'Gene Machines'

A reviewer discusses Richard Dawkins' The Selfish Gene, now in its 30th anniversary edition.

Full-text access for registered users only. Existing users login here.
New to GenomeWeb? Register here quickly for free access.

Richard Dawkins' book "The

Richard Dawkins' book "The Selfish Gene" has done a wonderful job in the education of the masses about genomics, but it's title unfortunately propagates confusion about the nature of genes. Moreover, the book advanced the mythology that natural selection works simply at the level of the gene and individual.

If total biomass and persistence are the best measures for the most successful species on the planet, it is clear that eusociality plays as much if not a greater role than the individual. This is well exemplified in social insects (e.g. ants, bees, wasps, termites), bats. dolphins, corrals and even by ourselves. An ant will easily sacrifice itself for her queen and colony, just as a young man with no offsprings will be willing to sacrifice himself for his king and country (or more realistically for his comrades with whom he does not share genes but rather learned values). Species are best able to undertake hunting or survive predation if they are organized in groups. Often, this means individuals in the group acquiring specialized properties that advance the interests of the group, and not necessarily those of the individual. Moreover, many of the most successful lifeforms are engaged in highly symbiotic relationships with other species, for examples as seen with flowering plants, trees, lichens, corrals and the ingested bacteria in animals.

The notion advanced in "The Selfish Gene" that an "individual" or even a "group" is merely the means by which a "gene" is able to survive and propagate itself over time is truly nonsensical. All genes are subject to random mutations that often are inconsequential, sometimes deleterious and rarely confer an improvement. A random mutation might lead to improved survival of the organism and its successful reproduction. However, such improvements are in the context of all of the other genes residing in the cell and these genetic changes may often only become advantageous with a changing environment.

Within a single eukaryotic cell, there may actual exist a diversity of oligonucleotide sequences that encode the same "gene." In the extreme case, mitochondrial genes accounts for a significant part of the mRNA in cells, and with about 1700 mitochondrial in the average cell, there are many versions of the "same" gene. In most animals, only a tiny fraction of the DNA actually encodes genes, sometimes less than 1%. Nevertheless, much of these non-coding DNA sequences are also highly conserved. Presumably, these non-coding DNA sequences are not as "selfish" as genes.

Moreover, the very concept of the gene as being the basic unit of natural selection arises from the bigotry of those that hold a strong genopocentric perspective. Genes are just blueprints for making functional proteins and RNA. The average protein is about 700 amino acids in length. With the degeneracy of the genetic code averaging to about 3 triplet nucleotide sequences per amino acid type, there are approximately 3 to the 700 power of possible gene sequences that will specify exactly the same protein. Therefore, the evolutionary pressure is not on the structure of the gene, but rather what kind of protein it will encode.

The survival advantages conferred to an organism (or more appropriately to its group) by any gene sequence is really manifested at the protein level. The precise amino acid sequence of a protein specifies what it can do, but the presence of other interacting proteins determines what it will do. Therefore, context is extremely important. Unless genes are physically connected to each other as direct neighbours, they cannot interact with or affect each other except through the actions of proteins.

Someone may wish to argue that the appearance of similar isoforms for a protein may be an example of successful propagation of a gene into a related family of genes. When it comes to metabolism, there may be a single gene that encodes the enzyme for each step in a metabolic pathway. However, in some bacteria, the oligonucleotides sequences that encode different enzyme activities in eukaryotes, such as for fatty acid synthesis, are reduced to a single gene. Genes that encode regulatory proteins by far outnumber those genes that specify metabolic pathway enzymes or structural proteins. However, signalling proteins such as protein kinases are extremely promiscuous and redundant in their interactions. Therefore, individual genes are not particularly critical for regulation, but rather it is important that there is a large pool of regulatory proteins that can coordinate the actions of thousands of other proteins.

It is also evident that most proteins are composed of functional domains, such as for catalytic activity (e.g. kinases, cyclases) and protein-protein interactions (e.g. SH2, SH3), which are encoded in the same gene. Therefore, does the "selfishness" reside within the portions of the gene specifying protein domains rather than whole protein, or does it permeate even further into the sequences encoding the common secondary structures of proteins (e.g. beta-pleated sheets and alpha-helices or zinc fingers)? Perhaps the oligonucleotide triplets in DNA is just the way that individual amino acids propagate selection for just themselves with this kind of logic.

Natural selection and biological diversity cannot simply be understood at the level of the gene. Rather, whether it is inside a cell, an organism or within a group, it is the complex interactions of the components of the system (be they macromolecules (e.g. DNA, proteins), tissues and organs, or individuals, respectively) in response to changing environments that drive evolution.

Thanks S. Pelech, that was

Thanks S. Pelech, that was terrific.

The 30th anniversary of its

The 30th anniversary of its publication was 2006.

The selfish gene is a

The selfish gene is a brilliant book! I think that no new information obtained through ENCODE project or other omics studies, undermine the basic premise of the book. The trouble is that certain socio-religious "memes" and some scientific "memes" cannot coexist in all survival machines. So we debate endlessly about the lack of co-expression, though many of us are comfortable with differential spatio-temporal expression of those very "memes"