Quotable Quote by Cairns-Smith
Graham Cairns-Smith is well known for a bizarre theory that the first living organisms were clay minerals of the origin of life. But not so well known is that he is driven to such outlandish ideas by the enormous chemical difficulties of mainstream theories of chemical evolution, such as the RNA World (see also The RNA World: A Critique):
The implausibility of prevital nucleic acid
If it is hard to imagine polypeptides or polysaccharides in primordial waters it is harder still to imagine polynucleotides. But so powerful has been the effect of Miller’s experiment on the scientific imagination that to read some of the literature on the origin of life (including many elementary texts) you might think that it had been well demonstrated that nucleotides were probable constituents of a primordial soup and hence that prevital nucleic acid replication was a plausible speculation based on the results of experiments.
There have indeed been many interesting and detailed experiments in this area. But the importance of this work lies, to my mind, not in demonstrating how nucleotides could have formed on the primitive Earth, but in precisely the opposite: these experiments allow us to see, in much greater detail than would otherwise have been possible, just why prevital nucleic acids are highly implausible.
Let us consider some of the difficulties:
First, as we have seen, it is not even clear that the primitive Earth would have generated and maintained organic molecules. All that we can say is that there might have been prevital organic chemistry going on, at least in special locations.
Second, high-energy precursors of purines and pyrimidines had to be produced in a sufficiently concentrated form (for example at least 0.01 M HCN).
Third, the conditions must now have been right for reactions to give perceptible yields of at least two bases that could pair with each other.
Fourth, these bases must then have been separated from the confusing jumble of similar molecules that would also have been made, and the solutions must have been sufficiently concentrated.
Fifth, in some other location a formaldehyde concentration of above 0.01 M must have built up.
Sixth, this accumulated formaldehyde had to oligomerise to sugars.
Seventh, somehow the sugars must have been separated and resolved, so as to give a moderately good concentration of, for example, D-ribose.
Eighth, bases and sugars must now have come together.
Ninth, they must have been induced to react to make nucleosides. (There are no known ways of bringing about this thermodynamically uphill reaction in aqueous solution: purine nucleosides have been made by dry-phase synthesis, but not even this method has been successful for condensing pyrimidine bases and ribose to give nucleosides (Orgel & Lohrmann, 1974).)
Tenth, whatever the mode of joining base and sugar it had to be between the correct nitrogen atom of the base and the correct carbon atom of the sugar. This junction will fix the pentose sugar as either the alpha or beta-anomer of either the furanose or pyranose forms (see page 29). For nucleic acids it has to be the beta-furanose. (In the dry-phase purine nucleoside syntheses referred to above, all four of these isomers were present with never more than 8 % of the correct structure.)
Eleventh, phosphate must have been, or must now come to have been, present at reasonable concentrations. (The concentrations in the oceans would have been very low, so we must think about special situations—evaporating lagoons and such things (Ponnamperuma, 1978).)
Twelfth, the phosphate must be activated in some way—for example as a linear or cyclic polyphosphate—so that (energetically uphill) phosphorylation of the nucleoside is possible.
Thirteenth, to make standard nucleotides only the 5′hydroxyl of the ribose should be phosphorylated. (In solid-state reactions with urea and inorganic phosphates as a phosphorylating agent, this was the dominant species to begin with (Lohrmann & Orgel, 1971). Longer heating gave the nucleoside cyclic 2′,3′-phosphate as the major product although various dinucleotide derivatives and nucleoside polyphosphates are also formed (Osterberg, Orgel & Lohrmann. 1973).)
Fourteenth, if not already activated—for example as the cyclic 2′,3′-phosphate—the nucleotides must now be activated (for example with polyphosphate; Lohrmann, 1976) and a reasonably pure solution of these species created of reasonable concentration. Alternatively, a suitable coupling agent must now have been fed into the system.
Fifteenth, the activated nucleotides (or the nucleotides with coupling agent) must now have polymerised. Initially this must have happened without a pre-existing polynucleotide template (this has proved very difficult to simulate (Orgel & Lohrmann. 1974)); but more important, it must have come to take place on pre-existing polynucleotides if the key function of transmitting information to daughter molecules was to be achieved by abiotic means. This has proved difficult too. Orgel & Lohrmann give three main classes of problem:
- While it has been shown that adenosine derivatives form stable helical structures with poly(U)—they are in fact triple helixes—and while this enhances the condensation of adenylic acid with either adenosine or another adenylic acid—mainly to di(A) stable helical structures were not formed when either poly (A) or poly(G) were used as templates.
- It was difficult to find a suitable means of making the internucleotide bonds. Specially designed water-soluble carbodiimides were used in the experiments described above, but the obvious pre-activated nucleotides—ATP or cyclic 2′,3′-phosphates—were unsatisfactory. Nucleoside 5′-phosphorimidazolides, for example were more successful, but these now involve further steps and a supply of imidazole, for their synthesis (Lohrmann & Orgel, 1978).
- Internucleotide bonds formed on a template are usually a mixture of 2′-5′ and the normal 3′-5′ types. Often the 2′-5′ bonds predominate although it has been found that Zn2+, as well as acting as an efficient catalyst for the template-directed oligomerisation of guanosine 5′-phosphorimidazolide also leads to a preference for the 3′-5′ bonds (Lohrmann, Bridson & Orgel, 1980).
Sixteenth, the physical and chemical environment must at all times have been suitable—for example the pH, the temperature, the M2+ concentrations.
Seventeenth, all reactions must have taken place well out of the ultraviolet sunlight; that is, not only away from its direct, highly destructive effects on nucleic acid-like molecules, but away too from the radicals produced by the sunlight, and from the various longer lived reactive species produced by these radicals.
Eighteenth, unlike polypeptides, where you can easily imagine functions for imprecisely made products (for capsules, ionexchange materials, etc.), a genetic material must work rather well to be any use at all—otherwise it will quickly let slip any information that it has managed to accumulate.
Nineteenth, what is required here is not some wild one-off freak of an event: it is not true to say ‘it only had to happen once’. A whole set-up had to be maintained for perhaps millions of years: a reliable means of production of activated nucleotides at the least.
Now you may say that there are alternative ways of building up nucleotides, and perhaps there was some geochemical way on the early Earth. But what we know of the experimental difficulties in nucleotide synthesis speaks strongly against any such supposition. However it is to be put together, a nucleotide is too complex and metastable a molecule for there to be any reason to expect an easy synthesis.
You might want to argue about the nineteen problems that I chose: and I agree that there is a certain arbitrariness in the sequence of operations chosen. But if in the compounding of improbabilities nineteen is wrong as a number that would be mainly because it is much too small a number. If you were to consider in more detail a process such as the purification of an intermediate you would find many subsidiary operations—washings, pH changes and so on. (Remember Merrifield's machine: for one overall reaction, making one peptide bond, there were about 90 distinct operations required.)
Cairns-Smith, A. G., Genetic Takeover and the Mineral Origins of Life (Cambridge, NY:Cambridge University Press, 1982).