Bioinformatic analysis
The first step of making the engineered Μ13 bacteriophage involved the discovery of the peptide sequence which would be able to bind to each of the volatile organic compounds.
Phage display biopanning experiments were considered for this purpose. After consulting with experts, we eliminated the option, due to the high cost of the purchase of the peptide library and the limited laboratory experience of our wet lab members.
Instead, we decided upon using molecular simulations (molecular docking), to discover the specific sequences. Soon after, our dry lab members started the process of the discovery of the sequences, as you can see on our “Dry Lab” page.
After going through many cycles of “design-build-test-learn”, five peptide sequences gave the most promising results:
DMWVK - Octadecanal
EIPWT - Perillaldehyde
EWWWA - Hippuric Acid 1
PCR strategy setup (Design #1)
After discovering our candidate sequences, the next step was to introduce them to the pVIII protein of M13 bacteriophage, near the N-terminal. In order to do so, we first translated the peptide sequences into DNA sequences that needed to be introduced to the M13KE genome.
We then needed to decide upon the strategy of introducing the 15bp long DNA sequences of the pentapeptides into the M13KE DNA. After conducting research and consulting our advisor and instructor, we decided to use PCR in order to remove the DNA sequence of the second, third, fourth and fifth amino acids (counted from the N-Terminal) and introduce the DNA sequence (that corresponds to the desired 5 amino acids) into the gene of pVIII protein. PCR presents the following benefits: It is a cost efficient method, many different primers could be used using the same reagents, and also PCR mutagenesis had already been used to modify pVIII protein of M13 bacteriophage in literature [1].
PCR for mutagenesis of the phage involved the usage of two overlapping primers. Both primers would include the PstI recognition sequence, while the forward primer would also include the coding sequence for our chosen amino acid sequence. The first amino acid of the final protein would still be Alanine (as in unmodified M13KE), and the following 4 amino acids would be replaced by a pentapeptide of our choice. The only limitation would be that the first base of the coding sequence for the second amino acid would need to be Guanine, since it is also the final base of the PstI recognition sequence (CTGCAG).
Since PstI was used to re-circularize the vector, the PstI restriction site in the original sequence needed to also be removed. The removal of PstI restriction site is done using site-directed mutagenesis using commercial kits.
Thus we needed to create two pairs of primers:
One pair that needed to remove the original PstI restriction site and the other pair that is used for the
introduction of the desired peptide sequence.
Codon choices were made in order to minimize hairpin, homo and hetero dimers, using the IDT OligoAnalyzer tool.
The primers that we constructed and ordered are shown below:
PstI removal primers:
5′-CAAGCTTGCATGCCAGCAGGTCCTCGAATTCACTG-3′ (Forward)
5′-CAGTGAATTCGAGGACCTGCTGGCATGCAAGCTTG-3′ (Reverse)
pVIII insert mutagenesis primers:
*same reverse primers for all of the mutagenic sequences
5′-CACCCTCTGCAGCGAAAGACAGC-3′ (Reverse)
5′-ATATATCTGCAGATTGGTGGTGGCAGCCCGCAAAAGCGGCCTTTAACTCCC-3′ (eicosane forward)
5′-ATATATCTGCAGATTGGTGGTGGCCCCCCGCAAAAGCGGCCTTTAACTCCC-3′
(octadecanal forward)
5′-ATATATCTGCAGTGTGGTGGTGGGCACCCGCAAAAGCCGCCTTTAACTCCC-3′
(perillaldehyde forward)
5′-ATATATCTGCAGAATGGTGGTGGGCTCCCGCAAAAGCGGCCTTTAACTCCC-3′
(hippuric acid forward)
Unmodified M13KE DNA with one PstI restriction site in lacZa gene
PstI restriction site is removed by the first PCR reaction and the DNA is recirculated.
M13KE with PstI restriction site removed
The second PCR reaction is used to create the modified sequence while simultaneously introducing two new PstI restriction sites for the DNA to be able to recirculate by restriction and ligation reaction. The restriction reaction with PstI, serves to remove part of the overhang that doesn’t belong to the modified bacteriophage, but only serves as a stabilizing factor for the mutagenic primers.
Two PstI restriction sites and the peptide sequence introduced by second PCR reaction
Modified M13KE DNA re-circulated by PstI restriction and ligation reaction.
After recircularization of the DNA, the modified M13KE DNA is ready to transform bacteria and to produce the engineered bacteriophage.
The success of the ligation and transformation can be determined by the presence of blue plaques in IPTG/X-gal agar plates. Τhe M13KE DNA contains the lacZa gene, so bacteria that have been infected by the M13KE DNA, will create blue plaques in the plate and will be differentiated from white colonies that belong to bacteria without.
Redesign of PCR primers (Design #2)
Since our PCR attempts with the primers we designed did not have the outcome we expected, we agreed upon redesigning our experimental setup. This process was conducted by both dry and wet lab members and was based both on advice from members of the genetics and biotechnology sector, including our advisor, and on a collective study of our failed PCR attempts.
From a holistic view of our results, we came to the conclusion that the most likely reason for failure was the absolute complementarity of the primers. This was based mostly on the fact that in the majority of our attempts, we observed an unexpected quantity of primer dimers and got no distinct bands of PCR product. The tendency of the already used primers to form hetero-dimers was re-calculated using “IDT oligoanalyzer” and was found to be indeed high, probably due to their very high complementarity. This means that the primers needed to be redesigned, taking into consideration their tendency to create hetero-dimers. The same mutagenesis strategy of removal and introduction of PstI sites was followed, to facilitate the substitution of the original sequence with the new.
Since we didn’t want to waste any more time with potentially unsuccessful future PCRs, we made three different pairs of primers with different sequence, length and GC content, aiming to simultaneously test all three. All the newly designed primers had a lower tendency to form hetero-dimers, and their sequences were:
PstI removal primers:
5’-CAAGCTTGCATGCCAGCAGGTCCTCGAATTCACTG-3’ (new forward 1)
5’-GTGAATTCGAGGACCTGCTGGCATGCAAGCTTGGC-3’ (new reverse 1)
5’-AGCTTGCATGCCAGCAGGTCCTCGAATTCACTGGC-3’ (new forward 2)
5’-GAATTCGAGGACCTGCTGGCATGCAAGCTTGGCGT-3’ (new reverse 2)
5’-CTTGCATGCCAGCAGGTCCTCGAATTCACTGGCCG-3’ (new forward 3)
5’-ATTCGAGGACCTGCTGGCATGCAAGCTTGGCGTAA-3’ (new reverse 3)
The primers of pVIII insert mutagenesis were kept the same as the initial ones.
Redesign of direct cloning setup (Design #3)
After a series of unsuccessful PCR attempts which are recorded in detail in the “Lab Notebook”, we decided that we needed to change our strategy. We considered inserting the desired sequence in the pVIII protein gene, by designing and ordering Gene Fragments. Specifically, the desired sequence would be incorporated in a larger insert fragment that corresponds to the M13KE sequence between two restriction enzyme sites. The only difference with the original sequence, would be a set of 15bp that represents the desired genetic modification.
We decided to design the shortest insert sequence possible for it to be cost-effective and for it to be produced faster. Thus we examined the M13KE genome, and we decided upon the enzymes BsrGI and KpnI. Fragments inserted with BsrGI and KpnI cloning sites have a length of 593bp.
Insertion cloning using BsrGI and KpnI.
In order to be able to differentiate between modified and unmodified M13KE DNA, we decided to introduce to the insert fragments a new HindIII restriction site through silent mutation inside “M13 geneIII”. Thus, genetically engineered M13KE DNA, when digested with HindIII, produces two fragments (4649bp and 2573bp), while non-engineered M13KE DNA produces one fragment of 7222bp.
M13KE DNA with insert fragment highlighted in blue and modified sequence highlighted in red. New HindIII restriction site highlighted in yellow.
Bands produced by digestion of M13KE DNA with HindIII. 1:Modified M13KE DNA produces two bands 2:Unmodified M13KE DNA produces a single band
Upon ordering the sequences from Twist Biosciences, we decided to add “adapters”, meaning that the length of the fragments increased to 614bp. The surplus of base pairs will later be removed through double digestion with BsrGI and KpnI.
The ordered sequences were identical with the M13KE genome between BsrGI and KpnI sites, with the difference of 15bp in the “M13KE gene pVIII”. This difference accounts for the altered pVIII sequence, modified accordingly to each VOC-detecting peptide.
In conclusion, we designed a 614 bp long cassette with BsrGI and KpnI recognition sites on its edges, and added an additional HindIII site in order to distinguish modified phages from unmodified M13KE. Modified phages that received the cassette would be cut twice by HindIII, resulting in two linear fragments, while unmodified phages would be cut only once. We added each VOC binding peptide’s coding sequence in the pVIII gene.
Furthermore, we also designed a larger (2624 bp long) cassette with HindIII and KpnI restriction sites, in order to test the impact of insert size on transformation efficiency. In this sequence, we removed the EcoRI and PstI restriction sites, and added the hippuric acid binding peptide (EWWWA) coding sequence. Again, the removed restriction sites would also serve as a way to differentiate between engineered and non-engineered phages. To improve DNA synthesis efficiency, we made silent mutations in order to remove some of the repeated sequences GGCTC and TCGCA that were present in the designed insert sequence (IDT’s recommendation).
Image of larger designed insert fragment of 2603bp highlighted in blue and modified sequence highlighted in red. EcoRI and PstI sites were removed.
Testing methodologies
In order to test the binding affinity of our modified phage and thus our modified protein with the corresponding ligands (VOCs), we came to an agreement to use two different testing methods.
1) Tryptophan fluorescence assay
After consultations from professors and members of the department of Chemistry of the University of Athens, we decided that this test was our best option considering the time and cost limitations. Other methods we examined were 1H NMR (Proton nuclear magnetic resonance), C-13 NMR (Carbon-13 nuclear magnetic resonance) and SPR (Surface plasmon resonance). We were discouraged from using 1H NMR, C-13 NMR and SPR for reasons stated below:
1H NMR, C-13 NMR : Due to the size of the phage, such a test would not work efficiently. The abundance of carbon molecules and protons of the phage would provide an image of a very dense and consequently uninterpretable spectrum. It would also be really difficult to gain access to such technology, since there is no NMR instrument in the department of biology.
However, this test could be possible in the case of instead of using the whole bacteriophage we used only the short peptide we had inserted in its capsid protein of pVIII. Such a test is accompanied by high cost of producing the short peptide and by long production time.Furthermore the inserted peptide sequence-VOC affinity test, does not ensure that the whole pVIII protein or the entire M13 Phage would exhibit the same or even similar behavior since the peptide won't be deployed in the same environment as it will when fused with pVIII protein.
SPR → We were informed that there is no such technology available in Greece at the time although this could have been theoretically a very reliable test.
On the other hand Tryptophan fluorescence assay is based on the principle that tryptophan, which is usually buried in hydrophobic sites of proteins or bound to ligands through hydrophobic interaction, is selectively excited at 295 nm and results in giving an emission spectrum with a 355 nm peak. It is also documented that structural changes near Trp residues, induced by ligand/protein interaction, can cause detectable alterations. These are commonly expressed as differences in fluorescence intensity or as wavelength shifts in the emission spectrum [2],[3]. Thus since the bacteriophage itself and most of the chosen peptides had Trp, the assay could be used for qualitatively testing the potential binding of pVIII with the VOCs.
While brainstorming we came in the place to answer the following questions:
1) Does the protein have one or more tryptophan(s) that are likely to be in different environments when the protein is free versus bound? → YES
2) Is there an absence of tryptophan(s) in the ligand? → YES
3) Does the ligand absorb at or near the excitation or emission wavelengths of tryptophan at the concentrations required for the titration study? → NO (maybe hippuric acid, but the inner filter effect is not significant at the concentrations we used)
4) Are the protein, ligand and the complex soluble at the concentrations required in the solution condition used for the titration? → YES (we increased the solubility of eicosane, perillaldehyde, octadecanal using DMSO).
All of our answers showed that the M13 phage, the VOCs and their complex, fill the criteria to try this testing method. Apart from that, it is a fast, low cost and easy method to depict whether our phages interact with the desired VOCs.
2) Colorimetric output of self assembled phages on a gold-coated silicon wafer
This testing procedure is at the crux of our project: In fact, it is our method of choice for utilizing M13 bacteriophages in order to detect Volatile Organic Compounds of interest. This method relies on specific micropatterns that arise when a gold-covered substrate (gold- coated silicon wafer in almost all cases) is submerged into a purified solution of M13 bacteriophages and then pulled at precise speeds (at the scale of micrometers per minute). The VOC detecting pattern is, in our case, a smectic helicoidal nanofilament pattern, which arises when the phage solution is between 4-6 mg/ml [4], [5]. The bacteriophages can be genetically engineered to express VOC - binding amino acid sequences on their pVIII capsid protein. When this interaction takes place, the phage bundles swell, decreasing the interspacing between them, causing optical effects like coherent scattering to take place and exhibiting distinctive color change [6]. This creates essentially a tunable and reversible structural coloration-based biosensor.
This method was chosen due to several of its advantages:
1) It is a gas sensing method, with extensive literature on VOC detection (not for our own compounds or purposes) [6].
2) It creates a tunable sensor, able to detect many compounds depending on the specific genetic modification of the bacteriophage.
3) Once created, the detection is fast, and since no chemical reaction takes place, highly repeatable. [7]
4) Performing it during project development would provide us with crucial information on its behavior as a finished product, and would evaluate our whole process at once.