Dear Daniela,
I am currently studying mRNA folding related to expression levels. What
we have seen is that MFOLD will often predict secondary structures that
covers SD or start ATG, but the expression levels can be fine. It is
also said that the individual bond should be of at least -6 deltaG to be
of any relevance, that is, to hinder translation, or translation
initiation (Maarten H. de Smit and Jan van Duin, 1994).
In your case it seems you get no protein at all so I don't think
secondary structure is the problem here. Btw, hos sensitive is your assay?
You also mention codon usage. What organism do the gene come from, and
in what organism do you try to express it? I've expressed different
human proteins and different bacterial proteins in E.coli without any
codon optimization, it has been of no problem.
One more thing, if you try to express a protein that needs disulfid
bridges to obtain right confirmation, and you use ELISA assay to detect
your protein, then missing disulfid bridges can be a problem. Antibodies
may not bind to insufficiently folded proteins, so in your case it can
be that the gene is expressed but because of misfolded protein it is not
detected. This is mostly a problem when trying to express proteins in
E.coli and not using a signal peptid (the protein is not being
translocated to periplasma).
Trond Erik
