Team:ZJU-China/Engineering
On this page, we introduce the engineering design cycle of how we developed Viruguard in chronological order. We made four rounds of improvement to Viruguard before we ultimately got a mature Viruguard product. We firstly constructed its virulence to cancer cells and then tried to shield normal cells from its virulence. Finally, an oncolytic virus, Viruguard, was produced, which had a high level of security and virulence respectively to normal cells and cancer cells.
Engineering success
The enginnering cycle
Synthetic biology is a subject that differs a lot from the conventional ones. In traditional life science subjects, researchers are supposed to use the scientific method to obtain accurate results, whose work mainly focused on making observations and performing experiments. However, synthetic biology goes beyond that, it is all about constructing and innovating a composite system from different parts and biobricks. Due to the distinctiveness of Synbio, an engineering cycle should be established from the beginning and undergoes continuous improvement with the results of experiments discoverd, which would finally give birth to a successful engineered product, namely a genetically engineered machine. And that was why we documented our engineering below which had helped us perfect Viruguard throughout the whole iGEM competition this year.
ViruGuard Development
Research 1: How can we use organism to kill the tumor?
In current hepatocellular carcinoma guidelines, traditional treatment including surgery, chemotherapy, and radiotherapy are still recommended more often than novel treatment. These methods directly eliminate or shrink the tumor in situ, and even with the development of medicine, limited changes could be made to these conventional therapies in a short term. In synthetic biology, however, we can engineer organisms according to our exact needs, which in this case, means HCC treatment. Thus, we set out to explore the appropriate viruses that possess the ability to infect cells and prokaryotic bacteria, which are widely used in synthetic biology, and could make it easier to conduct following experiments.
Imagine 1: Which type of organism is better for the project?
At the very first, we couldn't make our mind on what kind of virus to choose as our chassis organism.
Generally, we focused on the most normally used organism, Escherichia Coli, to be the possible chassis organism. As it has already been deeply and widely studied before, it might be easy for us to conduct transformation on the bacteria, and shape it according to our expectations. Besides, as E.coli couldn't infect human cells directly, the risk of safety incidence is relatively low. However, several problems had made us give up the idea. The difficulty of turning it into a eukaryon-infectable type of bacteria and the uncertainty of whether the procaryotic gene can function in the eukaryotes were the most challenging problems that we were incapable to solve.
Therefore, we broadened our vision and put the attention to the virus. Because of its natural characteristics, the viral gene could express in human cells, and there have already been some literature supporting the feasibility of virus treatment. Nevertheless, in previous studies, although researchers have extracted virus strains with relatively weak toxicity, current virus treatment still could pose grievous threats to animal models in experiments or patients in the clinical trials. So this is the first problem that we need to work out.
Design 1: How can we ensure Viruguard to kill cancer cells in a safe way?
The main focus of our design is on keeping the virulence of virus to kill malignant cells in a safe way.
We had thought of making our mechanism in a trascriptional, translational or epigenetical level and finally
we determined that the best solution of problem is using a peculiar promoter to regulate the expression of
gene-targeting siRNA. After we had a deeper understanding to
the RNAi mechanism, we noticed that it would not be functional if we merely tried to express the
siRNA sequence. Thus, we expressed the shRNAs in the cells which would further be processed into the
mature siRNAs.
Here are some results of function test of our designed shRNAs. From BBa_K3730023 to BBa_K3730030, they are all our shRNA primers used for RNAi killing mechanism.
BBa_K3730023 BBa_K3730024
BBa_K3730025 BBa_K3730026
To guarantee the security of our design, we decided to use tumor specific promoter to regulate the transcription of our designed shRNA sequence. We chose phTERT and psurvivin as the promoter to shoulder this responsibility because they have significant activity difference in the normal and cancer cells.
hTERT promoter Sequence 263bp
CAGCTGCGCTGTCGGGGCCAGGCCGGGCTCCCAGTGGATTCGCGGGCACAGACGCCCAGGACCGCGCTTCCCACGTGGCGGAGGGACTGGGGACCCGGGCACCCGTCCTGCCCCTTCACCTTCCAGCTCCGCCTCCTCCGCGCGGACCCCGCCCCGTCCCGACCCCTCCCGGGTCCCCGGCCCAGCCCCCTCCGGGCCCTCCCAGCCCCTCCCCTTCCTTTCCGCGGCCCCGCCCTCTCCTCGCGGCGCGAGTTTCAGGCAGC
BBa_K3730017
survivin promoter Sequence 397bp
ATTACAGGCATGCACCACCACGCCCAGCTAATTTTTGTATTTTTAGTAGAGACAAGGTTTCACCGTGATGGCCAGGCTGGTCTTGAACTCCAGGACTCAAGTGATGCTCCTGCCTAGGCCTCTCAAAGTGTTGGGATTACAGGCGTGAGCCACTGCACCCGGCCTGCACGCGTTCTTTGAAAGCAGTCGAGGGGGCGCTAGGTGTGGGCAGGGACGAGCTGGCGCGGCGTCGCTGGGTGCACCGCGACCACGGGCAGAGCCACGCGGCGGGAGGACTACAACTCCCGGCACACCCCGCGCCGCCCCGCCTCTACTCCCAGAAGGCCGCGGGGGGTGGACCGCCTAAGAGGGCGTGCGCTCCCGACATGCCCCGCGGCGCGCCATTAACCGCCAGATT
BBa_K3730018
Build 1: How can we build it?
After we assessed the difficulty of directly constructing a plasmid containing the tumor specific promoters and shRNA simultaneously, we decided to seperate them into two vectors to obtain a higher operability in our experiments, but still in the final virus package, shRNA would be put under the control of tumor specific promoter. We are going to build two vectors: the tumor specific promoter vector and the shRNA expression vector. We got the full sequence of htert and survivin promoter from human genome by PCR and replaced the original promoter sequence in pEGFP-C1 with the sequence that we got. As for the shRNA expression vector, we found it was hard to construct because of the difficulty of integrating the short sequence into the vector. Therefore, we made the decision to sythesize it directly as clonal plasmids.
Test 1: Do we get what we want?
The result satisfied our expectations generally. We transfected shRNA plasmids into HepG2 cells and analyzed the apoptosis rate to check the efficiency of the killing mechanism. The mRNA and protein expression level of vegfa and htert showed a decline in the context of shRNA that we designed. We also witnessed a dominant increase of induced apoptosis in shRNA groups by using the Annexin kit. These facts all showed that the shRNA played a role in inducing the apoptosis of HepG2. We transfected vectors containing two types of promoters into HepG2 and HEK293T cells. We observed that the fluorescence intensity is much higher in HepG2, indicating that these two promoters had an activity difference in normal and tumor cells.
Learn 1: What we did not know about our design?
As the mechanism we focused on was related to the HepG2 cell line and its specific growth characteristic, in vitro experiments was not enough to illustrate its killing efficacy to tumor in vivo. VEGFA is responsible for angiogenesis, which plays important role in the nutrition supply to tumor in situ, therefore the change led by anti-VEGFA shRNA couldn't be witnessed in culture dish. hTERT is responsible for the infinite proliferation ability of cancer cells. Though we had seen the increasing apoptosis rate in cell culture plate environment, the result was not convincing enough to demonstrate its function in human body. We will still need further in vivo experiments to confirm shRNA clinical function which may cause severe biosafety accident on current stage.
Improve 1: Is it enough for our design?
The first version of ViruGuard had just realized the expectation to be able to kill the tumor cell and it's safe to some extent. However, the result had shown that in normal cell situations, the promoter is still expressed at a relatively low level. There was still the possibility of potent risk to normal cells. Therefore, we had to continue to put restrictions on the killing mechanism activation. Therefore, we had to think about other ways to build a specific expression mechanism to reduce the risk.
Research 2: Another possible mechanism for improving specificity.
The way that we designed in the killing mechanism to limit the virulence was not enough for a treatment that would be possibly used for real patients. We must find other ways to make the virus distinguish the normal from the tumor cells. Therefore, we tried to find some other differences between normal and HCC cells. We mainly focused on the different expression level of proteins and cytokines in these two cell types to see whether their expression differences can be used to set up a second fence to protect normal cells from the virulence of Viruguard.
Imagine 2: Which type of "difference" can be used?
During the process that we tried to program the Viruguard to make it distinguish the normal cells and malignant cells we mainly focused the content in the cytoplasm including protein, cytokine and miRNA. We first tried to use protein and cytokine to help because of their essential role in regulating cell function. However, we realized that this could hardly be feasible, because only the genome of virus will enter into the cell to function. Thus, it would be difficult for endogenous protein to interact with it. With further brainstorming, we realised that there was a number of miRNAs existing in the cytoplasm, which also showed discrepancy in expression between the normal and the tumor cell. With further study we found that the miRNA could bind to its complementary sequence in certain parts, which had the potential to control the level of protein expression.
Design 2: How could we exploit it?
We hoped that our products would only grow 'healthily' in cancer cells while in normal cells, they were incapable to proliferate, so we tried to use the difference of miRNA expression in the cytoplasm to control it. From a database, we chose some differently expressed miRNAs:199a 195 22 and inserted their complementary sequences into the 3' UTR of the virus E1A gene, which is essential for the survival and multiplication of the virus. This design allowed the virus to act differently in these two types of cells, and that was how Viruguard played the role of "Viru" and "Guard" specifically.
Build 2: How can we build it?
We designed the suicide switch by inserting the complementary sequence of miRNA 199a,195 and 22 into the 3' UTR of the virus E1A gene. This design gave normal cells a weapon to escape the possible virulence from our product. E1A is an essential gene for virus proliferation. The combination between the free miRNAs and their complementary sequence in 3'UTR of the E1A would lead to its silence. And meanwhile, in tumor cells, due to the lower expression of those miRNAs, little effect will be put to viral viability and spread.
Test 2: Do we get what we want?
In order to examine the function of E1A-3'UTR, we transfected the 3UTR-eGFP plasmid vector into different cells to see whether eGFP expression level had coincident linear change with fluctuation of selected miRNA contents. The data showed a positive result, which indicated the feasibility and validation of our design.
BBa_K3730013
Learn 2: Could we further improve our specificity design?
We had learned that those lower expressed miRNAs in tumor cells could be used to increase the expression level of E1A. But it could always be better to boost the gap of E1A content in normal cells and cancer cells. Therefore, to eliminate the miniature inhibitory effect of free miRNAs on the oncolytic virus in cancer cells, we designed a special sequence called 'miRNA sponge' under the control of tumor-specific promoter PhTERT, which was composed of a complementary tandem sequence of miRNAs 199a, 195, and 22. The mRNA transcribed from sponge DNA can absorb free miRNAs mentioned above through complementary binding, which competitively snatched the free miRNAs that were originally supposed to bind to 3'UTR of E1A. And simultaneously, in normal cells, because of the effect of phTERT, the miRNA sponge has little expression, thus E1A expression will still be inhibited totally.
Improve 2: Is it enough for our design?
'miRNA sponge' is composed of several complementary tandem repeats of each miRNA flanking by two restriction enzyme sites in our vector and could act as a molecule sponge to
absorb the miRNA in the cytoplasma. The secondary structure and miRNA binding affinity of sponge were evaluated by RNA fold web server and miRNAsong to make sure the miRNA binding sites were exposed and it had strong attraction to free miRNA.
As the level of the miRNA in cancer cells is relatively low, this mechanism would largely increase the treating effect of Viruguard in HCC cells.
Besides, we used tumor specific promoter survivin for the expression of the sponge to
avoid the incidence of insecurity in normal cells.
BBa_K3730012
Neverthelessly, with the design of killing and specificity, the virulence control measures were still more like 'remedial measures' than 'preventive measures' because they all focused on the virulence restriction after the virus had entered the cells. Then it occurred to us that we could directly increase the specificity of the infection. We continued to improve our product.
Research 3: How to improve virus recognition ability to tumor cells
Now, our virus had a powerful weapon to kill liver cancer cells and an insurance to ensure that the virus could only replicate specifically in liver cancer cells which brought Viruguard to a new level. However, with further research conducted, we found that a better preventive measure to increase our safety. Let's start with how adenoviruses invade cells.
Adenovirus infects cells by two successive interactions with cells. Firstly, Ad fiber protein mediates its attachment to cells through the interaction with 46 kDa cell receptor CAR (coxsackievirus and Ad receptor). Then, the adenovirus penton base protein and the integrin αvβ 3 and αvβ 5 binds to promote virus internalization. That is to say, the infection efficiency of Ad infecting cells mainly depends on the expression level of CARs on the surface of target cells!
Surprisingly, most of cancer cells have low expression of CAR, and the expression of it is also significantly reduced in hepatocellular carcinoma. This may be one of the reasons that current adenovirus cancer therapy was not effective enough.
Many attempts have been made to solve the problem of low infection efficiency of virus to cancer cells. However, these methods generally focus on improving the efficiency of infection which may somehow ignore the specificity of infection. In this regard, we proposed a novel design that took both aspects into account.
Imagine 3: What can we use to reach the goal?
Our initial idea was to target Glypican-3 (GPC3, as a universal membrane marker of liver cancer cells), so that the virus can specifically recognize the protein and bind to liver cancer cells. When it comes to specificity, it is natural think of antibodies. In this way, single chain variable fragment (scFv) becomes a good choice. It is featured by low immunogenicity, high specificity and low molecular weight which ensured the minimum impact would be put on the viral protein structure when constructing the fusion protein. Therefore, we fused the fiber protein of adenovirus with the scFv of GPC3, so that adenovirus can target HCC cells and then enter the cells through secondary interaction
Design 3: How we use it?
Mouse derived monoclonal antibodies GC33 were humanized by using complementary determinant region transplantation (CDR grafting). Using humanized GC33 can make it avoid being recognized by human immune system in the early stage. The VH and VL sequences are as follows (blue is CDR region):
hGC33VHa:
QVQLVESGAEVKKPGASVKVSCKASGYTFTDYEMHWVRQAPGQGLEWMGALDPKTGDTAYSQKFKGRVTITADESTSTAYMELSSLRSEDTAVYYCARFYSYTYWGQGTLVTVSS
hGC33VLa:
DVVMTQSPLSLPVTPGEPASISCRSSQSLVHSNGNTYLHWYLQKPGQSPQLLIYKVSNRFSGVPDRFSGSGSGTDFTLKISRVEAEDVGVYYCSQNTHVPPTFGQGTKLEIK
Build 3: How we build it?
In order to prove that our fusion protein has little effect on the tertiary structure of the virus fiber protein, so that it will not affect the assembly of the virus and further affect the virus titer, we used predicted the three-dimensional structure of the fusion protein. The results showed that it had little effect on the structure of knob of fiber.
Before fusion of scFv and fibre protein, we were supposed to verify the activity and function of our scFv. Therefore, we chose prokaryotic expression to express scfv, and since Escherichia coli was the most common choice, it was selected for following experiments. We conducted a codon optimization for the amino acid sequence of scFv to ensure that it is suitable for expression in E. coli, and the gene sequence was synthesized by GenScript. We directly purchased fusion gene of scfv and fiber protein from Genscript as well.
Test 3: Do we get what we want?
We expected scFv to be expressed in a soluble form in E. coli, which is considered to be functional.The results of Western-blot showed that scFv was successfully expressed and had considerable soluble expression. Next, we verified the activity of scFv through a series of experiments, including Western-blot, ELISA, immunofluorescence and so on. The results met our expectations.
BBa_K3730035
Learn 3: What we did not know from our design?
In protein expression experiments, we managed to avoid rare codons in the protein, but failed, so we chose Rosetta competent cells. We will further select appropriate vectors, which contain appropriate antibiotic type, promoter type, fusion label, and etc, to make our plasmid better match competent cells (possibly with resistance), and have appropriate expression amount, soluble expression or inclusion body expression, and purification method. And optimal conditions for expression were also to be discovered, such as induction temperature, concentration, time, etc.
Improve 3: Is it enough for our design?
Although scFv did not affect the structure of adenovirus fiber protein in the tertiary structure prediction, further experiments were needed to confirm our prediction. In the design of this part, what we were most worried about was not the off-target effect caused by the specificity of the antibody, but the effect of modification of the virus capsid protein put on the vitality of the virus itself, which was very important to ensure that the virus could have an efficient killing effect.
Up to now, we have designed several parts to guarantee the safety of our product. However, we noticed that the current oncolytic virus treatment on the market always had the problem of early immunoclearance by the immune system, which could lead to compromising of therapeutic effect. Therefore, we had to find a way to temporarily allow our virus to escape from the clearance.
Research 4: The immunoreaction to the virus
We researched immunoreaction mechanism to adenovirus specifically in human body. We learned that CpG is an episode of DNA sequence in the virus genome, which could be recognized by the immune cell surface receptor TLR9. Once they bind to each other, TLR9 will switch into active form and trigger the downstream immunoreaction. If we were able to block this interaction, the immune cell would not eliminate Viruguard any more.
Imagine 4: What can we use to reach the goal?
In order to reach our goal of compromising immune resisitance, we had to find a competitive sequence to wrest the TLR9. As the traditional recognition method is based on DNA sequence complementary pairing, the most feasible way is to find a sequence which could competitively combine with TLR9.
Design 4: How we use it?
We got the TLR9i sequence from the manuscript published before. This special sequence acted in the same way as CpG sequence, which had higher affinity to the TLR9 receptor but it would not trigger the following downstream cascades of immune reactions. Besides, as the DNA-receptor combination is temporary and they will disassociate when the virus fully gets into the cell, the silencing of immune system will not affect the normal immunal function.
Build 4: How we build it?
In order to verify the function of TLR9i, we first constructed ODN2216+TLR9i and ODN2216+Ctrl plasmids.
CpG ODNs are synthetic oligonucleotides that contain unmethylated CpG dinucleotides in specific sequence contexts (CpG motifs). These CpG motifs could activate Toll-like receptor 9 (TLR9), leading to strong immunostimulatory effects. ODN 2216 is a class A CpG ODN and is a ligand of choice for human TLR9. We inserted ODN 2216 sequence into the plasmid, which could induce high IFN-α production from peripheral blood mononuclear cell (PBMC)
The following TLR9i sequence has a stronger affinity to TLR9, which prevents the activation of TLR9 and reduces the synthesis of IFN-α. In order to enhance the inhibitory effect of TLR9i, we insert four copies separated by short AAAAA linkers. ODN 2243, also known as ODN 2216 Control, is designed as a negative control for the TLR9 agonist ODN 2216.
In the adenovirus genome, we chose to insert an oligonucleotide consisting of four copies of TLR9i separated by spacer into the 3′ UTR.
ODN2216 sequence
5’-ggGGGACGATCGTCgggggg-3’ (20 mer).
ODN 2243 (ODN 2216 Control) sequence
5’- ggGGGAGCATGCTGgggggg -3’ (20 mer)
TLR9i sequence:
5’- ttagggttagggttagggttaggg -3’ (24 mer)
Test 4: Do we get what we want?
We transfected ODN2216+TLR9i, ODN2216+ODN2243 and empty plasmid into PBMC respectively. The supernatants were collected 24hrs and 48hrs after transfection and the levels of IFN-α produced in supernatants were measured using a sandwich-type enzyme-linked immunosorbent assay (ELISA) kit (Dakewe Biotech Co., Ltd.), according to the manufacturer’s instructions.
We found that PBMC transfected with ODN2216+TLR9i produced IFN- α significantly less than those transfected with ODN2216+OND 2243, which well verified the function of TLR9i. At the same time, PBMC transfected with empty plasmid produced least IFN- α among all other groups.
BBa_K3730007
References
[1] Gierach, I., Li, J., Wu, W. Y., Grover, G. J., & Wood, D. W. (2012). Bacterial biosensors for screening isoform-selective ligands for human thyroid receptors α-1 and β-1. FEBS open bio, 2, 247–253. https://doi.org/10.1016/j.fob.2012.08.002
[2] Gangopadhyay, J. P., Jiang, S., & Paulus, H. (2003). An in Vitro Screening System for Protein Splicing Inhibitors Based on Green Fluorescent Protein as an Indicator. Analytical Chemistry, 75(10), 2456–2462. DOI:10.1021/ac020756b
[3] Costa, S., Almeida, A., Castro, A., Domingues, L. (2014). Fusion tags for protein solubility, purification, and immunogenicity in Escherichia coli: the novel Fh8 system. Front. Microbiol., 5, 63, https://doi.org/10.3389/fmicb.2014.00063
[4] Costa, S. J., Almeida, A., Castro, A., Domingues, L., Besir, H. (2013). The novel Fh8 and H fusion partners for soluble protein expression in Escherichia coli: a comparison with the traditional gene fusion technology. Appl Microbiol Biotechnol, 97, 6779–6791. DOI 10.1007/s00253-012-4559-1
[5] Plasmids 101: Origin of Replication. Retrieved July 25, 2020, from [https://blog.addgene.org/plasmid-101-origin-of-replication]
[6] Gonzalez-Flo, E., Alaball, M. E., Macia, J. (2020). Two-Component Biosensors: Unveiling the Mechanisms of Predictable Tunability. ACS Synthetic Biology, 9(6), 1328-1335. DOI: 10.1021/acssynbio.0c00010
[7] UniProtKB - P10828 (THB_HUMAN). Retrieved October 22, 2020, from [https://www.uniprot.org/uniprot/P10828]
[8] Badran, A., Liu, D. (2015). Development of potent in vivo mutagenesis plasmids with broad mutational spectra. Nat Commun, 6, 8425. https://doi.org/10.1038/ncomms9425
[9] Chan, Y. K. , Wang, S. K. , Chu, C. J. , Copland, D. A. , & Church, G. M. . (2021). Engineering adeno-associated viral vectors to evade innate immune and inflammatory responses. Science Translational Medicine, 13(580), eabd3438.
[10] Mortezagholi, S. , Babaloo, Z. , Rahimzadeh, P. , Namdari, H. , Ghaedi, M. , & Gharibdoost, F. , et al. (2017). Evaluation of tlr9 expression on pbmcs and cpg odn-tlr9 ligation on ifn-α production in sle patients. Journal of Immunopharmacology, 39(1), 11-18.
FOLLOW US
