The mammalian and octopus camera eye - common design? - evolutionary mess?

Post of the Month: February 2011

by

Subject:    | Mammalian and Octopus eyes- a second look
Date:       | 28 Feb 2011
Message-ID: | l4hnm6950fi3i6aib87ggqd0c4kok70m92@4ax.com

For octopus, they produced a library of cDNA sequences from octopus eyes and ended up with some 2824 "nonredundant sequences" to study.  Of these, 1052 were matched to a protein.  Of those one thousand proteins, only 691 were proteins with a known function.

2) The authors used 3809 human eye genes from Body Map and 2430 human connective tissue genes from BodyMap compared with the 1052 genes from octopus that corresponded to a protein.  They found "162 genes are commonly expressed between octopus and human eyes whereas only 44 genes are commonly expressed between octopus eyes and human connective tissue." I do not understand this at all because of the 1052 genes involved there are 111 for enzymes and 39 for ribosomal protein and both octopus eyes and human connective tissue do glycolysis and Kreb's cycle and protein synthesis and all the basic cell biochemistry in common.  Certainly there are more than 44 genes shared!  But that is what the paper says.  Note: these numbers only indicate commonly expressed genes among those associated with known proteins.  That includes only about 1/3 of the total genes expressed in the cells.

I don't claim to have cleared up any confusion with this information. I only repeat what the authors write which, to my mind, only adds confusion.

3) The last common ancestor of the bilateria almost certainly did have a nice central nervous system with an anterior enlargement called a brain and specialized sense organs include photoreceptors that could be called eyes.  These are features of all bilateral animals.

For many decades, people have understood that there are basically two distinctly different types of photoreceptor cells: those derived from ciliated cells (like vertebrate rods and cones) and those derived from cells with microvilli instead of cilia (the "rhabdomere" type receptor) found in the protostomes like insects and molluscs.  This led to the notion that photoreceptors evolved independently in protostomes (the insect and mollusc line) and deuterostomes (the echinoderm and chordate/vertebrate line).  However it turns out the some clams and scallops have ciliary photoreceptors and some starfish have microvilli photoreceptors.  But the really big change was the finding that the gene, Pax6, is a master control gene for producing eyes and is widely conserved across all bilateral animals, protostome and deuterostome.  The mouse PAX6 can trigger eye development in the fruit fly!  That forces the conclusion that the production of eyes must have been present in the ancestral bilateria.

Unfortunately Pax6 is not expressed in the octopus.  That is just one of those flukes that constantly appear in biology: Pax6 is found in the non-camera eyes of scallops and the camera eyes of squid so the impression is that Pax6 should be there doing its thing.  Also, Pax6 does a whole bunch of things other than control eyes: it is involved in producing specialized olfactory neurons and is also found in the pancreas as well as many parts of the nervous system.

My impression is that you will never find a regulatory or control gene that has "this one specific function and only this", like "build an eye".  In order to connect genes with phenotype you need an ensemble of genes all interacting together in the proper pattern and sequence and no one specific gene has one specific function.  It is like saying "what is the specific function of this particular flip-flop or NAND circuit in the CPU of this computer?"  Well, it depends on what the computer happens to be doing at the moment.  There is a specific function but it is a very technical thing involved in the low-level details of the internal working of the computer.  Similarly, the "function" of a gene has a very technical meaning in the low-level details of the internal working of the cell but really cannot be related to the overall morphology or phenotype of the finished organism.  What we do, instead, is say that "disruptions of Pax6 cause really severe malformations of the visual system as its most immediately noticeable effect".  Therefore we call it the "eye control gene" even though it has all sorts of other important functions in other tissues that are just less evident on casual inspection.

There is no "camera eye" gene or gene set that can be found shared by cephalopod molluscs and vertebrates.  There are regulatory genes that, put together in the proper packages, produce complex structures.  Those we share with all animals.  How we put them together in different packages varies enormously across the animal kingdom.