Protein degradation during expression - (May/17/2012 )

I would like to ask for the help in order to solve my problem.
I'm expressing human protein (a fragment of it, to be precise). When I'm analyzing expression/Ni-NTA purification by sds-page, I see 2 bands instead of one. I have performed trypsin digest/peptide fingerprinting by mass-spec. Both bands belong to my protein. The difference between bands is ~5kDa. I cannot separate degraded fractions by subsequent size-exclusion chromatography. I have checked my protein by PeptideCutter. There is a number of peptidazes that cut. However, I found that Asp-N endopeptidase could cut ~5kDa at carboxy terminus which corresponds to what I see on sds-page.
I'm really annoyed by this degradation problem and do not really know how I can solve it.
May be I should mutate the "putative" recognition site for endopeptidase?
The cleavage site which is given at Expasy is:
Xaa-|-Asp, Xaa-|-Glu and Xaa-|-cysteic acid bonds
The sequence of my protein is:
MRKEAEKTAL STIAVATAKA KEQETILRTR ETMATRQEQI QVTHGKVDVG KKAEAVATVV
AAVDQARVRE PREPGHLEES YAQQTTLEYG YKER
I would be extremely helpful is anyone could help me to solve this problem.

add protease inhibitors to the expression medium.

Doesn't make sence to me, as pretease inhibitors will inhibit all bacterial proteases, thus, preventing cell from growing normally (I think cells will not grow/express at all). Does anybody else has another suggestion?

they shouldn't enter the cells but will inhibit the proteases that are released into the medium