Alison, 

Sometimes expression of recombinant proteins in a different system may lead to wrong localisation. The above in silico tools predict the localisation in the native system of the protein, not in a foreign host. More information on the expression system and the natural source of protein may help me to give a better suggestion. Also recommend you to try an alternate expression system without any signal peptide in the cloned construct..

Good luck.

Hi Ancy,

The majority of the fusion consists of two fluorescent proteins - derivatives of GFP and RFP - which in other constructs are both cytoplasmic. The other constituent is a Tobacco etch substrate (only 7 amino acids long). I have been expressing the construct in HEK cells.

Hi Alison!

Sometimes we have these kind of problems with our cytoplasmatic constructs. Adding a Nuclear Export Signal to the fusion proteins helps sometimes... Bence

Hi there,

the servers you have used are making predictions based on protein sequence, basically looking for NLS and NES. Have you got any other information on the protein you are interested in? The absence of nuclear localization signal in a protein sequence isn't enough to say that it should be cytoplasmic. There are cases where protein is nuclear or shuttling between nucleus and cytoplasm because taken in charge by a specific system (for instance ribonucleotide reductase small subunit in yeast). So in your case, it might be that your protein is nuclear because of the activity of host proteins...

Hi Alison,

In case the protein you are interested is of human origin, you can check the Human Protein Atlas:

http://www.proteinatlas.org/

for independent verification that the protein is expressed in the nucleus across different tissues and cell types.

Ilian