Removing annoying heavy chain detection in IP/blots

fred fred at
Wed Aug 21 17:27:35 EST 2002

> > When we do IPs and boil the beads in SDS to run a gel, we then probe the
> > blot for the protein/antigen of interest.  However, most of you should
> > that the IgG heavy (55 kDa) and light (25 kDa) chains often interfere
> > detection of proteins in this MW region.  Hence people often cross-link
> > their antibodies to Protein-A beads first.  We tried that, but still, a
> > small amount of IgG got thru somehow.  Would people care to comment on:
> >
> > - How to remove trace IgG from beads after cross-linking the antibodies
> > Protein A, but before using them for the IP.  We don't want to do
> > crosslinking as it might wreck the IgG binding to our target.  (We used
> > (dimethylpimelimidate) to cross-link).

> DMP is fine. wash thoroughly after crosslinking, using stringent buffers
to get rid
> of bound but not crosslinked IgG.
> We use short washing steps (in a small disposable column from Biorad)
> with 200mM Glycine/200mM NaCl/made pH2.0 with HCl. This breaks the
> IgG/ProteinA binding when it is not covalent (ie crosslinked). It is
> essential to make these washing steps fast, and immediately neutralize
> the material afterwards, as longer acid treatment will ruin much of the
> antibody. When we have 200ul of ProteinA-sepharose, we wash with 600ul
> of buffer (3 volumes), let the buffer drop through the column (won´t
> take longer than a minute or two), repeat once, then neutralize by
> adding 1ml of 500mM Tris pH8. Let drop through, and equilibrate the
> material with your buffer of choice, e.g. PBS.

Thanks for the suggestion.  We never considered going down as far as pH 2.0
!  My understanding was that at pH 3.0 glycine the Ab stays bound to Protein
A.  So your point is very helpful here.  I will give it a try.  I presume
you meant to wash the beads this way BEFORE IP. Then IP as normal.  However
I am still worried that the Ab may be damaged by this pH (afer all, it has
just gone thru a cross-linking exercise too).

> > - If we don't cross-link at all: Can you elute the antigen and LC off
the IP
> > beads with DTT, instead of SDS?  DTT will release the LC and should
> > the antigen fine, but will the HC stay on the Protein A column?  The two
> > are also held by disulphides, which the DTT will reduce, but will BOTH
> > stay on the protein A?  Or will one of them come free?

> In our hands, whenever you treat crosslinked material with SDS or
> DTT/mercaptoethanol, a fair amount of IgG chains will come off the
> column; this is most likely because there are only one to a few
> attachment points between ProtA-Sepharose and the IgG, so that only one
> chain will be bound. We routinely elute with the glycin buffer mentioned
> above, precipitate the eluted protein (see below), and redissolve the
> protein in sample buffer. This takes much longer, but gives very clean
> To precipitate eluted protein, make it 400ul with water, add 400ul of
> Methanol, vortex vigorously, add 100ul of chloroform, vortex again, spin
> in a microfuge for 2min full speed. Remove supernatant without touching
> the interphase (you may not see protein there, simply let 10-20ul of
> supernatant behind), add 300ul of Methanol, vortex, spin again, remove
> all supernatant (a pellet may not be visible), air dry pellet for at
> least 30min, redissolve in sample buffer and analyse on a gel.

Now we are talking about getting the antigen off the beads, leaving the IgG
behind if possible.  If you are right, that both SDS and DTT treatments of
the beads pulls off some IgG, then that may be the biggest part of our
initial problem (ie detecting unwanted trace HC and LC in the subsequent
blot), rather than inefficient crosslinking as we initially thought.  Are
you saying that to avoid the presence of trace eluted Ab, the glycine
elution (or the 3.5 M MgCl elution suggested by DK (thanks)) will have less
or no IgG contamination, compared with heating the beads in SDS or treating
with DTT?
I suspect this idea will help, but is much more work because the elution
volumes are larger than simple SDS.  To concentrate the eluates I routinely
use the Wessel & Flugge method you suggest, myself.  [Wessel,D. &
Flugge,U.I. (1984) A method for the quantitative recovery of protein in
dilute solution in the presence of detergents and lipids. Anal.Biochem. 138,
141-143].  It is the most powerful and simple one that I know!  However, I
find that more than half the people I show/teach it to cannot get it to work
properly or reproducibly.  There is nothing wrong with the method, but it
requires a lot more attention to detail than a lot of people apparently give
their experiments.  The last step is always the problem, making sure the
pellet dissolves in the SDS. We find we have to sonicate with a fine probe -
messy, and the SDS foams up and has to be centrifuged down to de-foam before

> > - Alternatively, has anybody ever tried this idea?  Incubate protein A
> > with antibodies (don't chemically cross link).  Do the IP, then elute
> > off the beads with SDS but no mercaptoethanol.  Then run the gel without
> > mercaptoethanol either.  In this way, the IgG should remian as a 160kDa
> > complex, leaving only nice clean target antigen/proteins to be detected
> > the 55 kDa region of the blot.  A nice idea, but does it work?

> Yes, it will work. However, not treating the sample with mercaptoethanol
> will result in aberrant mobility of many proteins, because they contain
> SS-bridges that will remain and keep the protein in a constrained
> conformation. If your protein does not contain SS-bridges, this is
> irrelevant, of course. Just keep in mind that certain marker proteins
> may not run as expected...

Thanks again.  Yes we are aware of the problems associated with this idea.
However in our case we are lucky.  All our target proteins for the IP are
monomeric and do not form S-S bonds.  We explored this earlier.  So we will
readily see our target protein OK.  As you point out, we may miss some of
the proteins that co-IP with our target, but we don't want them at this
Have you (anyone?) actually tried this idea?  Do you actually see the IgG
migrating at 160-175 kDa when heated in the absence of betaME?  Is there
actually no HC at 55 kDa left?  We don't want even a trace amount of HC.
You mentioned the problem of MW markers running abberantly in the absence of
mercaptoethanol.  We have been able to effectively deal with this by running
MW standards on the gel that  do contain betaME, alongside our test samples
that don't contain any.  The key trick here is that the betaME leaches
laterally in the gel (across at least one lane), so that two lanes between
your stds and samples must be blank (in terms of betaME) to prevent this
fully.  Otherwise, this works fine.  We fill those lanes with just sample
buffer without betaME.

> PS: there are secondary antibodies available that only recognize the
> light chain of the antibody; these are useful when your protein´s size
> is close to that of the heavy chains. So maybe your current protocol may
> be sufficient (as you say there are only little amounts of ab left).

This is a good idea.  I was unaware of this.  Often our secondary Ab will
not see the HC of the primary Ab anyway (which is what we want), but we
never seem to be able to predict this.  Your suggestion would help a lot.
So we will research it more...


More information about the Methods mailing list