The emergence of life for the first time on this planet constitutes the classic question of what came first; the chicken or the egg?! Did a self-replicating DNA system occur before transcription or translation evolved (the DNA World) or did a self-replicating RNA system first emerge (the RNA world) or did self-replicating protein system first emerge (the Protein World)…or … let’s just leave it there for now.
This much is clear; DNA was an evolutionary afterthought. RNA served a dual function of Genetic Material as well as enzyme. This begs yet again another chicken egg problem! How could RNA’s role as a catalyst evolve, if there was nothing yet available to catalyze except RNA? The answer could be The Metabolism-First World, which even preceded RNA. But that is a different story for another time.
In brief, “Retroelements” are relics of the so-called ancient “RNA-world”! Their sequences represent more than half the human genome. Evolution is essentially a saga of subduing and eventually co-opting or rehabilitating Retrotransposons/Retroviruse jumping as a means of gene regulation. The hero of that story is a heroine named Barbara McClintock. (Everybody please stand up and please reverently bow your heads at the mere mention of her name!)
That said mobile elements directly can still cause three kinds of problems:
1. A functional gene is interrupted and no longer functional
2. A second copy of a gene is relocated to a different region of DNA, perhaps even a different chromosome. Expression of this gene is now subject to different regulation due to its new location creating all sorts of potential problems including gene dosage effects.
3. The transposition (especially the terminal repeats) may include enhancers which now activate neighboring genes inappropriately.
Mobile elements indirectly contribute to two kinds of other kinds of genetic disorder commonly clumped together as “indel”s, separately considered as inversion and deletion events. Actually, mobile elements can also contribute to a third type of genetic disorder called translocations.
Of course, everything that was identified as potentially problematic is also potential grist for evolution’s mill. Retroelements are undoubtedly the origin of introns and multiple promoters and enhancers in eukaryotes.
The generation of antibody and T-Cell receptor diversity is a direct consequence of co-opting Retrotransposons/Retroviruse jumping to rearrange the DNA before mRNA transcripts are even generated. Ditto the 800 genes in humans and 1400 genes in mice for olfactory receptors.
It is even possible that commitment steps in cellular differentiation similarly employ DNA rearrangement rendering precluding totipotency in certain cell lineages (at least some of the time).
François Jacob (of lac-Operon fame) once remarked that “Nature is a tinkerer and not an inventor…” new sequences are adapted from pre-existing sequences rather than created de novo. Domains are the common material used by nature to generate new sequences, they can be thought of as genetically mobile units in a fluid genome. Protein domains often correlate strongly with exons in multiple eukaryotic genomes and new genes emerge as a result of gene duplication and exon shuffling. It would appear that eukaryotic genomes are pretty fluid!
All of the above heralds the RNAi story, which in fact was discovered in bacteria first! CRISPR in Bacteria and DICER in Eukaryotes can also bind transcribed single-stranded RNAs that fold back on themselves to generate small regions of double-stranded RNA, those so-called “stem-loops” that allow a host cell to defend against any “retroelement”. Boy does this is all beginning to sound repetitive!
It now turns out that Bacteria can also take ‘RNA mug shots’ of threatening RNA-viruses! Life needed to defend itself against parasitic RNA before needing to defend itself against parasitic DNA. It was an easy bet to wager an RNA version of CRISPR would be discovered in bacteria. Sure enough, just check out the latest version of SCIENCE.
http://science.sciencemag.org/content/351/6276/920.short
The field is moving so fast, it is difficult to keep up.
The latest from Nature:
http://www.nature.com/news/crispr-like-immune-system-discovered-in-giant-virus-1.19462
Viruses of viruses just like stacking Russian Matrushka Dolls?!
ITMT – Pandoraviruses of Amoeba are fascinating, but are they really a candidate for a “Fourth Domain” of Life, or are we witnessing the apogee of hype?
In brief, “Retroelements” are relics of the so-called ancient “RNA-world”! Their sequences represent more than half the human genome. Evolution is essentially a saga of subduing and eventually co-opting or rehabilitating Retrotransposons/Retroviruse jumping as a means of gene regulation.
Either you have a tendency to hyperbole or you believe some weird things. How can retroelements be relics of the RNA world, when they are bits of DNA and code for proteins? And are generally descended from viruses? And how can you reduce evolution to one sort of sequence that appears to play only a minor and local part, by all other measures than frequency in some genomes?
“indel”s, separately considered as inversion and deletion events.
You mean “insertion and deletion”. Inversions are something else.
John Harshman,
reYou mean “insertion and deletion”. Inversions are something else.
Hi John, you are correct, an inversion is not the same thing as an insertion. My bad… Although technically speaking, regardless of whether one invokes the replicative or the conservative model of transposition, both amount to the same thing.
More than half the human genome is left over from the RNA world. 🙂
TomMueller,
Hi again John
You ask
Either you have a tendency to hyperbole or you believe some weird things. How can retroelements be relics of the RNA world, when they are bits of DNA and code for proteins?
On more than one occasion you have corrected misconception on my part and I thank you. I am still sticking to my guns on this point.
What you say is only true of Class II Transposons. Class I transposons first transcribe the DNA into RNA before transposition and that same RNA (at least some if not all of the time) functions as a ribozyme. Those proteins you are on about were also evolutionary after thoughts which is why I got excited about this recent article to begin with. CRISPR reiterates the Class II transposon story.
Class II Transposons are probably descendants of Class I. If so – I agree with you the thesis becomes indeed “weird”, but not at all hyperbole.
best regards
TomMueller:
Bob?
keiths,
Hi Keiths – Thanx… I fixed that typo.
TomMueller,
Hi Tom,
I don’t think functioning as a ribozyme is a very sure indicator that the RNA’s ancestry is rooted in pre-coding pre-DNA world. ‘Orphan ribozymes’ are about as likely as ‘orphan proteins’, don’t you think?
Hi Allan,
RNA simultaneously acting as parasitic/selfish and as ribozyme seems to tell a nice story.
Remember the Über-enzyme, the ribosome is made of rRNA!
Meanwhile, it appears that some prokaryotic gene regulation constitutes proof-positive of the RNA-world-relic hypothesis. I am thinking of the riboswitch story in bacteria.
Whatchya think?
They do not. What?
I think the evidence and arguments in favor of there having once been an RNA-world where the genetic material was RNA, rather than DNA, is strong and convincing. But I don’t think transposons or retroelements are arguments or evidence in addition to that.
The primary reasons for thinking there was an RNA-world before DNA is that the biosynthetic pathways for DNA anabolism are additions and extensions to the pathways for biosynthesis of RNA (a single, universally conserved enzyme called Ribonucleotide Reductase converts all the monomers of RNA into DNA), that RNA is central to translation (mRNA, rRNA, tRNA), and a number of curious RNA-based coenzymes of very ancient origin, all of them involved in ancient and central metabolic reactions. Heck, ATP is a monomer of RNA, not DNA. All cellular life produces and can use ATP.
So there was most definitely a period in the early history of cellular life, when DNA synthesis had not yet evolved, but cells had RNA-based genomes instead of DNA. How that RNA world itself came about is unknown and subject to much speculation.
Adding a vote to the above – it doesn’t seem likely that transposons are relics of the RNA world, given the mechanism for their transmission primarily involves proteins. The case for self-splicing introns is less clear – they all do use/are ribozymes, and fairly sophisticated ones at that, so it’s possible they originated in the RNA world. But unlike the ribosome or RNAse P, they’re not universal in modern life, nor is their current activity essential like those others.
TomMueller,
Sure, I’ll buy genetic continuity of the ribosome to pre-protein-enzyme world. The rest, however … I’m sure RNA world would have had selfish RNA. But I’m less convinced that modern selfish elements are (necessarily) descendants of it.
david,
They seem localised to protein coding genes, however …
John Harshman,
Hi John,
Regarding my original contention about inversion vs. insertion.
My bad…Although technically speaking, regardless of whether one invokes the replicative or the conservative model of transposition, both amount to the same thing.
Your response:
They do not. What?
I have gone out of my way to be polite with you but it appears to me you may be recommencing your same rudeness as on sandwalk.blogspot. I hope I am mistaken, and we have “buried the hatchet”.
To answer your question. Addressing the phenomenon of transposition almost requires a flow-chart: Are we discussing Class I vs Class II Transposition? Are we considering a conservative vs. replicative model? Does it matter?
Let’s cut to the chase, from a transposon’s POV, insertion, inversion and deletion are equivalent events. Starting with last two: recombination between two transposable elements in the same orientation on the same chromosome leads to a deletion, whereas it results in an inversion if they are in opposite orientations. This is boilerplate…
Everything just said regarding inversion & deletion applies not just to recombination but also to transposition when different LTRs of different transposons are employed in the single transposition event.
I stand by my original contention: insertion, inversion and deletion are ALL equivalent events. Ditto tandem repeats.
Just the same, I should have identified the first half of “indel” as insertion and not inversion and for that I thank you.
Best regards
Rumraket,
Hi Rumraket
Re:
I think the evidence and arguments in favor of there having once been an RNA-world where the genetic material was RNA, rather than DNA, is strong and convincing. But I don’t think transposons or retroelements are arguments or evidence in addition to that.
I cite David’s mention of “self-splicing introns” and add thereto telomeres. I may be getting this wrong, but is there not a ribozymal component to telomerase? I am fascinated by the unstable stem-loop structures of the ribozyme story that seem to be mirrored in the DNA sequences of retroelements aka mobile elements are themselves inverted repeats, or flanked by LTRs which are inverted repeats where sometimes the inverted repeats have within nested inverted repeats. (long story there but let’s move on…) All these inverted repeats provide the stem-loop substrate for nucleic acid instability harking back to ancient RNA-World sequences’ autocatalysis and ribozyme function… or so I speculate.
I concede that at this point, I have crossed the line into speculation, explaining why I floated this trial balloon. In my defense, I would like to cite RNAi and telomeres in addition to introns… for starters.
Postscript – introns do in fact occur in two domains of life. Is it possible the intron-free status of Eubacteria is derivative?
david,
Allan Miller,
Hi David, Hi Allan
Re: David’s Adding a vote to [Allan’s] the above – it doesn’t seem likely that transposons are relics of the RNA world, given the mechanism for their transmission primarily involves proteins.
I think the invocation of protein a red herring to this discussion. We all agree that Ribosome as Ribozyme as relic of the RNA world seems reasonable enough. If I understand your line of logical reasoning correctly, then the necessary role of Ribosomal proteins in modern ribosomes would render even those lines of conjecture invalid… let’s just leave Protein (an evolutionary afterthought) out of this.
Meanwhile – another instance of stem-loop structure that has me scratching my head would be bacterial riboswitches, again an instance of those same sort of stem-loop structures that seem to herald back to an autocatalytic RNA world and again seem remarkably analogous to the stem-loop structures of retroelements. I don’t think I am guilty of circular reasoning, correct me if I am wrong. I just begin to wonder if so much smoke indicates some underlying fire.
TomMueller,
We can’t really ‘leave protein out of it’ if the issue is whether a particular element’s ancestry predates or postdates its origin! There is no conflict between ribosomal protein and the ancient (pre-protein) status of the ribosome. Logically, one is pretty much compelled to see the ribosome as predating ribosomal synthesis, at least by an hour or so! The sequence of ribosomal proteins is held in DNA, translated by ribosomes. It would appear that there has been coevolution between the products of primitive ribosomes and the ribosomes themselves, and several proteins are now essential.
There is no such logical compulsion in the case of other ribozymes.
TomMueller,
Sure, but loops will form pretty readily whenever there is bidirectional complementarity away from a centre (which will become the centre of the hairpin). You can get some complementarity stochastically, but you can get it readily by translocation of one strand’s antisense sequence onto the same strand. IOW, novel loops should not be too hard to generate. This may be simply a common structure readily adopted by the nucleic acids, rather than indicating ancient origin.
Hi Allan,
I agree with you on all counts. The modern ribosome as a ribozyme-protein hybrid as a product of co-evolution is exactly what I was thinking about for RNA mediated transposition.
Those novel loops are not incidental but essential to transposition. The question really boils down to whether or not those novel loops in transposition are substrates for RNA catalysis (at least in Class I transposons and associated retrovirus) and whether any such ribozymal function could then be attributed to ancient RNA ancestry… two big stretches, and I concede as much from the outset.
But wouldn’t that make a real neat story to tell our students! (OK – John Harshman’s accusation of “hubris” on my part may be on target.)
I am beginning to suspect (from my limited reading) that the role of RNA catalysis is emerging as a bigger player than ever suspected. And the story is getting more and more complicated, check out this alphabet soup of catalytic RNA:
http://www.nature.com/nature/journal/v480/n7376/full/nature10672.html
So I ask again, would any transposition of Class I transposons catalyzed by its own RNA also suggest the involvement of protein really was an evolutionary after thought, no differently than your ribosome? ditto intron splicing and telomerase activity? … and I certain this list is far from complete. At this point, I concede I am in far over my head.
So, before digressing onto riboswitches, the high priority question remains outstanding: Is transposition in RNA Class I retroposons dependent upon RNA catalysis? If so what does that mean and how far can we push that particular envelope?
This is where I could use some help… Maybe I misunderstood what you meant by “orphan”.
I believe they’re found frequently in rRNA and tRNA genes.
Both examples only use RNA as a guide, either to bind to a target RNA by base-pairing, or template DNA synthesis. No ribozymes involved. Tom Cech (and others) have argued these might represent hold-overs from the end of the RNA world, when proteins were prevalent but RNA was still around. I’m skeptical, because (1) the RNA in these enzymes is serving a simple task to which it is suited, evolution would favor its use even in a protein world (contrast this to the ribosome) and (2) phylogeny doesn’t support these enzymes/systems being as ancient as LUCA or serving a critical role in very early life.
Including both group I or II introns (which use different ribozymes), you can find some forms of self splicing intron in every domain of life. How far back they go is hard to tell, I don’t know what the current state of knowledge is regarding intron ancient phylogeny. Keep in mind that introns are mobile genetic elements, plus the bacterial ancestors of mitochondria probably carried them, so HGT will be prevalent.
The relative importance of the proteins involved is key – the ribosome is pretty clearly a ribozyme decorated with proteins. Many of the other examples we’re discussing the RNA is playing a passive role, mostly used as primer/template for proteins or binding to them. And unlike the ribosome, its not clear why these systems had to have evolved before protein existed, or that they did in fact evolve very early in life.
Stem-loop structures are incredibly easy to evolve. In fact, if you write a random RNA sequence >20 nt long and fold it at the mfold website, you’ll most likely find at least one stem-loop. These are just too common to indicate any tie to the RNA world, plus any which were that ancient would have to be under incredibly strong and constant selective pressure to remain recognizably similar to the forebears.
RNA didn’t stop evolving at the end of the RNA world, and this includes catalytic activity. Without a sophisticated ribozyme which functionally could be, but is not, replaced by a protein, plus strong evidence from phylogeny or biological necessity, I wouldn’t conclude something that involves RNA came from the RNA world. Some cases like self-splicing introns are borderline – sophisticated ribozyme, maybe pre-LUCA but not clear, could have basically done the same thing in an RNA world as it does now, but also easily could have evolved later. If there was evidence of telomerases where the RNA performed polymerization, or if a class I TE replicated itself without proteins (or where the proteins only helped structurally), that would be pretty interesting. It would still be odd given the lack of evidence that these systems evolved pre-LUCA, but I’d be inclined to believe they were derived from the RNA world. Let’s note I’m unaware of any evidence along these lines, and would be very surprised if there were.
I find the arguments by Nick Lane, that transposons, introns and other retroelements, are mostly a product of eukaryotes higher energy budget pr gene, persuasive.
If self-splicing introns really do derive from an RNA-world, my guess is that they have a viral origin. Introns and transposons are usually a form of junk-DNA and that would incur a non-negligible cost on prokaryote replication which is in my understanding thought to be extremely competitive. For those reasons it seems unlikely to me that retroelements would survive and proliferate as internal genetic parasites in an RNA-world.
But putting their persistence as a product of viruses deals with this problem. IIRC Eugene Koonin has made arguments to that effect too.
david,
OK, they interrupt transcripts involved in protein coding, then!
Worth a mention that Alu is derived from 7SL rRNA, and is taxonomically restricted. ie, at least some transposons clearly postdate RNA world.
Let’s be clear here,
Of course many retroelelements emerged post-RNA world. I never suggested otherwise. If my scribbles were unclear on that score, I apologize.
My contention is somewhat more nuanced:
I was wondering out loud whether the mechanisms of Class I transposition (and all its derivatives) indeed had their evolutionary origins in a RNA-world, even if the role of putative role of RNA as ribozyme has since become diminished and almost always eclipsed by protein.
Or as David phrased it better than I:
”Tom Cech (and others) have argued these might represent hold-overs from the end of the RNA world, when proteins were prevalent but RNA was still around.”
I love David’s metaphor of
… the ribosome is pretty clearly a ribozyme decorated with proteins. Many of the other examples we’re discussing the RNA is playing a passive role,
… not to mention, in the case of Class II transposons, RNA plays no more role at all anymore, presuming Class II indeed owes its origins to Class I.
The problem then becomes, as David again elucidates better than I ever could, even if that were true, whether or not we would be still able to detect any such a connection this late on in the story of eukaryotic evolution? … because
… RNA didn’t stop evolving at the end of the RNA world… plus any which were that ancient would have to be under incredibly strong and constant selective pressure to remain recognizably similar to the forebears… meanwhile Stem-loop structures are incredibly easy to evolve.
You have all provided me much to ponder and I thank you all and remain in your debt.
There is much more I need to read up on before revisiting this question. I intend to examine in more detail the telomerase story. ITMT – this blogsite is outstanding!
Best regards to one and all.
TomMueller,
Two steps have been added if so – transcription and reverse transcription! 😉
Of course, in RNA world (with, I reckon, a ds genome from the outset), transcription and reverse transcription’s analogues would both be simply template-directed RNA polymerisation. Depending on the order of acquisition of DNA and protein coding, ‘modern-style’ Class 1 mechanisms may conceivably have been available before protein. But either way, the modern version seems to piggy-back on the DNA-RNA shuttle.
I’m pretty sure RNA genomes had selfish elements, nonetheless.