Several possible reasons, honestly.  One, it could just be being degraded by your lysis buffer - you might try different/more protease inhibitors.  Two, it could be a translation problem by the yeast - you might try a different cell line, or a different vector, or whatnot.  Three, it could be the protein falling apart for no particularly good reason - that's probably most problematic, and you could monitor by doing SDS-Page at different timepoints post-purification, to see if the band gets any larger.  Four, it could just be a failure in your purification (assuming you have purified at all).  Five, it could be one of the many things I haven't thought of.

You might try a Western blot with a specific antibody for some sequence towards the middle of your protein, to make sure the top and bottom bands are actually some form of that protein (I assume you're just staining a gel for total protein content right now).  That would at least help narrow down the problem a bit.

Another option could be your lysis method is accidentally shearing your protein. Or your protein of interest may be thermo-unstable so you might try to work with your protein at lower temperatures.

In case you haven't tried it already, you might want to consider running a control of your parent plasmid (without the cloned gene) transformed into Pichia to check if the band corresponds to any native Pichia protein.

Temperature degradation is a big issue - Proteases become more active hence more difficult to control with inhibitors.  A western blot idea above is a great idea - that will define your bands as belonging to your target.   An IEF (little of the wall) will characterize proteins that are Glucoronidated similar to an antibody.  Last. Prokaryotic and lower Eukaryotic expressions systems are different - so don't expect the exactly the same products. Brings us back to the western blot to characterize an epitope.

Looking forward to a follow up

If you suspect glycosylation why not digest with endoH or PGNase F before you start changing lysis buffers etc. If it is indeed glycosylated, the higher band will be reduced to the size of the lower band. If there is no difference upon incubation with these enzymes, then you should look at the lysis buffers and time course expression advice given above.

A western blot, using an antibody specific to your protein, should pick up both the glycoslyated and non-glycosylated proteins.

Hi there,

As your protein is processed through the secretion pathway the bands you observe might correspond to the mature form (without signal peptide) and the proprotein form (with signal peptide). Are both forms collected from the culture supernatant? If so your mature protein might contain site for  kex2 protease cleavage (the one which is also in charge of eliminating the signal peptide).

Dear Liger

Thanks for the suggestion. Is it possible to get proprotein from (as you said with signal peptide) in supernatant? Actually, when I am collecting the supernatant, and running the SDS-PAGE, both the bands are observed. 

Thanks Lauren and Katt

Now, I am standardizing through western blot. I am planning to go for PNGase F digestion also. 

Thanks all for your valuable comments.

Hi again,

As I said if considering the supernatant only it's hard to imagine than the proprotein would be released from the membrane. That's why I emitted the hypothesis that your mature protein might contain a cleavage site for kex2 protease in addition with the one present between signal peptide and your protein sequence.