacles of sea anemones, which waft the embryo
and nutrients,’ says Leese. The oviduct also produces glycoproteins that
differ along its length; some bind to the zona pellucida but no one has
any idea what they are doing.

Only experiments that in the end use human embryos can answer these
questions, and lead to more effective treatments for infertility, and the
option of preimplantation diagnosis for couples who are carriers of serious
genetic disorders. Bob Edwards, who pioneered IVF with Patrick Steptoe and
Jean Purdy, says they met with much suspicion when they began their work
in the 1960s. When in the 1970s they developed an aspirator to retrieve
eggs from an ovary ‘a journal refused to publish the paper on ethical grounds’,
Edwards said. The birth of Louise Brown, the world’s first test-tube baby,
in 1978, did much to win acceptance for the technique, but it remains an
expensive and often unsuccessful technique, confined, on the whole, to infertile
couples who can afford the substantial fees. Our overwhelming ignorance
of the biological basis of human reproduction is largely to blame.

This article is based in part on the Patrick Steptoe Memorial Conference
held at Churchill College, Cambridge on 21-23 September 1989.