How specific is too specific?

By Paul Ko Ferrigno

It’s not often you read a paper that completely changes the way you think about a problem, but that’s just happened. Admittedly it was a review and that gives the author more scope to build a compelling argument, but I take my hat off to Nicole Baumgarth, author of How specific is too specific? B-cell responses to viral infections reveal the importance of breadth over depth (link is external). Now I have a new understanding of the role of the germinal centre in generating not just antibodies with greater affinities for a target, but also a broader range of antibodies that all recognise each target, because each of these also has a chance of binding to a new epitope on another pathogen – cross-reacting – thereby providing a deeper pool of potentially protective antibodies that can be rapidly deployed against future invasions.

I’ve wondered for a long time how many cells survive the selection in the germinal centres for B cells producing improved antibodies against new targets. I knew about somatic hypermutation and competition for antigen that is being presented on dendritic cells. But I hadn’t realised the answer could be that many, many B cells survive – perhaps all of them, just as long as they are presenting antibodies that exceed a certain threshold of binding affinity for the target. An interesting thought is this is a threshold that presumably can be determined by the abundance of the antigen in the circulation, which can be reflected by the dendritic cell population. But all of this is only possible or necessary because of the pre-existence of a subset of IgM-presenting B cells that have low affinity for self antigens – a complete contradiction to the idea that the immune repertoire is extensively quality controlled through negative selection against self antigens! The idea is that these are genomically-encoded binders with low affinity for self-antigens that are likely to be related to danger signals present on invading pathogens, making it likely that there will be rapid, innate response to invaders that would be mediated by what is usually thought of as the adaptive immune system.
So why care about the intricacies of antigen presentation and antibody optimisation? Well, one of the reasons we started making Affimer proteins is that we, and people we talked to, had experience of buying an antibody against one protein, and finding out that it actually reacted with multiple proteins. This is a problem that continues to dog both the antibody industry and its customers, and a great deal of effort is put into getting the dose of antigen right, the immunisation schedule right, and choosing the right time to end the study and collect the antibody: Probably all related to the processes outlined above.
The great thing about Affimer technology though is that we don’t have to worry about pre-existing cross-reacting-but-high-affinity antibodies, or about immuno-dominance driven by a peptide that is particularly well presented by dendritic cells, or immunisation schedules or even about the evolutionary advantage of having antibodies that cross-react with conserved epitopes. Affimer reagents are made in vitro without any design constraints around the target, although we can of course impose constraints upon the screen to drive Affimer binders towards particular performance characteristics, such as the ability to recognise their target in a biological sample type that is frequently used in the clinic, or in very high salt, low pH, or to distinguish between post-translationally modified proteins. For example, we have been able to screen for Affimer binders that bind tightly to a range of immunoglobulins from different animal host species (rabbit, mouse, goat, sheep, donkey, chicken…) and then select for those that only bound to the target and did not cross-react with other immunoglobulins – even very closely related isotopes from the same species. Some of these have already made it into our catalogue, such as AVA00055 (clone 14-B9) and AVA00056, which are specific for mouse IgG2b and do not bind to mouse IgG1, IgG2a, OgG3 or IgM. There is of course multiple choices for people wanting a specific mouse IgM binder, such as AVA00045 (clone 16-G4) or an Affimer against IgG2a (e.g. AVA00039, clone 13-A5).
The catalogue will continue to grow, but if there is a specific immunoglobulin you would like us to target, please get in touch. Our goal is to make life easier for all those people trying to specifically detect a mouse antibody in a human sample, or one immunoglobulin isotope at a time, or even to multiplex multiple antibodies in a single assay without worrying about cross-reactivity.
Further information