Construction and Test of the engineered bacteriophage
Multiple attempts utilizing different strategies to create the modified M13 bacteriophages were performed but unfortunately we didn't manage to isolate and test the M13 bacteriophage particles.
Nevertheless after chemical transformation of E.coli cells with the ligation mix of the cloning reaction. Blue plaques were observed and DNA was isolated. From the results of the restriction of the isolated DNA with HindIII we assume that we have produced the modified M13 bacteriophage for the detection of Hippuric Acid and Perillaldehyde as seen below:
Undigested isolated DNA from blue plaques of chemical transformation of ligation mix with E.coli cells.
Digested DNA with HindIII for confirmation of the modification of the bacteriophage.
Modified M13 DNA is expected to produce 2 bands at 4600 kb and 2600 kb when digested by HindIII. These bands are observed in the current electrophoresis in samples 4 (for Hippuric Acid) and 13 (for Perillaldehyde). It should be noted that the two bands did not coincide with the ones before HindIII digestion. Therefore, highlighted bands serve as an indication of the possibility of having successfully modified the bacteriophage since the highlighted bands seem to correspond to the 2600bp fragment produced by HindIII restriction while the other fragment of size of 4600bp is also present.
Data Analysis
Targeting VOC Specificity:
Our primary objective was to identify peptides with a high binding affinity for specific volatile organic compounds (VOCs) while maintaining low affinity for other non-target VOCs. In our pursuit of achieving this, we focused on finding peptides that possessed three critical characteristics:
1. High Affinity Score:
We sought peptides with a high affinity score towards our target VOC, ensuring effective binding.
2. Selectivity:
It was imperative for these peptides to exhibit selectivity when compared to other VOCs of interest.
3. Discrimination from unrelated compounds:
To distinguish the target VOC from other compounds, we emphasized the need for peptides with remarkable discrimination abilities.
Thus, we considered the presence of other VOCs commonly found in human sebum, making the task more intricate.
Iterative Search Process:
Transitioning from tripeptides to tetrapeptides, we embarked on a labor-intensive quest. With countless potential tetrapeptides, we had to devise a strategic approach. Given the extensive computation time required to analyze all tetrapeptides with our ligands, and our relatively low computational power, we decided to employ targeted searches based on patterns emerging from previously tested peptides. Using Python code, we identified trends, such as an overrepresentation of tryptophan (W) in eicosane results, and adjusted our strategy accordingly.
In addition to these criteria, we set specific "rules" for our target peptide:
1) It should initiate with Alanine.
2) The second position should contain A, V, D, E, or G.
Notably, we included Alanine as the first amino acid to mimic the cleaved pVIII protein realistically. We proceeded to explore pentapeptides and hexapeptides in our search, generating hexapeptide derivatives based on our top-performing tetrapeptides. Nevertheless, no significant correlations were found between tetrapeptides and their hexapeptide derivatives, necessitating the generation of random hexapeptides for further analysis.
Building the Peptide-VOC Database:
With extensive experimentation and filtering, we organized our results into a comprehensive peptide-VOC binding affinity database, consisting of a staggering 298,711 unique peptides. More information about it can be found on our Software page.
Data Analysis and Optimization:
To maximize the utility of this database, we conducted a series of data analyses:
Binding Affinity Distribution:
We plotted the binding affinities for each VOC and calculated the mean, providing insights into the overall binding characteristics.
Amino Acid Favoritism:
Investigating the top 10% of peptides for each VOC, we sought clues about amino acid preferences.
Optimal Peptide Length:
We determined the peptide length group (from tripeptides to heptapeptides) with the highest absolute binding affinity.
VOC Ranking:
We evaluated which of the four VOCs exhibited the highest absolute mean binding affinity with the tested peptides.
Hydropathicity Correlation:
Exploring the correlation between hydropathicity (using the Kyte-Doolittle scale) and binding affinity, we aimed to uncover any relationships, but found that the Pearson correlation was consistently close to zero for all four VOCs.
Experimental Validation:
Having distilled our results, we selected the most promising peptides for each VOC and examined their binding affinity with the remaining three VOCs. Subsequently, we rigorously tested the 4 selected peptides with a total of 22 common skin volatiles.
We went a step further by modifying the protein through the incorporation of the selected peptides, followed by testing the modified proteins against the wild type, providing valuable insights into the practical implications of our findings.
This comprehensive approach allowed us to navigate the complex landscape of peptide-VOC interactions and optimize the selection of peptides for a wide range of practical applications.
