• Overview
• Cas13a
• Readouts
• Targets
• Detection Chip
• Amplification

Overview

We successfully designed, constructed and characterized each module of our platform: the sample processing, the readout circuit, and the detection of pathogens.
Although did not fully integrate all parts together in the time frame of our project, we could connect each unit to the next, so that we are confident that our entire platform is functional. For example, we could achieve equally sensitive bulk detection of pathogen RNA from in vitro and in vivo sources, and were later able to detect in vitro RNA with lyophilized Cas13a on paper, therefore we believe RNA from lysed cells can be detected on paper. In this overview, we list our achievements and where we faced issues, and we present a summary of the characterization of each module in the sub-pages.

Our modular units and the integration between them are mostly validated. Green ticks indicated full validation, yellow ticks indicated partial validation.

What worked

• Demonstrated the functionality of Cas13a proteins, namely Lbu and Lwa.
• Constructed a functional fluorescence detector with high sensitivity and low production cost.
• Modeled the detection limit of our circuit and confirmed it experimentally (~10 nM RNA).
• Detected pathogen RNA sequence from in vitro and in vivo sources.
• Differentiated viral sequences from bacterial sequences.
• Used the RNase Alert and the Spinach aptamer fluorescence readout circuits.
• Used gold nanoparticles to detect general RNase activity.
• Detected pathogen RNA in bulk and on paper, from native and from lyophilized Cas13a.
• Amplified isothermally a target DNA from in vitro and in vivo sources.
• Combined isothermal DNA amplification and RNA transcription on paper, producing detectable concentrations of a pathogen RNA.
• Improved the biobrick BBa_K1319008 by adding a 6x His-tag and provided Cas13a Lwa as three different composite biobricks.
• Created and characterized a Lwa Cas13a coding sequence, submitted as BioBrick.
• Created and characterized a collection of degradation tags, submitted as BioBricks.

What presented issues

• Optimizing the purification protocol for Cas13a.
• Demonstrating functionality of Lsh Cas13a.
• Ruling out RNase contamination from heat-lysed in vivo samples.
• Detecting Qbeta RNA.
• Developing colorimetric read-outs.
• Optimizing the lyophilization and stability of Cas13a.
• Amplifying long sequences with RPA.

Discussion

Our project CascAID is a universal solution for low cost, point of care diagnostics of infectious diseases. Currently, the available diagnostic tools are based on PCR, antibodies or microbiological methods which all need trained personal and lab equipment. Therefore, these methods are cost and time consuming. This gives rise to the need of developing effective, affordable and portable devices.

In our project, we first successfully replicated the Cas13a-based detection of RNA pathogens that was demonstrated by Gootenberg et al. (2017). We thoroughly characterized the target detection limit for different bacterial and viral targets, from in vitro and in vivo sources, and proved the possibility to discriminate between viruses and bacteria with high specificity. We found that our detection circuit worked robustly across experimental conditions and experimenters, which proves that the readout is adapted for distribution and handling by non-trained users.
We laid the groundwork for colorimetric read-outs that will add another layer of amplification in our cascade detection (gold nanoparticles, intein-extein and ssDNA amplification). Those readouts should allow for a practical readability of the diagnosis by the user without the need of digital analysis. Additionally, their amplification scheme should also lower the detection limit of the Cas13a without the need for pre-amplification of the target.

However, we developed in parallel a scheme for amplifying the target using isothermal amplification (RPA) and transcription. This was motivated by our modeling work, which determined a detection limit in the range of the one found experimentally (around 10nM), and quantified the improvement we could expect from a cascade amplification, from in vivo DNA to RNA to readout. RPA worked well from in vivo and in vitro DNA sources, and the combination of RPA and transcription on paper was efficient enough to overcome our readout detection limit.

We built a fluorescence detector that is to our knowledge, the cheapest and most sensitive ever built by an iGEM team, and provides a reasonable alternative to commercial plate readers. We used it successfully to detect Cas13a activity on paper, from in vitro transcribed RNA pathogen. However, the product itself may need to be redesigned for market distribution: in general, a fluorescence detector is not necessarily user-friendly, the extraction of the RNA on chip needs to be optimized.

Nevertheless, we are glad to have created a functional platform that allows the detection of nanomolar concentrations of a pathogen's RNA within 30 minutes. With our modular approach, we have shown at least proof-of-concept results for each part, and are confident that no fundamental gap prevents our platform from being fully integrated.

Outlook

We still have some modules that need improvement in the future. We have therefore listed the following points that need to be optimized below.

• Positive and negative controls in the readout: We occasionally found that high target concentrations led to signals above the positive control (which could be due to the degradation and lesser activity of RNaseA used for this control) and that low target concentrations could lead to signals below negative control (which could be due to noise at low fluorescence intensities) For proper quantification of the percentage of cleaved RNaseAlert, the controls should be standardized.
• In vivo heat lysis: During our experiments, we realized that the direct use of the RNA extracted from E. coli using heat lysis can lead to RNase contamination. Although our Cas13a cleavage assays are performed in presence of RNase Inhibitor to suppress the activity of the RNases that could be present, we observed that the heat lysed samples show relatively higher fluorescence activity in comparison to the phenol-chloroform extracted samples.
• RNA extraction and amplification: The RNA extracted from Bacillus subtilis lead to unstable results, giving sometimes higher than positive control signals, sometimes very noisy kinetics. As this is a gram positive bacteria, we think some further characterization must be done on the efficiency of heat lysis of different type of cells. In all cases, the RNA should be amplified after lysis and before detection, as pathogens are often present below our detection limit of 10nM in real samples.
• Cost of the chip: At the moment, the cost of our reusable detection unit is less than 15$ per unit. We could still try to minimize the costs by reducing the chip size and making it fully recyclable. We should characterize the life-time of the detector, to see how its cost is buffered by the number of tests that can be conducted with one detector. However, at the industrial level one could easily reduce the overall cost of CascAID by scaling up of the production. A rough cost estimation for the setup of a 1000 reactions gave us a cost per single test of around 0.85 $.
• Lyophilization of Cas13a: the lyophilization protocol of the Cas13a has to be improved in order to make our paper chip portable and sustainable. We tried drying the Cas13a with the tardigrade intrinsically disordered proteins (TDPs) from Team TU Delft, as a cryoprotectant, but this lead to increased basal activity, rendering the detection less sensitive. Other cryoprotectants should be tried, and the stability of freeze-dried samples over a year should be assessed.
• Readouts with color and amplification: The colorimetric readouts need continued work, and possibly improved design, since we only managed to partially succeed with the assays.
• Handling of real patient samples: Due to the safety restrictions in our lab, and our lack of experience with clinical studies, we did not work with real-world samples. The next step of this project, before thinking of market distribution, would be to test the functionality of our platform outside of the lab and under real point-of-care conditions.
• Integration of all the modules of the platform: Although all our modules parts are functional, and locally integrated, we did not reach full integration into a unique object, which would eliminate the need for a lab environment. It would still need to be accessed that this diagnosis device works in a variety of environments, when handled by non or minimally trained users. However, we believe we only need more time to assemble a fully functional and integrated module system.