Australasian Science: Australia's authority on science since 1938

The Mouse Is Not Enough

The invasive nature of embryo retrieval has necessitated the use of a mammalian

The invasive nature of embryo retrieval has necessitated the use of a mammalian species that reproduces rapidly and is inexpensive to house – the mouse.

By Peter Pfeffer and Debra Berg

Fundamental differences in embryonic development mean that research using mice may not be reliably applied to other mammals, and that cattle embryos may be a better model for stem cell studies in humans.

New research has found that mice may not be the best model system for understanding early mammalian embryogenesis. Our study, published in Developmental Cell in February, identifies important differences in the timing of cell fate commitment during the development of mouse and cattle embryos.

It turns out that cattle embryos are a better model for understanding the earliest events in human development. This finding impacts on stem cell biology, explaining why the isolation of embryonic stem cells in humans and other mammals using methods based on mouse biology have failed, and highlights the need to broaden studies to cover more than a single model species.

Embryology is the study of how an animal develops from a single fertilised cell to an incredibly organised and complex multicellular entity. After fertilisation, mammalian embryos undergo several rounds of cell division without growth, forming a solid ball of cells. A cavity develops in the ball leading to a spherical layer of cells enclosing an inner clump or mass of cells attached to one side.

This stage is called the blastocyst, with the inner cell mass giving rise to the entire embryo. The outer layer of cells is called the trophectoderm (TE). This contributes to the placenta, which nourishes the embryo in the uterus. Thus a key event has occurred by the blastocyst stage: cells have decided whether they are part of the embryo or part of the placenta.

Two master genes are involved in this lineage decision in mice. The Oct4 gene controls the inner cell mass (embryo) lineage whereas Cdx2 determines the TE (placental) line. Oct4 and Cdx2 regulate hundreds of other genes required for the development of the respective lineages.

Importantly, Oct4 and Cdx2 turn each other off in a negative feedback loop. Thus, while initially both genes are active all over the embryo, by the blastocyst stage Oct4 is exclusively seen in the inner cell mass while Cdx2 is turned on only in the TE.

However, the functioning of the Oct4/Cdx2 regulatory circuit has only ever been observed in the mouse. Why not in other mammals?

Mammalian embryos are particularly difficult to study as they develop in the safe but inaccessible seclusion of the mother’s uterus. The invasive nature of embryo retrieval has necessitated the use of a mammalian species that reproduces rapidly and is inexpensive to house – the mouse.

Even more important was the need to be able to manipulate the genes that drive embryogenesis. Two breakthroughs a couple of decades ago made this possible in mice. One was the isolation of mouse embryonic stem cells; the other was pronuclear injections, whereby DNA can be directly introduced into mouse embryos at the one-cell stage.

Despite intense efforts, embryonic stem cells could not be isolated in any other mammal until very recently, and the pronuclear injection technique has been very inefficient in mammals other than mice. Thus mice not only became the mammalian embryological model system of choice but the only available system for functional studies.

The dependence on a single mammalian model system has made it difficult to judge whether what is seen during mouse embryogenesis is applicable to all mammals, including humans, or is a peculiarity specific to mice. Luckily, in many mammalian species, eggs can be artificially fertilised and grown in culture until the blastocyst stage. Furthermore, advances in DNA sequencing has enabled the actions of genes to be catalogued in a diverse range of mammals. Thus gene activities can be readily mapped up to the blastocyst stage.

This led to the observation that, unlike mice, Oct4 remains on in all cells of the blastocyst in cattle, pigs, rabbits and humans. However, why this should be so and what the difference means was a mystery that could only be solved by developing in one of these mammals a molecular and embryological toolkit equivalent to that used in mice.

We decided to test this using cattle embryo, and were in an ideal position to attempt this. The dairy industry is essential to New Zealand’s economy. To produce milk, cows need to calve every year so issues of fertility and embryo health take centre stage. Located in the middle of the best dairy pasture lands, the Ruakura research campus in Hamilton has a long tradition of research into cattle reproduction and in vitro embryo production. Secondly, one of our collaborators, David Wells, is a world expert in cattle cloning. Hence all we had to do to genetically manipulate cattle embryos was to stably introduce DNA into cattle cells that we could then use to make embryos via cloning.

At this time a major problem arose when objections from local activists delayed an environmental risk approval process to allow genetic modification of cattle. In desperation, we submitted a new proposal and built a small dedicated shed to house the cows receiving genetically modified embryos and bypass the stalled negotiations for their outdoor containment. After 6 months the lights turned green and we were set to go.

Much like electronic circuits, gene activity is controlled by regulatory regions on the same strand of DNA where the gene is found. The factors operating the switches are proteins called transcription factors.

We started with the knowledge that the Oct4 gene is switched off in the mouse TE whereas it remains on in cattle TE. Either of two scenarios could explain this:

• the transcription factors that switch Oct4 off in the TE of the mouse blastocyst may not be present in cattle; or

• the regulatory switches of Oct4 differ between mice and cattle.

To determine the answer we introduced the mouse Oct4 regulatory switch into cattle embryos. The result? Scenario 1 was true: Oct4 remained on in the TE (Fig. 1). We found that this inability of cattle embryos to turn off Oct4 is due to changes in the transcription factor Cdx2.

We then addressed scenario two by introducing the cattle Oct4 regulatory switch into mouse embryos. Unexpectedly, the mouse could not turn off the cattle Oct4 switch (Fig. 1) even though it contains the transcription factors necessary to turn off the mouse Oct4 switch. This means that scenario two applies as well: the cattle and mouse Oct4 regulatory switches differ.

In a series of experiments we pinpointed the critical differences within the switches to a set of binding sites for AP2-type transcription factors. Only mice contain AP2 binding sites and turn off Oct4, whereas cattle, pigs, rabbits and humans do not contain these sites and do not turn Oct4 off.

It follows that mice are the odd one out, having decided to rewire the molecular circuitry controlling one of the most important genes for embryology.

We speculate that there is a very good reason for this mouse invention. Unlike most other mammals, mouse embryos implant in the uterus immediately after reaching the blastocyst stage. The TE cells of the blastocyst embryo are responsible for implantation, but before fulfilling this function they need to firmly establish their molecular identity. How? By expressing Cdx2 and switching off Oct4.

Cattle represent the other extreme. Cattle embryos laze around in the uterus for 2 weeks after reaching the blastocyst stage before finally attaching to the uterus. (During this time the inner cell mass keeps developing to set up the blueprint for the fully formed three layered embryo.) Hence the cattle TE cells need not be in any particular hurry to establish their unequivocal “placental” identity. In molecular terms this means that there may be no need to rapidly and actively shut down Oct4 in the TE.

This line of thinking would predict that mouse TE cells are “committed” earlier than cattle TE cells. In mice, TE cells are committed to the placental lineage by the blastocyst stage. We explored the commitment of cattle TE cells by placing blastocyst-stage TE cells inside the two halves of uncommitted eight-cell stage embryos and recording whether, after further embryo growth, they would contribute only to placental tissue (implying they were committed to the TE fate) or whether they could also contribute to the inner cell mass and its lineage (Fig. 2). It turned out that cattle TE cells were not committed at the blastocyst stage, confirming our prediction and explaining why the mouse may have had to rewire the circuitry controlling Oct4.

The implications of this research reach beyond embryology. In stem cell biology, Oct4 is pivotal. This gene has to remain switched on for embryonic stem cells to maintain their pluripotent state. Such a state allows embryonic stem cells to develop into any adult cell, thus opening up a myriad of therapeutic applications.

Yet obtaining embryonic stem cells has been notoriously difficult in all mammals apart from the mouse. The new insights gleamed from cattle embryos are beginning to explain why – culture conditions suited to maintaining Oct4 in mouse embryonic stem cells will not be applicable to the embryonic stem cells of other mammals as the regulatory circuitry required to switch Oct4 on and off has changed.

Our research has shown that all mammals are not created equal. Even such a fundamental issue as the very first embryonic decision is varied.

Does this mean that cattle embryos have muscled mice out of the ring, being the better contender for representing the archetypical mammal? Definitely not. Instead this work has highlighted that it is important not to rely on only one representative of the furry class of animals.

Peter Pfeffer and Debra Berg are Senior Scientists at Agresearch in Hamilton, New Zealand.