Possible Pathways for the evolution of intracellular transport.
"A search of the professional literature and textbooks shows that no one has ever proposed a detailed route by which such a system could have come to be."
Some commenters here have argued that Behe is being intentionally deceptive they argue that there is an abundance of published material – shelf loads of it - that give a clear outline of how a pathway transporting a newly synthesized protein to an intracellular compartment could arise.
Some suggestions were given as to where I should start – (I acknowledge that they were probably hurriedly put together sources by scientists who are very busy doing more important work than arguing with me and I do appreciate the attempt to provide me with the references.)
This is sequence comparisons in the myosin superfamily looking at homologies between the different types of myosin molecules in different organisms.
This is the attempted production of a phylogenetic tree comparing different types of myosin molecule.
The Richards and Cavalier Smith Nature paper is similar and suggests that the most primitive eukaryotes had three types of myosin from which all eukaryotic myosins come but that does not really help me.
This gives 111 delightful titles but not really what I am after. They are mainly phylogenetic trees and studies of sequence similarities. Behe accepts that there are an abundance of this kind of study.
What I am after is a simple step by step process whereby a single transport system from protein translation completion to function in a separate compartment can arise. It does not have to be a DVD of the process happening just a suggestion of some of the useful steps along the way.
I am thinking of the kind of thing that Matt Inlay produced in response to Behe’s Immunology chapter (here) or Nick Matzke’s response to the flagellum chapter ( here)
This problem concerns the way proteins are targeted to the mitochondria. These organelles (again, they’re shown in your diagram) are responsible for supplying a major fraction of the cell’s energy needs. They are distantly descended from free-living bacteria that began a symbiotic relationship with an early eukaryote. As part of that evolutionary history, mitochondria still retain a small genome which encodes a few of the proteins required by the organelle. However, over evolutionary time there has been a general drift towards more and more mitochondrial genes being transplanted to the nucleus. Mitochondrial proteins produced from such nuclear genes somehow have to get to their correct organelle. How do they do that? It turns out that such proteins contain, right at the start of their amino acid sequence, a so called ‘targeting signal’ made of about the first ten or so amino acids and which docks with import machinery in the mitochondrion. A mitochondrial gene newly transplanted into the nuclear genome must acquire this signal or it risks turning into a pseudogene. So how easy is it to acquire a functioning targeting signal? Some years ago a clever experiment was performed to find out. It’s a neat example of how our intuitive ‘gut feelings’ about these issues can lead us badly off-course. The scientists took a gene for a mitochondrial protein, then replaced its normal targeting signal with random DNA sequences sized to encode between about ten and thirty amino acids. They then determined what fraction of these random sequences acted as functioning mitochondrial targeting signals for the protein.What do you think the answer was? One in ten million? Or some other Dembski number perhaps? Actually, they got a remarkable 3 to 5%! Subsequent work with more truly random and uniformly-length sequences increased this estimate still further. Evidently, it’s almost ridiculously easy to evolve working targeting signals. One more point is worth making here. Because the results were so striking and the way the experiment was conducted was so elegant, this work is rather well known in the field. It was published in 1987 – almost ten years before Behe wrote his book. Yet he tells us with a straight face that no experiments have been done to address the evolutionary origins of protein traffic!
Tony’s point here is that the ID code for the mitondrial car park is pretty easy to forge. The fellow checking the ID’s is a pretty sloppy fellow and a great variety of ID sequences will do.
However let us imagine that this putative mitochondrial gene is the very first one to complete the journey into the cell’s genome. Let us also assume that the appropriate insertion of DNA occurs of the correct length and with the approximately correct sequence. Is this all that is required for the newly made protein to find its way into the mitochondrion? Is it just a single rough ID sequence that is needed or are other modifications required in the mitochondrial genome and elsewhere in the cell?