Multiple Sclerosis (MS) is one of the most common disabling neurological diseases among young adults. Current diagnostic procedures take a long time and a lot of effort from patients. Thus, miRADAR aims to simplify the diagnosis of complex MS by building a paper-based test. Our test will consist of multiple parts which will all work in combination on a paper-based platform. In the proof of concept section, we discuss how all these parts work independently. We also explain how they will work in combination with each other in the future and what more work needs to be done.

Multiple Sclerosis (MS) is one of the most common disabling neurological diseases among young adults.¹ The body’s immune system damages the central nervous system, causing a wide range of symptoms including loss of motor function and severe fatigue. Accurate diagnosis of MS can allow for treatment to prevent further damage and progression of the disease. Unfortunately, diagnosis of MS asks a lot of time and energy from a patient and can be very complex. With the current diagnostic methods, 10-30 % of patients receive an accurate diagnosis only after months to years. This delays the possibility of receiving treatment which is essential to maintaining quality of life.

miRADAR aims to simplify complex diagnosis cases of MS with the help of synthetic biology. We worked on an accessible and accurate test to support neurologists in their work. Our test is based on miRNAs in the blood as novel biomarkers for MS. miRNAs play a vital role in various cellular processes and can become dysregulated when a patient becomes ill. The difference in miRNA levels can be detected which indicates the presence of a disease. Our simple, cell-free paper-based test aims to detect an MS-specific combination of miRNA concentrations.

Our Proof-of-Concept combines all aspects of the diagnostic test we studied throughout iGEM. We showed that a combination of miRNAs can be selected computationally to specifically diagnose MS patients. In addition, we studied the amplification of miRNAs and designed an RNA threshold system. We also successfully detected miRNAs using self-designed toehold switches in vitro. Testing this in vitro was important as our final goal is to make the test paper-based. These are key steps towards a fully effective, simple diagnostic test for MS. Moreover, the technological advances demonstrated in our project can provide the foundation for the development of modular diagnostic tests for a wide variety of miRNA-detectable diseases. We need to invest more time in making the test fully optimised and ready for use. Therefore, we have included a Future Focus for each part in our Proof-of-Concept.

Selecting a specific miRNA combination for MS diagnosis

We learned from Dr. Pablo Villoslada that the same miRNAs can be up or downregulated in multiple diseases. This means that the detection of just a single miRNA is not sufficient to give a specific diagnosis. Therefore, we aimed to select a miRNA combination that specifically indicates the presence of MS. This miRNA combination must distinguish between MS patients and healthy people, as well as between MS patients and patients with similar diseases.

We first used a machine learning algorithm to identify the miRNA combination that can distinguish relapsing-remitting MS patients from healthy people. The accuracy of the model was 0.69. These miRNAs were then searched in the Human miRNA Disease Database to find other diseases that are associated with the found miRNAs. Based on this analysis, we found 8 miRNAs of which the concentration differences can distinguish relapsing-remitting MS patients from healthy people and diseases similar to MS: hsa-miR-17-5p:9.1, hsa-miR-431, hsa-miR-494, HS-263.1, hsa-miR-106a:9.1, hsa-miR-191, HS-65, and hsa-miR-1287. Thus, this is the specific biomarker combination we aim to detect with our test.

Future Focus

From stakeholders, we learned that the required accuracy should be 0.95. To achieve this, more miRNA expression data for MS patients as well as healthy people is required. This data should come from a diverse population to ensure that the combination with the highest diagnostic value is found.

Amplification of selected miRNAs using NASBA

While miRNAs are present in the blood in the fM-pM range,² most toehold switches have a limit of detection in the high nM range.³ Therefore, an amplification step is required upstream of our toehold switch circuit. We chose to investigate the suitability of the isothermal amplification technique NASBA for our test. NASBA consists of two sub-reactions that allow for RNA amplification of selected RNA sequences. Based on earlier literature research, we found hsa-miR-484 to be significantly dysregulated during MS.⁴ We therefore chose hsa-miR-484 as our model miRNA. Thus, we performed the sub-reactions of NASBA and managed to successfully demonstrate the amplification of miRNA hsa-miR-484. The first step is reverse transcription (RT) of a target miRNA to DNA. By doing a PCR on the reaction product, we showed that miRNA miR-484 can be reverse transcribed when present at a concentration as low as 700 pM (Figure 1).

2% agarose gel showing the limit of a detection assay for the reverse transcription reaction of miRNA has-mir-484. Reverse transcription was performed in triplo with a hsa-mir-484 concentration range of 70 nM – 7 fM. PCR was performed afterwards to confirm successful RT with a known DNA template as positive control. Reactions without Reverse Transcriptase enzyme (No RT) and with miR-484 (No miRNA) were negative controls. The 1kb Plus DNA ladder (NEB) was used.

The second sub-reaction is the transcription of the produced DNA into RNA. We obtained amplified RNA products (Figure 2). This RNA can serve as the input for the next step of the diagnostic test, which is our threshold system.

2% agarose gel with samples of *in vitro* Transcription of DNA template (T) with T7 promoter and miR-484 sequence. Two template concentrations were tested (1 = 700 ng, 2 = 300 ng), as well as a negative control (3) without a template, and a positive control template (4 = Fluc control template from HiScribe T7 High Yield RNA Synthesis Kit (NEB)). Samples were loaded on gel directly after *in vitro* Transcription, after Dnase I treatment, and after RNA purification. The 1kb Plus DNA ladder (NEB) (L1) and low-range ssRNA ladder (NEB) (L2) were used.

Future Focus

We successfully amplified the miRNA sequence through two reactions. The next step is to integrate these reactions in a one-pot exponential amplification reaction and to optimise it for lower miRNA concentrations to cover the full range of concentrations of miRNA in blood. Then, the selected miRNAs could be specifically amplified from a blood sample.

RNA threshold system to distinguish between healthy and dysregulated miRNA levels

Due to dysregulation of miRNA levels in MS, there is a difference between the concentrations of specific miRNAs in the blood of MS patients and healthy people. We would like our test to be activated only if the concentration of MS-specific miRNAs meets a certain threshold, found to distinguish healthy people from MS patients. After amplification of the selected miRNAs, this concentration difference is expected to be enlarged due to exponential amplification. Our detection system gives a linear output meaning that as the miRNA level increases, the output will increase. However, we aim to make a paper-based test which gives a binary output (i.e., either ON or OFF). Therefore, we designed an RNA threshold system based on the toehold-mediated strand displacement (TMSD) system.

The TMSD system uses the combination of four RNA strands: an input RNA strand (A), which is our amplified miRNA, an inhibitor strand (B), a trigger strand (C), and an output RNA strand (D). In principle, B is bound to C to prevent an output signal. There is an excess of B in the reaction to which the amplified A binds until a certain threshold concentration. Above that concentration, A starts displacing C as it has a higher affinity towards B. Then the displaced C can trigger the detection system.

To test this threshold system, we used fluorescent output. We showed that the trigger strand C can bind and activate output strand D, and that inhibitor strand B can bind C and thereby decrease the fluorescent signal. In addition, when we modelled this system, we predicted an optimal ratio of 3:1 for B: CD for the desired threshold (Figure 3).

TMSD-based RNA threshold system. Top left: Fluorescence signal over time for C, D and CD. Top right: The fluorescence signal over time after the addition of B to an incubated mixture of A, C, and D. Excitation = 470 nm, emission = 505 nm. Bottom: The dose-dependent curve for a concentration of B of 0.1, 0.5, and 2.5 \muM. C and D are kept constant at 0.5 \muM.

Future Focus

The threshold system requires further optimisation of the strand ratios in the reaction mixture. Once the system is optimised, the fluorescent output signal can be replaced by a toehold switch or toehold switch circuit, that is triggered if the amplified miRNA concentration is high enough.

Accurate detection of miRNA combinations using toehold switch gates

In our final test design, the threshold system can activate our toehold switch system to obtain a visible output. miRADAR uses toehold switches to detect trigger RNAs. For our Proof-of-Concept, we studied the performance of our self-designed toehold switches when targeting miRNAs directly in vitro. We showed that a de novo toehold switch can detect our model miRNA, hsa-miR-484, in a reaction volume as small as 2 µL. The lacZ coding sequence is translated upon activation of the toehold switch, which leads to enzymatic conversion of yellow substrate to purple product (Figure 4).

Cell-free test (PURExpress) of toehold constructs in 2 uL reaction volumes. To the samples in the top row, no trigger miRNA was added, in the bottom miRNA was added. A red box is placed around the AND-gate toehold switch, that detects two miRNAs. The negative and positive control are on the far left and right respectively.

Future Focus

Computationally, we found eight miRNAs that distinguish MS patients from mimic disease patients and healthy individuals. This suggests that our toehold switch circuits should become larger. To maintain accurate detection for large circuits, we made an algorithm to create optimal toehold switch circuits for a selected number of input miRNAs. A future step would be to test these larger circuit designs in the lab and to further optimise the model for large circuit design. In addition, different types of toehold switch AND gate designs could be tested and compared, such as the conventional AND gate from (Green et al., 2017),⁵ or the LIRA system from (Ma et al 2022).⁶

Accurate detection of miRNAs in a cell-free paper-based test

We aim to make our test accessible and easy to perform. Therefore, we perform our test cell-free and on paper. To carry out our toehold switch reactions, we require various necessary components, such as ribosomes, amino acids, enzymes and building blocks. We used the PURExpress system reaction mixture to perform the test cell-free and showed that it can be applied to paper discs, freeze-dried, and reactivated with a trigger miRNA solution after storage. (Figure 5)

We performed our cell-free test on paper discs. Left: schematic overview of the components added onto our paper disc, which is then freeze-dried and reactivated by a miRNA sample. Right: cell-free test on paper discs without miRNA added (-), with a construct with only the lacZ gene (+), and with our de novo toehold switch (B) and trigger miRNA.

Future Focus

From our discussions with Dr. Wim de Kieviet, we understood that our test should be easily distributed and stored for a long period. Therefore, the shelf-life of our diagnostic test remains to be investigated in the future. In addition, we can explore the use of other cell-free systems to optimise our test conditions.

Summary

In a nutshell, the miRNAs are selected (Figure 6.1) and amplified using NASBA (Figure 6.2). The amplified miRNAs go into the threshold system (Figure 6.3). After they reach a certain threshold, the trigger RNAs are released from the threshold system. These released trigger RNAs can activate the toehold switch system (Figure 6.4). The activation of the toehold switch leads to the translation of the lacZ gene, which allows for the colour change of the paper-based test (Figure 6.5).

Altogether, our results show that miRNAs can be specifically selected and amplified and can activate a toehold switch in a cell-free paper-based test. This provides a foundation for a modular and accessible diagnostic test platform based on miRNA-biomarkers.

(1)

Multiple Sclerosis National Institute of Neurological Disorders and Stroke. https://www.ninds.nih.gov/health-information/disorders/multiple-sclerosis (accessed 2024-09-29).

(2)

Gines, G.; Menezes, R.; Xiao, W.; Rondelez, Y.; Taly, V. Emerging Isothermal Amplification Technologies for microRNA Biosensing: Applications to Liquid Biopsies. Molecular Aspects of Medicine 2020, 72, 100832. https://doi.org/10.1016/j.mam.2019.11.002.

(3)

Park, S.; Lee, J. W. Detection of Coronaviruses Using RNA Toehold Switch Sensors. International Journal of Molecular Sciences 2021, 22 (4), 1772. https://doi.org/10.3390/ijms22041772.

(4)

Regev, K.; Healy, B. C.; Paul, A.; Diaz-Cruz, C.; Mazzola, M. A.; Raheja, R.; Glanz, B. I.; Kivisäkk, P.; Chitnis, T.; Jagodic, M.; Piehl, F.; Olsson, T.; Khademi, M.; Hauser, S.; Oksenberg, J.; Khoury, S. J.; Weiner, H. L.; Gandhi, R. Identification of MS-Specific Serum miRNAs in an International Multicenter Study. Neurology(R) Neuroimmunology & Neuroinflammation 2018, 5 (5), e491. https://doi.org/10.1212/NXI.0000000000000491.

(5)

Green, A. A.; Kim, J.; Ma, D.; Silver, P. A.; Collins, J. J.; Yin, P. Complex Cellular Logic Computation Using Ribocomputing Devices. Nature 2017, 548 (7665), 117–121. https://doi.org/10.1038/nature23271.

(6)

Ma, D.; Li, Y.; Wu, K.; Yan, Z.; Tang, A. A.; Chaudhary, S.; Ticktin, Z. M.; Alcantar-Fernandez, J.; Moreno-Camacho, J. L.; Campos-Romero, A.; Green, A. A. Multi-Arm RNA Junctions Encoding Molecular Logic Unconstrained by Input Sequence for Versatile Cell-Free Diagnostics. Nature Biomedical Engineering 2022, 6 (3), 298–309. https://doi.org/10.1038/s41551-022-00857-7.

Implementation