When designing peptides as immunogenic antibody generators, the first and foremost thing to remember is that these peptides mimic linear consecutive epitopes of the target protein. This means that smaller epitopes become more important.
As a trivial example, let’s consider the ends of the peptide. The N-terminal amino group carries a +1 charge and the C-terminal carboxyl group has a -1 charge. When the sequence is located within the protein, these end-point charges do not exist. Thus, in order to better mimic the native protein, these charges should be blocked. The terminus at which the peptide is conjugated to the carrier molecule however is not a problem – this can be left with a charge.
A conformal or discontinuous epitope concerns the three dimensional structure of the protein, where amino acids from different parts of the protein sequence may form an antibody target. These epitopes cannot be mimicked using peptides. Thus, antibodies raised against synthetic peptides will target linear continuous epitopes. What then is the optimal length of a peptide that is to be used for immunisations? It should not be too short, since the yield of specific antibodies may be affected. 10-20 amino acids have proven to work very well, and when we design peptides they are typically 13-14 AAs in length. Longer peptides will work, but they are perhaps not cost effective. Furthermore, although longer peptides have more structure, there is nothing to say that this structure will be beneficial. Considering the fact that anti-peptide antibodies target linear epitopes, a shorter peptide should suffice – as experience shows us that it does.
The desired or unwanted potential cross reactivity is a point of concern for many researchers. From experience we know that it may only take a difference of one amino acid in order to obtain an antibody specific for the target protein. We also know that an antibody may cross react with a seemingly very different sequence. It may seem like a daunting task to take potential cross reactivity into account when designing peptides, but it should be considered and the pragmatic approach is to simply create the best conditions for the desired outcome.
The key to a successful peptide design, with respect to potential cross reactivity (desired or not), is to consider only a limited set of proteins. Simply put, as more sequences are analysed, the signal-to-noise-ratio, SNR, dwindles to a point where no sequence will stand out as being better or worse than any other. If one can reduce the number of proteins for which cross reactivity should be avoided (or obtained), the SNR will improve and this is something that can be taken into consideration in the design work.
When you have decided on a set of proteins the standard method is to create a multiple alignment in order to reveal similarities. But this does not necessarily provide the information you require, or worse, it may give you a sense of false security. It is important to keep in mind that synthetic peptides mimic short linear epitopes. Thus a given sequence could be potentially cross reacting regardless of where it is found. For example, if there is a significant similarity between the C-terminus of the target protein and the centre of one of the non-target proteins, this will not show in an alignment. In our proprietary software PeptideCAD™ however, these similarities are found.
BLAST is a wonderful tool, but perhaps overrated as an all important design parameter for peptides that are to be used as antigens for antibody production. Although short sequences like peptides can be used, “The program compares nucleotide or protein sequences to sequence databases and calculates the statistical significance of matches. BLAST can be used to infer functional and evolutionary relationships between sequences as well as help identify members of gene families”. As mentioned before; peptides mimic short linear epitopes and thus statistics becomes the main reason for treating BLAST results with some scepticism.
Consider the case where your species of interest, and its neighbours, are poorly represented in the database. You will only get a few hits, all insignificant, which you might want to interpret as your sequence being unique and likely to generate antibodies specific for the target protein. But since most proteins are not considered in this search, the result is inconsequential and you may very well end up having issues of cross reactivity – or misinterpret data since you feel confident that there are no potentially cross reacting proteins. Conversely, the case when you study a protein in a model organism such as Arabidopsis or Drosophila, any sequence you BLAST will show significant hits other than that which comes from the target protein.
Innovagen was founded in 1992, at a time when anti-peptide antibodies were not as common as they are today. Since then we have provided the research community with numerous antibodies raised against synthetic peptides. You will benefit from our software PeptideCAD™, our peptide design expertise and our commitment to quality, both crucial to any anti-peptide antibody project, when you turn to us for your polyclonal or monoclonal antibody requirements.
+46 46 286 20 60
+46 46 18 81 71