Chemical Forums
Specialty Chemistry Forums => Biochemistry and Chemical Biology Forum => Topic started by: curiouscat on June 29, 2016, 04:10:26 AM
-
I am planning on expressing a recombinant plant protein in Yeast using a synthetic gene (approx 1700 bp) that is made from a codon optimized DNA sequence. The vector is a pUC57 plasmid.
How does one go about selecting the best promoter to maximize expression levels of the protein? Any heuristics / guidelines?
-
PMC4128558 "Different Expression Systems for Production of Recombinant Proteins in Saccharomyces cerevisiae"
I have never worked with yeast systems. However, the paper above does discuss promoters, among other factors.
-
PMC4128558 "Different Expression Systems for Production of Recombinant Proteins in Saccharomyces cerevisiae"
I have never worked with yeast systems. However, the paper above does discuss promoters, among other factors.
Thanks @Babcock. Which systems have you worked with? E Coli?
I suppose promoter choice matters in almost all hosts?
-
Yes, I have always expressed in E. coli. Indeed, the promoter is a key issue. In my experience and in talking informally with colleagues, sometimes one does not see good expression even with a strong promoter, and tracking down the problem can be difficult. Since the time I did any cloning myself, people have been paying greater attention to abundance of certain tRNAs versus codon usage (which is not to say that it was completely ignored before).
-
Since the time I did any cloning myself, people have been paying greater attention to abundance of certain tRNAs versus codon usage (which is not to say that it was completely ignored before).
Isn't that what's called "codon optimization"? Or are you referring to something different?
To me, of all the things one can do for protein expression improvement, codon optimization seems the part I understand to some extent.
The stuff that I'm having a hard time understanding are how exactly to exploit the other two strategies i.e. (a) Promoter choice & (b) Plasmid copy number
Those both seem like an arcane art.
-
https://www.emdmillipore.com/US/en/product/Rosetta™%28DE3%29-Competent-Cells---Novagen,EMD_BIO-70954
Everything I know comes from E. coli, but some of the ideas may carry over to yeast. Some strains of E. coli have additional tRNA molecules for codons that are rare in E. coli but common in eukaryotes (Rosetta, see above). Before such strains were available, the only way I can think of to deal with a rare codon would be to perform a silent mutation on the gene. With respect to promoter choices, the need for a strong promoter is obvious, but it should also be a promoter that can cleanly be turned on and off, meaning that expression should not be leaky in the absence of an inducer. That is partially because the protein might be toxic to the host. In E. coli this sometimes means overproducing a repressor protein, such as the gene product of lac i. I am not sure how to control plasmid copy number, although I believe that there is a relationship between copy number and levels of expression (not necessarily a linear one). I can waive my hands and argue that if the plasmid copy number were too high, it might hurt one's ability to turn off expression in the absence of an inducer (see above).
This might be a good place to start:
http://blog.addgene.org/plasmids-101-the-promoter-region
-
Why are you using yeast as an expression system instead of bacteria?
-
Why are you using yeast as an expression system instead of bacteria?
Why are you using yeast as an expression system instead of bacteria?
Actually the choice isn't fixed yet. The plan is to measure expression levels in both yeast & E Coli and then decide.
Some points in favor of Yeast:
(a) The immediate precursor molecule is Farnesyl Pyro Phosphate (FPP) which gets produced via the Mevolonate Pathway that's present in yeast natively but not in E Coli. E Coli do produce FPP but via the MEP pathway which so far as I understand is not as productive as the pathway in yeast
(b) One commercially scaled up molecule Artemesinin has also this FPP precursor in common and there the successful process is yeast based. So there's reason to think that on-net yeast has proven more viable. But this is just a crude analogy.
PS. Maybe I misunderstood your question. By "bacteria" did you have something other than E Coli in mind? Which ones? I could have missed an obvious choice.
-
https://www.emdmillipore.com/US/en/product/Rosetta™%28DE3%29-Competent-Cells---Novagen,EMD_BIO-70954
Before such strains were available, the only way I can think of to deal with a rare codon would be to perform a silent mutation on the gene.
I think the approach I am thinking of (which I think obviates this problem but I could be wrong) is to get the entire natural gene de novo synthesized but with a pre-processing step on the sequence that codon optimizes for the expression host's tRNA distribution.
That way a rare codon is swapped out for the better alternative codon specific to the host organism.
Does that make sense?
-
Why are you using yeast as an expression system instead of bacteria?
Why are you using yeast as an expression system instead of bacteria?
Actually the choice isn't fixed yet. The plan is to measure expression levels in both yeast & E Coli and then decide.
Some points in favor of Yeast:
(a) The immediate precursor molecule is Farnesyl Pyro Phosphate (FPP) which gets produced via the Mevolonate Pathway that's present in yeast natively but not in E Coli. E Coli do produce FPP but via the MEP pathway which so far as I understand is not as productive as the pathway in yeast
(b) One commercially scaled up molecule Artemesinin has also this FPP precursor in common and there the successful process is yeast based. So there's reason to think that on-net yeast has proven more viable. But this is just a crude analogy.
PS. Maybe I misunderstood your question. By "bacteria" did you have something other than E Coli in mind? Which ones? I could have missed an obvious choice.
Ok, so will the application will be in small molecule biosynthesis rather than in producing large amounts of the protein for purification?
By bacteria, I did mean E. coli. E. coli are much more convenient for producing and purifying large amounts of recombinant protein, but for biosynthetic applications, yeast may indeed be better.
-
Some of your terminology ( pre-processing step) is unfamiliar to me. However, if you meant that a synthetic gene will be constructed with codons that are not rare in the host, then that sounds like a reasonable approach. Just for fun I looked up the over expression of farnesyltransferase by the Casey laboratory (doi: 10.1073/pnas.241407898), and I found this: "As in the GGTase-I expression construct, the first 13-aa codons of the α subunit were changed to improve translation and maintain the natural amino acid sequence." What I was told long ago was that rare codons near the N-terminus could hurt expression, and this passage seems to be consistent with that.
-
Some of your terminology ( pre-processing step) is unfamiliar to me. However, if you meant that a synthetic gene will be constructed with codons that are not rare in the host, then that sounds like a reasonable approach.
Artificial gene synthesis and codon optimization are quite common now for protein expression studies (e.g. see http://www.genscript.com/codon-opt.html). I have heard conflicting reports of how helpful it is versus just using E. coli strains supplemented with tRNAs for rare codons (e.g. the rosetta strain you mentioned).
What I was told long ago was that rare codons near the N-terminus could hurt expression, and this passage seems to be consistent with that.
There are some reports that this has more to do with mRNA secondary structure than rare codon usage: http://science.sciencemag.org/content/342/6157/475
-
Ok, so will the application will be in small molecule biosynthesis rather than in producing large amounts of the protein for purification?
Yes. Right.
Although the option to immobilize the enzyme and use it to process synthetic substrate (FPP) remains open. That would then involve protein synthesis.
Basically whatever option maximizes end product yield and we won't know until we try.
MW 250 is a "small" biomolecule, right?
By bacteria, I did mean E. coli. E. coli are much more convenient for producing and purifying large amounts of recombinant protein, but for biosynthetic applications, yeast may indeed be better.
Interesting. Can you elaborate on why E Coli are better for one job and yeast for the other, in general?
-
Interesting. Can you elaborate on why E Coli are better for one job and yeast for the other, in general?
For recombinant protein expression and purification E. coli are more convenient because they grow faster and you can generally get very high yields of protein from them. Plus, the field has much more experience using E. coli for protein expression that there are many more resources available to help optimize expression (e.g. specialized strains or plasmids to help with rare tRNA supplementation, folding, etc.).
I don't have much experience with biosynthesis, though I see a lot of papers using yeast for these purposes. Maybe this is because people have accumulated more experience working with yeast for industrial purposes (e.g. brewing) than bacteria?
-
This may be tangential to your application, but to overproduce a small molecule (and biochemists will call a molecule with MW = 250 a small molecule) you may need to consider the ideas of metabolic control analysis. Metabolic control analysis looks at the flux through a biochemical pathway, almost from an engineering perspective. Just doing a very cursory search, I found this: "Increasing the Flux in Metabolic Pathways: A Metabolic Control Analysis Perspective" by David A. Fell.
BIOTECHNOLOGY AND BIOENGINEERING, VOL. 58, NOS.2&3,APRIL 20/MAY 5, 1998
-
This may be tangential to your application, but to overproduce a small molecule (and biochemists will call a molecule with MW = 250 a small molecule) you may need to consider the ideas of metabolic control analysis. Metabolic control analysis looks at the flux through a biochemical pathway, almost from an engineering perspective. Just doing a very cursory search, I found this: "Increasing the Flux in Metabolic Pathways: A Metabolic Control Analysis Perspective" by David A. Fell.
BIOTECHNOLOGY AND BIOENGINEERING, VOL. 58, NOS.2&3,APRIL 20/MAY 5, 1998
Thanks @Babcock!
This is *exactly* what I had wanted to do. But I felt totally inadequate regarding the tools needed to model (say) a yeast network. What I'd love to do is to predict, at least order of magnitude, what sort of expression levels I might expect even before I start any experiments.
e.g. FPP goes to squalene and that will be a major drain on the precursor pool. I'd love to model, for example, how much improvement in flux I can get out of blocking squalene synthase.
I'll read that paper you linked to.
Meanwhile if someone knows of any good tools to do this sort of modelling in I'd love to hear.
-
Here's a review that might be useful to take a look at:
http://www.sciencedirect.com/science/article/pii/S0958166913000554
For example, it has a few paragraphs discussing how to optimize tune protein expression levels.