Have you checked for rare codon usage, too?

Have you checked for rare codon usage, too?

I will assume you are working on a bacterial host.

If at all possible, get a positive control (an unrelated gene cloned in the same plasmid, known to be well expressed) that you can transform and culture in parallel with your construct to discard culture conditions, faulty cell bank preparation, etc. as the source of your troubles.

Also, check for the expression of your gene by Western blotting. An absolute lack of signal despite reasonable sensitivity is sign of trouble; re-check your construct; ensure that it is in-frame with any foreign start codons or N-terminal extensions if those were used. Also, are you using the RBS provided by the vector? If not, check the construct using something like the RBS calculator (https://www.denovodna.com/software/).

As to your specific question yes, stem-loop structures such as those might act as rho-independent transcriptional terminators, and a protein translated from a prematurely terminated mRNA will not only be incomplete -and likely misfolded- but will be tagged for degradation by the SsrA–SmpB system. There are online termination predictors you can use, such as http://rssf.i2bc.paris-saclay.fr/toolbox/arnold/ or http://lin-group.cn/server/iTerm-PseKNC/ (these two use completely different algorithms) to screen for cryptic terminators in the sequence of your designed gene.

If nothing else works, the protein might not be folding properly and being degraded (Western blotting might give you hints there). In that case, you might have to resort to cell-free transcription/translation systems.

Hope it helps!

Dear Alejandro Martin,

Thank you for your detailed answer.

I am working on Baker's yeast as a host. I have integrated this construct into the genome and got the sequencing result of the plasmid, which is 100% correct. I have used the native promoter (i.e. from the S. cerevisiae genome) upstream of the artificial DNA sequence. I have also made another plasmid with the same promoter and its own gene [and both constructs are tagged with mCherry]. When I integrated the native one, I could see fluorescence signal (btw, the max. dyad symmetry level for it is -8.4). That's why I think my protein of interest might be not stable. It is a nice idea to check protein levels by Western Blotting as you suggested. I will try it to see whether the protein is degraded (I will detect it by anti-mCherry Ab).

I guess RBS calculator is to check expression in prokaryots. So, in my case it won't be helpful. But thank you for the link.

Do you know whether there is a way of enabling (i.e. forcing) yeast to properly fold the protein of interest in case there is a folding problem?

Cheers,

Arman

Hi Arman,

there is a way to screen various conditions for allowing optimum folding of your protein of interest. The REFOLD database (http://pford.info/refolddb/) and the iFOLD Protein Refolding System 2 could be interesting for you. You can check the following publication for details: Article Poly(ADP-Ribose) Polymerase Is a Substrate Recognized by Two...

I would also recommend using E. coli as an expression system.

The answer by Alejandro Martin is excellent. It can give you several ideas to check. Codon usage, as Adam B Shapiro pointed out, is also an important factor to consider.

Good luck!

Cheers, Christian

Following

Baker's yeast, oops! :-) Yes, those links were all for G- bacteria.

Make sure, this being an artificial gene and all, that you have not inadvertently introduced signals that make the cell recognize part of the mRNA as an intron. Also, remember that if the designed final structure of the protein is held together by disulfide bridging, the protein must be targeted to the secretory pathway. Codon usage is also important, as pointed out by Adam B Shapiro and Christian Q. Scheckhuber. Premature termination of translation due to rare codons is known to target the mRNA for degradation via NMD in mammalian cells (honestly, don't know in yeast).

Apparent lack of expression owing to misfolding is a hard problem to tackle. If this is a de novo design there is a chance that no matter how careful the design, folding dG will actually be positive, and in this case there will be nothing to do because you can't beat thermodynamics. Assuming this is not the case, there is literature on co-expression of chaperones, prolyl isomerases and the likes in yeast to enhance the production of recombinant proteins, but those tend to be protein-specific. It might be easier, as Christian suggests, to simply express the protein in a bacterial system and screen in vitro for refolding conditions.

By the way, if the control gene has a stem-loop with a free energy of -8.4 kcal/mol, doesn't this go against your hypothyesis that dyad symmetry motifs in the coding gene are affecting expression? After all, the most stable stem-loop of the artificial gene is only -7.8 kcal/mol, according to the table shown above.

Dear Christian Q. Scheckhuber ,

Thank you very much for your answer!

I have checked The Refold Database. However, my protein of interest doesn't appear in their system. I will check the article you have mentioned. Thank you for the link!

Indeed it would be nice to use E. coli, however I would like to show the functionality of this artificial protein in yeast. If you have any further suggestions, I would really appreciate.

Cheers,

Arman

Dear Alejandro Martin ,

Thank you for your valuable contributions.

As I have mentioned in the previous comment to Christian, I would like to check the functionality of this artificial protein in yeast. That's why I am using yeast as host.

You have said "By the way, if the control gene has a stem-loop with a free energy of -8.4 kcal/mol, doesn't this go against your hypothyesis that dyad symmetry motifs in the coding gene are affecting expression? After all, the most stable stem-loop of the artificial gene is only -7.8 kcal/mol, according to the table shown above."

Control gene has the minimum -8.4 kcal/mol and the artificial has -17.8 kcal/mol (I think when you click it becomes visible properly (instead of -7.8)). If it was -7.8, then it would contradict my hypothesis.

Cheers,

Arman

Hello Arman,

at the moment I have no further ideas how to address your question. However, you can find a lot of literature on how to optimize the production of recombinant/artifical proteins in yeast from the lab of Jens Nielsen (Göteborg). I hope this helps you.

Cheers, Christian