MrDarwin

Every cell in the body has a complete set of chromosomes and genes, even though the vast majority of those genes are unnecessary and will never be used by those cells. I'm not talking about "junk DNA", I'm talking about genes that operate only in certain tissues or organs, or only at certain times (e.g., during embryonic development). Only the cells of the gonads, which produce the gametes (sperm and egg) for reproduction, need to have the full set. Most cells in the body are very specialized and have no use whatsoever for the vast majority of genes they contain, yet every time they reproduce, they reproduce each and every one of those genes, even the unneeded ones.

This is not only wasteful of resources (why use up raw materials to make something you don't need?), but also introduces more opportunity for replication errors (which cause diseases like cancer). And unless they occur very early in development (i.e., during the first few divisions of the embryo) these replication errors can't even be considered a source of beneficial mutations; since they occur in somatic cells they cannot be passed on to the next generation.

It would make far more sense for cells to lose the genes they do not need as they become more specialized during embryonic development. This would be easy enough to accomplish by segregating genes for similar functions together on the same chromosome, which can be discarded when not needed (e.g., the genes guiding early embryonic development to differentiate cells into tissues and organs could be dropped once embryonic development is completed).

A non-biological analogy would be every library in the world having a copy of every book ever printed. But of course it's neither efficient nor sensible for an entomology library to have books on quantum physics, or a botany library to have books on stellar evolution, any more than it is for a physics library to have a book on the wildflowers of South Africa.

pz

There is also precedent for this. Cells of the immune system do remodel their genome, throwing out bits and splicing together chunks to make custom immunoglobins. I have a vague recollection that at least some nematodes do something similar, restricting fates of some cells by discarding some genes.

I can also think of a great advantage to differentiation by gene excision: cancer resistance! One of the things that causes cancer is a loss of constraint on the cell by deregulation -- neoplasias are basically de-differentiated, reacquiring the properties of a more generic, pluripotent cell. Excising genes would make this virtually impossible. Cells would be locked into a terminal fate, and any attempt to backtrack to an earlier state would lead to failure as the cell tried to rely on missing chunks of code.

MrDarwin

Red blood cells are another example I didn't think of when I first posted this. Red blood cells have no genes at all because they have no nucleus (however, this becomes a problem when blood is needed in a hurry due to traumatic blood loss, as they cannot reproduce themselves and must be manufactured by specialized cells in the bone marrow).

Okay, so we know that some specialized cells do lose genes, so I suppose that's "good design"; but why not more of them?

lpetrich

But red blood cells lose ALL their genes; the big issue is how to get rid of only those genes that are unnecessary, while keeping the necessary ones.

Though doing so may require that the genes be divided into groups, such as one for all cells, one for skin cells, one for muscle cells, one for nerve cells, one for bone cells, one for gland cells, etc. Or else some way of tagging genes with tags that represent these categories.

[ October 16, 2002: Message edited by: lpetrich ]</p>

Peez

Another example is the Barr Body. Female mammals have two "X" chromosomes while males have one "X" and one "Y". This means that female cells potentially produce twice as much of the proteins that are coded for on the "X" chromosome as males do (the "Y" chromosome does not code for the same proteins). Males produce enough of these proteins, so why don't females produce too much? Too much of certain proteins can be a big problem. For example, Down's syndrome is caused by too much of a particular protein being produced, because of an extra copy of one chromosome.

Female mammals inactivate one of their "X" chromosomes and stow it in the margin of the nucleus. This forms the "Barr Body", which I believe was once used to test for sex in Olympic athletes. This supports the notion that it is possible, in principle, to get rid of extra DNA in somatic cells. It would be a cinch for the IPU to design such a feature in, but it would be tricky to evolve it by natural selection. I haven't given it much thought, but getting all the relevant loci segregated on certain chromosomes before the evolution of a system to dump locally useless DNA seems unlikely.

Peez

MortalWombat

Just to nitpick: among vertebrates, only mammals have anucleated red blood cells; birds, reptiles, amphibians, and fish RBCs have nuclei. And it's not the lack of a nucleus per se that prevents mature red blood cells from reproducing themselves. Avian red blood cells for instance, although containing nuclei, are terminally differentiated and thus do not replicate.

[ October 16, 2002: Message edited by: MortalWombat ]</p>

MrDarwin

Thanks for the nitpicking. This merely emphasizes my point by presenting a more precise example of suboptimal design: why do birds have nucleated red blood cells, when the cells do not divide and apparently do not need any genetic information? We already know that mammalian red blood cells do quite well without a nucleus. If we acknowledge that enucleate red blood cells are an "optimal design" feature, this certainly suggests that the nucleated blood cells of birds are a suboptimal design.

GeoTheo

suboptimal design is a stupid concept in my opinion. It reminds me of Plato's concept of "forms" like there is some ultimate essence of how these organisms are supposed to look like floating around in the universe some where.
An organism is an organism. Neither optimal nor sub-optimal. All you can say is that they show historical constraint.

MrDarwin

GeoTheo, I agree with your last sentence entirely. But while you may consider suboptimal design a "stupid concept", there are many people who use "optimal design" along with similarity of design as cornerstones of "intelligent design" theory. As I've tried to point out, this theory has serious problems accounting for examples of suboptimal design or dissimilarity of design. They have yet to come up with a cohesive theory of constraint of design, which is easily accounted for in evolutionary theory as being due to past history, common descent, and shared genealogy.