More Recent Comments

Sunday, January 27, 2019

Yeast loses its introns

Baker’s yeast (Saccharomyces cerevisiae) is one of the best studied eukaryotes. Its genome is just slightly larger than the largest bacterial genome and it was the first eukaryotic genome to be sequenced (Mewes at al., 1997). It has about 7000 genes in total and 6,604 of these genes are protein-coding genes but only 280 of these genes contain introns.1 The rest have lost their introns over the course of several hundred million years of evolution (Hooks et al., 2014).

We know that introns have been lost in yeast because the genes of related species have lots of introns. The common ancestor of all fungi undoubtedly had genes with multiple introns because the available evidence indicates that introns invaded eukarotic genes very early in the evolution of eukaryotes. The fact that most introns have been purged from the yeast genome suggests that introns are not essential for gene function. In other words, introns are mostly junk.2

What about the remaining introns? There’s an ongoing attempt to build a completely synthetic yeast genome by synthesizing artificial chromosomes and part of the process is to eliminate all the junk DNA by cutting out all unnecessary introns (Richardson et al. 2017). So far it looks like most of the introns can be removed without any obvious effect on the viability of the cells but there are a few introns that are essential and others that affect growth under certain conditions (Parenteau et al., 2008; Parenteau et al., 2011). Some of the essential introns contain genes for small noncoding RNAs (e.g. snoRNA) and that’s why they can’t be deleted but the role of other essential introns is still a mystery. The nonessential introns that affect growth appear to be required to regulate growth under starvation conditions (Morgan et al., 2019; Parenteau et al., 2019).

If we assume that the common ancestor of yeast and other fungi had three introns in every protein-coding gene then there would have been about 18,000 introns in the common ancestor. Most of them (>98%) have been eliminated by evolution without serious consequences but these results indicate that some of them have secondarily acquired an important role in gene expression.

I'm writing this post because of two recent Nature papers on the function of some of the remaining introns in yeast genes (Morgan et al., 2019; Parenteau et al., 2019). The important lesson is that >98% of the ancestral yeast introns have been eliminated by evolution, which is clear evidence that the vast majority of introns are junk. They have no function. The Nature papers confirm a previous observation that a subset (~30) of the remaining introns in yeast produce an RNA that confers a growth advantage on yeast cells under limiting growth conditions. This result has been widely misinterpreted by the legitimate media and by creationists who claim that it cast doubt on junk DNA.

Here's a video from the Nature website that ignores the big picture (introns are junk) and gives the impression that while the role of introns is a mystery there are, nevertheless, several adaptationist explanations that may account for their existence. The Nature video focuses on the new function of a small number of yeast introns and suggests that it can account for the prevalence of introns in humans and the incredible complexity of life. I contacted the author of the video and he declared that he was not an expert on the subject but simply reporting on what the two Nature articles said. When I asked him where he got the information for context—information that was not in the papers—he declined to comment. He also did not respond to my request to discuss this with me on my blog. That's too bad because I think it's important to try and understand how misinformation gets propagated by one of the leading science journals in the world.

Here's a link to an Intelligent Design Creationist blog post: As Predicted by Intelligent Design, “Junk” Introns Are Actually Functional. Again, the author of this blog post ignores the big picture; namely, that 98% of the yeast introns have been removed without affecting the viability of the cells. Instead, the anonymous author says,
ID proponents have long predicted that functions would be uncovered for such non-coding DNA. In The Myth of Junk DNA, Jonathan Wells reviews various functions discovered for introns. He points out that introns play vital roles in alternative splicing, where exons of a single gene can be mixed and matched such that one gene can give rise to many different proteins ...

Wells cites many studies that provide strong evidence that introns are involved in alternative splicing, helping to create diverse proteins required by cells.

It’s very difficult to argue that functions for introns are just anomalies in a sea of junk. These new Nature papers further confirm that this is the case.
Actually, the Nature papers strongly imply that functions for introns are just anomalies in a sea of junk but you have to understand the big picture in order to see this!

Image credit: Differential interference contrast image of yeast cells from Wikipedia.

1. Most genes have only one intron but some have two for a total of 295 introns in the yeast genome.

2. Introns take up about 23% of the human genome. If they are mostly junk as all the evidence suggests then this is a significant contribution to the total amount of junk DNA in our genome. The fact that some organisms can get along without most of their introns (e.g. yeast) plus the fact that some species have much smaller introns (e.g. pufferfish) strongly supports the idea that introns are mostly junk.

Hooks, K.B., Delneri, D., and Griffiths-Jones, S. (2014) Intron Evolution in Saccharomycetaceae. Genome Biology and Evolution 6:2543-2556. [doi: 10.1093/gbe/evu196]

Mewes, H., Albermann, K., Bähr, M., Frishman, D., Gleissner, A., Hani, J., Heumann, K., Kleine, K., Maierl, A., and Oliver, S. (1997) Overview of the yeast genome. Nature, 387:7-8. [doi: 10.1038/42755]

Morgan, J.T., Fink, G.R., and Bartel, D.P. (2019) Excised linear introns regulate growth in yeast. Nature [doi: 10.1038/s41586-018-0828-1]

Parenteau, J., Durand, M., Véronneau, S., Lacombe, A.-A., Morin, G., Guérin, V., Cecez, B., Gervais-Bird, J., Koh, C.-S., and Brunelle, D. (2008) Deletion of many yeast introns reveals a minority of genes that require splicing for function. Molecular biology of the cell, 19:1932-1941. [doi: 10.1091/mbc.E07-12-1254]

Parenteau, J., Durand, M., Morin, G., Gagnon, J., Lucier, J.-F., Wellinger, R.J., Chabot, B., and Elela, S.A. (2011) Introns within ribosomal protein genes regulate the production and function of yeast ribosomes. Cell, 147:320-331. [doi: 10.1016/j.cell.2011.08.044]

Parenteau, J., Maignon, L., Berthoumieux, M., Catala, M., Gagnon, V., and Elela, S.A. (2019) Introns are mediators of cell response to starvation. Nature. [doi: 10.1038/s41586-018-0859-1]

Richardson, S.M., Mitchell, L.A., Stracquadanio, G., Yang, K., Dymond, J.S., DiCarlo, J.E., Lee, D., Huang, C.L. V., Chandrasegaran, S., and Cai, Y. (2017) Design of a synthetic yeast genome. Science, 355:1040-1044. [doi: 10.1126/science.aaf4557]


Mikkel Rumraket Rasmussen said...
This comment has been removed by the author.
Georgi Marinov said...

It's even worse than that for the "introns are hugely importat" crowd -- people have looked at intron positions across eukaryotes as a whole, and it appears that the LECA was actually very intron rich (because you see a lot of those introns found in the same positions across the whole eukaryotic tree).

But within each eukaryotic supergroup there are all sorts of lineages where introns have been more or less eliminated (almost completely in some).

It's not just yeast that has lost them

And, of course, if LECA was intron rich, given that LECA was a "primitive" single-celled eukaryote, there is a lot of explaining to be done if the claim that having a lot of introns is necessary for or a hallmark of organismal complexity is to be defended

Larry Moran said...

Did you read the papers? I think they did the experiment.

The Other Jim said...

Can you post a citation for this work? I teach a 2-hour lecture in a course where I discuss genome structure and these sorts of issues. A reference for this would be great!

SPARC said...

I wonder if anybody ever checked the available databases but if there really was a significant number of functional sequences in junk DNA why aren’t there tons of them listed in OMIM. And why didn’t ENU mutagenesis, gene trapping and transposon mutagenesis in mice didn’t result in thousands of phenotypically visible consequneces resulting from mutations in junk DNA?

Georgi Marinov said...

There are more papers than these, but this is copy-paste from something I wrote once upon a time:

Fedorov A, Feisal Merican A, Gilbert W. 2002. Large-scale comparison of intron positions among aminal, plant, and fungal genes. \textit{Proceedings of the National Academy of Sciences}. \textbf{99:}16128-16133.

Rogozin IB, Wolf YI, Sorokin AV, Mirkin BG, Koonin EV. 2003. Remarkable Interkingdom conservation of intron positions and massive, lineage-specific intron loss and gain in eukaryotic evolution. \textit{Current Biology}. \textbf{13}:1512-1517.

Roy SW. 2006. Intron-rich ancestors. \textit{Trends in Genetics} \textbf{22}:468--471.

Csuros M, Rogozin IB, Koonin EV. 2011. A Detailed History of Intron-rich Eukaryotic Ancestors Inferred from a Global Survey of 100 Complete Genomes. \textit{PLoS Computational Biology} \textbf{7}(9):e1002150

Georgi Marinov said...

Most GWAS hits are in non-coding space though. Not in proper "junk DNA", obviously, but it appears that regulatory elements are quite important overall (if not necessarily critically important individually)

SPARC said...

If such hits were of greater relevance wouldn't one have to expect them to be discovered more frequnetly in ENU screens which actually don't look for lethal phenotypes but screen for things like gradual hearing loss, impaired vison, dismorphologies etc.
E.g., screens done by the German Mouse Clinc

lantog said...

I'll be interested to see what the mechanism is for the growth advantage. I suspect that the introns are not providing some extra function that protects cells. Its more the case that removing an intron throws off the regulation of a protein which changes some genetic network. If thats the case then hitting the intronless yeast with a mutagen might restore good growth with a point mutation or 2 that tweaks the network

Unknown said...

I think your discussion shows a lack of appreciation of the power of natural selection for organisms with such immense effective population sizes as yeast. Tiny selective pressures can be effective. I suspect that the experimental tests for neutrality here have explored only a tiny fraction of the range of substrates, temperatures, light, etc that yeast might experience in the wild. Probably far less than 2% of the ecological space has been tested, yet 2% of the introns have a detectable effect even in that minuscule laboratory exploration of the ecological space. I think it is premature to say definitely that 98% of the introns are junk. I think it will be very hard, in principle, to pin down the correct number.

You've always been careful to consider the rate of false positives in experiments that report evidence for function. You should be equally careful to consider the rate of false negatives, which could be quite high, because even tiny selective effects could be important when effective population size is huge.
L. Jost

John Harshman said...

I think you have mistaken the point. It isn't that 98% of yeast introns are junk. It's that 98% of the introns have been lost. You can't test them in different situations because they aren't there at all. Clearly, if they were lost they couldn't have been doing anything that subjected them to positive selection. My suspicion is that they were slightly deleterious, perhaps because of selection favoring faster replication, as in bacteria. But they could as easily have been lost through completely neutral evolution.

Joe Felsenstein said...

I'd say that "they can't have been doing anything that subjected them to purifying selection".

Unknown said...

Yes John, you are right, I misread the post! Apologies.
L. Jost

Georgi Marinov said...

Somewhat related:

"Individual long non-coding RNAs have no overt functions in zebrafish embryogenesis, viability and fertility"

SPARC said...

Thanks for the link. One may wonder if editors and reviewers of other journals would have been as cautious regarding the original title of the article and its original conclusions as eLife (you'll find the decision letter and the author response at the end of the page):

1) The title is not fully supported by the data. "Largely dispensable" implies that some phenotypes have been observed. Also, the extrapolation to all lncRNAs suggested in the title is overreaching based on 25 genes. A less general title, e.g. "Zebrafish embryogenesis, viability and fertility are not overtly affected by loss of embryonically expressed long non-coding RNAs", would be more appropriate.

Marcoli said...

There is this other article that claims a function for a small class of what some apparently have called junk DNA: The authors say that pericentromeric satellite DNA is found to be essential for bundling chromosomes into a heterochromatic structure called the chromocenter in mice and in Drosophila, and that not doing so leads to cell death.
They say that pericentromeric satellite DNA was considered 'junk', but here it is found to not be. I am not sure if this DNA was ever seriously considered to be junk.

Larry Moran said...

No knowledgeable scientist ever claimed that centromeres were junk DNA. In fact, they make up one of the largest components of functional DNA in our genome.

What's In Your Genome? - The Pie Chart

Centromeres contain a lot of DNA and not all of it is essential for survival. The core sequence consists of 171 bp. repeats called alpha-satellite DNA surrounded by other satellite DNA such as AT repeats. There's lots of evidence that these AT repeats have some sort of function but the exact role has always been unclear. This paper provides evidence that there are proteins that bind to the pericentric satellite DNA and assist in packaging.

Marcoli said...

I only have the abstract to go on. There they say: "Here, we provide evidence that pericentromeric satellite DNA, which is often regarded as junk, is a critical constituent of the chromosome..." I don't know if that DNA is considered part of the centromere proper or is peripheral to it.

Mikkel Rumraket Rasmussen said...

By now I think it should become common practice for reviewers to ask for a citation if authors declare that their analyzed locus was "previously dismissed as junk", and if the authors can't find a literature reference wherein someone predicts or dismisses the particular locus as junk, then the authors should be required to remove the sentence.

Donald Forsdyke said...

Larry writes: "The fact that most introns have been purged from the yeast genome suggests that introns are not essential for gene function. In other words, introns are mostly junk." However, the fact that many introns can be dispensed with merely suggests, as indicated above, that the loss favours something else, such as rapid replication. This observation greatly focuses attention on the fewer genes that have, despite the intense pressure, retained their introns. Is there some condition under which the utility of the surviving introns can be demonstrated?

A good first start, as in the discussed papers, is to study the stationary growth phase when one would expect SOS signals for expediting a return to exponential growth. Indeed, the authors demonstrate a general utility of surviving introns that is not related to host gene functions.

This reminds one of the early ideas of Darryl Reanney who suggested an SOS invocation of recombination repair mechanisms that would benefit by having structural features in introns that might have been less accommodated by exons. It would be interesting to see if there is anything particular about the location of the surviving introns that might facilitate their general, non-gene-specific, function.

Larry Moran said...

If you read the papers on the stable introns you'll see that they bind to spliceosomes and slow down RNA processing. That's presumably how they confer a selective advantage under low growth conditions. It doesn't matter which introns are retained in various species as long as there's a sufficient number of them.

Several of the essential introns contain genes for various RNAs (e.g. snoRNA) and that's why they weren't lost.

I stand by my statement that the loss of introns in yeast strongly suggests that they are dispensable (i.e. junk). If fact, they are probably slightly deleterious but they are only going to be selected against in large populations.

Larry Moran said...

It seems to me that most reviewers don't care what you write in the introduction and the conclusion.

A Salty Scientist said...

I think the concerted loss of ~95% of introns in yeast pretty strongly suggests that most are slightly deleterious. Half of the remaining introns are in ribosomal protein encoding genes, and my guess is that these are maintained either due to a positive effect on fitness during stress (by slowing RP expression), or due to regulatory constraint (losing some introns may lead to unbalanced expression across RP genes).

Joe Felsenstein said...
This comment has been removed by the author.
Joe Felsenstein said...

Of course an intron being slightly deleterious when lost does not mean that all of its sites are nonneutral.

A Salty Scientist said...

Oh sure, the sites necessary for splicing are decidedly non-neutral, and it should not have been surprising to see them show up in disease GWAS. I imagine that perfect or near perfect intron deletion is relatively rare, which could explain why it was able to happen in something like yeast with a large population size, short generation time, and the ability to self (maintaining homozygosity).

A Salty Scientist said...

I misread your sentence, Joe. My hypothesis is that the majority of introns are slightly deleterious (i.e. parasitic DNA), and that they are maintained because they are hard to get rid of without causing coding region deletion, frameshifts, etc.

Larry Moran said...

lantog, did you read the papers? Do you disagree with the mechanism they proposed?