Native genes use low frequency codons as well. People noticed that these are usually in between domains. One theory is that this is to slow down the ribosome and allow the domain to fold properly before more amino acids are added to the chain. This is why people started somewhat recently to talk about codon harmonization rather than optimization.
For example: Article Heterologous Protein Expression Is Enhanced by Harmonizing t...
tanks for the answers but let me explain a little more:
i am trying to express and purify hemaglutinin protein (60kDa) from influenza virus in E.coli when i check the frequencies of a same codon in human which is a natural host for virus i see 0.2 compared to 0.06 in E.coli.
we had some problems using His-tag for purification and when checked with western blot most of the protein expressed in bacteria is trapped in cell pellet.
i was wondering if protein couldn't fold correctly due to low codon frequencies in E.coli, thus making aggregates.
Codon misfolding might occur by many different reasons, and codon usage is only one of these (I'm not sure its a primary reason). There are ways to unfold and refold your protein (usually by adding urea and then dialysing it out). I would give it a try.
This problem happens quite often. It can sometimes be improved by reducing the temperature of the cell culture during induction. It is also worth investigating using different expression strains, such as Rosetta, which expresses rare codon tRNAs.
Besides strategies providing tRNAs to heterologous genes in E. coli, I would also consider doing some protein engineering. The Fleischman/Tawfik labs at Weizmann developed a computational tool, PROSS (Protein Repair One-Stop Shop), which could help you to express your target protein in E. coli in soluble form: https://www.cell.com/molecular-cell/pdfExtended/S1097-2765(16)30243-X.
I have not used the software, but there are various citations to it: http://pross.weizmann.ac.il/bin/steps