The first step of any anti-peptide antibody project is the design of the peptide to be used for immunisations, and it is a key to success. When using peptides as antigens, the importance of the peptide design cannot be stressed enough. Since there are many aspects to consider, the task can seem daunting, but this guide will pave the way for you and provide insight in some key elements as well as some basic calculations.
When there is prior knowledge of the target protein, you can define regions that should be avoided - or domains that are to be targeted. In the design process, targeting the relevant section of the protein is the first and foremost criterion. Since you will be in the better position in order to identify such regions this should be done by you, even if you decide consult an expert to aid you in the overall design work.
No matter how good your peptide is in all other aspects, if it’s insoluble you still can’t use it for antibody production. As far as we know, there are no reliable algorithms for predicting solubility, but if there are a few charged residues (Arg, Asp, Glu or Lys) it will probably be fine. You can also enter your sequence here for a report from our peptide property calculator PepCalc.com (opens a pop.up windows):
Making antibodies against neo-epitopes, post translational modification sites, exon bridges, the choice of possible sequences are limited. When you are bound by restrictions such that you must use an insoluble sequence, one will have to resort to adding spacers and hydrophilic residues. It is a strategy that is often employed, but only when necessary, and you may want to do this in consultation with the lab that is to synthesize the peptide for you.
Peptides that are ~4 kDa are usually large enough on their own to illicit an immune response, but smaller ones need to be conjugated to a larger carrier molecule such as Keyhole Limpet Hemocyanin, Ovalbumin or Thyroglobulin in order to work well for immunisations.
The most common strategy is to add a Cys to one of the termini of the peptide and use its thiol group for conjugation to a carrier molecule via a maleimide cross linker. In many cases choosing location, N- or C-terminus, has little impact but the choice can potentially make a significant difference for the outcome of the project. This is particularly true for the production of phospho specific antibodies or other site specific modifications.
In all cases, the addition of the Cys does not affect the specificity of the resulting antibody. This strategy works very well, except for the case when there is a Cys already present in the interior of the peptide sequence.
Regardless of conjugation strategy, there are reasons for wanting to avoid the presence of Cys in the peptide sequence. The function of Cys in a protein is often to stabilise the structure by the formation of disulphide bonds. It is likely that the area surrounding a Cysteine has some structure that cannot be mimicked with a peptide antigen.
There are scenarios when it is desirable to obtain an antibody that targets an epitope spanning a Cysteine. In such cases one may consider other coupling strategies for the conjugation to a carrier molecule. These options typically are amine or carboxyl reactive, and it is desirable to have a single such functional group present in the peptide sequence.
If the 3D structure of the protein is known, then verifying whether or not a particular sequence is exposed on the surface becomes trivial. In most cases however this information is not available. The most common strategy then is to look at hydrophilic regions. When gauging the solubility of a peptide one looks only at the peptide sequence itself. In order to make a prediction of whether or not this peptide might be located on the protein surface, one has to look at a larger region surrounding the peptide. If it is a hydrophobic region then the peptide may be located inside a pocket inaccessible to an antibody. If it is a hydrophilic region the peptide is more likely to be exposed on the surface. By choosing the location carefully, the odds for obtaining a more versatile antibody are improved.
A synthetic peptide conjugated to a carrier molecule will nearly always illicit an immune response, even if it scores low in an antigenicity algorithm. What matters is obtaining an antibody that works in the intended application with the target protein. As far as we know there are no studies made that shed any light on whether or not these algorithms of antigenicity have any bearing on success rate. Nevertheless, used with some discretion, a prediction of antigenicity may very well be included in the final selection of suitable sequences.
The typical length of a peptide used for antibody generation is about 10-25 AA. We tend to choose sequences of about 15 AA and this has proved to work very well. One could argue that longer sequences have more structure and are more similar to the target protein. On the other hand, the structure of the free peptide may be quite different from the same sequence as it sits within the protein. If we consider that antibodies raised against peptides should target short linear and continuous epitopes, then it makes sense to use shorter peptides. If the goal is to obtain antibodies that depend on a point mutation, for example an antibody that can discriminate between the phosphorylated and non-phorsphorylated state of a specifc residue, then a longer sequence would expose more trivial epitopes while a shorter sequence would direct the response to epitopes spanning the target site.
In principle, any peptide can be made for you. However, if there is a choice, all things being equal, then the risk for potential delays can be minimized by avoiding obvious issues. For example, even though a sequence is deemed to be soluble, as a whole, it can contain a very hydrophobic stretch, which may be an issue in the production of the peptide antigen. The question of whether or not there are peptide chemistry issues is usually better left to the lab that should make the peptide.
You would like the anti-peptide antibody to be specific for the target protein with no cross reactivity with any arbitrary proteins. However, this is not possible to use as a design criterion since any random sequence you choose is likely to generate relevant hits when running a BLAST search. If you get no relevant hits, beware! Did you turn off the low complexity filter? Are you using the relevant similarity scoring matrix? Are all proteins for the organism included in the database that is searched?
Still, the risk for potential cross reactivity is a design consideration that should be addressed. You should do what can be done, and this is to avoid similarities with a limited number of specific sequences for which you know beforehand that cross reactivity will be a serious issue. It is also wise to avoid repeats, and other sequences of low complexity, GTP binding sites, RGD recognition motifs, etc.
Innovagen was founded in 1992. We have designed numerous antibody generating peptides in that time. We use our own proprietary software PeptideCAD™ as a supportive tool for our experienced personnel in their selection of immunogenic peptide antigens.
+46 46 286 20 60
+46 46 18 81 71