It is not trivial to go from an idea to a fully-functioning product over the course of one summer. Therefore, we have envisioned the next series of experiments that would be performed to develop our project into something suitable for clinical trials, incorporating the ideas of constant feedback between modelling and wet lab to ensure an optimal system.
We have already shown that our pTet-eYFP works as we designed, to fit our model. This is the basis of having the correct kinetics for this diagnostic system. To develop it further, we propose five design-build-test cycles.
Design-Build-Test 1 - Full experimental proof-of-concept
We add IPTG to induce the production of TetR, which binds to the tet operator. We would test this by looking at the level of induction needed from our construct to reduce the fluorescence. This would allow a more accurate figure for our modelling of the amount of TetR needed and could be put into our model to see if this changes any parameters going forward.
Once we have established the amount of TetR needed, we would create cells as a double transformation with a TEV plasmid, and induce expression of TEV with arabinose. TEV can then cleave the TetR.
If TetR is cleaved then YFP will start to be produced.
We can relatively quantify the amount of YFP produced, comparing different levels of induction of TEV.
Design-Build-Test 2 - Proof-of-concept system in cell lysate
Instead of inducing TetR production, we add a calculated amount of purified TetR to the system and allow it to bind. This will check that our purification doesn’t affect the efficacy of TetR.
We induce the TEV protease by the addition of arabinose, as before.
TEV protease can then cleave TetR, as before.
We can quantify the fluorescence and compare all of the data to the system in the cells, and use this to feed into our DNA system modelling. We can also optimise our lysate.
Design-Build-Test 3 - Test with cruzipain and production of TEV protease
We would then construct a plasmid with TEV protease under the tet operator and promoter, and add this to a cell lysate with an appropriate amount of TetR, as determined in the above experiment cycles.
We would then add cruzipain to our system, to check that the TetR can be cleaved by cruzipain in the same way as it was cleaved by TEV protease.
The output would be the TEV-mediated cleavage of something like a quencher-fluorophore system, such as BBa_K2450501.
If cleavage occurs, then all the above steps were successful, and we will prove that the system works for cruzipain as an input.
Further DBT cycles
A repeat of cycle 3, in a blood sample, testing our output bivalirudin
We would then repeat the experiments again, starting with a freeze-dried kit. This would confirm whether our kit will be able to travel in this form, and that the addition of blood will be sufficient to rehydrate the kit. At this point we can also experimentally test the levels of cruzipain our device can detect, and see if it agrees with our model.
In our project, the central sfGFP-dark quencher part has been shown to have repressed fluorescence which is relieved on cleavage by TEV, and this fluorescence is unaffected by the presence of the TorA leader sequence, SpyCatcher and His tag. This represents the first design built test cycle, proving the concept of relieved repression at this point, and as will be seen, this will be used throughout our suggestions for future experiments. We have also designed the parts that would be used to target proteins to the membranes, and will explain the future experiments to test these.
Design-Build-Test 1 - Membrane Insertion and OMV Packing Efficiency
Next DBT- design, build and test membrane insertion efficiency, and the packing efficiency into the OMV. Express the parts BBa_K2450401 and BBa_K2450451, which have SpyTag at the N terminal and in the middle of the protein respectively. Localisation of the fluorescence of these parts should be seen in the outer membrane, which can be performed by microscopy. Once it is confirmed that both these parts can be expressed in the membrane, the sfGFP should be removed by PCR, and subsequent experiments carried out in the absence of the fluorophore to avoid interference with other fluorescence experiments. The SpyCatcher-sfGFP part (with the dark quencher removed during PCR) should then be expressed in the same E. coli strains. These will then be transported into the OMVs which can be extracted by ultracentrifugation using the protocol shown here. The fluorescence of these OMVs can be compared, and the highest fluorescence will be the most efficient at targeting protein to the OMVs. This targeting protein can then be used in all subsequent experiments.
Design-Build-Test 2 - OMV Lysis Method
After identifying the most efficient targeting protein, it is then possible to start assaying OMV lysis. We have proposed using a detergent, such TritonX or some other ‘soft’ detergent. The detergent should be able to be added in a powdered form, should not denature proteins, and should be cheap. It may be that future teams would like to test other methods, such as increase in temperature or sonication, which are tested in the same way. In assaying the OMVs, TEV protease should be added first and left, to ensure no fluorescence change occurs and that the OMVs are intact. If this is not the case, then the TEV protease + OMV mixture should be left until the fluorescence plateaus, and this taken as the new zero. The method being assayed should then be used (e.g. adding the detergent) and the fluorescence should increase as the OMVs are lysed and the TEV protease gains access to the SpyCatcher-sfGFP-TEV cleavage site-dark quencher. This assay has the advantage that if the lysis technique begins to lyse proteins, this will be seen as a decrease in fluorescence, and can be observed by its self by lysing the OMVs in the absence of TEV protease.
Design-Build-Test 3 - Standardisation of TEV Cleavage
The next DBT cycle should be standardisation of TEV cleavage of the SpyCatcher-sfGFP-TEV protease cleavage site-dark quencher part (BBa_K2450501) in OMVs. OMVs containing this part should be generated as in the OMV lysis assays above. The OMVs should be lysed using the chosen lysis method, left for a while, and the TEV protease added after incubation, to ensure a standard response by the TEV protease. Then, the Michaelis-Menton curves at several concentrations of TEV should be produced. The timepoint at which there is the most separation between the curves should be selected as the timepoint to measure the fluorescence of future cleavage experiments. A standard curve should be produced against known concentrations of TEV at this timepoint T. This will allow the future quantification of the amount of TEV produced. It should be noted that the standardisation of fluorescence levels of GFP as is being attempted by the interlab challenge would be particularly helpful, as it would allow different groups to share and directly compare standard curves and data.
Design-Build-Test 4 - Testing TEV Protease Output
Once the fluorescence response of the quenched fluorescence OMVs has been standardised against known concentrations of TEV, the output of functional TEV protease can be tested. Specifically, these experiments should answer the question as to whether the feedback loop works and improves the response time, or if a linear amplification step would be more effective. For a full explanation of how the feedback loop and linear amplification step works, see the design page. For this, two sets of split TEV constructs should be generated. One with TEV protease cleavage sites which would create a TEV protease loop, and another with a different protease cleavage site, such as purified cruzipain, which would create a linear amplification step. To see a discussion of the pros and cons of a feedback loop verses a linear amplification step, see the Protein Based Model.
For this, the two halves of the split TEV should be loaded into outer membrane vesicles from different cell lines. Each should have a fluorophore, which have absorption and emission spectra that allow for FRET between the two. In this way, if a problem occurs, it can be determined as to whether the dimerization has failed or the subsequent complex is not active. It would also be possible to determine the proportion of dimerised versus monomeric split-TEV fragments in solution, which would inform on the kd of the complex. The kd of the active complex is an important variable in the modelling of the OMV system, and one that can be changed, and should be experimented with. One suggestion is to use stronger coiled coils as described by De Crescenzo et al (2003). We chose coils from fjksghk as they had previously been described for use in protease cleavage detection, however this does not preclude further optimisation.
Upon lysis of the OMVs, the system should then be incubated independent of activator to check it cannot autonomously produce output. Over multiple experiments the probability of a false positive should be deduced. After incubation, a known concentration of active TEV should be added. If the feedback loop is effective, the fluorescence produced by this system will be higher than that of the linear amplification method after a given period of time. This time period is discussed in the DNA based system.
Ultimately, future experiments in this area of the project should focus on two areas: the feasibility of OMVs as a protein chassis, and the properties of the feedback loop versus linear amplification. In both areas, the BBa_K2450501 part plays a central role in probing properties and efficacy. It is our hope that, separate from the overall aim of this project, that these ideas can be used in future research.
Phan, J., Zdanov, A., Evdokimov, A. G., Tropea, J. E., Peters, H. K., Kapust, R. B., … Waugh, D. S. (2002). Structural basis for the substrate specificity of tobacco etch virus protease. Journal of Biological Chemistry, 277(52), 50564–50572. http://doi.org/10.1074/jbc.M207224200
Wehr, M. C., Laage, R., Bolz, U., Fischer, T. M., Grünewald, S., Scheek, S., … Rossner, M. J. (2006). Monitoring regulated protein-protein interactions using split TEV. Nature Methods, 3(12), 985–993. http://doi.org/10.1038/nmeth967
Walker, M., Kublin, J. G., & Zunt, J. R. (2009). NIH Public Access, 42(1), 115–125. http://doi.org/10.1086/498510.Parasitic
Alves, N. J., Turner, K. B., Medintz, I. L., & Walper, S. A. (2016). Protecting enzymatic function through directed packaging into bacterial outer membrane vesicles. Scientific Reports, 6(1), 24866. http://doi.org/10.1038/srep24866