Tailing of a monoclonal antibody by SEC-HPLC

[ameeler]

I've been using the same type of Tosoh TSKgel G3000SWxl column (part #08541) for a few years on the same monoclonal antibody. Recent column lots have shown significant tailing. When I use an older lot, there is no tailing (same instrument, mobile phase, sample, etc). The tailing is more severe with neat injections than with dilute sample injections (same amount of protein loaded). When speaking with Tosoh, they mentioned that the silica resin lots have changed. Has anyone else had this problem?
I'm not sure the best way to approach this issue, short of changing the method (which would be a big pain at this point for the project).
I have tried conditioning the column with mg of protein, but that did not seem to help. Any suggestions?

[Rafael Chust]

If Tosoh changed the silica, then you have a big problem!

Tailing in SEC columns could come from overload (maybe you can make a test decreasing the sample concentration) or competitive separation mechanisms - silica can have metals and other impurities that might cause interaction to your sample.

You still have the possibility to search another stationary phase. I know Phenomenex has a guaranteed replacemente for Tosoh SW column.

[tom jupille]

I'll second Rafael's comments. You have already explored the other possible fix with a big protein injection. I would expect that Tosoh would try hard to help you; maybe you need to push them harder.

[ameeler]

I've just followed up with Tosoh. It is possible that hydrophobic sites on the resin from the coating process can link up together causing some hydrophobic interaction with the protein. The base material of the resin changes from lot to lot, so basically, I am stuck with a bunch of columns that cause tailing for now (not a good thing for a cGMP transfer of the method).
I will try to add 5% IPA and look into other column vendors.

Thanks again for the suggestions.

[Rafael Chust]

Now you might understand why some companies bang so much with ultrapure silica!

[HW Mueller]

This is the first time I see something possibly pertaining to a long standing problem I had with a TSKgel SuperSW 3000 column (the one which no longer baseline separated an antibody from its Fab after contacting TFA).
Certainly would like to see more info on this, if you have it, maybe we should continue on this privately? (e-mail Hans.W.Mueller@strz.uni-giessen.de)