How to Build a Marsupial

A. L. Keyte, K. K. Smith. Developmental origins of precocial forelimbs in marsupial neonates. Development, 2010; 137 (24): 4283

Marsupials—kangaroos, koalas, wombats, and their relatives—are an exotic group of mammals. This is partially because, unless you live in Australia, you’re unlikely to see many outside of a zoo.  But marsupials are also unique because of their mode of development.  Most mammals are eutherians, or mammals with placentas. The placenta is an organ that connects a fetus to the uterus of its mother, allowing for the exchange of nutrients and waste. This means babies can develop for a longer period of time inside the mother than other animals (if you’re an elephant, this can be as long as 60 months).  Marsupials lack a placenta; so their offspring are born much earlier than eutherian mammals.  So without any help from the mother, the newborn has to migrate to the mother’s pouch, where it can suckle milk and continue development.

The reason that marsupial newborns can accomplish this is that they have unusually strong and well-developed forelimbs. Below are the images of a newborn mouse (a eutherian) on the left and a newborn opossum (one of the only non-Australian marsupials) on the right.  Even though the newborn opossum is at a much earlier stage of development, the arms are similar in proportion to the mouse:

In a recent issue of the journal Development, Keyte and Smith set out to find whether changes in the timing of genes might correlate with the early development of marsupial forelimbs.  They compared the development of the opossum Monodelphis domestica with the mouse Mus musculis, looking at a number of genes known to be important to limb development. This included the limb-patterning gene Sonic Hedgehog (shh), forelimb gene T-box 5 (Tbx5), hindlimb gene T-box 4 (Tbx4), and the Tbx downstream targets, fibroblast growth factor 10 (Fgf10) and 8 (Fgf8).  When Keyte and Smith synched up the two animals development, they discovered that almost all of these genes are turned on earlier in the opossum than in the mouse (in the figure below the term “FL” before the gene name implies expression of the gene in the forelimb, while “HL” means gene expression in the hindlimb):

In addition to changes in the timing of gene expression, there is also early movement of nerve cells and myocites (primitive muscle cells) into the limb bud.  Although it might seem like a lot of changes have to occur to create this advanced forlimb, the authors argue that most of the observed changes can be explained by a single event.  A change in the timing (or heterochrony) of Tbx5 expression would lead to early expression of Fgf10 since Tbx5 is known to regulate Fgf10. Fgf10 controls expression of Fgf8, so a mutation that turns Tbx5 on early will set the whole cascade of the gene network going.  The developing forelimb bud itself encourages the migration of nerves and myocites into the limb, creating a complex foreleg early in development.  Thus, a few small changes can have a large affect on the growth of the animal limb.

It might seem initially surprising that genes in the hindlimb also turn on earlier in opossum than in mouse, even though the opossum doesn’t also develop large, functional hindlimbs at birth. Since Tbx5 and Tbx4 are closely related genes, it is possible that the two are controlled by a similar upstream genes.  This would make it difficult (or perhaps impossible) to evolve early expression of forelimb genes without also patterning early development of the hindlimbs.  Such complications between forelimb and hindlimnb adaptation have been noted before.  In Stephen Jay Gould’s essay, “The Panda’s Thumb”, he notes how pandas have an elongated wrist bone (or radial sesamoid) to use as a false thumb to eat bamboo, but they have a similarly elongated bone (a tibial sesamoid) in their foot:

In a panda's foot, the counterpart of the radial sesamoid, called the tibial sesamoid, is also enlarged, although not so much as the radial sesamoid. Yet the tibial sesamoid supports no new digit, and its increased size confers no advantage, so far as we know…Repeated parts of the body are not fashioned by the action of individual genes—there is no gene "for" your thumb, another for your big toe, or a third for your pinky. Repeated parts are coordinated in development; selection for a change in one element causes a corresponding modification in others. It may be genetically more complex to enlarge a thumb and not to modify a big toe, than to increase both together.

Interestingly, it appears that the lack of hindlimb growth doesn’t come from a lack of gene expression.  Instead there simply aren’t enough cells in that region to provide the raw material necessary to build the hindlimb. The concept that there are tradeoffs regarding how many “traits” a body can build at once might deserve more consideration then it has gotten, but there are known examples.  For instance, some beetles that develop elaborate horns have reduced antennae, wings, or even eyes, because there simply isn’t enough cellular material and energy to build all of these things (e.g. Emlen and Costs 2001). Keyte and Smith note that the last gene in the hindlimb pathway Fgf8, is the only gene that does not turn on significantly earlier in the opossum than in the mouse.  This “pause” in the hindlimb gene network might be the result of a lack of cellular material, or be an additional cause of the normally developed hindlimb. 


Ultimately, like Osterauer et al. (see my post “How to Build A Snail”), this is a clear example of how a small change in development can have a large impact on an animal.  The only minor issue I had with it was its constant referral to how the development of the marsupial has changed so dramatically from the mouse.  But if you look at the evolutionary tree of mammals (I whipped one up below for you to look at), it seems most likely that placental mammals evolved from pouched marsupials, who evolved from egg-laying monotremes (and before that, reptiles).  So if anyone’s development has radically changed, it would be the mouse, and not the opossum.

Papers Cited

Emlen, D. J. (2001) Costs and the diversification of exaggerated animal structures. Science 291: 1534-1536.

Gould, S.J. (1980) The Panda's Thumb, New York: W. W. Norton, ISBN 0-393-01380-4

Tullio, A. N. Accili, D. Ferrans, V. J. Yu, Z. Takeda, K. Grinberg, A. Westphal, H. Preston, Y. A. Adelstein, R. S. (1997) Nonmuscle myosin II-B is required for normal development of the mouse heart. Proc Natl Acad Sci USA 94:12407–12412