That's a pretty low salt concentration, I would probably increase it to 300-500mM NaCl. Also, are these tryptic peptides or another enzyme for producing glycopeptides? Larger peptides will have a higher propensity to have hydrophobic residues so very little non-ionic detergent such as 0.1% DDM (dodecyl maltoside) might help. DDM is acid labile so you can add your typical 0.1%-2% formic acid to resuspend your peptides and the detergent will break down and should not disrupt your LC-MS/MS runs.

if the components sticking to your enrichment fraction are proteins and you are trying to enrich glycosylated peptides, maybe a simple 10 kDa spin filter might do the job.

No. The component sticking are peptides which are non-glycosylated.

Hi,

which lectin do you mean and what are the chromatography conditions (sampling and elution buffer, is it specific-sugar elution)?

Maybe RP fractionation of SPE cartriges after lectin enrichment.

I'd rather prefer to increase the ionic strenght after elution of non-adsorbing fraction, this should be enough to elute non-glycosylated proteins (binding probably weaker) and next elute glycosylated proteins with the sugar

How much salt are you using in your binding buffer? You can get away with 0.5M NaCl or KCl. I recommend KCl. The salt is critical for buffering interactions between charged moieties on peptides/proteins and any charge on your stationary phase resin. Eluting with a competitive sugar as opposed to with a nonspecific eluent such as acetic acid will reduce the amount of nonspecifically bound components in your enriched sample.

i use 10mM HEPES +.15M NaCl for binding and washing... and elute in the sugar made in same buffer (500mM lactose) for MAL-II.

That's a pretty low salt concentration, I would probably increase it to 300-500mM NaCl. Also, are these tryptic peptides or another enzyme for producing glycopeptides? Larger peptides will have a higher propensity to have hydrophobic residues so very little non-ionic detergent such as 0.1% DDM (dodecyl maltoside) might help. DDM is acid labile so you can add your typical 0.1%-2% formic acid to resuspend your peptides and the detergent will break down and should not disrupt your LC-MS/MS runs.

I agree with Albert's suggestions of higher salt and non-ionic detergent (good call) as your best bet to reduce non-specific binding. That said, I wonder if the problem is not more fundamentally related to the difficulties of lectin enrichment. Have you tested your lectin with a standard MAL-II binding glycoprotein (e.g., bovine fetuin) in the presence of BSA? If that works, can you digest fetuin and enrich its alpha-2,3 sialylated glycopeptides from its non-glycosylated peptides? If neither works, it suggests that the association constant of the MAL-II (being apparently low) may limit the specificity of your affinity experiment.

Another thing I wanted to add is that lectins bind in a 3D context which is why lectins are usually multimeric complexes when they bind glycoconjugates (for example MAL-II is dimeric). I'm not exactly sure how you're running your MAL affinity chromatography but by digesting your proteins into peptides you're essentially making the mixture more 'complex' (for example one glycoprotein vs 10-15 peptides (maybe 1 or 2 are glycosylated). Personally when I enrich for glycopeptides using lectins I normally perform an overnight batch bind (4oC) for the following reasons: 1) it will increase the time for the lectin to bind your glycopeptides with good mixing; 2) it will reduce the amount of non-specific binding with the longer incubation; 3) by performing gravity chromatography you are also *indirectly* doing something similar to a 'size-exclusion separation' i.e. larger proteins will occupy more space than smaller ones and so when I re-pack the mixture (beads + peptides) into a column I make sure I pack it into a long thin column rather than a wider column, this will slow down the flow rate and allow any unbound lectins to bind remaining glycopeptides. I presume these are agarose conjugated lectins. But before you re-start the experiment I would do just like what Ben mentioned which is run control fetuin and elute using free sugar to ensure that the lectin is performing as expected. Usually when lectins are conjugated to agarose their multimeric forms can be destroyed in the coupling process and can leech off in the process (run a Coomassie gel). If there is some leeching observed, perform a pre-elution with free sugar to remove leeched lectin, then equilibrate with your Hepes buffer. Hope this helps.

In this case, the use of RP-HPLC may be recommended.

Don't know enough about your end goal, but some thoughts: If your contaminant is sticking to the column matrix, remove it by passing your sample through unconjugated column matrix. If your contaminant is sticking to the lectin on the column, use different lectin columns in series (of course the lectins have to bind your protein(s) of interest). Chances are your contaminant won't stick to two different lectins the same way if it is not glycosylated. If your contaminant is sticking to your glycosylated proteins, try using a chaotropic wash of the loaded column before elution. High or low pH, removable detergents, chaotropic salts, urea, quanidine, glycerol, arginine... all these methods people use to avoid protein aggregation. You can't know what will work unless you try it, so try some of these conditions on a small scale by using small batches of lectin beads in microfuge tubes, loaded with your sample, then washing with trial conditions, spinning down and eluting with gel sample buffer (if running a gel is your endpoint). Good luck.