Public DNA/protein databases now contain enormous amounts of DNA sequence information from a vast array of species. These databases contain the necessary software to transcribe, this information into the amino acid sequences of the corresponding proteins.

A team of scientists from CSL, The RIKILT-Institute of Food Safety, IFA Tulln and r-Biopharm have been investigating an approach (http://www.srmtest.com) to develop immunotests for specific cooked animal tissues, (Specified Risk Material (SRM)), based on the sequence information of proteins in these databases.

Local databases of bovine, ovine, caprine and porcine protein sequences have been constructed from the public protein databases and proteins have been identified from these that appeared unique to the particular tissues of interest, (brain, eye, tonsil, ileum, spleen). The protein sequences have been cut in silico into 10mers and the sequence of these short decapeptides used to search the main public databases (a BLAST search) to determine whether these were unique to the protein. Selected decapeptides that were confirmed as unique were then examined in silico using software tools to determine whether they were both resistant to specific protease digestion and were potentially good antigens.  Peptides that met these criteria were synthesised and the synthetic peptides used to raise monoclonal antibodies for immunoassay development. The planned assay format involved first digestion of the cooked meat sample with the selected protease leaving the target peptide sequence in tact. The digested peptides were then assayed using the developed immunoassay.
Studies identified nine proteins in the animal databases that appeared specific to the tissues of interest. Twelve peptide sequences that retained theoretically specificity to one of the four SRM tissues; brain, eye, ileum or tonsil, were selected from these proteins. Only two of these peptides were theoretically also specific for the animal species, based on the information in the databases at the time. However later database interrogation showed one of these peptides also matched both the bovine and ovine protein, due to additional sequence information entering the public databases.
The peptides were commercially synthesised and used to raise monoclonal antibodies. Only 30% of the peptides produced antibodies specific to the free synthetic peptide, as measured by competitive ELISA. Of these, one antibody, specific to eye tissue, showed specificity for cooked eye tissue after digestion with pepsin. The assay was able to detect cooked eye tissue from a range of species at the 1% contamination level in the presence of other cooked meat products in a competitive ELISA.
While there are still a lot of questions to be answered in this approach to antibody production the study has so far shown that an in-silico approach to identifying targets for specific immunoassays is possible. With the growth of the public DNA/protein databases the identification of useful targets is becoming more specific, and in-silico digestion of peptides to identify proteolytic methods of antigen release from complex cooked matrices has been shown to work. Problems still remain regarding efficiency of antibody production to peptides by traditional means, avidity of the subsequent antibodies and possible loss of antigenicity of peptides due to severe heat damage to the protein sequences. However some of these problems will be addressed in the present SAFEED –PAP project.