Perspectives in Biology and Medicine

A POLYPHYLETIC VIEW OF EVOLUTION: THE GENETIC POTENTIAL HYPOTHESIS CHRISTIAN SCHWABE and GREGORY W. WARR* Introduction Modern theories of evolution contain an implicit assumption about die origin oflife which runs through nearly all writings on evolution and which carries the original idea of Darwin—that is, that life forms can develop into new species—one step further to the view that all life forms, extant and extinct, are related by common ancestral relationships . The following quote is from a recent Science article [I]: "The systematic comparison ofevery newly determined amino acid sequence with all other known sequences may aUow a complete reconstruction of the evolutionary events leading to contemporary proteins." As it makes clear, this view assumes a single ancestral gene and a single ancestral cell from which all other genes, cells, and organisms have arisen. This monophyletic view has been supported by the finding that the genetic code is all but unitorm. In this paper we advance an idea on evolution that dispenses with the assumptions of die central monophyletic dogma and places more emphasis on deterministic chemical principles. This genetic potential hypothesis is in harmony with existing data and, as we shall show, is strengthened by findings that are incompatiblewith a monophyletic dieory of evolution. The hypothesis is supported not only by our data on relaxin sequences from various species but also by other examples from comparative studies of proteins. Existing Models Proteins are thought to evolve mainly through gene duplications and continuing random processes of mutation that may be neutral, deleteThe experimental aspects of this work were supported by NSF grant PCM-8302194 and NICHHD grant HD-10540 to C. S., and NSF grant PCM-8108872 to G. W. W.»Department of Biochemistry, Medical University of South Carolina, 171 Ashley Avenue , Charleston, South Carolina 29425.© 1984 by The University of Chicago. AU rights reserved. 003 1-5982/84/2703-0398$01 .00 Perspeaives in Biology and Medicine, 27, 3 ¦ Spring 1984 \ 465 rious, or beneficial, to the extent that the extra gene copies may tiius be converted into new genes encoding molecules widi new functions [2, 3]. The neo-Darwinian argument is diat by chance some such mutations (diat by definition give rise to greater complexity ofan organism) wül be adaptively advantageous [4]. Once a new function is attained and the gene fixed, structural constraints will limit further mutations to those positions of the protein that are occupied by amino acids that are not required for the proper folding or for the biological activity of die molecule. These subsequent neutral mutations are inconsequential as far as species development is concerned and are therefore diought to occur randomly and independendy of generation frequency as a function oftime [5]. The rate ofmutation acceptance in a protein as a stricdy time-dependent function is known as die evolutionary clock hypodiesis. That the clocks for different protein families run at different speeds is explained by die number ofunessential sites for mutation in die particular molecule [5]. A variant of the molecular clock model is die neutralist dieory of Kimura [6], which invokes statistical arguments to show diat neutral mutations can spread through a population by a purely random (genetic drift) process [7]. Under this theory, in a population of size N and effective size Ne, a fraction VaJV of neutral mutations wül spread dirough a population and become fixed, and die average length of time for diis process for a single mutant gene is approximately 4Ne generation times. However, we need to have a very much better measure of the rate of neutral mutations (if such truly occur) in structural proteins before we can begin to fit die observed data to die predictions of this dieory and make a reasonable test. It also has to be pointed out, however , that this neutralist dieory makes many assumptions diat are in principle difficult tojustify. A clear example of this is what exactly constitutes a neutral, or nearly neutral, mutation. The hystricomorph insulins provide an interesting case study in die difficulties ofdefining functional fitness of molecules in different species. The alternative to die neutralist and molecular clock views of protein evolution is that of die positive selectionists who argue that mutation fixation...