My POI is a human protein. However we when we had the DNA sequence synthesized we codon optimised for S.cerevisae (for protein expression). However, I now wish to do some interactome studies in a human cell line. Do I need to have the sequence codon optimised? Is there anyway to do this without spending a lot of time/money? My sequence has ~71% identity to the equivalent human codon optimised sequence. Do you think it's worth having a go without optimising first? Anyone have any similar experience?
If you optimised the coding sequence for S. cerevisae, usually you can try the expression in human cell line without re-optimising it. Codon usage in yeast and human is not that different in comparison with these used in E. coli. The sequence identity is not really a issue, what you need to check is the rare codons that the human cells don't use them. Check your codon usage against these used in human cells http://www.kazusa.or.jp/codon/cgi-bin/showcodon.cgi?species=9606, if the codons in your sequence are all used in human cells (even higher or lower), you should be ok.
If it's a human protein in human cells, the so-called "optimization" can do more harm than good in my experience. There is a lot that the great optimizers do not understand, such as pause / folding signals and who knows what else might be going on.
If you consider the cost of labor and indirects, it's going to be most cost-effective to have it synthesized from scratch. But you could clone it from cDNA I suppose, especially if someone in your lab already has suitable human cDNA.
Whether it's worth "trying" the yeast version depends on your purpose. Can you define specifically what you need the protein to do, and can you measure it easily? Then it's probably worth trying the yeast sequence and getting to your specific test as quickly and painlessly as possible.
Or do you need to be worried about subtle systemic effects that for example a partially misfolded protein, or an unnatural expression level might have on your experiment? Do you need to worry about "unknown unknowns" that might affect your result without you even knowing it? In those cases, I'd rather play it safe and get the authentic human sequence.
Thanks for your advice Huanting. I checked on this server
http://genomes.urv.cat/CAIcal/E-CAI/
My understanding is that the CAI values indicate the percentage of codon usage in the organism chosen. The values range between 0.252 and 1.0 (although my start ATG and stop TAA are coming back with 0.0 but I think this may be because they do not have synonymous codon). So I think that there aren't any codons present that are not used in human cells from what I can see.
Thanks John. No afraid no human cDNA in the lab. I am looking to do either transient or stable transfection of my POI with a GFP and biotin ligase tags...so aspects such as protein folding, expression levels, and localisation are already relatively compromised and unknown at this stage, until I can look at the transfected cells under the microscope.
- Badges
- Science method