Antibody sticks to antigen affinity column

WS via (by novalidaddress from
Sat Mar 6 19:28:31 EST 2010

Hi Dima,

I forgot to mention that previously, I had coupled the ligand to BSA,
previously immobilized to sepharose, with a similar overall ligand
density (about 50µmol in 10ml of sepharose). There I could elute about
15% specific IgY in a similar exploratory experiment, So I doubt
unspecific binding of IgY. However, I also observed some ion exchange
properties of the BSA-derived material, so instead we decided to try
direct coupling of the ligand to a (appropriately modified) sepharose.
At the moment, I favor your "too high coupling density" theory. I
haven't tried stronger chaotropes like NaI and NaSCN yet, so I am
curious, what results I will get on monday. NEt3, pH 10.5 is an
additional another option. And making a smaller column which might be
better suitable for method development. In worst case, I'd revert to
the BSA based column.

Concerning IgY, it's incomparably cheap compared to mice and rabbits
and we usually get good results. Sheep and goats are reasonable
alternatives, of course, and we also use them. However, I already have
about 500mg of crude IgY waiting in the fridge for purification, so
changing the animal is not an option anymore.

Many thanks for your help!


> Sounds like any IgY bind strongly to your column. Which means
> you are doing non-specific absorption chromatography rather than
> affinity. Two possible reasons: 1) your conditions or 2) your ligand.
> #1 is unlikely unless you keep screwing up royally, #2 is very likely
> if your ligand is at very high coupling density (you hint at that with
> your stated high theoretical capacity) and #2 is possible if your
> ligand is unusual in that IgY bind to it strongly (is it something
> strongly hydrophobic or soemthing likely to bind complex sugars?).
> More exotic case is that you did not block sites left over after coupling
> chemistry and you are doing covalent chromatography.
> I'd try: batch chromatography versus column AND lower coupled
> ligand concentration 10 and 100-fold and see what happens. If
> I can use some derivative of that ligand (e.g., deoxy form or
> something), I'd try coupling that, too.
> As a general comment, forget chickens, go with rabbits or goats.
> Way easier and higher affinity antibody.
> DK

More information about the Methods mailing list