Team:AFCM-Egypt/Engineering
Introduction
This year challenged us to think of new ways of engineering our project under heavy constraints. In order to work on our engineering we decided to go by the book and follow the engineering design cycle, so in our journey we went into many phases of trial & error. However, the way we implemented the Design, Build, Test cycle was heavily impacted by our surrounding circumstances.
Figure. illustrates the iterations that our team went through in the
process of engineering our vector.
Nevertheless, we attempted to adapt to the new norm by thinking outside of the box and building upon/designing tools and resources to aid us in the conceptualization of our project.
Throughout this season, we kept revisiting our design, consulting various experts, resources and taking the results of our modeling and simulations into account. We went through multiple iterations in order to finally reach a successful engineering of our project. In this page, we depict the multiple iterations of our design as guided by our engineering efforts. To view our finalized suggested design, please consult our design page.
Before we jump into details, i’d like to summarize all what you are going to see in the next
pages. This year we decided to use a replicon-based plasmid, and a replicon is a region of DNA
or RNA, that replicates from a single origin of replication. Which is an essential component in
multiple viruses such as alphaviruses like EEV & SFV to replicate their structural genes.
A replicon mainly consists of four non structural proteins (NSPs) that are the main part that
cause the replication.
This year we decided to use replicons for their self amplification abilities & high expression
of a desired gene or construct.
First we are going to show you our initial design and ideas followed by improvements and
additions that were added to them later on after integrated practices meetings
Finally we are going to show what we are planning to assemble to get experimental data so we can
get the best efficacy possible out of our vaccine.
Now let’s Dive in.
Designing of parts and Fragments
Initial Design
We planned to create two vectors
1.The first vector was made as a modular design that represented our circuit in which its main parts were
| Biobrick | Type | Role | Length |
|---|---|---|---|
| BBa_K3504000 | Coding | Nsp4 for replicon | 1605 |
| BBa_K3504001 | Coding | Nsp4 for replicon | 2382 |
| BBa_K3504002 | Coding | Nsp4 for replicon | 1671 |
| BBa_K3504003 | Coding | Nsp4 for replicon | 1821 |
| BBa_K3504004 | Regulatory | Subgenomic promoter | 30 |
| BBa_K3504005 | DNA | Malat-1Triplex | 132 |
| BBa_K3504006 | Regulatory | mi-RNA FF6 | 64 |
Figure 1. Initial circuit design
Figure 2. Initial circuit SBOL design
Replicon design & Optimizations:
The main component of our replicon-based plasmid is the 4 non structural proteins by which we can attain self-amplification of a downstream construct which would highly increase the expression of our vaccine using the parts (BBa_K3504000,BBa_K3504001,BBa_K3504002,BBa_K3504003). Which allowed an increased expression which was proven from literature and when combined with the part BBa_K3504004 as a Subgenomic promoter demonstrated the following expression results.
Figure.3 Number of positive strand RNA Before and after self replication
We also found that we can improve our EEV alphavirus replicon amplification ability by addition of selective mutations to the non-structural proteins all of which are shown in Table(2) & Figures (4-9) below.
| Mutation Region | Nucleotide Mutations | Amino acid Mutations |
|---|---|---|
| Nsp2 | A2619G | G110G |
| Nsp2 | G4576C | G763R |
| Nsp3 | A4951G | K94E |
| Nsp3 | A5399G | S243G |
| Nsp3 | G5436T,G5584A | E255D,V305M |
Initial Switch design
We designed an on-switch that was designed to further increase expression, stability and provide protection from silencing. This switch had an “OR” logic gate to work this on switch was activated by either one of two miRNA:
-
the first of was mir-FF6 which was chosen according to data provided from literature that showed that it had highest expression in usage on cases of an on switch demonstrated in the figure 10.
- The second of them we choose it to be a prognostic marker for TNBC, our chosen miRNA was mir-373 and we choose it after comparing it with multiple candidates as we found it to be highly upregulated in TNBC patients and for it’s high stability as it had a minimum free energy of -36.40 kcal/mol.
Figure 10. miRNA Fold change between various miRNAs, Showing that all of them can act as
an OFF motif but only few can actually act as an ON motif.
Figure11.TNBC prognostic marker mir-373 RNA fold
And this switch contained the part BBa_K3504005 (Malat 1 triple helix) as a stabilization domain to replace the function of the Poly A tail. Finally we used multiple terminator sequences to be used as a degradation domain.
Current design & improvements
Figure12. Final circuit Design
Figure13.Final Sbol Design for the circuit
Parts optimization
MiRNA Optimization & binding site interaction
Methodology
First of all, we obtained DNA sequences of miRNA templates used in pPRIME-based cloning from
supplementary information on nature website which are FF3, FF4, FF5 and FF6.And by using RNAfold
Webserver tool, RNA sequence of the miRNAs was obtained in addition to their folding and
modeling with the energy of the thermodynamic ensemble (Kcal/mol).
Consequently, we selected the best base-pair probabilities and by using another tool called RNA
reserve and complementary sequence generator, The basepair probabilities’ complementary sequence
was gained.
Eventually, The IntaRNA interaction tool was used to determine the stability and binding energy
of the miRNAs with the circuit.
Results
| miRNA | Sequence of miRNA | free energy of the thermodynamic ensemble | Best stability for RNA interaction | Sequence of binding site |
|---|---|---|---|---|
| FF3 | UCGAGUGCUGUUGACAGUGAGCGCACGAUAUGGGCUGAAUACAAA UAGUGAAGCCACAGAUGUAUUUGUAUUCAGCCCAUAUCGUUUGCC UACUGCCUCGGAG | -50.98 kcal/mol | -6kcal/mol | UCGAGUGCUGUUGACAGUGAGCGCACGAUAUGGGCUGAA UACAAAUAGUGAAGCCACAGAUGUAUUUGUAUUCAGCCC AUAUCGUUUGCCUACUGCCUCGGAG |
| FF4 | UCGAGUGCUGUUGACAGUGAGCGCCGCUUGAAGUCUUUAA UUAAAUAGUGAACCACAGAUGUAUUUAAUUAAAGACUUCA AGCGGUGCCUACUGCCUCGGAG | -51.23 kcal/mol | -3.14 kcal/mol | AGCUCACGACAACUGUCACUCGCGGCGAACUUCAGAAAU UAAUUUAUCACUUCGGUGUCUACAUAAAUUAAUUUCUGA AGUUCGCCACGGAUACGGAGCCUC |
| FF5 | TCGAGTGCTGTTGACAGTGAGCGAAGCACTCTGATTTGA CAATTATAGTGAAGCCACAGATGTATAATTGTCAAATCA GAGTGCTTTGCCTACTGCCTCGGAG | -53.24 kcal/mol | -10.99kcal/mol | AGCACUCUGAUUUGACAAUUAU |
| FF5 | TCGAGTGCTGTTGACAGTGAGCGAACCAAAGAGAT TCCTCATAAATAGTGAAGCCACAGATGTATTTATG AGGAATCTCTTTGGTTTGCCTACTGCCTCGGAG | -53.34 kcal/mol | -3.92kcal/mol | CGCUUGGUUUCUCUAAGGAGUAUUUAU |
Figure: This Design->Build->Test Cycle Illustrates the process of SGP
optimization & selection
Cse promoter Optimization
Methodology
Using RNAfold web server, we could predict secondary structures of single stranded RNA or DNA sequences and obtain the free energy of the thermodynamic ensemble. Not only those but also we could obtain a graphical output for the inserted sequence.
We used 3 sequences for SGP with 3 different lengths. Each one of them has 6 expressions; 3 expressions are measured by mVenus and the other 3 are measured by mKatewe and could obtain the energy for each one of them by using RNAfold web server.
Eventually, we correlated the expressions of the sequences with the stability and another time with the length by using the Pearson Correlation Coefficient Calculator tool.
For instance, by using RNAfold web server to obtain the free energy of the thermodynamic ensemble for SGP41 with sequence (CCTGAATGGACTACGACATAGTCTAGTCCGCCAAGGCCACC) it was
-13.82 kcal/mol. And the same for each SGP till we get all free energies. Then we correlate them with their expression using the Pearson Correlation Coefficient Calculator tool and again with the length
Figure (1)
Figure (2) The RNAfold of the sequence which has the least energy which is -13.82 kcal/mol.
| Sequence | Length of sequence | Stability | Expression mVenus | Expression mKate |
|---|---|---|---|---|
| CCTGAATGGACTACGACATAGTCTAGTCCGCCAAGGCCACC | 41 | -13.82 | 0.05 | 0.13 |
| CCTGAATGGACTACGACATAGCCACC | 26 | -3.18 | 0.07 | 0.18 |
| CCTGAATGGAGCCACC | 16 | -1.02 | 0.04 | 0.16 |
| CCTGAATGGACTACGACATAGTCTAGTCCGCCAAGGCCACC | 41 | -13.82 | 0.33 | 0.35 |
| CCTGAATGGACTACGACATAGCCACC | 26 | -3.18 | 0.04 | 0.55 |
| CCTGAATGGAGCCACC | 16 | -1.02 | 0.025 | 0.667 |
| CCTGAATGGACTACGACATAGTCTAGTCCGCCAAGGCCACC | 41 | -13.82 | 0.04 | 0.667 |
| CCTGAATGGACTACGACATAGCCACC | 26 | -3.18 | 0.04 | 0.69 |
| CCTGAATGGAGCCACC | 16 | -1.02 | 0.02 | 1.15 |
| R value for stability -0.1648 | R value for stability 0.3094 | |||
| R value for length 0.3159 | R value for length -0.3479 |
Results
The end result of the correlation coefficient between the stability and the expression by mVenus was R value equals -0.1648 and between stability and expression by mKate was 0.3094. On the other hand, the correlation coefficient between the length and the expression by mVenus was R value equals 0.3159 and between length and expression by mKate was -0.3479.
Project improvements based on Integrated practices
We have been so focused on the self-amplification property that we missed another important feature as after our integrated practices meeting with Dr.Noreen Wauford co-author of a promising platform for RNA regulation called PERSIST have pointed us to the fact that we need to consider the recognition and attack of immunity against our vector and mining literature for more ways to increase our circuits efficiency we came with the following decisions and conclusions that furthermore increased the total efficiency of our replicon vaccine.
we decided to add two protective mechanisms the first of which was to protect our vaccine from recognition and attack by Cytotoxic lymphocytes which could lead to the degradation of cells expressing our vaccine, so we added glycine-alanine repeats which has been proven to prevent this presented by the part BBa_K3504007 (Gly-Ala repeats) as we found that it would help protect our circuit from innate immunity attack.
Figure14.CTLs degradation rate after addition of Gly ala vs Before adding it
Figure15.Gly-Ala repeats amino acid sequence
The second was designing an OFF-Switch based on a modular NOR gate which depends on two miRNAs as its inputs, the first of them being mir-126, which would help protect our vector from being recognized and attacked by innate immunity if transfected to antigen presenting cells like DCs as these cells express high level of mir-126 which would activate the switch and stop the replication in these cells while preserving it in other cells.
And the other input being synthetic microRNA FF4 which can be administered to stop the replication and work as a safety switch.
But We choose this microRNA between multiple choices as it experimentally showed strong downregulation when used as off switch and furthermore for its highly stable structure as seen in figure 10
With their binding sites being The part BBa_K3504008 (Mir-126 binding site) and the BBa_K3504019 (FF4 Binding site).
Figure: This Design->Build->Test Cycle Illustrates the Immune evasion steps through Integrated
practices.
Finally we decided to remove the ON-switch as its ability to protect from silencing was replaced by Gly-Ala repeats and the OFF-switch.
Assembly of plasmids
WFirst we decided to assemble two variants of our vaccine, the first one to be made using Equine Encephalitis Virus (EEV) Non Structural Protein replicon parts and the second using Semliki Forest Virus (SFV) parts with minor differences in design.
We decided to use Golden Gate assembly cloning system in the assembly of EEV plasmids while using Simple restriction-enzyme ligation system in SFV Plasmids.
We separated the circuit to 8 fragments
The first fragment contains Origin of replication and the AmpR and promoter.
The second fragment contains CMV promoter , FF4 binding site & mir126 binding site.
The 3rd to 6th fragments contained NSP1,NSP2,NSP3,NSP4 in that order with every fragment containing only one NSP.
The 7th fragment contained SubGenomic promoter, B-defensin, Badri and most importantly the chosen epitopes to our vaccine.
The 8th and final fragment contained reporter genes and enhanced p2a+gsg And when designing our reporter gene, we found that we could enhance its function by adding p2a cleavage site separating the vaccine from the reporter gfp. And even further more increase its cleavage ability by adding 3xflag & v5tag upstream of it.
In order to choose between multiple design variations and select the best and most efficient one we assembled multiple plasmid designs to experimentally & computationally compare between them due to the lack of lab availability due to Covid-19 pandemic.
The factors we had to compare between were :-
- Different versions of Multi-Epitope vaccine
- Subgenomic promoter optimization
- Off switch variants & the efficacy of mir-126
Figure: This Design->Build->Test Cycle Illustrates the process of alphavirus type selection
Multi-epitope vaccine optimization:
First we are planning to design multiple versions of our vaccines to validate the best form of the vaccine to be used as personalized immunization techniques according to using pSFV3 plasmid assembled with different versions of the vaccine using in them SGP15 as the Subgenomic promoter.
Multi-epitope Vaccine V1
Multi-epitope vaccine v2
Multi-epitope vaccine v3
Nor Gate testing & Mir126 characterization
In efforts to characterize mir126 effects we are planning to assemble two plasmids one with the mir126 binding site and one without it to experimentally show the effectiveness it adds to the circuit.
Sgp Optimization
Finally we are planning to assemble 2 plasmids to experimentally decide which subgenomic promoter would show higher expression abilities in which would be identified using reporter genes.
However using mathematical modelling and simulation techniques we found out that SGP15 had higher expression values which was confirmed from literature presented in Figure16 to be confirmed after experimental testing .
Figure 16.Sgp 30 expression vs SGP15
References
-
DiAndreth, B., Wauford, N., Hu, E., Palacios, S., & Weiss, R. (2019). PERSIST: A programmable RNA regulation platform using CRISPR endoRNases. bioRxiv.
-
Wagner, T. E., Becraft, J. R., Bodner, K., Teague, B., Zhang, X., Woo, A., ... & Sanders, N. N. (2018). Small-molecule-based regulation of RNA-delivered circuits in mammalian cells. Nature chemical biology, 14(11), 1043-1050.
-
Hofacker, I. L. (2003). Vienna RNA secondary structure server. Nucleic acids research, 31(13), 3429-3431.