Project Engineering

Drylab: Toehold Switch Testing - Design

1 / 4

Design:

We aimed to design a set of toehold switches able to produce a clear colour change in the presence of upregulated miRNA biomarkers for latent tuberculosis() (LTB). These toehold switches would act in a cell-free system capable of being freeze dried onto a paper strip and activated with the addition of a blood sample containing LTB biomarkers.

From our research and engagement with stakeholders, we found that point-of-care (POC) testing would be most accessible for areas with high tuberculosis (TB) burdens and risks for developing active tuberculosis (ATB). In trying to formulate a test that could be as cheap, easy to produce on a large scale and accessible we landed on our initial method of using toehold switches in a paperstrip based test.

Toehold switches are a simple, easy to produce single strand of RNA able to detect a biomarker and express a specific reporter gene once detected.

Using miRNA optimised toehold switch structures from previous research (), we custom designed toehold switch DNA sequences specific to each of the four miRNA biomarkers we were testing for. A standard toehold switch sequence comprises: the toehold switch forming region, the 21 nucleotide linker sequence, and finally the reporter gene (Figure 1).

Figure 1. A diagram showing the standard secondary structure of a toehold switch, alongside the dot-bracket notation for this structure at the bottom of the image. The red toehold domain alongside the ascending stem is complementary to the trigger sequence, this is where the trigger binds and unwinds the stem structure, opening the toehold switch. We will be referring to this as the trigger binding sequence of the toehold switch. The toehold switch forming region refers to the entire sequence forming the toehold switch from the Toehold Domain to the Descending bottom stem. Source

While the linker sequence can prevent the toehold switch from forming functionally detrimental off-target structures, it also introduces additional amino acids to the N-terminus of the reporter gene that we hypothesised could be inhibitory to transcription initiation or interfere with reporter gene function. This is because the linker sequence could contain a rare codon, or the additional amino acids could interfere with protein folding. We decided to exclude the linker sequence from our toehold switch sequences to see if they would remain functional without them.

Drylab: Toehold Switch Testing - Build

2 / 4

Build

Build:

To build our toehold switch sequences we first started with a base for the toehold switch forming region. This consisted of a trigger binding region complementary to half the trigger miRNA, stem forming regions complementary to the other half of the trigger, a 5 nucleotide stem forming region, a start codon embedded between these two stem forming regions and finally the stem loop containing the ribosome binding site (RBS). To create toehold switches for each of the four miRNAs we were detecting we simply changed the miRNA complementary regions to be specific to the selected miRNA. This means that other than the miRNA complementary regions of the toehold, every structure and sequence of the toehold forming region for each toehold switch is the same (reference to a figure?).

Three reporter genes were chosen to test with our toehold switches: LacZ, mRFP and NanoLuc luciferase. We wanted to see if changing the reporter gene would make a difference in the function of a toehold switch, or if a toehold switch could be used with almost any reporter gene. LacZ was chosen as the reporter gene that would work on the paperstrip test as it would produce a clear colour changing reaction with a number of substrates. mRFP and LacZ were chosen to quantitatively test the amount of readout our toehold switches would produce within a certain time period.

Wetlab: Toehold Switch Testing - Design

1 / 4

Design:

How would we design toehold switches to be tested in the lab (reporter gene that is more sensitive e.g.) , what experiments will be done , controls for testing to see if it works.

The purpose of the design is to construct a functioning toehold that detects specific miRNA biomarkers that are present in latent TB. We have tested our toehold in the NEBExpress® Cell-free E. coli Protein Synthesis System. As our reporter gene we have chosen to use mRFP,as it is easily detected by the naked eye and its length is relatively small.

In order to determine if our toeholds work as intended we have planned to do the following tests

Test for leaky expression - we would police the toehold in a system with IPTG where no miRNA is present. Then, we would observe under UV light to detect any mRFP expression. If none or little expression is observed, then the level of leakage would be deemed acceptable and we would move on to test 2
Test for miRNA detection - we would place our toehold in a system with target miRNAs and incubate for 3 hours. After that time, we would measure the absorbance value of each vile with the toehold and its complimentary miRNA. If the mRFP expression is visible with a naked eye and cleary discernable from the background noise using a spectrophotometer, we would move on to test 3
Test for specificity - we would put our toehold in a system with wrong, but similar miRNA (up to 80% similarity) and check for mRFP expression via spectrophotometry. If the level of expression is similar to the one observed in test 1, we would conclude that the toehold is specific.

Wetlab: Toehold Switch Testing - Build

2 / 4

Build

Build:

For the assembly of the toehold we chose the Golden Gate cloning method, as this method allows for a scar-less and quick assembly of our construct, requiring only a single reaction to be carried out. As our plasmid backbone we have chosen pJUMP29A1 as it is a suitable plasmid for level 1 cloning via golden gate assembly. For our E.Coli strain we have chosen the XL1-BLu for plasmid propagation and BL21-de3 for testing. We chose XL1-Blu because it is a commonly used strain for the purposes of plasmid propagation and because it lacks endonucleases, which increases the quality of the DNA. Bl21de3 was chosen because it is a common strain used for protein expressions and lacks several proteases, which will ensure that mRFP signal is not diminished. We have also chosen a cell free system by NEB, since we want our test to be cell-free and easily accessible.

The pJUMP29A1 was linearised with the BsaI restriction enzyme, after which the inserts were added to the reaction tube. PCR was then run under specific settings for 3 hours and the results of that cloning were sent off for sequencing in order to be verified. Portions of the results were also subjected to analysis via gel electrophoresis to check for length of the sequences.

Wetlab: Toehold Switch Testing - Test

3 / 4

Test

Test:

Our initial test with the IPTG revealed that no leaky expressions were observed in any of the toehold switches. We continued on with the test 2 - the test for miRNA detection.

Toehold Switch leakiness Test example for 1306-5p:

Test 2 ended with no mRFP expression being observed, even after another hour of incubation. Since test 2 failed, we did not move on to test 3.

Figure 5: Measurements were carried out using the FLUOstart OMEGA spectrophotometer from BMG Labtech. The spectrophotometer was tested on known concentrations of Ogarne G protein,which yielded expected results.

Wells A1-A6 were our test chambers. Where A1 and 2 had negative and positive control respectively, and wells from A3-6 each of the 4 toeholds with their respective miRNA, except well A5, which only had toehold 3 due to its respective miRNA not expressing in cloned cells, killing them. Our measurements can be used by other teams who have toeholds that did not work to see if there is an underlying cause in the methodology or toehold switch design. They could compare the absorbance values with the ones that we got and then try to follow the steps outlined to potentially fix the non-functional toehold.

Wetlab: Toehold Switch Testing - Learn

4 / 4

As our experiments did not generate expected results, we went over our lab notebook looking for issues with our protocols and human error that may have occurred. We didn’t find anything that seemed incorrect or improperly performed, so we began to consider if the issue was with the toehold switch sequences themselves.

We sequenced our constructs for toehold switch composite parts with mRFP made in the lab. We checked if the resulting sequencing matched the individual sequences that we ordered to see if we correctly constructed them. The results showed no differences between the sequenced mRFP toehold switches and the designed composite parts.

When our toehold switches showed no sign of leaky reporter gene expression it suggested they were not functional, toehold switches normally have some background leakage (5). This was further suggested by the second test where no mRFP was expressed in the presence of the correct miRNA trigger. To try and learn more about this, we turned to modelling for a more detailed look on the predicted secondary structures of the toehold switches.

We compared the secondary structures of functional toehold switches, from previous iGEM teams and research papers, with our sequences. We highlighted key parts in their structure, as well as differences in modelling methods. Some previous teams only modelled the secondary structure of the toehold switch as we had initially done. However, others included 50 nucleotides after the toehold switch, including the linker sequence and partially the reporter gene, as suggested in literature (50). After applying the latter method to our sequences, we noticed two of them formed off-target secondary structures that could prevent them from functioning. The miRNA binding regions of the two sequences were binding with the reporter gene, potentially preventing the miRNA binding to and activating the toehold switch. While this is likely not the primary reason for no mRF expression, we next wanted to focus on methods of correcting this to match functional toehold switches, comparing behaviour after this problem with secondary structure has been resolved

Drylab: Toehold Switch Testing - Test

3 / 4

Test

Test:

Preceding labwork, we used modelling to test the structure and functionality of our toehold switches from three different angles: off state (no miRNA trigger bound), on state (miRNA trigger bound) and finally cross talk between different toehold switches and miRNA triggers in a multiplex system. Please see our Model page for a lot more detail about this.

For this modelling we focused purely on the toehold switch region (not including any part of the reporter gene) to see if the sequences we designed formed the correct secondary structure of a toehold switch. NUPACK was used for all modelling of secondary structures and multiplex system cross talk.

In the off state, all of our sequences formed the expected secondary structure of a toehold switch. All miRNAs correctly bound to each respective toehold switch in the on state, and this binding opened up the toehold switch structure. Crosstalk modelling revealed many unexpected interactions we did not consider, for example, two of the same miRNAs being able to bind together in a stable complex. From cross-talk modelling we were able to identify miRNA hsa-miR-7850-5p combined with hsa-miR-6529-5p as the best combination of biomarkers with the least crosstalk.

Drylab: Toehold Switch Testing - Learn

4 / 4

Learn

Learn:

Overall based on this initial modelling out toehold switches were structured and behaving as they should, however crosstalk modelling did introduce more considerations and room for optimisation.

We learnt it was important to consider the interactions between the different miRNAs themselves. miRNAs that form strong interactions with other miRNAs may be less able to bind to a toehold switch and therefore activate the switch, this may limit the readout certain miRNA combinations can produce.

Changing concentration and temperatures could have a big effect on the amounts and types of binding occurring in a multiplex system. We did not know what concentrations of toehold switch and miRNA we would be able to achieve in a test, nor what would be the most optimal. Therefore this modelling is likely not accurate and may be showing results that are more ideal than is actually the case. A major improvement would be repeating this modelling with precise concentrations of each component in the test to get a more accurate model of the behaviour of the test.

Toehold Switch troubleshooting - Design

1 / 4

Design:

After learning that off-target secondary structures in two toehold switches were potentially preventing miRNA triggers from binding and activating the toehold switches, we decided on two paths forward:

Reintroducing the linker sequence we decided to not include in our initial designs. This would introduce the additional codons and amino acids that we were previously looking to avoid.

Optimising the codon sequences in the reporter gene to maintain the amino acid sequence but limit off-target binding occurring. This avoids the concerns we had over the linker sequence, but we were a lot more limited with what we could actually change without coding for a different amino acid entirely.

Lastly, we researched how the spacer sequence in the Ribosomal Binding Site of E.Coli affects protein expression to see if we could optimise the linker sequence, and therefore translation. The spacer sequence within the RBS is the space between the shine dalgarno sequence (AGGAGG) and the start codon (usually ATG). From a literature review, we found that:

The optimal length of the spacer sequence is 6-9 nucleotides
It’s preferable to have a high content of A and U’s because they bind less strongly (2 hydrogens bonds) compared to C and G’s as they bind more strongly (3 hydrogen bonds)
If we were to say that the start codon was position 0 then position -3 (within the spacer sequence) being an A is highly favourable whilst a C is unfavourable
AUG should be the start codon because other start codons like CTG show less protein expression. Avoid having other start codons found before the main start codon (the one found right before the protein sequence) because this could prevent translation
Ensure there is no secondary structures formed within the spacer sequence

We tried applying as many of these principles in the design of our linker sequence as possible.

Toehold Switch troubleshooting - Build

2 / 4

Build:

To create a new linker sequence for these two toehold switches (hsa-miR-7850-5 toehold switch and hsa-miR-1306-5p toehold switch) we first took a linker sequence from a previous working toehold switch: BBa_K3453080 and input this into our sequences. We then began an iterative process of modelling the structure and making changes to the linker sequence to remove the off target binding remaining. This involved changing specific nucleotides within the linker sequence, replacing as many as possible with As and Ts to match with the linker sequence research we discussed above. This process created two different linker sequences, one optimised for hsa-miR-7850-5p toehold switch and the other optimised for hsa-miR-1306-5 toehold switch.

Whilst attempting to eliminate the off target structure between the mRFP and the miRNA binding region in the two previously mentioned toehold switches, we found that we were unable to maintain the exact coding sequence. As an alternative method we decided to alter the amino acid sequence, replacing the original amino acid with one with as similar structure and properties as possible. An example is shown below with hsa-miR-7850-5p toehold switch in Figure 6.

Figure 6

The two regions of the sequence binding together to form the off target structure are highlighted in orange, the other colours are irrelevant to this discussion. The coding mRFP coding sequence contributing to the structure is ATG GCT which we changed to ATG GGG. This changed the amino acid sequence from methionine and alanine, to methionine and glycine. Glycine is structurally very similar to alanine, only containing an additional methyl group on its side chain, and has similar properties of being non-polar and hydrophobic.

Toehold Switch troubleshooting - Test

3 / 4

Test:

Using NUPACK (8)(9) we modelled the secondary structures of our improved sequences and compared them to their original states. Overall, the addition of a linker sequence seemed like the most successful optimization, we were able to completely eliminate the trigger off target binding with hsa-miR-7850-5p toehold switch being shown as an example below in Figure 7.

Figure 7. Lefthand side: inactivated hsa-miR-7850-5p toehold without a linker sequence, showing the secondary structure between the trigger binding region and the beginning of the reporter gene. Middle: inactivated hsa-miR-7850-5p toehold switch including the built linker sequence, the secondary structure between the trigger binding region and reporter gene is no longer present, and the predicted probability of these regions remaining unpaired is high. Righthand side: activated hsa-miR-7850-5p toehold switch including the built linker sequence. The miRNA still binds as expected, and the secondary structure formed around the RBS in this state is of a low probability.

While we were able to make improvements to the off target binding via codon optimisation of the reporter gene, we could not completely eliminate it using this method for both hsa-miR-7850-5p toehold switch and hsa-miR-1306-5p toehold switch. The main outcome was exposing more nucleotides and reducing the probability of the structure forming as shown in figure 8.

Figure 8. The left structure shows hsa-miR-1306-5p toehold switch without codon optimisation, with the off target structure between the trigger bind region and reporter gene. The right structure shows hsa-miR-1306-5p toehold switch codon optimised. Although the off target binding remains, the probability of the off target structure forming is lower than the original version of the toehold switch.

Due to time shortages we couldn’t test this in the lab. However, if we did have time we would test our new toehold switches using the same method as before, and also use a positive control of the working toehold switch from which we derived our linker sequences: BBa_K3453080. If the positive control did not work, it shows that there is a problem with our lab protocol, if it did work then it was a problem with the sequence design.

Toehold Switch troubleshooting - Learn

4 / 4

Learn:

While researching more into toehold switch design and properties, we began to have doubts about how suitable this detection method would be for our POC diagnostic test. An example of this was the time required for a readout, which could take up to 2 hours (12). This incubation time would limit the extent our test could be deployed in POC environments, reducing the number of patients that can be tested within a time frame, and also possibly be a deterrent to taking the test. Another example is the leaky reporter gene expression of toehold switches (5), which limits the accuracy of our test as if given a long enough time every test may eventually give a positive readout. For these reasons we decided to look into other detection methods and test formats e.g. a lateral flow test.

Cyclic Chain Displacement Reaction - Design

1 / 4

Design phase for CCDR.

Design:

We assessed other detection methods and decided on changing our detection method from toehold switches to Cyclic Chain displacement Reaction . This method has many benefits such as being done in 15 minutes total, was tested in a multiplex system before with different miRNAs , has good sensitivity with as low as 0.123 pM for miRNA-223 and also amplifies the miRNA signal whilst detecting it too The purpose of this design is to use the CCDR isothermal amplification method to increase the miRNA concentration, whilst also forming a DNA1-DNA2 hairpin duplex which can be detected in a LFT. The design includes miRNA binding to DNA-1 hairpin to expose the complementary region to DNA-2 hairpin so that a hairpin duplex can form and the miRNA is recycled. We aimed to design DNA hairpins using the literature paper that created this method, but adapt it to our miRNAs, and test the design to see if the predicted structures are similar to the ones in the paper. Their template design for the crRNA was made specific to hsa-miR-200b and hsa-miR-223 mRNA detection, and we used their DNA design to construct our own hairpins (13).

Cyclic Chain Displacement Reaction - Build

2 / 4

Build phase for CCDR.

Build:

We used Zhou.P et al design for DNA hairpins as a template to base our hairpin designs to LTB-associated miRNAs. There are two DNA hairpins involved in the process. DNA-1 hairpin has a complimentary region to the miRNA to allow binding when present and it also has a complimentary region to DNA-2 hairpin which is hidden within the sem loop, thus can only be exposed once the hairpin opens up to allow DNA1-DNA2 binding and displacement of the miRNA.

We used NUPACK (8)(9) to demonstrate how the DNA sequences extracted from the paper will bind to each other so we can understand how and why these sequences were constructed this way for us to effectively use them as templates (13)

Below shows how we designed our sequences, using DNA3 and DNA4 for hsa-miR-7850-5p as an example:

The sequence construct for DNA 1 from 5’ to 3’ starts with a poly-A-tail to allow binding to biotin. Then there is the reverse complement of the miRNA sequence , followed by a conserved region which is complementary to DNA 2 structure. This conserved sequence is taken from the template and within inactivated DNA hairpin 1 it’s enclosed in the hairpin loop. Lastly, it ends with an 11 nt long sequence which is complementary to the last 11nt sequence in the complementary region which forms the stem in the hairpin loop.

The sequence constructed for DNA 2 (5’ to 3’) started with a poly-A-tail to allow binding to FAM. This is followed by the reverse complement of the 11nt end sequence from DNA 1 (but does not include the first 2 nucleotides). Next is the reverse complement of the conserved sequence (taken directly from the template of DNA 2 in the paper). Then the reverse complement of the miRNA for only 9nt which is in the hairpin loop and lastly to form the stem the matching reverse complement of the conserved sequence is added (again taken directly from the paper in DNA 2). This was repeated for the other two miRNAs we wanted to test.

Cyclic Chain Displacement Reaction - Test

3 / 4

Test phase for CCDR.

Test:

We used NUPACK (8)(9) to model our sequence interactions.The settings on NUPACK were the standard with 37 degrees celsius, MFE for structure was ticked on and the miRNA was added first and then the DNA 1 hairpin loop and then DNA2 + DNA1 when testing their interaction.Our positive control would be the 2 DNA hairpins created for miR-223 designed by Zhou.P et al.

Please note that NUPACK did not have a setting where you can test RNA-DNA interactions so these interactions are between DNA-DNA which can hinder the accuracy of our modelling but can be used as a good starting point for testing if your sequence design is correct. Due to time shortage we couldn’t model using a different software.

Table 1: Comparison of DNA3 and DNA4 with the positive controls (DNA1 and DNA2 - as listed from the literature)

Cyclic Chain Displacement Reaction - Learn

4 / 4

Learn phase for CCDR.

Learn:

When comparing DNA3 with the control primary DNA hairpin, we found that they had the same secondary structure and both had an MFE<0 , meaning both structures are favourable. The equilibrium probability was high for both, showing stable structures. When comparing how these primary DNA hairpin loops interact with their miRNA , we found they both had the same secondary structure and the miRNAs bind fully with all equilibrium probabilities very high, which suggests the DNA hairpins were designed correctly to be complementary to their respective miRNAs. The MFE for both interactive structures are less than 0 , which shows the interaction to be favourable.

When comparing DNA4 with the control secondary DNA hairpin, we noticed that the control’s stem length was 11nt whereas the ones we designed had 12nt; hence, even though we followed the template exactly. We further analysed DNA4 hairpin and found that two bases within the loop of DNA4 were complementary and thus had a strong binding probability and would be part of the stem. The loop sequence is complementary to miRNA-7850-5p so we can’t change the base to prevent this interaction.

DNA 4: 3' TTGAGTGTGGT ATACAGGTT ACCACACTCAA AACCTGTAT AAAAA 5'

The bases in bold are found in the loop and they bind as seen in figure

When analysing how DNA4 interacts with DNA3, results showed the interaction to be favourable because the MFE < 0, suggesting that DNA4 and DNA3 were designed correctly. The control’s interactions and our designed hairpins both had their complimentary sections bound with a high equilibrium probability which is what was expected, however the secondary structure differs for both in the sections

DNA 3: 5' AAAAAA CCAGGCCACAC TATGTCCAA TGGTGTGAGTT TTGGACATAGT 3'

The sections in bold were not designed to interact, however in the hairpins we can observe strong partial binding between DNA3 and DNA4. However, we do not predict this to be a problem with testing because our controls have a similar observation (just not the same structure) and they were functional.

The next step would be to test our designed sequences in the lab. We will measure the fluorescence of the FAM which is conducted in a cell free system using the FLUOstart spectrophotometer. The first test would be with DNA hairpin pairs + no miRNA present so we can learn how much background leakage our designs would have (because we expect them to be somewhat leaky as most DNA hairpins). We then would move onto test 2 where we test our DNA hairpin pairs with their corresponding miRNAs. The DNA hairpin loops remain at constant concentrations and the miRNAs have increasing concentrations. We would expect to see an increase in fluorescence recorded as the miRNA concentration increases. This is important to learning if our system functions correctly. We would then move onto test 3 which includes testing the DNA hairpin pairs with the alternative miRNAs and measuring the fluorescence whilst comparing it to the negative control (when no miRNA is present) to test cross-reactivity as this is important for implementing our test into a multiplex system.

Crispr/cas13a - Design

1 / 4

Design phase for Crispr/cas13a.

Design:

We were originally considering moving onto a CRISPR-Cas13a detection mechanism rather than CCDR but wanted to do a concise literature review before going ahead with CRISPR and found the CCDR better suited our test plan. However, we still want to include how we designed our crRNAs for our four miRNA sas CRISPR-Cas13a has many different applications apart from just diagnostics and it would be useful for future iGEM teams to see how we designed it if they wanted to use this detection method.

The purpose of this design is to make a crRNA that is complementary to a miRNA associated with LTB. When the crRNA's spacer sequence binds to the miRNA , it activates Cas13a to trans-cleave a Biotin-hairpin probe at the rU site to release an initiator sequence which opens a FAM-hairpin probe. Thus, a biotin-dsDNA-FAM product is formed which can be detected in a lateral flow test. We aimed to design crRNAs using the literature paper that created this method but adapt it to our miRNAs and test the design in NUPACK software to see if the predicted structures are similar to the ones in the paper. Their template design for the crRNA was made specific to SARS-CoV-2 mRNA detection, and we used their DNA template to construct our own DNA template (14).

Crispr/cas13a - Build

2 / 4

Build phase for Crispr/cas13a.

Build:

We used authors He Sun et al design of crRNA as a template to base our designs of crRNAs specific to LTB-associated miRNAs. crRNAs are short RNA molecules designed to have a direct sequence (DR) and a spacer sequence from 5' to 3'. The DR for LwaCas13a interacts with LwaCas13a and is conserved in the crRNA, whereas the spacer sequence is complementary to the biomarker sequence so will differ in design for each crRNA.

DNA template of crRNA (for SARS-CoV-2 mRNA):

5' GCTGTAGTTGTGATCAACTCCGCGAACCGGTTTTAGTCCCCTTCATTTTTGGGGTGGTCCCCCTATATAGTGAGTCGTATTAATTTC 3'

T7 Promoter( 3' to 5') GGGATATCACTCAGCATAATTAAAG

When creating the DNA template, the miRNA's reverse sequence (with U's swap to T because it's DNA) is inserted , then the lwacas13a conserved direct sequence is added with the underline sequence forming a hairpin stem. Lastly, the reverse complement is added which binds to the T7 promoter to initiate transcription of the crRNA (14).

crRNA sequence :

5' GACCACCCCAAAAAUGAAGGGGACUAAAACGGUUCGCGGAGUUGAUCACAACUACACAGG 3'

SARS-CoV-2 RNA: 3' CCAAGCGCCUCAACUAGUGUGAUGGUCG 5'

From 5' to 3' , the crRNA first has the direct sequence and then the spacer sequence. The miRNA binds to the spacer sequence which will activate the Cas13a sequence. We used this as a template to design our own crRNA. Below shows the crRNA design process for hsa-miR-7850-5p as an example (crRNA 3)

DNA template of crRNA (for hsa-miR-7850-5p):

5' GTTTGGACATAGTGTGGCTGGGGTTTTAGTCCCCCTTCATTTTTGGGGTGGTCCCCCTATATAGTGAGTCGTATTAATTTC 3'

T7 Promoter( 3' to 5') GGGATATCACTCAGCATAATTAAAG

crRNA sequence :

5' GACCACCCCAAAAAUGAAGGGGACUAAAACCCAGCCACACUAUGUCCAAAC 3'

hsa-miR-7850-5p: 3' GGUCGGUGUAUACAGGUUUG 5'

Crispr/cas13a - Test

3 / 4

Test phase for Crispr/cas13a.

Test:

We used NUPACK to model our sequence interactions and compared that to the SARS-CoV-2 crRNA template we followed to verify we built it correctly. The settings on NUPACK were the standard with 37 degrees celsius, MFE for structure was ticked on and the RNA was added first and then the crRNA when testing their interaction. Below shows an example of our modelling for crRNA 3 and we used the crRNA designed for SARS-CoV-2 mRNA as our "control crRNA".

Crispr/cas13a - Learn

4 / 4

Learn phase for Crispr/cas13a.

Learn:

From the modelling and literature we learnt that our designed crRNA + miRNA interaction have a MFE <0, specifically , showing the interactions are favourable. hsa-miR-1306-5p with its respective crRNA, with a MFE of -49.05 kcal/mol specifically. The control crRNA had a MFE of -56.49 kcal/mol. When compared to the control, the direct sequence structure was the same in structure and the spacer sequence was similar in its interaction with miRNAs. When analysing the control crRNA's secondary structure there was loose binding between multiple nucleotides between the spacer sequence and miRNA towards the 3' end of the crRNA and a hairpin loop is formed unlike crRNA 3 which has no additional hairpin loop observed. To verify results, testing in the lab in a cell free system would be needed but as we didn't go through with this detection method of crRNA we did not do this.

Research Engineering Cycles - Engineering Cycles Summary

1 / 4

These engineering cycles are concept-driven with a focus on how we made a final decision in research via analysing multiple peer-reviewed literature resources, consulting experts within the area and weighing the pros and cons of each method to optimise or reconstruct our project plan. Each idea is listed as a numbered stage.

Research Engineering Cycles - Stage 1: PET-degradation

2 / 4

Our initial research focused on the degradation of polyethylene terephthalate (PET) through the application of specific enzymes to develop a sustainable and biodegradable alternative to conventional plastics. This strategy entailed the utilisation of thermophilic or mesophilic hydrolases to effectively dismantle the ester bonds present in PET over a designated timeframe (15)(16). Furthermore, we intended to employ directed evolution within protein engineering to enhance the PET-degradation capacity of known hydrolases documented in existing literature (17). It is noteworthy that prior research efforts have primarily concentrated on augmenting enzymatic activity and thermal stability, while additional factors such as protein solubility, substrate specificity, pH adaptation, and solvent tolerance have received comparatively limited scrutiny (18).

Additionally, we planned to implement ancestral sequence reconstruction methodologies to produce proteins exhibiting unique and improved stability and activity profiles. Our final objective involved evaluating the synergistic potential of optimal thermophilic bacteria to construct a thermophilic microbe-enzyme system (19) designed for scalable biodegradation.

However, this proposed methodology faced a significant challenge: the difficulty in obtaining quantitative results in laboratory settings. This challenge arises from the variability of degradation benchmarks for plastics, which differ between controlled laboratory environments and natural habitats, especially in real-world pH or waste-stream conditions, thereby hindering our ability to reliably ascertain the efficacy of our method in enhancing plastic biodegradation.

Research Engineering Cycles - Stage 2: Tay-Sachs disease

3 / 4

Our next potential area of research we considered was Tay-Sachs disease (TSD), with a specific focus on tackling the heterodimer challenge as a means to find a cure. TSD is distinct from other lysosomal storage disorders due to the defective heterodimer HexA enzyme, and researchers have faced considerable challenges in synthesising and transporting both of its subunits. A significant hurdle in delivering the HexA enzyme is its necessity to traverse the blood-brain barrier, as the accumulation of GM2 in the brain leads to neurological symptoms.

If we had pursued this research initiative, we would have concentrated on reconstructing and developing the enzyme. Our aim would have been to improve its kinetic favorability, thermal stability, and overall effectiveness in breaking down GM2. We would have intended to create a more efficient enzyme for enzyme replacement therapy (ERT), where we could replace the mutated HexA enzyme with a modified version designed to replicate its functionality. Utilising yeast as a host organism to transform recombinant plasmids containing the gene needed for our modified HexA enzyme would have been a critical part of our strategy (20).

Nevertheless, we recognised several notable challenges that still exist in this field. Despite the promising advancements for such a rare disease, we would have needed to carefully assess whether our enzyme would indeed be a significant enhancement over the current HexM. This was particularly important since we would have been producing it in yeast that involves separate post-translational modifications. Moreover, we would have had to weigh the cost-effectiveness of alternative methodologies, such as IVF, particularly in certain countries, as this could have influenced the viability of our approach.

Research Engineering Cycles - Stage 3: Infecheck-Latent tuberculosis diagnostic

4 / 4

When we initiated our work on Tay-Sachs disease, our primary emphasis was on enzyme replacement therapy, aiming to address the deficiency of Hexosaminidase A at a molecular level. Although this endeavour was fundamentally therapeutic, it provided us with invaluable experience in enzyme dynamics, biomarker identification, and the design of biological assays. The expertise gained—specifically in recognising and responding to distinct molecular signals—has proven directly relevant to the field of diagnostics.

As we deepened our understanding, we recognised that while Tay-Sachs represents a devastating and rare condition, the methodologies we were developing could be adapted to confront a far more widespread issue: latent tuberculosis (LTB). With approximately 2 billion individuals living with latent TB, effective treatment could significantly prevent the progression to its active form, thereby curtailing transmission and mortality rates. Transitioning our focus from enzyme replacement to diagnostics allowed us to leverage our molecular biology expertise within a diagnostic framework, employing similar principles of specificity and biological recognition. This strategic pivot enabled us to build upon our foundational work while redirecting our efforts towards a disease that impacts nearly a quarter of the global population, thereby enhancing both the relevance and potential impact of our initiative.

In our current project, we are developing a point-of-care diagnostic test for latent tuberculosis, the concealed manifestation of the world's most lethal infectious disease. Despite the significant prevalence of LTB, existing diagnostic methods are inadequate—often expensive, reliant on laboratory infrastructure, and incapable of differentiating between latent and active tuberculosis (21) . This inadequacy contributes to delayed diagnoses, extended treatment regimens, and ongoing transmission (22).

Infecheck aims to resolve this issue through the early detection of LTB. We are designing an economical, paper-strip diagnostic test that utilises a simple finger-prick blood sample to identify upregulated miRNAs specific to latent tuberculosis. We have investigated two promising methodologies for exosomal miRNA detection: Toehold Switches and CRISPR-Cas13a (23). We envision our final product as a lateral flow test employing microfluidics for sample extraction. This test will deliver straightforward, interpretable results without the necessity for laboratory equipment. By facilitating the prevention of progression from latent to active TB, our approach seeks to alleviate patient suffering and diminish the spread of tuberculosis.

LFT into Toehold Switch - Summary

1 / 6

Although we did not end up choosing toehold switches as our detection method, we still thought about its implementation. For toehold switches, its current application into diagnostic devices include having a reporter gene as an enzyme which can act on a chromogenic substrate to see a colour change. This occurs in a cell free system that is freeze dried onto a paper based-strip, with the sample applied directly on top of the cell free system (24). From our research into LFT uses for miRNA detection, we wanted to adapt toehold switches in a LFT format rather than just a paper-strip test.

Advantage	Explanation	Reference
Validation system with a control line	The strip includes a control line that must appear for the result to be considered valid which is independent of the test result. This means that if the test line is negative, the presence of the control line confirms that the assay reagents have flowed properly and that the strip mechanics functioned.	(25)
Higher specificity as a dual detection system of toehold switches and protein detection via antibodies/aptamers	A dual recognition element lateral flow assay that uses an antibody paired with an aptamer provides quantitative validation that sensitivity and specificity is increased and false positives are decreased.	(26)
Cheaper and more feasible way of transforming the test into a multiplex system	One paper specifically states that lateral flow assays are the cheapest, fastest and easiest to use paper-based assays which makes it ideal for our test considering our goal for our test to be field-deployable	(27)
The reporter gene choice selection varies more	In traditional toehold switches, the reporter gene is something that produces visible colour or fluorescence – for example, GFP or LacZ, because the output must be immediately observable. However, the advantage of using a lateral flow test with a toehold switch is that the reporter gene we select does not need to do this. Lateral flow tests instead use antibodies, aptamers or streptavidin to capture or bind reporter tags and use labelled particles to generate the visible readout.	(26)(28)
Real-time Connectivity:	There is literature about an app that adds real-time connectivity to the diagnostic test by measuring the gold nanoparticle absorption to give quantitative results. An iPhone was positioned in a small cardboard darkbox to take a photo of the gold-nanoparticle lateral-flow strip under consistent illumination. The photo was then imported into the open-source R-package GNSplex, which includes a Shiny web-app for analysis. The app isolates the test and control lines, corrects background and calculates normalised intensity ratios for each strip. The ratios are then fitted with a simple linear regression model which relates intensity to known digoxigenin concentrations, enabling the conversion of any measured ratio to an estimated concentration. GNSplex automatically provides standard deviation, confidence intervals, Pearson correlation, and can compute analytical limits.	(29)

LFT into Toehold Switch - Stage 1

2 / 6

When researching how a lateral flow test commonly works, our first idea was to have a typical sandwich immunochromatography assay for detecting proteins translated from the toehold switch and incorporating the GFP Dipstick ELISA (Enzyme-Linked Immunosorbent Assay) for the rapid determination of GFP (30). From our experience at the mini jamboree hosted by Imperial iGEM, we got feedback that RNA can be unstable thus will need to be freeze-dried to have a longer shelf life, so the reaction for our toehold switch would be best of freeze dried on the conjugate pad so the product can have a longer shelf life.

The miRNA present in the sample will activate the toehold switch to produce GFP. It will be bound to anti-GFP antibodies that are conjugated to gold nanoparticles. These GFP-antibody - gold nanoparticles conjugated complexes flow to the test line. At the test line there is an immobilised anti-FAM antibody which binds to the complex At the control line, it's not specific to the complex, only the anti-FAM antibodies (30).

Limitation:

Dual-detection may increase specificity, but it does not advance sensitivity of the test and is not cost-effective with the use of three antibodies detecting a protein.

LFT into Toehold Switch - Stage 2

3 / 6

We then examined literature papers on current research into lateral flow test detection for the reporter genes firefly luciferase, GFP and B-lactamase to see if there is an alternative cheaper method that we can use. Firefly luciferase and GFP in LFT literature was often used as increased sensitive labels but not as a detected product. One literature paper showed B-lactamase detection in a LFT. On the conjugate pad, there is the substrate cefotaxime (or can use any substrate for B-lactamase). When the reporter gene is not produced, the substrate will bind to an anti-cefotaxmin antibody (conjugated to gold nanoparticles) and the complex will flow to the control line as it binds to an immobilised antibody which is specific to the conjugated antibody. When the B-lactamase is produced, it cleaves the substrate so no binding to the conjugated antibody occurs. No complex will be formed so the unbound mAb can flow to the test line and bind to the immobilised substrate coupled to BSA (31).

This method is an improvement of our original method because the number of antibodies has been cut down from 3 to 2, with still a sandwich format at the test line and measures enzymatic activity to verify toehold switch produced the enzyme and it folded correctly to be functional. We wanted to optimise this design to make it cheaper.

LFT into Toehold Switch - Stage 3

4 / 6

We then researched possible replacement for antibodies, which we found mostly to be DNA/RNA detector probes which are normally complimentary to other DNA/RNA molecules, not proteins. However, Aptamers are short, single-stranded DNA or RNA molecules that fold into unique three-dimensional structures and can target proteins. Because of their shape, they can bind tightly and specifically to target molecules (proteins, small molecules, etc.), much like antibodies do. We can replace antibodies with aptamers. They are selected from large random libraries of nucleic acids through a process called SELEX (Systematic Evolution of Ligands by Exponential Enrichment).The nucleic acid nature of aptamers also enables to develop simplified LFA design via adding a short nucleic acid sequence region (e.g. 5'-AAAAAA-3'; complementary capture sequence of an additional sequence, e.g. 5'-TTTTTT-3') at the end of the detection aptamer to serve as a universal capturing agent for the control line design (32).

Benefits of aptamers: they are synthetically produced (no animals needed), can be chemically modified easily, are often more stable than antibodies (less sensitive to temperature or storage), can be tailored to bind many types of targets and cheaper to manufacture (32).

Disadvantage: Aptamers composed of natural DNA or RNA nucleotide sequences are vulnerable to the nucleases, particularly RNA aptamers, and they are complicated to design with all designs needing experimental testing. The test's sensitivity depends on the aptamer design(32).

LFT into Toehold Switch - Stage 4

5 / 6

miRNAs need to be amplified for sensitive detection and the toehold switch detection system on its own does not do this, so we explored if we could change the enzyme that is used in lateral flow tests commonly and is very sensitive. From our miRNA detection method literature search, we explored CRISPR-Cas13a use in lateral flow test and found a detection system that took 10 minutes and has a sensitivity of 0.33pM. miRNA binds to a crRNA which activates Cas13a to trans-cleave a U-U in a LOCK-U DNA

loop to form a double-stranded product. The fuel chain triggers a strand displacement reaction and releases the sequence P1 from the double strand (P1 originally being in the hairpin loop). Therefore, lots of P1 is produced which can be detected in a lateral flow test using MnO2 Nss-BP (33).

Having the toehold switch synthesise Cas13a upon miRNA detection and is then activated once miRNA binds to the crRNA allows for a dual-detection system that is more specific and sensitive, does not require an enzyme supply chain so reduced costs, detects nucleic acid rather than protein and is detected using DNA capture probes and MnO2 Nss-BP which is easier to make compared to aptamers.The limitation of combining the two is that Cas13a is a very large protein so it will be very time-consuming to synthesise in a toehold switch so results will have a long readout.

LFT into Toehold Switch - Stage 5 (Final Decision)

6 / 6

Due to time being a limitation of the previous method, we swapped the method where the enzyme being produced from the toehold switch is exonuclease III which is much smaller, takes 30 minutes to be detected, has a sensitivity of 0.5pM, and isothermally amplifies the miRNA so it can recycled It's detected using DNA probes and is sensitive as it using gold nanorods (34).

Firstly, the miRNA acts as a trigger as it is complementary to a section in a hairpin DNA probe to open it up. Once hybridised, exonuclease III will cleave the complex to produce ssDNA fragments. The miRNA is not cleaved so it is recycled to repeat the hybridisation and cleavage process, thus amplifying the product used in a LFT.The ssDNA product is complementary to the gold nanorods so will bind here first at the conjugate pad. On the test line there is an immobilised capture DNA probe with biotin modified at the end. The capture DNA probe is complementary to the ssDNA product, thus forming a coloured sandwich on the test line when the ssDNA product is present. On the control line, there is a control DNA probe which has the same sequence as the ssDNA product whilst also having biotin modified at the end. The gold nanorods are complimentary to this sequence so will always bind here (35).

HUMAN PRACTICES - Cycle 1 - Design

1 / 4

Cycle 1: Mini Jamboree Education event

The Mini Jamboree was an event held by the Imperial 2025 iGEM team which presented the opportunity for UK-based iGEM teams to present their project and network with experts in a range of scientific fields. We treated the Imperial Mini Jamboree as a practice for the real jamboree and took the opportunity to present our project in its exact stage at the time and receive feedback to improve upon. We were only given a short slot of 10-15 minutes so we needed to prioritise what would be said. To begin with, we had many team meetings, heavily involving all the sub teams, allocating slots for each part of our project to prepare slides for. We discussed at length what should be included in five sections: introduction, current technologies, research, target demographic and plan for the future. Following this, the appropriate corresponding team leads made in-depth scripts that summarised all of their work up until this point that coincided with the five sections we had decided on. Research and Lab team were responsible for the Research section , Human Practices and Ethics Team were responsible for the Target Demographic section and the Business team were in charge of the Plan for the Future section. From there, these scripts went through a rigorous process of summarising, revising and rehearsing before they were condensed down to only the most essential points which could be made in short digestible dialogues for our mini jamboree audience.

HUMAN PRACTICES - Cycle 1 - Build

2 / 4

Cycle 1: Mini Jamboree Education event

Following this, our business lead organised the scripts generated by the different subteams into a sensible order and made a google slides PowerPoint to support our presentation. This presentation highlighted the key information we wanted to get across and functioned as a base we could go off whilst presenting at the conference. We used our team colours pink and green (at the time) to give continuity between the different parts as well as keeping our logo on every side. Team members made visual representations such as toehold switch and CRISPR mechanisms to go along with the narratives we were explaining to facilitate understanding. Due to not all of our members being able to attend due to logistical reasons, we asked team leads to write up a question and answer section based on questions we thought we would likely receive and to provide this to our fellow team members who would be presenting on our behalf.

HUMAN PRACTICES - Cycle 1 - Test

3 / 4

Cycle 1: Mini Jamboree Education event

Going to the Jamboree was a good opportunity to meet other UK-based IGEM teams, such as Oxford and Imperial iGEM Team, and take inspiration from some of their presentations to improve our own project as well as to network with industry experts who may be valuable to our test. We promoted the event on our project instagram and used it as a learning experience. Our presentation was 10 minutes long and consisted mostly of our project idea and the research behind it, rather than the results of our test. We were not as prepared as we hope to be as we did not memorise the script , thus were less confident when presenting. We did a Q&A afterwards and some examples of questions we received included: how we would store and package our test with RNA being regarded as unstable, how our table is important for multidrug-resistant TB and the therapeutic side of TB etc . After our presentation, we received praise for our presentation and had people interested in iGEM and how to get involved, as well as how we came up with the project idea.

HUMAN PRACTICES - Cycle 1 - Learn

4 / 4

Cycle 1: Mini Jamboree Education event

HUMAN PRACTICES - Cycle 2 - Design

1 / 4

Cycle 2: LBTS6 Educational Presentation

Since our decision to compete for the education special prize, we knew we wanted to complete an educational presentation at a school to encourage the younger generation to get involved in synthetic biology. Our PI kindly put us in touch with a sixth form biology teacher in the local area. Ethan and Maria scheduled a meeting with the biology teacher, where we pitched the educational presentation we intended to give, convincing him of the benefits it would be to his students' learning. From this, we managed to secure a 75-minute window to deliver our session. We made a plan of what we wanted the educational presentation to cover, encompassing TB as a disease, the components of diagnostic tools, as well as the Human Practices aspects of our project. From there, we made a rough draft of our full presentation in Microsoft PowerPoint equipped with the corresponding script written in the notes section. Due to the younger audience, we tried to make this session a lot more interactive compared to our Mini Jamboree presentation with many planned activities. We also explained scientific concepts in more detail to get through to the kids and related this all to their upcoming UCAS deadlines as a motivating factor to participate.

HUMAN PRACTICES - Cycle 2 - Build

2 / 4

Cycle 2: LBTS6 Educational Presentation

We then adapted our PowerPoint presentation into Canva to be more visually appealing and professional-looking to our audience. We used our new team colours (pink-purple and blue) to represent our project and divide up the two main sections (TB as a disease and where we come in with diagnostics). Ethan, Maria, Zara and Matty all agreed to be involved in the presentation, with Maria being in charge of the beginning section about TB, then Zara explaining the mechanism of some diagnostic tools, Ethan talking about our project and the Human Practices and Matty helping out with classroom engagement. We had multiple meetings, both in-person and online, running through the presentation and refining the script to be fully prepared for our session. We incorporated a mentimeter into our slides to add a more fun, interactive aspect to it all. We wanted to frame it in a format where we prompt the kids to think about scientific concepts and then explain them in greater detail in relation to our project. For example, we planned to ask them what the world's most deadly infectious disease was and give clues to throw them off and we prompted them to think about categories of diagnostic tests before explaining the ones we chose for our project. Maria also prepared and printed physical resources for a matching activity we planned to consolidate knowledge we taught the kids about ELISA, lateral flow tests and PCR and encourage group work.

HUMAN PRACTICES - Cycle 2 - Test

3 / 4

Cycle 2: LBTS6 Educational Presentation

The majority of our presentation went smoothly and we were able to engage with our audience. They seemingly enjoyed all the activities we planned. During most of the more challenging activities, our team giving the presentation would break off and help out certain tables, using real-life examples to aid them in finding the correct answer, e.g., when working out different categories of diagnostic tools, we related this to existing ones that our audience would be familiar with e.g., pregnancy and Covid tests. Building off the feedback from our Imperial Mini Jamboree, we tried to integrate the Human Practices throughout the presentation, focusing on vulnerable communities. We also added many parts where we referenced our findings brought up by different stakeholders to give our project a more personal touch and set it apart. Additionally, we believe relating science to real-life situations enriches learning. We took photos (with permission) of this event to promote it on our website. We did have some problems with connectivity in the school, but luckily, we had accounted for this and brought pens and wrote on the whiteboard instead. Unfortunately, we were unable to get all the way through our presentation, missing out some of the section specifying the categories of our diagnostic in relation to Human Practices. This was a time management related issue as we overprepapered an extra long presentation with many activities and resources which we could not conceivably get through within the allocated time.

HUMAN PRACTICES - Cycle 2 - Learn

4 / 4

Cycle 2: LBTS6 Educational Presentation

This experience was a vast improvement from our first presentation we gave, we were significantly more organised, had good reception and interaction from the audience and had a more integrated approach to our Human Practices. We attribute this success to better organisation and assessment of our audience. We received positive feedback from both the teachers and students alike. The main things we need to improve upon for the final session we wish to conduct with our fellow undergraduates are: time management, we must be realistic with the material we prepare and the time it will take to get through as well as talking about the categories of our test from a Human Practices perspective as this was the part of our presentation we missed out and have not spoken publicly about yet.

References:

Engineering Cycle:

LAB:

iGEM. Design Stage [Internet]. Igem.org. 2025 [cited 2025 Oct 1]. Available from: https://technology.igem.org/engineering/design
iGEM. Build Stage [Internet]. Igem.org. 2025 [cited 2025 Oct 1]. Available from: https://technology.igem.org/engineering/build
iGEM. Test Stage [Internet]. Igem.org. 2025 [cited 2025 Oct 1]. Available from: https://technology.igem.org/engineering/test
iGEM. Learn Stage [Internet]. Igem.org. 2025 [cited 2025 Oct 1]. Available from: https://technology.igem.org/engineering/learn
Green Alexander A, Silver Pamela A, Collins James J, Yin P. Toehold Switches: De-Novo-Designed Regulators of Gene Expression. Cell. 2014 Nov;159(4):925–39.
McSweeney MA, Zhang Y, Styczynski MP. Short Activators and Repressors of RNA Toehold Switches. ACS Synthetic Biology. 2023 Feb 21;12(3):681–8.
Team:Exeter/Toehold Design - 2015.igem.org [Internet]. Igem.org. 2015. Available from: https://2015.igem.org/Team:Exeter/Toehold_Design
M. E. Fornace, J. Huang, C. T. Newman, N. J. Porubsky, M. B. Pierce, N. A. Pierce. NUPACK: analysis and design of nucleic acid structures, devices, and systems. ChemRxiv, 2022.
J. N. Zadeh, C. D. Steenberg, J. S. Bois, B. R. Wolfe, M. B. Pierce, A. R. Khan, R. M. Dirks, N. A. Pierce. NUPACK: analysis and design of nucleic acid systems. J Comput Chem, 32:170–173, 2011. (pdf)
B. R. Wolfe, N. J. Porubsky, J. N. Zadeh, R. M. Dirks, and N. A. Pierce. Constrained multistate sequence design for nucleic acid reaction pathway engineering. J Am Chem Soc, 139:3134-3144, 2017. (pdf, supp info)
B. R. Wolfe and N. A. Pierce. Sequence design for a test tube of interacting nucleic acid strands. ACS Synth Bio, 4:1086-1100, 2015. (pdf, supp info, supp tests)
Hoang Trung Chau T et al. Developments of Riboswitches and Toehold Switches for Molecular Detection—Biosensing and Molecular Diagnostics. International Journal of Molecular Sciences [Internet]. 2020 Jan 1;21(9):3192;1-20. Available from: https://www.mdpi.com/1422-0067/21/9/3192/htm
Zhou P et al. Cyclic chain displacement amplification-based dual-miRNA detection: a triple-line lateral flow strip for the diagnosis of lung cancer. Chemical Communications [Internet]. 2021 Jan 1 [cited 2025 Sep 30];57(92):12301–4. Available from: https://pubs.rsc.org/en/content/articlelanding/2021/cc/d1cc05442b
Sun H et al. A novel CRISPR/Cas13a biosensing platform comprising dual hairpin probe and traditional lateral flow assays. Sensors and Actuators B: Chemical [Internet]. 2024 Oct 9;423:136752; . Available from: https://www.sciencedirect.com/science/article/abs/pii/S0925400524014825
Yan ZF, Feng CQ, Zhou JQ, Huang QS, Chen XQ, Xia W, et al. Complete degradation of PET waste using a thermophilic microbe-enzyme system. International journal of biological macromolecules. 2024 Mar 1;260:129538–8.
Wei R, Oeser T, Schmidt J, Meier R, Barth M, Then J, et al. Engineered bacterial polyester hydrolases efficiently degrade polyethylene terephthalate due to relieved product inhibition. Biotechnology and Bioengineering. 2016 Feb 4;113(8):1658–65.
Then J, Wei R, Oeser T, Barth M, Belisário-Ferrari MR, Schmidt J, et al. Ca2+and Mg2+binding site engineering increases the degradation of polyethylene terephthalate films by polyester hydrolases fromThermobifida fusca. Biotechnology Journal. 2015 Jan 19;10(4):592–8.
Huang D, Zhang L, Sun Y. Rational Design of Disulfide Bridges in BbPETaseCD for Enhancing the Enzymatic Performance in PET Degradation. Molecules. 2023 Apr 17;28(8):3528–8.
Barth M, Honak A, Oeser T, Wei R, Belisário-Ferrari MR, Then J, et al. A dual enzyme system composed of a polyester hydrolase and a carboxylesterase enhances the biocatalytic degradation of polyethylene terephthalate films. Biotechnology Journal. 2016 Jun 21;11(8):1082–7.
Tropak MB, Yonekawa S, Karumuthil-Melethil S, Thompson P, Wakarchuk W, Gray SJ, et al. Construction of a hybrid β-hexosaminidase subunit capable of forming stable homodimers that hydrolyze GM2 ganglioside in vivo. Molecular Therapy - Methods & Clinical Development [Internet]. 2016;3:15057. Available from: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4774620/
Zhang H, Guan W, Zhou J. Advances in the Diagnosis of Latent Tuberculosis Infection. Infection and Drug Resistance. 2025 Jan 1;Volume 18:483–93
Price C, Nguyen AD. Latent Tuberculosis [Internet]. PubMed. Treasure Island (FL): StatPearls Publishing; 2024. Available from: https://www.ncbi.nlm.nih.gov/books/NBK599527/
Zhu Y, Lai J, Yang X, Wang S, Gu D, Huang Y, et al. CRISPR/Cas13a-driven lateral flow assay for preamplification-free and ultrasensitive miRNA-21 detection. Biosensors & bioelectronics [Internet]. 2025 Autumn;288:117850. Available from: https://pubmed.ncbi.nlm.nih.gov/40780174/
Pardee K, Green AA, Takahashi MK, Braff D, Lambert G, Lee JW, et al. Rapid, Low-Cost Detection of Zika Virus Using Programmable Biomolecular Components. Cell. 2016 May;165(5):1255–66.
Koczula KM, Gallotta A. Lateral flow assays. Essays In Biochemistry [Internet]. 2016 Jun 30;60(1):111–20. Available from: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4986465/
Le TT et al. Dual Recognition Element Lateral Flow Assay Toward Multiplex Strain Specific Influenza Virus Detection. Analytical Chemistry. 2017 Jun 9;89(12):6781–6.
Omidfar K, Riahi F, Kashanian S. Lateral Flow Assay: A Summary of Recent Progress for Improving Assay Performance. Biosensors [Internet]. 2023 Aug 23 [cited 2025 Sep 26];13(9):837. Available from: https://pubmed.ncbi.nlm.nih.gov/37754072/
Kang X, Zhao C, Chen S, Yang S, Zhang X, Xue B, et al. A Novel Approach Using LuxSit-i Enhanced Toehold Switches for the Rapid Detection of Vibrio parahaemolyticus. Biosensors [Internet]. 2024 Spring;14(12):637. Available from: https://pubmed.ncbi.nlm.nih.gov/39727902/
Ruppert C, Phogat N, Laufer S, Kohl M, Deigner HP. A smartphone readout system for gold nanoparticle-based lateral flow assays: application to monitoring of digoxigenin. Microchimica Acta. 2019 Jan 19;186(2).
Abcam. GFP ELISA Kit (ab171581) | Abcam [Internet]. Abcam.com. 2023 [cited 2025 Sep 25]. Available from: https://www.abcam.com/en-us/products/elisa-kits/gfp-elisa-kit-ab171581?srsltid=AfmBOoqrkFx2xAgi36qziDjCG5L5SWAoOS6OYVdW1M21sX-QJ
Boutal H, Moguet C, Pommiés L, Simon S, Naas T, Volland H. The Revolution of Lateral Flow Assay in the Field of AMR Detection. Diagnostics [Internet]. 2022 Jul 1 [cited 2025 Sep 25];12(7):1744. Available from: https://www.mdpi.com/2075-4418/12/7/1744/htm
Wang T, Chen L, Chikkanna A, Chen S, Brusius I, Sbuh N, et al. Development of nucleic acid aptamer-based lateral flow assays: A robust platform for cost-effective point-of-care diagnosis. Theranostics [Internet]. 2021 Mar 5;11(11):5174–96. Available from: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8039946/
Wang M et al. A lateral flow assay for miRNA-21 based on CRISPR/Cas13a and MnO2 nanosheets-mediated recognition and signal amplification. Analytical and Bioanalytical Chemistry. 2024 Apr 17;416(14):3401–13. Available from: https://link.springer.com/article/10.1007/s00216-024-05290-0
Nafari NB et al. Recent advances in lateral flow assays for MicroRNA detection. Clinica Chimica Acta. 2025 Feb;567:120096. Available from: https://www.sciencedirect.com/science/article/pii/S0009898124023490
TAKALKAR S et al. Gold Nanoparticle Coated Silica Nanorods for Sensitive Visual Detection of microRNA on a Lateral Flow Strip Biosensor. Analytical Sciences. 2016;32(6):617–22. Available from: https://link.springer.com/article/10.2116/analsci.32.617