This series of posts is intended as a basic introduction to the science of evolution for non-specialists. You can see the introduction to this series here. In this post we discuss the origin and diversification of mammals, including the genomic evidence for a transition away from yolk-filled eggs to embryonic nourishment through the placenta.
As you will recall from the last post in this series, crown-group tetrapods are a highly diverse group, ranging from amphibians, to reptiles, to birds, and mammals. Since the latter group includes our own species, it would be remiss to not discuss at least a few of the features that distinguish mammals from other tetrapods, and their evolutionary history.
Stem-group mammals: the synapsids
Mammals are the only living representatives of a group called synapsids, a group that parted ways with the dinosaur/avian lineage (the sauropsids) and went on to diversify beginning in the late Carboniferous period, around 325 million years ago. The numerous synapsids known from the fossil record are stem-group mammals – organisms related to mammals that branch off from the lineage leading to crown-group mammals (the last common ancestral population for all living mammals, and all of that population’s descendant species). The fact that some of these stem-group, extinct synapsids were once known as “mammal-like reptiles” and others as “reptile-like mammals” reveals the transitional nature of their features – they blur the distinction between “reptiles” and a “mammals” in the way we are now familiar with: through the gradual accumulation of traits characteristic of crown-group mammals, and in a branching pattern that indicates the order in which those characteristics were acquired. Examples of such acquired traits include jaw morphology (including the co-option and repurposing – i.e. exaptation – of jaw bones for a specialized hearing function in the inner ear), the development of hair, and lactation (the secretion of milk for feeding young).
From egg to placenta
Sometimes I encounter non-biologists (and even some biologists) who are surprised to learn that “live birth” is not a defining characteristic of mammals (or, more precisely, ofcrown-group mammals). The reason for this is because a lineage of egg-laying mammals, the monotremes, has living representatives. The fact that egg-laying mammals exist in the present day means that they are part of the crown group (by definition), and as such any characteristic that they lack cannot be a defining feature of the crown group:
Crown-group mammals include egg-laying mammals (monotremes) as well as non-egg-laying mammals (marsupials and placental mammals). Of the features shown on this phylogeny, only lactation is a characteristic common to the entire crown group.
In egg-laying mammals, such as the platypus and the various species of echidna, young hatch from the egg and then are nourished by their mother’s milk, which is secreted from a patch on the skin, and lapped by the young. In marsupials, pregnancy is much shorter than in placental mammals, and after birth the (still very much embryonic) young crawl to a protected pouch where they nurse at a teat in order to complete their development. In marsupials a brief connection is formed in utero between the embryo and the mother through the yolk sac membrane to form a “yolk sac” placenta. Yes, it is somewhat confusing that both marsupials and “placental” mammals – i.e. eutherians – both have a placenta. The difference is that placental mammals form their placenta from a different membrane – the chorioallantoic membrane.
While monotremes and marsupials are not stem-group mammals (since their lineages persist to the present day) we can appreciate their features in exactly the same way that we have done for stem groups. (Put another way, if the monotreme and marsupial lineages had in fact gone extinct, we would call eutherians “mammals” and monotremes and marsupials would be stem groups on the eutherian lineage.) Similarly to what we have seen with stem groups, monotremes and marsupials demonstrate that the eutherian state was arrived at over time, and through a series of gradual steps. Though in reality this process was a gradient, we can arbitrarily denote some “stages” along the way:
- Egg laying with after-hatching lactation (monotreme state)
- Short gestation with a yolk-sac placenta with post-birth lactation (marsupial state)
- Gestation with a combination of yolk-sac and chorioallantoic placentas (that extended gestation time) with post-birth lactation
- Reduction of the yolk-sac placenta in favor of the chorioallantoic placenta (with further extended gestation time) with post-birth lactation
- Long gestation with a chorioallantoic placenta with post-birth lactation (the eutherian condition).
Along the way, metatherians and eutherians would shift away from yolk-based nutrition for their embryos in utero, and shift towards nutrition based on their placentas. In doing so, the biochemical machinery for yolk-based nutrition would be predicted to become less and less important – and eventually, useless altogether. At the genetic level, genes required for yolk production would eventually no longer contribute to the survival or reproduction of the organism – meaning that they were no longer under selection, and now free to accumulate mutations without consequence to the organism.
Seeking the dead among the living
In practical terms, when a gene is no longer under natural selection, it is then maintained only by the overall precision of DNA replication during cell division. While quite accurate, DNA replication is not perfect. For sequences under natural selection, mutations are removed from the population if the individuals carrying those variants cannot reproduce at the same frequency as their non-mutated relatives. For genes no longer subject to natural selection, mutations will accumulate slowly over time.
For marsupial and placental mammals, one such gene, named vitellogenin, is one that would be expected to be released from selection after the establishment of a placenta. Vitellogenin is vitally important to the formation of egg yolk, since it acts as a major carrier of nutrients from the liver to the forming egg yolk in egg-laying organisms. In 2008, a research group went looking for the remains of vitellogenin sequences in placental, marsupial, and monotreme mammals. Monotremes, as you would expect, have a functional vitellogenin gene sequence, since they are egg-laying mammals. While marsupials and placental mammals do not have functional vitellogenin gene sequences, they do have the (heavily) mutated remains of vitellogenin sequences, indicating that these lineages once did have functional biochemical machinery to transfer nutrients in bulk to egg yolk. This observation makes perfect sense in light of a phylogenetic prediction that placentals and marsupials share a common ancestral population with monotremes (and of course other tetrapods) – with egg-laying as the ancestral state that was subsequently lost in the marsupial - placental common ancestral population:
So, genome sequencing allows us to test specific predictions about what we should find (based on phylogenies assembled using anatomical and morphological features). In this case, the presence of vitellogenin sequences in placental mammals (including the human genome) that cannot function to make egg yolk is a striking example of a confirmed evolutionary prediction (and one that continues to be highly problematic for antievolutionary groups). In this context, however, the loss of vitellogenin was but one small, anticlimactic step along the way to the metatherian and eutherian lineages, which, in contrast to the monotremes, remain highly successful to this day.
In the next post in this series, we’ll explore the diversification of placental mammals, including the lineage leading to our own species: the primates.