The history of gene duplication Phylogenies are not just useful for studying morphological traits and geography but they also are essential tools for making sense of the evolutionary history of genomes. As already discussed, trees can be used to infer ancestral gene or protein sequences. While beyond the scope of this primer, statistical analyses of gene sequence evolution along the branches of a tree can provide evidence that selection has acted to shape molecular variation. Here, however, we will discuss gene duplication. When biologists began sequencing genomes they were surprised to find that many genes have closely related genes within the very same genome. We now understand that during evolution genes often duplicate – an ancestral genome with one copy gives rise to a descendant genome with two copies of a particular gene. Over time, repeated duplications can result in gene families: sets of related genes that have similar, but often somewhat diverged, functions. It is not necessary to go into the molecular mechanisms of duplication, nor to discuss the fascinating natural history of gene copies and their long-term evolutionary fate. But it will be useful to consider the way that gene duplication shapes gene trees and, correspondingly, how phylogenetic analysis of gene families can shed light on the history of gene duplication. There are parallels between the biogeography and gene duplication. In the same way that a species within a particular landmass can undergo lineage splitting to yield two daughter species, so to can a gene within a species’ genome give rise to two descendant genes. Furthermore, the splitting of geographic areas during vicariance affects the species living in those areas in much the same way that the splitting of population lineages affects the gene copies that “occupy” those populations. We will address these two concepts in turn. First, we will consider the duplication of genes within a single lineage. Then we will discuss how lineage splits interact with gene duplication to shape the topology of gene trees. Imagine a gene, A, that exists as a single copy in the genome of all organisms in an ancestral population. Through an error in DNA replication, or the action of a transposable element, or some other molecular mechanism, a second exact copy of the original gene is generated somewhere else in the genome. Since it is an exact copy it is not fruitful to worry about which is the original gene and which is the copy. Let’s call the duplicate genes A1 and A2. Gene duplication isa lineage branching event in that we have gone from one ancestral gene to two descendant genes. As with lineage splitting, following gene duplication the two gene copies will accumulate mutations independently and will gradually diverge in sequence. The duplication will only persist in the long run if it first arises in an individual that leaves offspring and if eventually it comes to be fixed in the population lineage. Imagine that the after being fixed, gene A2 undergoes yet another gene duplication to give rise to genes A2a and A2b. Once this second duplication goes to fixation, what will be the relationship among the three genes? Since A2a and A2b share a more recent common ancestor (A2) than either does with A1 (A), the correct tree is the one shown in figure x.If we correctly inferred this (rooted) gene tree we would immediately see that genes A2a and A2b represent a more recent gene duplication, whereas A1 vs. A2 was a more ancient gene duplication. Now we have the opportunity to test your tree-thinking skills: what would the gene tree look like if, after these duplication events had happened, the lineage split to give rise to two living species, X and Y? Species X and Y would each have three gene copies, A1, A2a, and A2b, meaning there would be six tips. But how would they be related to each other? One way to think through this problem is to first draw the population lineages as though they were hollow tubes. Then you can draw the gene tree inside these tubes making sure that all gene copies present in an ancestral population make it into the two species lineages. Then use tree thinking skills to “unfold” the gene tree, labeling the genes based in which species they came from. As shown in the figure, three nodes, marked with a circle, correspond to a lineage-splitting event (X versus Y), whereas two nodes, marked with squares, correspond to the two gene-duplication events. Before discussing alternative possible histories of gene duplication and lineage splitting, we should clarify some widely-used terminology applied to genes. Because all these genes descended relatively recently from a common ancestral gene they are all homologous genes or homologs. Pairs of genes that occur in different species whose last common ancestor corresponds to a lineage-splitting event are orthologous genes or orthologs. For example, XA1 and YA1 are orthologs, because they both descend from a node that corresponds to the split of the X and Y lineage. In contrast, pairs of genes (in the same or different organisms) that descend from a gene duplication event are paralogous genes or paralogs. For example XA1 is paralogous to XA2a because the last common ancestor of these two genes was the root node, which corresponds to the A1-A2 gene duplication event. A1 A2a A2b A1 A2a A1 A2a A2b A2b XA1 YA1 XA2a YA2a XA2b YA2bThe concepts of orthology and paralogy relate to the process that caused the existence of distinct gene lineages: population splitting or gene duplication. It does not directly relate to the role that a gene plays in the development of an organism – its function. When looking between species, orthologs have a more recent common ancestor than paralogs. For example, XA1 is more closely related to YA1 than to YA2a or YA2b. Because gene functions generally change slowly, it is more likely that orthologs share functions than paralogs. However, this is not a rule. Supposing that YA1 acquired a novel function, while the other genes retained an ancestral function, then XA1 could be functionally more similar to YA2a than to YA1. This is another manifestation of the principle that trees depict relationships not similarity (Chaps 3-4). The preceding scenario explained the occurrence of three gene copies in species X and Y via two gene duplication events that predated all lineage splitting events. Now consider a different scenario where