This is something that I have also been thinking about. Whether there can be a ribozyme (X) capable of replicating itself by using another X as a template strand, which produces the complementary X’ which can be used as a template to make more X. The question I have is, why don’t we see this in biology or among viruses? If it was a stable state that preceded the current state of life where proteins do the job of template directed polymerization of RNA and DNA, then why doesn’t a ribozyme RNA/DNA polymerase exist anymore in some form or another in nature (or maybe it does, I am not aware)?
One could argue that proteins do the job more efficiently such that any old ribozyme was superseded and lost. That could be true, but that also raises the question for why we do still have a couple of very important ribozymes still remaining: Self-splicing introns, spliceosomes that likely evolved from self-splicing introns, RNAse-P that is involved in the maturation of tRNA, various others that can perform (self-) cleavage and/or ligation of RNA, and finally ribosome, which is also composed of proteins but the catalytic PTC site is entirely RNA and a ‘naked’ (protein free) ribosome is capable of peptide bond formation.
One side note: it’s a curious fact that, while prokaryotes only have one RNA polymerase, eukaryotes have multiple types of RNA polymerases that produces distinct classes of RNA. RNA Pol I transcribes rRNA except for 5S RNA, Pol II transcribes mRNA, and Pol III transcribes tRNA, 5S RNA, RNAse-P and spliceosomal RNA. I don’t know why this is he case.
So, ribozymes mostly perform cleavage/ligation on polynucleotides and mediate peptide bond formation, but none perform template directed nucleotide polymerization. If proteins are so much better catalysts such that they replaced these, why didn’t proteins replace the other ribozymes that we still have, such as the ribosome?
My guess (it’s a guess admittedly so I may be wrong) is that a ribozyme transcribing itself wouldn’t work. Let’s suppose there is a ribyzome X, which transcribes another X as a template strand, producing the complement X’ and that is transcribed to make more X. If there is any degree of error, this ribozyme would quickly destroy itself via a run-away feedback loop. One error could make the X ribozyme of the next generation more error prone, and producing more error prone X ribozymes, until any degree of fidelity (or the catalytic function itself) is lost and self-replication grinds to a halt. Why doesn’t this also happen with the protein based replication of RNA and DNA one might ask? After all, you can have the same feedback. Protein polymerases make an error in the coding strand for the polymerase itself, which makes the protein polymerase more error prone, producing more errors later. This doesn’t happen because (as I discussed in a previous comment) the translational system and the genetic code buffers against errors, and proteins themselves often tolerate errors (robustness). Again, take this with a grain of salt, I am open to hear any other suggestions or disagreements.