overexpress his-tag protein, wrong size?! - (Nov/17/2005 )

Hi all, I have been trying to overexpress and purify a truncated protein using pTrsHis-Topo. it is a truncated (about 1/2) of the whole protein because I couldnt get the whole protein to overexpress. When I IPTG (0.5mM) induce it and run it on SDS-PAGE, I see a large band overexpressed in the induced lane (as compare to non induced), and when I purify the protein using Ni-NTA column, I can see the same band in the elute. But, its not the right size at all! The estimated size including his tag should be about 15 to 16 KDa, but my band is close to 25kDA. When I perform a Western, the band appeared so it seems to be the right protein. So why the wrong size? I have sequenced the construct and it came back no problem. In addition, I received another construct from another lab that used a pET vector and it also contains a truncated protein. the difference in that my is a Clostridium protein, and the other one is Bacillus. Their construct overexpressed a protein at the predicted size! anyone ever had similar situation as mine? are my protein some how forming a dimer or aggregated? It is not a membrane protein.

perhaps the truncated form makes some sort of complex with a native e coli protein?

perhaps is you expressed it in another host strain, and ecoli with a different phenotype, you might help the situation?