What makes a mammal a mammal? In grade school, we were taught that all mammals have three distinguishing characteristics: fur, milk and live birth. But there is a problem: not all mammals have all three of these features. Monotremes (the group that includes the platypus and the spiny echidna) have fur and milk, but they do not give birth to live young. They lay eggs!
So, there seems to be a sort of gray area between mammals and non-mammals. Is there something wrong with our definition of mammals, or do we have a deeper problem? Perhaps there are other gray areas that we need to worry about.
Another good example of a gray area is the group of organisms called vertebrates. Vertebrates have a “back bone” – the series of bones (or cartilages, in some vertebrates, such as sharks) called vertebrae. They also have a segmented body plan, a series of slits in the region of the throat called the pharynx (fish have turned these into the famous “gill slits”), and a whole suite of other features. But there exists a small number of organisms that have the same suite of features, without the vertebrae.
In this case, our naming system seems to do a better job of handling the problem. Vertebrates are part of a larger group of organisms called chordates. Chordates have most of the vertebrate features just listed, but they do not have vertebrae. Instead, they have a springy, flexible rod along their backs, called the notochord. Examples of non-vertebrate chordates include the hagfish, the sea-squirt (which looks kind of like a soft hollow bulb of flesh attached to a rock) and a sort of brainless, fish-shaped thing called a lancelet.
We know that vertebrates are inside the chordate group because vertebrates start life with a notochord – although it is replaced by vertebrae during their embryological development. Clearly, the ancestor of all vertebrates was a chordate that had evolved an additional series of bones or cartilages in its back. Notice, however, that non-vertebrate chordates are in fact classed with the invertebrates, even though they are more closely related to the vertebrates. There is still something wrong with our naming scheme!
So we have at least two examples of very large, important groups of organisms – mammals and vertebrates – that have these “exceptional” cases. In fact, there are many more examples.
What we think of as discrete groups are actually parts of a continuum. As I discussed in my last post, large suites of characteristics are slowly assembled in evolutionary time: first one character, then the next, until we have the whole suite. Our problem is that often the intermediate forms in the spectrum die off or become rare, so we humans form the impression of discreteness. Our pattern-seeking minds give these groups more independent substance than they really have, and our naming system often follows these impressions. We end up with artificially constructed groups of organisms, where evolution actually tells a different story.
Now, all of this was a bigger problem in the middle of the twentieth century, before the invention of a new way of naming groups, called cladistsics. By focusing on ancestral relationships rather than large, constant suites of characteristics, cladistsics aims to restore some evolutionary realism to our scheme of nomenclature. Many of the sorts of problems that we are discussing here are in fact already resolved by a cladistic approach to grouping and naming species. I’ll have a lot more to say about cladistics in future posts.
Still, we are left with a huge new question: If all life evolves piecemeal, why do we have large groups with robust suites of characters? Why do some groups seem so different from others, and why are these differences so persistent throughout a whole group? Why do mammals seem so mammaly? Is it just a mater of intermediate forms becoming extinct, or is there something deeper at work?
Two recent articles from the journal Molecular Biology and Evolution begin to address these issues. One deals with major changes that occurred in the evolution of the mammalian genome. The second describes changes to the epigenetics of the vertebrates.
In the first paper (Holloway, Bruneaau, Sukonnik, Rubenstein & Pollard, 2016), the authors performed searches for important characters in the whole genomes of five placental mammals (humans, mice, cows, dogs and elephants), two marsupials (opossum and wallaby) one monotreme (platypus) and eight non-mammalian vertebrates for comparison (two birds, a lizard, one frog and four species of fish). This is a really well-chosen set of species, by the way. Their goal was to find genes or stretches of DNA that showed evidence of rapid evolution in the ancestor of therian mammals (the ones that have live birth), but which then became very conserved – largely unchanged since the establishment of the mammalian lineage. They called these stretches of DNA “TSARS”, which stands for “Therian Specific Accelerated Regions”.
They had two nifty findings: the first, which they emphasize most, is that TSARS tended to be associated with regulation of other genes. Some coded for proteins that activate other genes, and some represented regions of DNA that regulate nearby genes. However, very few coded for enzymes that have a more direct effect on the body – that is, very few enzymes e.g. for breaking down sugars or binding to hormones.
The second nifty finding was simply that there were so many TSARS! They found 4,797 regions of DNA that matched their criteria. This is an impressive suite of genes that unites all therian mammals, and seems to provide defining characteristics for them. An analysis of the function of many of these genes showed that they are involved in the development of mammal-specific characteristics, such as the uterus or our acute visual system. At a first glance, it is tempting to say that something special happened in the evolution of the first therian mammal that really sets this group apart from all others – and particularly excludes monotremes. Maybe our conjecture of gray areas and arbitrary groupings isn’t so hot after all?
Similar results were found in the second article that I want to discuss (Keller, Han & Yi, 2016), although the subject matter was quite different. In this case, the authors did not look at the genes themselves. Instead they looked at the epigenetics of the whole genome. Epigenetics is the study of modifications of the genome, especially by methylation. Methylation is a process of adding a small molecular tag directly to some of the nulceotides in our DNA. Usually, methylation of a gene results in the gene being turned off. Methylation is an important mechanism for transforming one kind of cell into another – a process called differentiation. When stem cells turn themselves into neurons or muscle cells, or other specific kinds cells, we say they “differentiate”, and they do it by methylating various genes. They turn off the genes that are not needed in the new cell-type.
This study looked at methylation events in three vertebrates (humans, chickens and a fish), two non-vertebrate chordates, and a sea urchin, which is a non-chordate invertebrate. The selection of species was more limited that of our first article because whole genome-methylation maps are harder to come by in the literature, but the results, again, are striking. Vertebrates have an intrinsically different pattern of methylation compared to invertebrates. The pattern is complex, so the following is a gross simplification. Invertebrates tend not to methylate many stretches of DNA, but they do methylate both regions that code for proteins, and regions that promote the expression of nearby genes (promoters). By contrast, vertebrates seem to have more methylation over-all, but they tend to methylate genes more often than promoters. Suffice it to say that vertebrates regulate gene expression between cell types very differently than invertebrates do.
Once again, we seem to have important characteristics that define a clear difference between two large important groups of species. But I think that in both cases, we would be rash to jump to the conclusion that vertebrates and mammals are each defined by some special evolutionary event.
In the case of the vertebrates, the evidence is that the patterns of methylation are sort of like our patterns of anatomical or physiological differences: they get built up slowly over evolutionary time. Something unites the vertebrates, true, but the in-between groups (non-vertebrate chordates) seem to have a pattern somewhat intermediate between vertebrates and other invertebrates.
In the case of the mammals, there seems to be a clear difference across the boundary between therian mammals and everything else – but this is the only boundary that they looked at! Note that this is not a shortcoming of their study. They looked at therians because they were looking for genes that make therians different, not evidence that therians had some special kind of evolution. But the fact remains that if they if they looked at the ancestor of some other group (say, the group of all mammals including monotremes, or the group of all birds), they might find genes there too that have evolved rapidly and then became conserved. Or maybe not! Who knows until the study has been done?.
Whatever we find, it is remarkable that we can distinguish large coherent groups of organisms. The gaps that separate them may be caused by special evolutionary events, or by the extinction of intermediate forms, but either way we have some fascinating questions to play with.
Cited Articles (Thanks!)
- Holloway, Bruneau, Sukonnik, Rubenstein and Pollard, 2016. Accelerated Evolution of Enhancer Hotspots in the Mammal Ancestor. Mol. Biol. Evol. 33(4):1008-1018
- Keller, Han and Yi, 2016. Evolutionary Transition of Promoter and gene Body DNA Methylation across Invertebrate-vertebrate Boundary. Mol. Biol. Evol. 33(4):1019-1028