Natural Selection and the Origin of Life

NATURAL SELECTION AND THE ORIGIN OF LIFE GORDON ALLEN* At least four conferences on the origin oflife have been convened and the proceedings published in the last fifteen years [1-4], yet central and difficult questions remain unanswered [5]. Those who explain the event in terms ofknown phenomena believe either in unrealistic quantities of time and reagents, or in unspecified constraints that limited random reactions to a small and appropriate range. Solutions proposed for one set of problems often raise new problems ofequal or greater magnitude. Ifthe probability of life's origin is low according to present concepts, this foUows chiefly from the apparent need ofthe first form oflife for a self-copying mechanism, which caUs in turn for a highly specialized macromolecule. These requirements could be bypassed if reproduction by the environment could take the place ofreproduction by self. Reproduction by the environment can, in fact, be postulated in a form that satisfies the elementary requirements of natural selection. Such selection could, therefore, have guided evolution during the development ofmore complicated genetic mechanisms. My first purpose is to review and to integrate some earlier proposals for the origin oflife that canbe interpreted as reproductionby the environment [6-9]. This is not enough, because when self-copying is avoided in this fashion, two capacities essential in organic evolution appear also to be bypassed: the capacity to mutate and the capacity to accumulate biological information. My second purpose is to explain how these two capacities may exist in the absence ofcopying. This thesis admittedly conflicts with what many geneticists regard as self-evident truth. It wiU also be necessary to indicate how evolution could proceed along a continuous path from such beginnings to more familiar forms. FinaUy, I shaU urge a new direction for research into the origin oflife. * National Institutes ofHealth, 9000 Rockville Pike, Bethesda, Maryland 20014. IO9 Reproduction by the Environment The paramount importance of self-copying in biology has been challenged by a number ofauthors [e.g., io, il]. At least three have independently proposed reproduction without copying as akey mechanism in the origin oflife. Under these proposals, the structure ofreproducing catalysts was determined not by the catalysts, but by the chemical environment in which they arose. The catalysts were not necessarüy present at the final steps ofsynthesis and it therefore seems fitting to characterize the process as reproduction by the environment. The three independent contributions that wül be described below came from three disciplines and presented quite different aspects ofwhat was essentiaUy a single hypothesis. Yeas [6], a cell physiologist, pictured the prebiotic ocean as a homogeneous reaction mixture that graduaUy assumed the character ofa single metabolizing system integrated by "a kind of'natural selection' based on reaction rates." Primitively, sunlight or electric discharges generated simple compounds like cyanide, formaldehyde, and glycine. These would slowly revert to the primary reagents in a continuous cycle, but a small proportion would react with one another and produce new species, some more stable or of higher molecular weight. The secondary compounds would participate in further reactions, also cyclic. Primary valence-bond catalysis is frequent among organic compounds, and a product of one cycle could catalyze a rate-limiting reaction in another cycle. The two cycles would thus become interlocked, the second partly depending on the first. In rare instances, a compound of the second cycle would also catalyze synthesis in the first cycle so that an increment in either would produce further increments in both. The quantity of matter passing through these two cycles would then increase at the expense of other reactions. "EventuaUy (and automaticaUy), a system ofinterlocking cycles wUl be selected which operates at the maximal rate." As new compounds appeared, catalysts ofa peptide nature would be expected, and when these molecules started to agglomerate, discrete organisms would emerge. Calvin [7] described linear reaction sequences leading to famüiar compounds with catalytic properties. He began with a simple model of an autocatalyticreactioncompeting with otherreactions forthe sameprimary compound. Spontaneous formation of the catalyst may be unlikely, but when a single molecule appears, the autocatalytic reaction wiU begin to no Gordon Allen ¦ Natural Selection Perspectives in Biology and Medicine · Autumn 1970 speed up, making more of the catalyst...