Minimum distance of epitopes for protein detection with antibodies?

Deitiker, Philip R via (by pdeitik from
Wed Aug 3 12:14:33 EST 2011

Linear epitopes minimally occupy about 5 residues. There is the occasional binding of 4 at low affinity. The better epitopes are around 6 to 8 residues in length. As originally defined by MZ. Atassi back in the 70s these antibodies are attracted to loops and bends on the extremes of smallish proteins, but this also is a place where evolution is fast (less immunological suppression between species). Anti-peptide elicited antibody binding can occur anywhere on the molecule provided the region can unfold long enough for the antibody to bind. 

Discontinuous epitopes, as you appear to be talking about are frequently oriented toward clefts within molecules, and the space they inhibit has to be factored in as a consequence of the shape of the binding. As a consequence the binding on flat surfaces may block fewer other sites than the binding in a cleft. As stated above, the size of a binding site can vary depending on the affinity to core structural epitopes of the antigen. Some antigens may need only a few contacts, while others may need more of lower affinity. Theoretical studies of discontinuous epitopes are perilous. I have a recent publication on this matter (a rebuttal) which you can look up. What you need to do is take a look at the 3D structure and the binding site of interest, first. From there you might need several designs. In some instances, as Atassi did, you can simulate a discontinuous site with a continuous peptide, elicit cross reactive antibodies and even make monoclonal antibodies. Really the best way to go about doing this is to generate mAbs, purify fAb from an mAb and co-crystallize antigen-fAb complexes for X-ray crystallography. This will allow you to see where the Ab is binding and how much space it takes up. The final thing is to look at antigen flexibility and access (CD spectrum is one way), and see what conditions the antigen can open-up (phosphorylation, pH, temperature, solvents).
I have to reiterate, discontinuous recognition is very tricky, antigens can conform in shape to antibodies on a linear epitope, and so actual contact can be less (all things being held equal). But unless the discontinuous site is close to a disulfide or other covelant linkage between strands, the motion of the molecule may deserve considerably more contact to create that conformation, and of course this also means that contact in some areas is fluid. 

-----Original Message-----
From: methods-bounces from [mailto:methods-bounces from] On Behalf Of DK
Sent: Tuesday, July 26, 2011 10:32 AM
To: methods from
Subject: Re: Minimum distance of epitopes for protein detection with antibodies?

In article <593d475a-c530-4d82-adca-6365ad76787e from>, WS <novalidaddress from> wrote:
>Dear experts,
>do you have any ideas what is the minimum distance (e.g. in amino
>acids in an ideally stretched polypeptide or absolutely in nm) that 2
>antibodies can bind and do not compete for each other?

Back of the envelop estimate: 

Fab is about 40-45 nm "thick". Alpha helix is 0.15 nm per residue. So 
it makes it 25-30 residues for alpha helix. Beta stand is as 
"stretched" as they come and is ~ 0.5 nm per residue, so 8-10 
residues is geometrically possible. 

Methods mailing list
Methods from

More information about the Methods mailing list