 The evolution of the vertebrate eye, a molecular perspective. Let's start by talking about the living fossil, platyneris de maryli. It's a polychaete worm, same family as earthworms. It possesses a prototypical pigment cup eye. It might win a prize for the ugliest whatever. Even HR Geiger would be inspired by this guy. Here's the larval eye at 24 hours traced from EM micrographs. The yellow are raptomeric photoreceptor cells. The green are pigment cells. These two cells are the simplest possible visual system. This eye senses light direction, even color. It directly couples to other cells, no nerves at this stage. No lens, brain, nerves, and only a one cell retina. Here's the same eye at 72 hours in the proto adult. The development of structures has led to enhanced directionality. This eye can be used for locating. Nerves will begin to grow toward the raptomere because of its surface markers. No lens, single cell retina, a single optical nerve, no brain. Here's the adult pigment eye. It's really just a cluster of the larval eye parts created by repeated duplication. This eye has no lens, a multicellular retina, nerves, and a simple brain. This is the stereotypical progression in embryonic development in bilaterians. It also gives us an idea of the evolutionary progression from the ur bilaterians. All of these forms exist in animals extant today. In fact, here we see a cellular level view of the retinas of the Aschidian, hagfish, lamprey, and nastostoma. These closely related organisms show a great deal of diversity in eye structure, but all clearly evolve from two basic cell types, the raptomeric and ciliary photoreceptor cell. In vertebrates, this correlates to rod and cone cells respectively. With this close up, you can begin to see why cone and rod cells got their names. Each of these terminates in a simple synaptic terminal, which is where the retina meets a nerve ending. All sensory tissues have these types of terminals, suggesting that the complex eye could have evolved from a simple eye spot. It does not require the special growth of new neural pathways in development. Let's take a closer look at the photoreceptor cells. Why are they special? Well, they contain chemicals called opsoons. Ciliary photoreceptors contain a broad diversity of these, of which the best known is rhodopsin, derived from vitamin A. This is why your mother told you to eat your carrots, if you want to see well in the dark. The emergence in bilaterians, that's animals with bilateral symmetry, of these specialized light sensing opsoons provide a test for evolution. What can we learn about the eye from examining the data produced by a field called comparative genomics? If the theory of common descent is true, we should see a pattern of genetic similarity that matches the stages of development of eye complexity. I've used one of my favorite metagenomic tools, homology. It will search the genomic databases at the NCBI for homologous genes between organisms. It will calculate the variation in protein sequence between different organisms. If the modern theory of evolution is true, I would expect to see relatedness among all the animals tested. I would also expect to see that genes that are present in the simplest visual system would be highly conserved in the complex vertebrate eye. We should also be able to tell based on the accumulation of heritable changes which genes are most ancient and which are most recent. We should also be able to find plausible alternative functions for the ancestors of the visual genes. Part one, relatedness. If we start with comparing the chimpanzee visual system to that of humans, almost no differences can be detected. We need to zoom out a little to see any differences at all. Here is a comparison of several important visual system genes across a variety of mammals versus the human protein sequence. Blue is dog, red is cow, and yellow is mouse. Note that all three mammals have about the same proteins for eye formation, as we might expect. This data suggests that pen 2, IP, and opsin 4 are more recent in humans or have weaker selection than the truly essential genes like PACS 6, which is the homeobox gene that determines where eyes are formed. Genes such as rhodopsin, opsin 1, and transducin are all very highly conserved, and these are the earliest visual system genes. Part two, primitive visual genes are highly conserved. Here's transducin, the protein that transmits the signal from the opsins. Note that it is a very highly conserved between humans and fish, and that we also find a very similar protein sequence in a variety of plants. This gene must be a very early component of the signaling from opsin activation. Here's guanine nucleotide binding protein, GNBP, also a signaling component of the simplest visual system, and we find it in vertebrate rods and combs, as well as in chickens and fruit flies minimally changed. Note that genetic change corresponds to the evolutionary distance as based on morphology and fossil evidence. Part three, we should see which genes are ancient and which are most recent. We've already seen some ancient genes shared between plant leaves and vertebrate eyes. What genes are more recent? Here is Pax 6, which is incredibly conserved between vertebrates as far away as fish and chickens, but more distantly related in fruit flies. I would predict that Pax 6 had a predecessor in early bilaterians, but that it was modified after the vertebrates and invertebrates parted ways. This would make sense considering the differences in eye morphology between flies and humans. Part four, all genes should have a plausible explanation for alternative functions. Well, here's a list of genes and their paralog functions. Opsins, for example, have function in circadian rhythms in single celled organisms, and they're responsible for some blue-green phototaxis analogy. Transducens and GNBPs are both signal transduction cascade proteins. They interact with a number of different membrane pound receptors. Atonal Pax 6 and Fox A2 are all in the family of basic helix loop helix transcription factors, and they're all used in the formation of multiple sensory systems. Cryptochromes are found in all varieties of animals, but they are related to bacterial photoliase, which repair UV-induced DNA damage using actual light energy. They are also involved in circadian rhythms. I know plausible explanations have been given for the evolution of the physiology of the vertebrate eye, but I wish to advance a theory on the genomic evolution of that eye. One plausible scenario is that in the genome of an early heterotrophic eukaryote, a cryptochrome became fused with a transmembrane domain containing a transduce in binding site. This one change produces the first opsin and provides a way for the cell to react to light, which would have been an increase in fitness for a photosynthetic cell. These genes were inherited by all subsequent eukaryotes, but the evolution of specialized eye structures was via the transposition of the opsin gene with a basic helix loop helix promoter sequence from another gene. These specialized eye structures would also be selected for and inherited. That gives rise to a whole family of genes specialized for light capture, signal transduction, and eye structural elements. In conclusion, the vertebrate eye is marvelous and complex, but it's also highly possible by incremental changes, by natural processes we can observe today.