In ID circles we often discuss the sheer rarity of biologically relevant polypeptides with respect to combinatorial sequence space (and the related conundrum of macromolecular interdependence). It has often been argued that this represents a potent challenge to chemical origin-of-life models of an order substantially greater than the challenge it presents to biological evolution. There is one related problem in this regard which is often overlooked, and I want to briefly explore it in this blog entry.
When it comes to polymerization of amino acids to form proteins, two things must be borne in mind with regards to the formation of peptide bonds.
- Peptide bond formation is an endothermic reaction. This means that the reaction requires the absorption of energy: It does not take place spontaneously.
- Peptide bond formation is a condensation reaction. It hence involves the net removal of a water molecule. So not only can this reaction not happen spontaneously in an aqueous medium, but, in fact, the presence of water inhibits the reaction.
There is also the added problem of interfering cross-reactivity (the probability of interfering cross-reactions between the chemical groups on the various amino acid side chains is quite high).
But this is only the peak of the proverbial ice berg. The difficulties associated with synthesizing peptides (altogether with appropriate homochirality and all) are only half the story. There is also the problem of breaking the peptide bonds in order to generate a range of amino acid sequences in view of finding some with meaningful activity. I mentioned previously that the formation of a peptide bond requires a loss of a water molecule and the input of energy. On the flip side of the coin, then, breaking these bonds requires the addition of a water molecule and involves an energetically favorable reaction. But here’s the thing: Although this entails a net release of energy, the reaction involves high activation energy. But the activation energy for hydrolysis of peptide bonds is such that spontaneous hydrolysis under ambient conditions is not something which occurs readily.
In view of the difficulties associated with the making and breaking of peptide bonds, a very bleak picture is painted for the exploration of amino acid sequences in the pre-biotic context. Given that the conditions required for the making and breaking of peptide bonds are really quite different from one another, if naturalistic origin-of-life scenarios are to have any traction, it would entail that a location be required in which the conditions can vary significantly, alternating between conditions suitable for peptide bond formation and breaking. And this of course is compounded by the fact that the reactions, when they do occur, are likely to be slow and inefficient. Even granting that volcanoes and ocean vents might have provided the necessary changing conditions, it still stands to reason that the production of different polypeptides cannot have exceeded the rate of change of environmental conditions. This would dramatically limit the potential number of polypeptides which could have been produced in the prebiotic world, thus placing considerable restraints on the probabilistic resources at one’s disposal for the formation of multiple biologically relevant (and functionally interdependent) polypeptides.
In view of the reasons articulated above (and many others), the proteins-first model of the origin of life may be taken as essentially dead in the water. Not only are there the substantive challenges of even forming biologically relevant polypeptides. But even supposing that such prebiotic polymers could be produced in this way and useful sequences were happened across, the polymyers have to be able to reproduce with reasonable integrity. But there does not appear to be any way in which a polypeptide can determine a peptide sequence in some fashion analogous to that of base pairing of nucleic acids. How would these proteins be replicated in order to facilitate the workings of natural selection?
In view of the obvious closed-loop “catch-22” paradox of DNA making proteins and proteins making DNA, there is, of course, the fashionable scenario of the RNA world: That is to say, the possible role of RNA as the earliest hereditary macromolecule. This is seen to follow from the realization that RNA not only has information-carrying capacity, but also possesses catalytic capability. Proposed evidence for this notion included the fact that RNA makes up a large proportion of ribosomes (the protein factory of the cell). Furthermore, in eukaryotes (organisms with nucleated cells), components of genes which don’t code for proteins (called “introns”) are spliced out of an RNA transcript before translation. RNA molecules are involved in many of the RNA-splicing processes, and it has been documented that some RNA introns have self-splicing capability: that is to say, they can excise themselves, though at a slower rate than proteins can do it. Further observations which were taken as evidence for the plausibility of the RNA world thesis included the existence of RNA viruses, which use RNA as their genetic material which is translated directly into proteins.
Leaving aside the problems of attaining an RNA-based replicase (for that discussion, see Signature in the Cell), the problem is that the difficulties outlined above with regards the formation of polypeptides are really quite trivial in comparison to the difficulty of obtaining polynucleotides, in part because of the different kinds of bonds which need to be made and broken and the very different reaction conditions which are necessary at each stage. Nucleotides are composed of three chemical subunits – a ribose sugar, a phosphate group, and a nitrogenous base. Not only do these components need to be present and react together in an appropriate fashion in order to produce one nucleotide, but these nucleotides then have to be polymerized, a process which requires a series of endothermic condensation reactions, thereby requiring a high-energy condensing agent in order to perform them. In order to obtain nucleosides (i.e. base and ribose), one would need to begin with a mixture of nitrogenous bases and ribose and an appropriate condensing agents. To obtain nucleotides requires the mixing of nucleosides with phosphate and a different condensing agent.
The scenario for self-replicative capability of polynucleotides is more optimistic than that for polypeptides. But this is by no means trivial. At the heart of Darwinian rationale lies the concept that evolution must strike a balance between reliable reproduction of a species on the one hand, and opportunistic variation on the other. A poor replicator is much more likely to degrade through inaccurate copying than to be enhanced by evolution. There thus exists a threshold before the cumulative improvement of a replicator can occur by selection. A replicator must already have a reasonably good performance before it can even improve on that performance. At this point, however, we are running perilously close to yet another catch-22 conundrum: If (as I think is a legitimate assumption), this threshold performance level may be only attained with a sequence substantially longer than the minimum required for folding, one is faced with the even greater improbabilities of attaining such a replicator by a blind search.