The Panda’s Thumb
Twenty years ago, I spent a summer at the University of Colorado, fortifying myself for the transition from high school to college. Amidst the various joys of snowcapped peaks and sore asses from trying to “set a trot,” I well remember the highlight of my stay—George Wald’s lecture on the “Origin of Life.” He presented with infectious charm and enthusiasm the perspective that developed in the early 1950s and reigned as an orthodoxy until very recently.
In Wald’s view, the spontaneous origin of life could be considered as a virtually inevitable consequence of the earth’s atmosphere and crust, and of its favorable size and position in the solar system. Still, he argued, life is so staggeringly complex that its origin from simple chemicals must have consumed an immense amount of time—probably more time than its entire subsequent evolution from DNA molecule to advanced beetles (or whatever you choose to place atop the subjective ladder). Thousands of steps, each requiring the one before, each improbable in itself. Only the immensity of time guaranteed the result, for time converts the improbable to the inevitable—give me a million years and I’ll flip a hundred heads in a row more than once. Wald wrote in 1954: “Time is in fact the hero of the plot. The time with which we have to deal is the order of two billion years…. Given so much time, the ‘impossible’ becomes possible, the possible probable, and the probable virtually certain. One has only to wait: time itself performs the miracles.”
This orthodox view congealed without the benefit of any direct data from paleontology to test it, for the paucity of fossils before the great Cambrian “explosion” 600 million years ago is, perhaps, the outstanding fact and frustration of my profession. In fact, the first unambiguous evidence of Precambrian life appeared in the same year that Wald theorized about its origin. Harvard paleobotanist Elso Barghoorn and Wisconsin geologist S. A. Tyler described a series of prokaryotic organisms from cherts of the Gunflint Formation, rocks nearly two billion years old from the northern shore of Lake Superior. Still, the gap between the Gunflint and the earth’s origin spanned 2½ billion years, more than enough time for Wald’s slow and steady construction.
But our knowledge of life continued its trek backward. Laminated carbonate deposits, called stromatolites, had been known for some time from rocks of the Bulawayan Series, 2.6 to 2.8 billion years old, in Southern Rhodesia. The laminations resemble patterns formed by modern blue green algal mats that trap and bind sediment. The organic interpretation of stromatolites won many converts after Barghoorn and Tyler’s Gunflint discoveries removed the odor of heresy from belief in Precambrian fossils. Then, ten years ago in 1967, Barghoorn and J. W. Schopf reported “algalike” and “bacteriumlike” organisms from the Fig Tree Series of South Africa. Now the orthodox idea of slow construction spanning most of the earth’s history began to crumble, for the Fig Tree rocks, based on dates available in 1967, seemed to be more than 3.1 billion years old. Schopf and Barghoorn dignified their discoveries with formal Latin names, but their own characterizations—algalike and bacteriumlike—reflected their doubts. In fact, Schopf later decided that the balance of evidence stood against the biological nature of these structures.
The recent announcement of 3.4-billion-year-old life is not a startlingly new discovery, but a satisfactory culmination of a decade’s debate about the status of life in the Fig Tree. The new evidence, gathered by Andrew H. Knoll and Barghoorn, also comes from cherts of the Fig Tree Series. But now the evidence is close to conclusive; moreover, recent dates indicate a greater age of 3.4 billion years for the series. In fact, the Fig Tree cherts may be the oldest appropriate rocks on earth for the discovery of ancient life. Older Greenland rocks have been too altered by heat and pressure to preserve organic remains. Knoll tells me that some unstudied cherts in Rhodesia may range back to 3.6 billion years, but eager scientists will have to await a political denouement before their arcane concerns attract sympathy or ensure safety. Still, the notion that life has been found in the oldest rocks that could contain evidence of it forces us, I think, to abandon the view of life’s slow, steady, and improbable development. Life arose rapidly, perhaps as soon as the earth cooled down sufficiently to support it.
The new fossils from the Fig Tree Series are far more convincing than the previous discoveries. “In younger rocks [they] would without hesitation be called algal microfossils,” Knoll and Barghoorn claim. This interpretation rests upon five arguments:
1. The new structures are within the size range of modern prokaryotes. The earlier structures described by Schopf and Barghoorn were disturbingly large; Schopf later rejected them as biological, primarily on the basis of their large size. The new fossils, averaging 2.5 micrometers in diameter (a micrometer is a millionth of a meter), have a mean volume only 0.2 percent as large as the earlier structures now considered inorganic.
2. Populations of modern prokaryotes have a characteristic distribution of size. They can be arranged in a typical bell-shaped curve, with the average diameter most frequent and a continual decrease in number towards larger or smaller sizes. Thus, prokaryotic populations not only have a diagnostic average size (point 1 above), they also have a characteristic pattern of variation about this average. The new microfossils form a beautiful bell-shaped distribution with limited spread (range from 1 to 4 micrometers). The previous, larger structures exhibited much greater variation and no strong mean.
3. The new structures are “variously elongated, flattened, wrinkled, or folded” in a manner strikingly similar to Gunflint and later Precambrian prokaryotes. Such shapes are characteristic of postmortem degradation in modern prokaryotes. The larger, earlier structures were distressingly spherical; spheres, as a standard configuration of minimal surface area, can be easily produced by a host of inorganic processes—consider bubbles.
4. Most convincingly, about one quarter of the new microfossils have been found in various stages of cell division. Lest such a proportion caught in flagrante delicto sems unreasonably high, I point out that prokaryotes can divide every twenty minutes or so and take several minutes to complete the process. A single cell might well spend one-fourth of its life making two daughters.
5. These four arguments based on morphology are persuasive enough for me, but Knoll and Barghoorn add some biochemical evidence as well. Atoms of a single element often exist in several alternate forms of different weight. These forms, called isotopes, have the same number of protons but different numbers of neutrons. Some isotopes are radioactive and break down spontaneously to other elements; others are stable and persist unchanged throughout geologic time. Carbon has two major stable isotopes, C12 with 6 protons and 6 neutrons, and C13 with 6 protons and 7 neutrons. When organisms fix carbon in photosynthesis, they use preferentially the lighter isotope C12. Hence, the C12/C13 ratio of carbon fixed by photosynthesis is higher than the ratio in inorganic carbon (in a diamond, for example). Moreover, since both isotopes are stable, their ratio will not alter through time. The C12/C13 ratios for Fig Tree carbon are too high for an inorganic origin; they are in the range for fixation by photosynthesis. This, in itself, would not establish the case for life in the Fig Tree; light carbon can be fixed preferentially in other ways. But combined with the evidence of size, distribution, shape, and cellular division, this additional support from biochemistry completes a convincing case.
If prokaryotes were well established 3.4 billion years ago, how much further back shall we seek the origin of life? I have already pointed out that no suitable (or at least accessible) older rocks are known on earth, so for now we can proceed no further from the direct evidence of fossils. We turn instead to the second front-page item, the claim of Carl Woese and his associates that methanogens are not bacteria at all, but may represent a new kingdom of prokaryotic life, distinct from the Monera (bacteria and blue green algae). Their report has been widely distorted, most notably in the New York Times editorial of November 11, 1977. The Times proclaimed that the great dichotomy of plants and animals had finally been broken: “Every child learns about things being vegetable o
r animal—a division as universal as the partition of mammals into male and female. Yet…[we now have] a ‘third kingdom’ of life on earth, organisms that are neither animal nor vegetable, but of another category altogether.” But biologists abandoned “the great dichotomy” long ago, and no one now tries to cram all single-celled creatures into the two great groups traditionally recognized for complex life. Most popular these days is a system of five kingdoms: plants, animals, fungi, protists (single-celled eukaryotes, including amoebas and paramecia, with nucleus, mitochondria, and other organelles), and the prokaryotic monerans. If methanogens are promoted, they will form a sixth kingdom, joining the monerans in a superkingdom, Prokaryota. Most biologists regard the division between prokaryotes and eukaryotes, not between plants and animals, as the fundamental partition of life.
Woese’s research group (see Fox, et al., 1977 in the bibliography) isolated a common RNA from ten methanogens and from three monerans for comparison (DNA makes RNA, and RNA serves as the template upon which proteins are synthesized). A single strand of RNA, like DNA, consists of a sequence of nucleotides. Any one of four nucleotides can occupy each position, and each group of three nucleotides specifies an amino acid; proteins are built of amino acids arranged in folded chains. This, in a compressed phrase, is the “genetic code.” Biochemists can now “sequence” RNA, that is, they can read the entire sequence of nucleotides in order down the RNA strand.
The prokaryotes (methanogens, bacteria, and blue-green algae) must have had a common ancestor at some time near the origin of life. Thus, all prokaryotes had the same RNA sequence at one point in their past; any current differences arose by divergence from this common ancestral sequence, after the trunk of the prokaryotic tree split up into its several branches. If molecular evolution proceeded at a constant rate, then the extent of current difference between any two forms would directly record the amount of time since their lineages split from a common ancestor—that is, the last time they shared the same RNA sequence. Perhaps, for example, a different nucleotide in the two forms at 10 percent of all common positions would indicate a time of divergence a billion years ago; 20 percent, two billion years, and so on.
Woese and his group measured the RNA differences for all pairs of species among the ten methanogens and three monerans and used the results to construct an evolutionary tree. This tree contains two major limbs—all the methanogens on one, all the monerans on the other. They chose their three monerans to represent the greatest differences within the group—enteric (gut) bacteria versus free-living blue-green algae, for example. Nonetheless, each moneran is more similar to all other monerans than any moneran is to any methanogen.
The simplest interpretation of these results holds that methanogens and monerans are separate evolutionary groups, with a common ancestry preceding the appearance of either. (Previously, methanogens had been classified among the bacteria; in fact, they had not been recognized as a coherent entity at all, but had been regarded as a set of independent evolutionary events—convergent evolution for the ability to make methane). This interpretation underlies Woese’s claim that methanogens are separate from monerans and should be recognized as a sixth kingdom. Since good monerans had already evolved by Fig Tree times, 3.4 billion or more years ago, the common ancestry of methanogens and monerans must have been even earlier, thus pushing the origin of life even further back toward the beginning of the earth itself.
This simple interpretation, as Woese and his group realize, is not the only possible reading of their results. We may propose two other perfectly plausible hypotheses: (1) The three monerans that they used may not represent the entire group very well. Perhaps the RNA sequences of other monerans will differ as much from the first three as all the methanogens do. We would then have to include the methanogens with all monerans in a single grand group. (2) The assumption of nearly constant evolutionary rates may not hold. Perhaps the methanogens split off from one branch of monerans long after the main groups of monerans had branched from their common ancestor. These early methanogens may then have evolved at a rate far in excess of that followed by moneran groups in diverging from each other. In this case, the great difference in RNA sequence between any methanogen and any moneran would only record a rapid evolutionary rate for early methanogens, not a common ancestry with monerans before the monerans themselves split into subgroups. The gross amount of biochemical difference will accurately record time of divergence only if evolution proceeds at reasonably constant biochemical rates.
But one other observation makes Woese’s hypothesis attractive and inspires my own strong rooting for it. The methanogens are anaerobic; they die in the presence of oxygen. Hence, they are confined today to unusual environments: muds at the bottom of ponds depleted of oxygen or deep hot springs in Yellowstone Park, for example. (The methanogens grow by oxidizing hydrogen and reducing carbon dioxide to methane—hence their name.) Now, amidst all the disagreement that afflicts the study of our early earth and its atmosphere, one point has gained general assent: our original atmosphere was devoid of oxygen and rich in carbon dioxide, the very conditions under which methanogens thrive and for which the earth’s original life might have evolved. Could modern methanogens be remnants of the earth’s first biota, originally evolved to match its general condition, but now restricted by the spread of oxygen to a few marginal environments? We believe that most free oxygen in our atmosphere is the product of organic photosynthesis. The Fig Tree organisms were already indulging in photosynthesis. Thus, the golden age of methanogens may have passed long before the advent of Fig Tree monerans. If this reverie be confirmed, then life must have originated long before Fig Tree times.
In short, we now have direct evidence of life in the oldest rocks that could contain it. And, by reasonably strong inference, we have reason to believe that a major radiation of methanogens predated these photosynthesizing monerans. Life probably arose about as soon as the earth became cool enough to support it.
Two closing thoughts, admittedly reflecting my personal prejudices: First, as a strong adherent to exobiology, that great subject without a subject matter (only theology may exceed us in this), I am delighted by the thought that life may be more intrinsic to planets of our size, position, and composition than we had ever dared to imagine. I feel even more certain that we are not alone, and I hope that more effort will be directed toward the search for other civilizations by radio-telescope. The difficulties are legion, but a positive result would be the most stupendous discovery in human history.
Secondly, I am led to wonder why the old, discredited orthodoxy of gradual origin ever gained such strong and general assent. Why did it seem so reasonable? Certainly not because any direct evidence supported it.
I am, as several other essays emphasize, an advocate of the position that science is not an objective, truth-directed machine, but a quintessentially human activity, affected by passions, hopes, and cultural biases. Cultural traditions of thought strongly influence scientific theories, often directing lines of speculation, especially (as in this case) when virtually no data exist to constrain either imagination or prejudice. In my own work (see essays 17 and 18), I have been impressed by the powerful and unfortunate influence that gradualism has exerted on paleontology via the old motto natura non facit saltum (“nature does not make leaps”). Gradualism, the idea that all change must be smooth, slow, and steady, was never read from the rocks. It represented a common cultural bias, in part a response of nineteenth-century liberalism to a world in revolution. But it continues to color our supposedly objective reading of life’s history.
In the light of gradualistic presuppositions, what other interpretation could have been placed upon the origin of life? It is an enormous step from the constituents of our original atmosphere to a DNA molecule. Therefore, the transition must have progressed laboriously through multitudes of intervening steps, one at a time, over billions of years.
But the history of life, as I read it, is a series of stable states, punctuated at rar
e intervals by major events that occur with great rapidity and help to establish the next stable era. Prokaryotes ruled the earth for three billion years until the Cambrian explosion, when most major designs of multicellular life appeared within ten million years. Some 375 million years later, about half the families of invertebrates became extinct within a few million years. The earth’s history may be modelled as a series of occasional pulses, driving recalcitrant systems from one stable state to the next.
Physicists tell us that the elements may have formed during the first few minutes of the big bang; billions of subsequent years have only reshuffled the products of this cataclysmic creation. Life did not arise with such speed, but I suspect that it originated in a tiny fraction of its subsequent duration. But the reshuffling and subsequent evolution of DNA have not simply recycled the original products; they have produced wonders.