One of the “holy grail”
goals of reproductive biology research has just been reported by
Dr Tomohiro Kono and coworkers at Tokyo University. A live mouse
was produced from the parthenogenetic development of an
unfertilized egg without any contribution from a paternal genome.
Parthenogenesis is not uncommon in non-mammalian species. However,
current knowledge indicates that in mammals, the presence of
differences in gene expression between the two parental genomes
restricts normal term development to embryonic genomes that derive
from both parents. This latest result actually provides strong
support for this theory as well, since artificially manipulating
the expression of two key imprinted genes was required to allow
term development of the parthenotes and the efficiency of this
process was greatly reduced. Nevertheless, this stunning result
not only provides a viable methodology for producing uni-parental
mammals it also reveals surprising information about the
plasticity of gene regulation during development.
The main problem
with the development of uni-parental mammals such as
parthenogenetic mice is the presence of uni-parentally expressed
epigenetically imprinted genes. During mammalian evolution, some
genes have become restricted in their expression to chromosomes
that derive from either the male or female germline. This is
thought to have been brought about due to competing evolutionary
interests on the part of male and female individuals in
controlling gestation. The result is that a functional genome can
only be created by a combination of maternal and paternal
chromosomes – which is, of course, the normally occurring scenario
during sexual reproduction. Experimentally derived embryos
harboring either all maternal or all paternal genomes exhibit a
variety of developmental abnormalities and do not survive to term.
In theory, if one could alter the requisite gene expression of one
set of chromosomes to epigenetically complement the other set then
normal development could take place. The current parthenogenetic
result was obtained by creating a unique maternal genome with
forced “paternal” expression at two key imprinted genes. This was
combined with a second normal maternal genome to create an
embryonic genome that was capable of supporting term development.
Prior work by the
Kono lab had indicated that the genome of immature “non-growing”
oocytes (obtained from new-born mice) could support longer and
more appropriate parthenogenetic development when combined with a
normal mature oocyte genome. This was thought to result from the
fact that the genomes of such non-growing oocytes are naïve to
some of the imprinting-related changes that occur later in
development. Therefore, gene expression patterns in these immature
eggs might be more similar to paternally-derived genomes than to
the normal maternally-imprinted genome present in mature oocytes.
A second manipulation involved creating an oocyte genome in which
a known imprinted “maternal-only” gene was deleted. The H19 gene
is normally expressed only on maternally-derived chromosomes. This
is due to the absence of methylation at a regulatory site upstream
of H19 on chromosomes from the maternal germline. Embryos with
entirely maternal chromosomes would exhibit abnormally increased
expression of H19 from both active chromosomes. Deleting the H19
locus mimics the situation in a paternally-derived chromosome
where H19 expression is lost due to methylation of this regulatory
site. By combining a non-growing oocyte genome with an H19-deleted
“paternal-like” chromosome with a normal oocyte genome, single
allele expression of H19 was achieved leading to improved
parthenogenetic development (up to day 17.5 – almost to term).
In the final
experiment, just reported in the journal Nature, another genetic
manipulation was added to this system which altered the expression
of a second imprinted gene. Conveniently, this actually involved a
minor change to the H19 deletion. The regulatory site upstream of
H19 is in fact involved with the regulation of two differentially
imprinted genes, H19 and the nearby IgF2 gene which encodes for
insulin-like growth factor 2. As stated above, on paternal
chromosomes, the regulatory site is methylated leading to
suppression of H19. Suppression results because a required
enhancer protein cannot bind to the methylated regulatory site.
However, when the enhancer protein is present (on maternal un-methylated
chromosomes), it blocks other regulatory proteins from binding
near IgF2 and promoting its expression. Thus, this site
simultaneously regulates (in reciprocal fashion) the imprinted
expression of both H19 and IgF2. On paternal chromosomes, H19 is
“off” and Igf2 is turned “on” while on maternal chromosomes, the
enhancer binds, H19 is activated but IgF2 expression is blocked.
This type of reciprocal control has been observed at other
imprinted loci. By ingeniously deleting the entire H19 region
including the enhancer binding site, a new chromosome was created
with a “forced paternal” expression pattern at both genes – H19
was simply absent and enhancer site deletion relieved the
restriction on IgF2 expression.
Using nuclear
transfer, Kono and colleagues created almost 600 hybrid oocytes
with genomes derived from a combination of non-growing eggs
harboring the double H19/Igf2 modification and normal mature
oocytes. Upon activation, the resulting diploid genome directed
development leading to the production of 371 morulae/blastocysts
which were transferred to surrogate mothers. 28 pups were isolated
just prior to parturition on day 19.5 of development and to the
great surprise of the researchers, 10 of these were living and 2
were recovered as viable, overtly normal pups. Genetic analysis
confirmed the parthenogenetic origin of these mice. One of these,
a female named “Kaguya”, exhibited good post natal development,
reached adulthood, and has even successfully produced a first
litter of normal pups.
As stated, this
result was, in fact, surprising to the researchers who most likely
expected this second genetic modification to have a more
incremental effect on development. However, the most surprising
finding came from microarray analysis of gene expression in day
12.5 parthenogenetic embryos. Embryos resulting from the H19/IgF2
double modification exhibited differential expression of a wide
range of genes compared with standard parthenogenetic embryos. In
fact, the H19/IgF2 embryos displayed an almost normalized gene
expression pattern including normal expression at every imprinted
gene analyzed. Since H19 or Igf2 expression alone would not be
expected to directly affect imprinted expression of other genes
this was a stunning result. It is currently unclear how this
alteration in H19 and IgF2 expression could somehow essentially
normalize gene expression in what were still expected to be
epigenetically very abnormal embryos. The authors theorize that
providing correct expression of these key developmental genes
created a more normal gestational state leading to correct
signaling and downstream regulatory behavior resulting in the
normalized gene expression and, in some cases, normal development.
Some genes did exhibit abnormal expression patterns and the
majority of pups exhibited growth retardation and developmental
abnormalities so the true nature of the relationship between gene
expression and development in this system remains to be
determined. However, this result indicates that epigenetic control
of gene expression during development is perhaps even more plastic
than was previously thought.
What does this
result mean for reproductive biology and assisted reproduction?
For one thing, the technique will allow for a tremendous amount of
powerful research into the precise nature of epigenetic regulation
during development. This will hopefully fill many gaps in our
basic understanding of how gene expression drives development and
how perturbations to expression result in developmental deficits
and abnormalities – an important issue to ART manipulations. This
result already provides evidence of considerable plasticity in
developmental gene expression. This would suggest that pessimistic
predictions about epigenetic gene expression problems negating any
chance of success with nuclear transfer protocols (such as similar
pessimistic predictions about the impossibility of a mammalian
birth from parthenogenesis) may be poorly founded. This result
provides hope that similar manipulations could be used to address
the epigenetic aberrations observed following other nuclear
transfer “cloning” scenarios.
It is highly
unlikely that the actual techniques involved with the current
mouse research could be transferred to the human. This result
required genetic engineering to produce the knockout mouse line
used for the H19/IgF2 donor eggs. Obtaining a similarly modified
human genome would not be possible with conceivable clinical
techniques. However, the basic principle of creating a modified
genome with a “reversed” expression pattern at key imprinted genes
might be achieved by other means. The desired genomic
manipulations could be introduced in human embryonic stem cells
followed by differentiation into a functional oocyte harboring a
“paternal” genome pattern. Also, combining an oocyte with a source
of cytoplasm that would bring about a re-setting or reversal of
existing imprints might be attempted. Such techniques could
eventually allow for same sex partners to conceive a child with
their combined genomic contribution.
It remains to be
seen if such techniques will ever be developed and applied.
However, the birth of Kaguya is without question a result of great
significance to the advancement of basic research and development
in mammalian reproduction and to those who wish to apply this R&D
to the improvement of human assisted reproduction treatment. We
salute Dr Kono and his team.
Further
Reading:
Kono T, Obata Y,
Wu Q, Niwa K, Ono Y, Yamamoto Y, Park ES, Seo J, Ogawa H (2004)
Birth of parthenogenetic mice that can develop to adulthood.
Nature 428:860-864.
Kono T, Sotomaru Y, Katsuzawa Y, Dandolo L (2002) Mouse
parthenogenetic embryos with monoallelic H19 expression can
develop to day 17.5 of gestation. Dev Biol 243:294-300.
Kono T, Obata Y,
Yoshimzu T, Nakahara T, Carroll J (1996) Epigenetic modifications
during oocyte growth correlates with extended parthenogenetic
development in the mouse. Nature Genet 13:91-94.
Hubner K,
Fuhrmann G, Christenson LK, Kehler J, Reinbold R, De La Fuente R,
Wood J, Strauss JF 3rd, Boiani M, Sholer HR (2003)
Derivation of oocytes from mouse embryonic stem cells. Science
300:1251-1256.
