.png)
Introduction: What are biosensors?
Biosensors are described as analytical devices used to detect biomarkers, or analytes, which can be anything from toxic chemicals, to small molecules, microbes or even peptides and nucleotides (Bhalla et al., 2014). These devices are made up of key parts. Firstly, biological components which uphold the biosensing capacity, for example antibodies detecting antigens or aptamers binding to their specific targets. Then, physical or chemical information transducers which signal the presence of a target. Finally, the signal is processed and we obtain an output (Figure 1). Traditionally these kinds of biosensors rely on electrochemistry, and we often think of them in the context of diabetes and Continuous Glucose Monitoring Devices, allowing patients with the condition to control their blood sugars. However, biosensors have very diverse applications which can range from disease monitoring to environmental monitoring, and there is a continuing need to make cheap and reliable, as well as simple biosensors that can be used across a wide range of applications.
Figure 1: Schematic of biosensors. Image taken from Verma and Gahlaut (2019)
Synthetic biology biosensors
Synthetic biology mainly relies on exploiting the power of nature to create useful tools, thus this approach could provide a new generation of biosensing systems that present great advantages over traditional devices. The first of these advantages lies in the fact that they are custom-made biological tools which allow for highly specific target identification. Moreover, there are two approaches in designing biosensors. On the one hand, they can be created by genetically engineering live cells to create Whole-Cell Biosensors (WCB) that are able to respond in real-time to their environment and detect specific analytes (Chen et al., 2023). On the other hand, we can generate cell-free biosensors which are essentially solutions consisting of all the required machinery, proteins and genetic circuits that allow for the detection of our targets (Zhang et al., 2020). Ultimately, both approaches produce a distinguishable qualitative or quantitative signal. They allow for quick and accurate sensing of targets, which as discussed is crucial for fulfilling the demand across a range of applications.
The rationale behind the workings of such biosensors is quite simple. They have an input, which can be anything from environmental signals to light or even small molecules, this input gets converted into a signal that gets processed by a genetic network which can be designed to perform diverse functions depending on our applications (Wang et al., 2023). In a similar way to electrical circuits, we can add in different ranges of signal processing steps. As a result, our biosensor is able to generate an output, for instance a change in colour by production of pigments visible to the naked eye, emission of fluorescence that we can see and quantify with the right equipment or even liberation of gases that can be monitored using ultrasound (Figure 2). There is a level of adaptability and it is ultimately up to us to decide on the purpose of these biosensors.
Figure 2: Schematic of synthetic biology based biosensors. (diagram from Xinyi Wan, University of Edinburgh, PhD Thesis)
Bacterial biosensing
Biosensors uphold great promises for bacterial biosensing whether it is from water, soil, human or food samples. It all starts with sample collection, treatment of the samples, and then detection can either be done directly on the bacteria by detecting exposed biomarkers or unique DNA and RNA sequences, or by indirect detection of unique toxins, peptides, or volatile organic compounds (VOCs) produced by the bacteria (Mazur et al. 2023).
Point-of-care (POC) devices
POC devices are medical diagnostic tools designed to provide rapid and convenient testing and analysis of patient samples. These devices aim to bring testing closer to the patients and aim to reduce the need for the sample to be sent into a central laboratory, thus allowing quicker decision making by healthcare providers. POC devices play a crucial role in improving patient outcomes, especially in scenarios where time sensitive results are required. A great example of those would be lateral-flow tests which can give simple informative results in less than 30 minutes without requiring the assistance of trained experts. Key features of POCs include: quick turnaround times of test results, great accessibility and ease of use in various healthcare and nontraditional settings and research areas, and crucially they must be user friendly even to those who are not laboratory staff (Mazur et al., 2023). These devices allow for great diagnostic ranges, for example diagnostic of infections, monitoring chronic conditions, and screening for patients before sourcing a study.
Paper-based biosensors
The vast majority of us will have been exposed to and acquainted with the use of paper-based biosensors during the COVID-19 pandemic, in the form of lateral flow tests. In simple terms, it corresponds to a sheet of paper where we have an ensemble of biological elements that can react with each other and generate our desired output upon contact with the target. They are cost-effective, provide rapid responses (usually between 5-30 minutes), are a simple portable solution, and can be adapted for remote or resource limited settings. Indeed, many of these advantages have been seen through the use of paper-based biosensors during the COVID-19 pandemic. These paper-based cell-free biosensors, and can be used alongside genetic networks to generate a whole new range of specific functions.
The workflow behind paper-based biosensors follows as described in figure 3; firstly we have our platform which is the microchannel paper, it has great microfluidic capacities and is designed to use a special paper which contains microchannels and spots. These channels are created using a technique such as printing or wax patterning, and within those channels we can embed our capture molecules as well as our synthetic gene network and cell-free extracts. The capture molecules are substances that can specifically bind to the bacteria of interest, and they are immobilised within the paper's channels. The liquid sample is inserted into the biosensor, and if the sample contains the bacteria, it will be able to be processed by our synthetic genetic circuit. The liquid flows along the paper channels carrying the bacteria with it, and as it comes into contact with the capture molecules our systems work and a signal is generated (Pardee et al., 2014).
Figure 3: Assembling a paper-based synthetic biology biosensor. (diagram taken from Pardee et al., 2014)
Why are they relevant to us?
There already exists a few golden standard techniques such as ELISA, qPCR, FTIR, etc., however while these techniques are very robust, accurate and sensitive to strain level, they can often also be costly, time consuming, require centralised laboratories, trained personnel, extensive sample pretreatment and multi-step processing. There is now a need to develop point-of-care devices that are fast, cheap, portable, and do not require any specialist training. This is especially important as low-income regions too often struggle to access adequate diagnostic tools for the detection of pathogens, ultimately leading to higher mortality rates (Pardee et al. 2014).
We will focus our attention on the article ‘A low-cost paper-based synthetic biology platform for analysing gut microbiota and host biomarkers’ (Takahashi et al, 2018). This study is of great importance, especially considering that the microbiome make-up is key to understanding health and diseases. This study was focused towards trying to detect the hypervariable regions of the 16S rRNA genes in order to perform taxonomic profiling of the microbiome. The study aimed to develop an approach that is affordable, on demand and allows for simple analysis of the microbiome from stool samples. In addition they aimed to develop a platform that could accurately identify species-specific mRNA from 10 different bacteria as well as the mRNA of 3 key biomarkers involved in inflammation (calprotectin, CXCL5 and IL-8) and one cytokine (oncostatin M) which helps to predict the efficacy of TNF-α therapies in IBD patients. Furthermore, they pushed their research for rapid and inexpensive detection of toxin mRNA in the diagnosis of C. difficile infections.
As a result of the study, a platform was successfully developed for analysis of the gut microbiota for clinical research and adaptability and low resource settings.Their device allowed for the orthogonal detection of species-specific mRNA from 10 different bacteria associated with gut health and disease with 3 fM limit of detection (LOD) achieved. In other words, this tool was able to discriminate between different bacteria and accurately report which ones were detected without having any cross-talk or reporting the wrong species. On top of a great accuracy, it is important to consider the LOD, which refers to the lowest amount of a substance (here mRNA) detected by the sensor. The lower the LOD, the more sensitive the biosensor is, making it efficient in detecting extremely small amounts of the bacteria's genetic material. A LOD as low as the one they achieved is reflective of great sensitivity. For comparison, traditional methods like qPCR often have LODs in the range of picomolars (pM) or femtomolars (fM). As a result, this biosensor is great for accurately detecting bacteria even if they are present in very low concentrations!
The Takahashi group was able to develop a simple approach to study relative abundances of bacteria in the stool samples via a semi-quantitative determination of the concentrations of each target mRNA. Although this semi-quantitative approach initially only provided them with a rough idea of mRNA amounts rather than exact concentrations, the group compared the samples with known referencing standards and created a standard curve which allowed them to estimate mRNA concentrations in the samples.
Results were even validated by a comparison with RT-qPCR which showed very similar performance. Finally, they were also able to identify different toxin mRNA expression levels from pathogenic C.difficile strains that were otherwise indistinguishable by using standard DNA-based qPCR diagnosis.
The system they generated is a translation-based biosensor which relies on a technology called “Toehold-switch sensors” that control the translation of genes. In simple terms, toehold switches correspond to a strand of RNA that can form a hairpin structure based on complementary base pairing. This RNA strand has a region specifically complementary to our target sequences and a region that encodes for the protein we want to express for the detection - in this case a green fluorescent protein (GFP). Due to the secondary structure of the RNA, the GFP cannot be expressed as it is made inaccessible to the transcriptional machinery. The only way of expressing the GFP is by exposing the toehold switch to the target mRNA sequence from the bacteria we want to detect. The two strands will be able to bind to each other, causing a conformational change in the toehold switch which allows GFP to be transcribed and expressed (Figure 4). The advantage of this technique is that RNA folding is universal and many computer design softwares can predict the 3D structure which facilitates their design. However, on the other hand RNA folding is highly sensitive to physiological conditions, so unless the right conditions are met, the ability of the biosensor to function might be hindered.
Figure 4: Toe-hold switch mechanism. (Diagram taken from Takahashi et al., 2018)
Advantages of this approach
Some of the main advantages of this approach are that it provides us with a great ease of use with simple results being observed. Moreover, the biosensors can be adapted for low resource settings, as the reactions that are happening do not require highly specialised equipment. For example, the GFP detection can be monitored on affordable and easy to build portable electronic readers that can quantify the changes in absorbance. The sample amplification of the mRNA is enabled through Nucleic acid sequence based amplification (NASBA) by using simple isothermic incubators, so there is no real need for thermocyclers with greater powers such as the ones which are required for qPCR (Kia et al. 2023). Additionally an important aspect of this is the cost, using this technique the mRNA can be quantified in around 3-4 hours for $16 per transcript using commercially available kits, with the potential to decrease this cost to $2 per transcript by using in-house cell-free extracts. Compared to qPCR which costs significantly more, it is clear that this is a more cost effective alternative (Takahashi et al., 2018).
All of this could be applied to a broad range of studies, including skin samples. This technology is easily adaptable to target other microorganisms, such as fungi, bacteriophages and human viruses. It could also be adapted for point-of-care use and at home monitoring for patients to respond to one of the greater goals: being able to regularly monitor disease progression in patients.
Limitations of the study
The first limitation that they encountered upon initial design was the low limit of detection of the toehold switch: with the sensor alone, the limit of detection was in the 10-30 nM range. They therefore decided to pair their approach with NASBA which is a technique done during the sample processing step to amplify the bacterial RNA prior to detection. This technique is very robust, particularly well-suited for RNA, and presents less risk of DNA contamination, therefore minimising the risk of a false positive result. As a result, this step greatly improved the sensitivity of this device to 3 fM.
Additionally, as the 16S sensors they first generated presented significant crosstalk in closely related bacteria and were therefore not suitable for discriminating amongst highly related bacterial species, they chose to target different mRNA sequences which are unique to the desired bacteria based on a bioinformatics pipeline they generated and obtained a perfectly orthogonal detection of the different bacterial species. Their sensors were however not yet optimised to identify down to strain level.
Moreover, despite requiring easily accessible equipment, their approach still involves some sample processing steps and is still not yet fully adapted for at-home testing, and therefore is not yet at the point-of-care application levels. However, we can still get a lot of important and valuable information in a more accessible way than traditional techniques.
Pioneering skin microbiome biosensors
There is very little research around biosensors in the field of the skin microbiome. However, translating this methodology and finding a wide array of specific targets for skin microbiome studies is achievable. An interesting avenue in the future would be introducing ribocomputing circuits for bacterial interaction studies to allow us to track the social interactions of bacteria in different dysbiosis conditions. Nowadays, more advanced paper-based biosensors allow for detection of signals directly with smartphones which can save the data into a cloud to monitor changes over time (Kim et al., 2021). The rapid advances in synthetic biology and growing interest in the skin microbiome will certainly open doors to a future where these proof-of-concept studies will serve as a basis to create wearable devices capturing in real time the changes of our beloved skin microbial communities.
References
Bhalla, N., Jolly, P., Formisano, N. & Estrela, P. (2016) Introduction to biosensors. Essays in Biochemistry. 60 (1), 1-8. 10.1042/EBC20150001.
Chen S, Chen X, Su H, Guo M, Liu H. Advances in Synthetic-Biology-Based Whole-Cell Biosensors: Principles, Genetic Modules, and Applications in Food Safety. Int J Mol Sci. 2023 Apr 28;24(9):7989. doi: 10.3390/ijms24097989. PMID: 37175695; PMCID: PMC10178329.
Damour, A., Robin, B., Deroche, L., Broutin, L., Bellin, N., Verdon, J., Lina, G., Leclère, F. M., Garcia, M., Cremniter, J., Lévêque, N. & Bodet, C. (2021) Phenol-soluble modulins α are major virulence factors of Staphylococcus aureus secretome promoting inflammatory response in human epidermis. Virulence. 12 (1), 2474-2492. 10.1080/21505594.2021.1975909.
Kia, V., Tafti, A., Paryan, M. & Mohammadi-Yeganeh, S. (2023) Evaluation of real-time NASBA assay for the detection of SARS-CoV-2 compared with real- time PCR. Irish Journal of Medical Science. 192 (2), 723-729. 10.1007/s11845-022-03046-2.
Kim, S., Lee, M. H., Wiwasuku, T., Day, A. S., Youngme, S., Hwang, D. S. & Yoon, J. (2021) Human sensor-inspired supervised machine learning of smartphone-based paper microfluidic analysis for bacterial species classification. Biosensors & Bioelectronics. 188 113335. 10.1016/j.bios.2021.113335.
Mazur, F., Tjandra, A. D., Zhou, Y., Gao, Y. & Chandrawati, R. (2023) Paper-based sensors for bacteria detection. Nature Reviews Bioengineering. 1 (3), 180-192. 10.1038/s44222-023-00024-w.
Nai, Y. H., Doeven, E. H. & Guijt, R. M. (2022) An improved nucleic acid sequence-based amplification method mediated by T4 gene 32 protein. PloS One. 17 (3), e0265391. 10.1371/journal.pone.0265391.
Pardee, K., Green, A. A., Ferrante, T., Cameron, D. E., DaleyKeyser, A., Yin, P. & Collins, J. J. (2014) Paper-Based Synthetic Gene Networks. Cell. 159 (4), 940-954. 10.1016/j.cell.2014.10.004.
Takahashi, M. K., Tan, X., Dy, A. J., Braff, D., Akana, R. T., Furuta, Y., Donghia, N., Ananthakrishnan, A. & Collins, J. J. (2018) A low-cost paper-based synthetic biology platform for analyzing gut microbiota and host biomarkers. Nature Communications. 9 (1), 3347-12. 10.1038/s41467-018-05864-4.
Wan, X. (2019) Synthetic biology enabled cellular and cell-free biosensors for environmental contaminants. The University of Edinburgh.
Wang, C., Zeng, H., Liu, K., Lin, Y., Yang, H., Xie, X., Wei, D. & Ye, J. (2023) Biosensor-based therapy powered by synthetic biology. Smart Materials in Medicine. 4 212-224. 10.1016/j.smaim.2022.10.003.
Zhang L, Guo W, Lu Y. Advances in Cell-Free Biosensors: Principle, Mechanism, and Applications. Biotechnol J. 2020 Sep;15(9):e2000187. doi: 10.1002/biot.202000187. Epub 2020 Jul 23. PMID: 32667120.
Comments