More Recent Comments

Monday, April 13, 2009

A Breakthrough in Gene Expression?

When we teach protein synthesis in undergraduate molecule biology classes we cover the main mechanisms regulating the rate of translation.

One of them is the influence of codon bias among synonymous codons. We've known for 35 years that rare codons are translated more slowly that the common codons. Highly expressed genes have a pronounced codon bias in favor of the most common codons. As a result of this phenomenon, it is not true that every codon for leucine, for example, is equal. Some are better than others in some genes. Synonymous codons are not always neutral in their effect. (For a complete description of this phenomenon see: Silent Mutations and Neutral Theory.)

We also teach about the influence of messenger RNA secondary stucture. The classic examples in the E. coli ribosomal protein genes are in all the textbooks, as are the examples of attentuation—especially in the Trp operon. Again, this stuff was standard fair in textbooks and courses beginning in the 1970's.

A press release caught my eye: Penn biologists discover how 'silent' mutations influence protein production. "Cool," I thought, "maybe this is something that I'll have to put into the next edition of my textbook."

Here's the breakthrough.
For biologists, these results fundamentally change the understanding of the role of synonymous mutations, which were previously considered evolutionarily neutral. ....

The silent mutations changed the amount of fluorescent protein by as much as 250-fold, without changing the properties of the protein. Codon bias, the probability that one codon of three adjacent nucleotides will code for one amino acid over another, was previously thought to be the cause for protein expression variance, but it did not correlate with gene expression in these experiments.

"At first we were stumped," Plotkin said. "How were the silent mutations influencing protein levels? Eventually, we looked at mRNA structure and discovered that this was the underlying mechanism."
Imagine that. They've rediscovered what most of my students have been taught for 35 years!

[Image Credit: The figure is from page 706 of my textbook. Similar figures are in all biochemistry and molecular biology textbooks. The figure shows the secondary mRNA structure around the initiation codon of the S7 ribosomal protein gene in E. coli. The secondary structure inhibits translation initiation. Although in this case the actual codons are not involved in the formation of double-stranded regions, in other cases they are.]


  1. It does sound like a fairly mundane result from the media report. I've yet to fully read the article (hopefully in a journal club soon - the NMD and long mRNA-like ncRNA paper comes first though) I think the exciting thing is that they've managed to quantify the influence of secondary structure vs codon-usage which I think is an important result.

  2. Sheesh, anyone who does a lot of protein overexpression already knows it very well. It's simply a common knowledge among protein crystallographers. That's why synthetic gene services are becoming so popular - more often than not, simply changing bunch of codons almost randomly greatly boosts expression of proteins that can't otherwise express decently (even with the rare tRNA co-expression).

  3. They've rediscovered what most of my students have been taught for 35 years!

    Yes this information has been around for decades. Indeed, in my class discussing allelic variations that are silent in a survey of gene sequences, we consider this as a reason why these "silent" mutations may not be neutral. However, the 96 well plate showing variations in GFP fluorescence makes a much stronger case than for the effects in rare codons on translation efficiency than anything previous I am aware of.

    If the goal is to effectively teach the concept, I think you would be wise to use this work. If the goal is to teach the history of molecular biology, then I guess you should stick with what you have. Now I guess Ill get the hell off your lawn.

  4. I just received this paper in my e-alerts today, and my initial impression is that the press release was poorly researched, and even more poorly worded (this speaks directly to your original article about science journalism, Larry). The authors used this technique to investigate the precise mechanism by which synonymous mutations effect gene expression. In their analysis they show that the range of expression did not correlate with codon bias, as one might expect. It actually correlated more closely with mRNA stability near the ribosomal binding site. I do not think this had been previously verified, although I think it was posited.

    So as far as I can tell, the authors were under no delusions that they had uncovered a novel role for synonymous mutations. They were merely delving deeper into the mechanism by which the subset of mutations that alter expression work. The first sentence of the intro reads "The theory of codon bias posits that preferred codons correlate with the abundances of iso-accepting tRNAs and thereby increase translational efficiency and accuracy."

    Just my impressions from a quick scan anyway.

  5. The first sentence of the intro reads "The theory of codon bias posits that preferred codons correlate with the abundances of iso-accepting tRNAs and thereby increase translational efficiency and accuracy."The authors imply that this theory is wrong. They are wrong. There are plenty of examples where expression of a gene in E.coli was increased dramatically by supply of tRNAs. Yet, anyone who found better expression with synthetic gene as opposed to co-expression of native gene with rare tRNAs (and there are many) also knows that this is not the only factor.

    Still, upon reading the paper, I have to admit it's a decent work. While I do not think that it is groundbreaking (as per Science's supposed standards), the authors do show that predominant factor (at least 50% of the total) ifluencing expression level is simply a theoretical RNA structure near trasnlation initiation (see Fig.3C). And this is definitely a useful and new observation. At least for expression of GFP gene in E.coli... :-)

    To contrast and compare, for D.discoideum (which has exceptionally AT-rich genome), it was reported about 10 years ago that the main determinant of the expression level was Dicty-like codon usage in the first ~ 10 codons after ATG, the rest mattered much less.

    Also, it's not inconceivable that for other genes that have peculiar "problems" with NA sequence, other factors, like rate of transcription or RNA stability end up more important than the rate of translation initiation.

  6. "The authors imply that this theory is wrong. They are wrong."I didn't read it that way. I thought they were giving a nod to what was already known, and trying to contribute a bit of research to the pile. In any case, I agree with you that the paper is of limited scope - a fairly interesting result, but hardly revolutionary. The main point to all this remains - most journalists do not care about getting the scientific details correct, only about selling it as the next great thing. The press release was completely misleading and useless.

  7. I think I agree with Paul that this does seem significant. If you are expressing transgenes in various systems it may be important to know the different contributions of codon usage vs mRNA structure for optimal expression. A slight difference in expression level can be important in some cases. Or perhaps, studying aspects of translation, you want optimal translation initiation or perhaps you want optimal elongation. These would then be different mRNAs.
    Perhaps I'm biased but it even seems Science worthy. OK the press release was worded poorly.
    Are they going to test this library in a eukaryotic cell now and get another Science article?
    abstract on Pubmed