Stirring up the primordial soup

I've discussed in the past about the primordial soup. Some scientists believe that this soup was a 'letters soup' indeed. They call it the RNA world, where RNA was both a genetic repository and a catalytically active molecule. In modern day 'soups' (our cells), the RNA is an intermediary between DNA (genetic info store) and proteins, but there are some species of RNA that have a catalytic activity -- the ribosomic RNA is the responsible of forming the peptide bond.

How could both functions of RNA coexist and allow a sort of molecular evolution to take place (and eventually lead to the first living cell)? It's supossed that a particular kind of RNA existed in those times, a ribo-RNA-polymerase. Taking a RNA molecule as template, it should be able to catalyse the synthesis of an exact copy (with a small margin of error, without mutation there can't be evolution). Most mechanisms proposed so far involve the formation of an intermediary double stranded RNA: the polymerase should create first a complementary copy of the target molecule and then another complementary copy of this one (therefore with the same sequence as the original template) should be synthesised.

This poses a problem: the formation of double stranded RNA will act as a sink of active RNA molecules, because it's more stable but catalytically inactive. William Taylor has thought about this issue and came up with a possible solution ¹: RNA polymerases add nucleotides to the growing strand by the 3' end. Why not the 5'? If a polymerase followed a template reading the sequence from 3' to 5' and adding new nucleotides also from 3' to 5' (instead of from 5' to 3'), then the two strands would have complementary sequence but would not be able to hybridise. And this mechanism will leave the new molecule available (even before it has been completed) for another polymerase to start synthesising the complementary copy of it, thus achieving the replication of the original without a dsRNA intermediate.

However, I think that if today's RNA polymerases do synthesise with that particular polarity, thus creating perfectly complementary strands, it would be for a very good reason. Maybe the Watson-Crick pairs can't form if the two strands are not anti-parallel (this will make the proposed mechanism not suitable because no sequence complementarity could be achieved). Maybe the energy needed to catalyse a 5' extension is much higher. I don't know about these issues, but more in depth thermodynamic studies should be done before giving more relevance to Taylor's proposed mechanism. But still is a jolly good idea.

---

1. William R. Taylor Stirring the primordial soup Nature 434, 705 (07 April 2005)