Pairwise epitope binning

Dear Arnoud,

First of all, thank you so much for the SPRpages. There is hardly anyone at my institute working with SPR (yey!) and those who do use it for quite different purposes than me. You have no idea how extremely helpful this website has been. I recently also ordered the SPRpages book, I can really recommend it for anyone who wants to learn SPR and really understand the system. Thank you!!

Concerning my question - it's about so called pairwise epitope binning. I have some mouse mAbs (IgG) and two different Antigens (18kDa). I want to check whether binding of one mAb to an antigen inhibits binding of the other one. It seems a quite straigth forward task, but I am unsure about how to best do the referencing and the immobilisation strategy? I would like to use a capture approach, using anti-mouse IgG for directed capture of the first mAb and also for more flexibility of the assay. The injection would be mAb 1 + antigen + mAb 2.

From a previous assay I still have Fc1 and Fc2 immobilized with ca. 12000 RU of anti-mouse IgG (CM5), reference is Fc2-1. I inject everything BUT the antigen to Fc1 and Fc2, the antigen only to Fc2. This means that in the Fc2-1 sensorgram I can only see the antigen + mAb2. Of course for this kind of assay it is important to make sure that all unoccupied anti-mouse IgG on the chip is blocked by some irrelevant IgG, otherwise an association of mAb2 might be due to association to anti-mouse IgG and not due to association to antigen. However, with so much anti-mouse IgG immobilized on the chip, it seems impossible to completely block it.

In order to circumvent the "impossible to block" anti-mouse IgG, should I just immobilize less? Like 500 RU anti-mouse IgG instead of 12000RU?

The other possibility could be to immobilize anti-mouse IgG only to the active, but not the reference flow cell. I would then inject all proteins over both Fcs. However, in the little booklet that comes with the anti-mouse IgG (from Biacore) it says that the reference should also be immobilized with anti-mouse IgG. Or can I also make the reference Fc with an irrelevant IgG?

Do you have any experience or tips with this kind of assay? I would very, very much appreciate any kind of help!

Best regards,
Mitz

Dear Mitz,

Thank you for the kind words. Much appreciated.

I think the main problem lays in the blocking of Fc-1 (reference) for the second antibody. I don't think that there is fools prove solution for this. One thing you could try is to split the experiment in two:
1) Capture Ab1 on Fc2 – block Fc1 – inject Ab2 over Fc1-2 (gives you non-specific binding to FC1)
2) Capture Ab1 on Fc2 – block Fc1 – capture the antigen on Fc2 - inject Ab2 over Fc1-2

Second option to immobilize only anti-mouse Ab to Fc2 and an irrelevant to Fc1 is a good option. As long as the response on the Fc1 is low you can say that the response on Fc2 is specific.

The other –most common– approach is to immobilize the antigen. If you have a four channel instrument both antigen can be immobilized + 1 reference. Then inject the antibodies pairwise to check if the block each other.

I have two publications for you to show the last option.
1. Abdiche, Y. N., Lindquist, K. C., Stone, D. M., et al.; Label-free epitope binning assays of monoclonal antibodies enable the identification of antigen heterogeneity. Journal of Immunological Methods (382) 1: 101-116; 2012.
2. Abdiche, Y. N., Malashock, D. S., Pinkerton, A., et al.; Exploring blocking assays using Octet, ProteOn, and Biacore biosensors. Analytical Biochemistry (386) 2: 172-180; 2009.

Kind regards
Arnoud

Dear Arnoud,

Wow! Thank you so much for your quick and helpful answer and for your input.

I think I'd prefer the option of immobilizing the anti-mouse IgG to Fc2 and an irrelevant IgG to Fc1. Curious to see if it will work, it seems the easiest option. I would, however, infact go for a much lower immobilization of anti-mouse IgG (and the other IgG) than maximum. That way I might be able to almost saturate the surface already with my first mAb, considering the Rmax and Rligand equation.

The other option you mention, with antigen immobilized, is actually something I have done before and it did work to some extent. I was however thinking that if I see that mAb 2 doesn't bind when mAb1 is already bound to antigen, that this may be also just due to it not having any "space" anymore cause mAb 1 is all over the place and not necessarily because its epitope is specifically covered. Also, a problem with immobilizing my antigens is that they are a pain to regenerate (easily damaged). I have therefore previously captured them but the capture was unstable due to the kind of tag I used, very annoying. I am now also adding a different kind of tag, so I'll definitely also check this approach again in the future!

Thank you very much for your help and I'll gladly write an update once I have made this assay work

Best regards,

Mitz

Hi Mitz,
Some clarification on direct immobilization and capturing. The direct immobilization of the antigen is a random process (amine coupling) so the antigen is attached to the surface in all kinds of different orientations including the possible binding sites. Since this is more or less a random process (amine coupling is mainly via lysine residues) most of the binding epitopes will be available.
Capturing however will orientate all the analyte molecules in the same manner (one binding site).
Binding of a 'large' antibody to the smaller antigen always has the possibility that the antibody will get in the way of a second antibody binding to the same antigen on a different epitope. This will not be different between direct immobilization or capturing the antigen. But I don't think that the first antibody is all over the place because that implies non-specific binding. Therefore it is always a good idea to reverse the Ab1 and Ab2 injection and make sure the results are the same for the epitope binning. As an alternative Fab-fragments can be used but that will be a lot of work if you may to test.
Looking forward to your update.
Arnoud