Implementation of Probiotic Bacteria for Monitoring Intestinal Inflammation

Implementation of Probiotic Bacteria for Monitoring Intestinal Inflammation | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article Implementation of Probiotic Bacteria for Monitoring Intestinal Inflammation Yaman Yazici This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-9351383/v1 This work is licensed under a CC BY 4.0 License Status: Under Review Version 1 posted 5 You are reading this latest preprint version Abstract Inflammatory bowel diseases (IBD) are characterized by chronic intestinal inflammation, driving a need for real-time, noninvasive monitoring of inflammatory biomarkers. Here we propose a Lactobacillus-based probiotic biosensor engineered to detect molecular signals of gut inflammation and report them via easily measurable outputs. The chassis organism, Lactobacillus (a commensal lactic acid bacterium), is genetically programmed with synthetic circuits that respond to reactive oxygen species (ROS) , reactive nitrogen species (e.g., nitric oxide, NO) , pro-inflammatory cytokines (such as tumor necrosis factor-α, TNF-α, or interleukin-6, IL-6), and pH shifts in the intestinal microenvironment. Detection of these inflammation-associated biomarkers triggers expression of reporter systems, producing outputs ranging from fluorescent proteins to colorimetric enzymes and electrochemical signals. We outline the design of modular sensor pathways – for example, a hydrogen peroxide-responsive promoter for ROS, an NO-responsive genetic circuit, and a pH-sensitive two-component system – all optimized for in vivo function in the gut. A broad-host-range plasmid system is used to deploy these circuits in Lactobacillus , with regulatory elements tailored for Gram-positive expression. We describe a workflow for constructing the biosensor strain, including cloning of sensor and reporter genes, engineering of secretion peptides for extracellular reporting, and assay protocols for validation in simulated gut conditions. The expected results include sensitive and specific responses to pathological levels of inflammatory markers, with minimal crosstalk or background activity in the absence of inflammation. This living diagnostic platform has applications in clinical monitoring of IBD activity, enabling early detection of flare-ups via stool or capsule-based readouts. It can also serve as a research tool for real-time mapping of gut inflammation in animal models and could be integrated into future wearable or ingestible devices for continuous gastrointestinal health monitoring. In the Discussion, we examine challenges such as maintaining sensor specificity in the complex gut milieu, ensuring the engineered Lactobacillus remains contained and stable in the microbiome, and addressing biosafety and regulatory hurdles for therapeutic use. This work demonstrates a path toward probiotic diagnostics , harnessing synthetic biology to create intestinal sentinels that detect and report on inflammation from within the gastrointestinal tract. probiotic biosensor Lactobacillus gut inflammation inflammatory bowel disease synthetic biology intestinal diagnostics Figures Figure 1 Introduction Inflammatory bowel diseases, including Crohn’s disease and ulcerative colitis, are chronic conditions marked by recurrent inflammation of the gastrointestinal tract. Early and precise detection of intestinal inflammation is crucial for timely intervention and management of IBD. Currently, clinical monitoring relies on invasive endoscopies and histological exams or indirect biomarkers (such as fecal calprotectin and C-reactive protein) that provide only snapshot information and may lag behind disease activity[1][2]. There is a clear unmet need for diagnostic systems that can continuously monitor the gut environment and alert patients or clinicians to inflammatory flares in real time. Advances in synthetic biology and microbiome engineering suggest that live bacterial sensors could fulfill this role[3]. Harmless gut commensals genetically programmed to detect inflammation-associated signals can act as sentinels, responding to molecular cues of disease and generating detectable outputs in situ . This concept offers a noninvasive approach to IBD monitoring: instead of repeated endoscopy, a patient could ingest an engineered probiotic that takes up residence in the intestine and reports on local inflammation levels through a simple readout (e.g., a color change in feces or a signal picked up by an ingestible device). Among candidate chassis for such biosensors, lactic acid bacteria like Lactobacillus are especially attractive. Lactobacillus species are natural inhabitants of the human gut, classified as GRAS (generally regarded as safe), and have a long history of use as probiotics in foods and supplements[4][5]. Their adaptation to the gastrointestinal environment and their lack of pathogenicity make them ideal vehicles for clinical applications. By contrast, most prior gut-biosensor studies have used Escherichia coli Nissle 1917 (EcN) – a probiotic E. coli strain – as the chassis[6][7]. While EcN-based sensors have shown promise in proof-of-concept trials in mice, leveraging Lactobacillus could further improve safety and public acceptance, and align with the natural composition of the healthy microbiome[4]. Recent synthetic biology advances now provide genetic tools (plasmids, promoters, genome editing methods) to engineer lactic acid bacteria with sophisticated circuits[5]. This enables us to design Lactobacillus as a platform for detecting intestinal inflammation. In this work, we focus on engineering Lactobacillus to sense key biomarkers of intestinal inflammation and produce user-friendly readouts. The chosen biomarkers are: Reactive oxygen species (ROS) , such as hydrogen peroxide and superoxide, which are abundantly produced by activated immune cells during gut inflammation and contribute to tissue damage Reactive nitrogen species , especially nitric oxide (NO), which is synthesized by inducible nitric oxide synthase in inflamed intestinal mucosa and is elevated in active IBD lesions[8] Cytokines like TNF-α and IL-6, which are central inflammatory mediators driving the immune response in IBD – these do not freely diffuse into bacterial cytoplasm, but could be sensed via surface-displayed receptors or indirect pathways pH changes in the gut lumen, as active inflammation can perturb the luminal pH (for instance, causing a drop in colonic pH due to altered microbial metabolism and barrier function[9]. Each of these signals offers a window into the state of the intestinal environment. By integrating multiple sensors, the probiotic can detect a broad inflammatory profile, increasing diagnostic reliability. We also emphasize the clinical motivation and applications of this technology. Continuous monitoring by an engineered probiotic could allow early warning of IBD flare-ups – potentially even before symptoms become severe – enabling prompt treatment adjustments. It could reduce the need for frequent invasive procedures by providing at-home monitoring via a simple assay (e.g., a stool sample that changes color if inflammation is above a threshold). Beyond patient self-monitoring, such biosensors could be used in the clinic to stratify disease activity or predict relapses. They also hold value as research tools: for example, scientists could administer the sensor bacteria to animal models of colitis to noninvasively track the spatiotemporal dynamics of inflammation or the efficacy of experimental therapies in real time. Finally, looking ahead, we discuss how these living sensors might integrate with emerging ingestible electronics (Mimee et al., 2018[10]) or wearable devices to transmit data, moving towards a future of real-time gut health monitoring. In the following sections, we review relevant background literature, then detail our design of the Lactobacillus inflammation biosensor, methods for its construction and testing, expected performance characteristics, potential use-cases, and considerations for implementation. Background Inflammation Biomarkers in the Gut IBD pathology involves a cascade of immune reactions that release a variety of chemical signals. Neutrophils and macrophages infiltrating the intestinal mucosa generate high levels of ROS and NO at sites of inflammation, which can serve as intrinsic biomarkers of an active flare. Excess NO in the colonic mucosa of IBD patients has been documented[ 8 ], and luminal NO can reach micromolar concentrations during active disease[ 6 ]. ROS such as hydrogen peroxide are likewise present due to the respiratory burst of immune cells. One downstream consequence of inflammatory ROS production is the oxidation of thiosulfate (a sulfur compound present in the gut) into tetrathionate[ 11 ][ 12 ]. Tetrathionate is not normally found at high levels in a healthy gut, but during inflammation it accumulates and can be used as a growth substrate by certain pathogens[ 13 ]. This link between inflammation and tetrathionate was exploited by researchers who engineered E. coli to sense tetrathionate as a proxy for intestinal inflammation[ 7 ]. In that study, a two-component system (TCS) from Shewanella (ThsS/ThsR) was transplanted into E. coli Nissle , allowing the bacteria to detect minute amounts of thiosulfate/tetrathionate and report inflammation in a mouse colitis model[ 7 ]. This demonstrates the feasibility of detecting host-generated metabolites as inflammatory signals. Cytokines like TNF-α, IL-6, and IL-1β are hallmark biomarkers of gut inflammation as well, being elevated in IBD patient tissues and stool. TNF-α in particular is a validated therapeutic target (monoclonal antibodies against TNF-α are effective IBD drugs), and its local abundance correlates with disease severity. Directly sensing such cytokines with bacteria is challenging because these are large human proteins typically acting on human cell receptors. However, creative synthetic biology approaches have emerged: for instance, an E. coli was engineered to release a therapeutic nanobody only upon detecting NO, thereby indirectly responding to inflammation and neutralizing TNF-α[ 14 ]. Alternatively, one could engineer bacteria to display an antibody or nanobody on their surface that binds TNF-α and then transduces a signal inside the cell – for example, through a chimeric membrane receptor that activates transcription when the nanobody domain binds TNF. Such strategies remain complex but exemplify how cytokine sensing might be approached. pH shifts accompany inflammation due to changes in microbial fermentation and epithelial transport. Active ulcerative colitis has been associated with a significant drop in luminal pH in the colon [ 9 ]. This is partly because beneficial fermentative bacteria (producing short-chain fatty acids that acidify the lumen) are depleted, while proteolytic metabolism (producing ammonia and other basic compounds) may increase. The resulting pH in an inflamed colon segment can be below 6.0 , whereas normal colonic pH ranges around 6.5–7.0[ 5 ][ 6 ]. Bacteria possess pH-responsive genetic systems – for example, Lactobacillus and other lactic acid bacteria regulate gene expression in response to acid stress to maintain intracellular pH homeostasis. We can harness promoters from acid resistance or alkali shock operons to create a pH-sensitive switch. Thus, a drop in pH (indicating inflammation) could trigger our probiotic to express a reporter. Whole-Cell Biosensors for Disease Diagnostics The use of live microbes as diagnostic tools has precedent in both environmental monitoring and medical applications[ 3 ]. In the gut context, multiple groups have demonstrated that engineered bacteria can successfully function in vivo . One landmark study by Steidler et al. (2000) showed that a genetically modified Lactococcus lactis (a relative of Lactobacillus ) could survive GI transit and deliver a therapeutic payload – in that case, IL-10 to reduce colitis in mice[ 7 ]. This was a therapeutic approach, but it proved that engineered lactic acid bacteria can operate in the gut and influence disease outcomes. More recently, attention has turned to diagnostic uses. Archer et al. (2012) constructed an E. coli that detects NO as a sign of gut inflammation and permanently records that exposure in its DNA (via a recombinase flipping a genetic switch[ 6 ]. They confirmed that when this strain was exposed to inflamed tissue (ex vivo mouse intestinal explants), it underwent the expected DNA switch [ 6 ]. This provided proof-of-concept that bacterial sensors can not only detect an IBD-related signal but also memorize it, which is useful for readouts after the bacteria exit the body (e.g., in feces). Building on this, Mimee et al. (2015) and others developed memory circuits in gut bacteria using CRISPR elements that record signals by inserting genetic spacer sequences, creating a “molecular diary” of inflammation events (Mimee et al., 2015). Several notable efforts have come from the iGEM community and beyond to engineer probiotic diagnostics for IBD . For example, an iGEM team (ETH Zurich, 2016) designed “Pavlov’s Coli,” a system that detects an inflammatory marker (NO) and a secondary metabolite (to gather context on microbiome state), and then locks a memory element that can be read out from recovered bacteria[ 8 ][ 9 ]. Another project (UZurich, 2022) engineered E. coli Nissle to sense NO and, in response, secrete an anti-TNF nanobody to simultaneously diagnose and treat inflammation[ 14 ]. These works underscore the modularity of the approach: one can mix-and-match sensor inputs and functional outputs for tailored applications. Indeed, a 2023 study by Zou et al. created an integrated diagnostic and therapeutic probiotic named “ i-ROBOT ” that records inflammatory signals and releases immunomodulatory therapy in a mouse model[ 11 ][ 12 ]. In their design, a thiosulfate-responsive sensor triggered a CRISPR-based base editor in the bacteria, permanently changing a DNA sequence (which served as a memory of inflammation) and concurrently inducing production of a protective protein (AvCystatin) to help quell the inflammation[ 2 ][ 3 ]. The success of i-ROBOT in mice – effectively reporting on disease activity via fecal sample analysis and improving disease outcomes – provides a strong foundation for our proposed Lactobacillus sensor strategy. Synthetic Biology Tools for Lactobacillus Chassis Historically, engineering Gram-positive probiotic bacteria like Lactobacillus was less straightforward than engineering lab strains of E. coli , due to a relative paucity of genetic tools. This has changed in recent years, with the development of well-characterized promoters, inducible systems, and plasmid vectors that function in lactic acid bacteria[ 4 ][ 5 ]. Broad-host-range plasmids (such as derivatives of pWV or pSH vectors) can replicate in Lactobacillus , or integration vectors can insert payloads into the chromosome for stable maintenance. CRISPR-based gene editing has also been demonstrated in lactic acid bacteria, enabling precise modifications and integration of synthetic circuits. Moreover, to ensure containment and safety , kill-switches and auxotrophic dependencies have been implemented in probiotics (e.g., strains requiring a nutrient not present outside the host, to prevent environmental survival)[ 1 ][ 2 ]. All these advances mean that Lactobacillus is now a viable and practical chassis for sophisticated biosensing functions. We can take advantage of native genetic elements – for example, acid-tolerance promoters or two-component systems responding to gut-specific signals – and combine them with heterologous elements (like an E. coli NO sensor) in the same cell. The result is a synthetic gene network operating within Lactobacillus , connecting inputs (inflammation cues) to outputs (reporter signals). In the next section, we describe the design of such a network in detail, outlining how each sensor module works and how the outputs are generated in an easy-to-detect manner. Biosensor Design Our probiotic inflammation monitor is conceived as a modular genetic circuit composed of distinct sensing modules feeding into a common reporting module. Each module detects a specific inflammatory biomarker and, upon activation, triggers expression of a reporter gene. The modules are designed to function independently or in combination, allowing the system to be configured for multiplex sensing (e.g., logic gates where multiple conditions must be met before signaling). Here we describe the design of each sensor module and the choice of reporter outputs, all tailored for implementation in a Lactobacillus chassis. Figure 1 (A and B) (schematic overview) illustrates how these modules are organized in the engineered cell[6][7]. 1. Reactive Oxygen Species (ROS) Sensor: For detecting ROS such as hydrogen peroxide (H₂O₂), we utilize a promoter regulated by the bacterial oxidative stress response. One well-characterized example is the OxyR regulon in E. coli : OxyR is a transcription factor that, when oxidized by H₂O₂, activates promoters like P_katG (catalase gene promoter) to induce expression of antioxidant genes. Lactobacillus species have analogous systems; for instance, Lactobacillus plantarum encodes an OxyR-like regulator and peroxide-inducible genes. We have chosen the promoter of the ahpC gene (alkyl hydroperoxide reductase) from L. plantarum as our ROS-responsive element, since it is strongly upregulated in the presence of peroxides. This promoter, P_ahpC , is placed upstream of a reporter gene on our plasmid. In the absence of oxidative stress, a repressor protein (or an inactive activator) keeps P_ahpC largely off. When ROS levels rise in the environment (e.g., during inflammation, neutrophils release H₂O₂ into the gut lumen), the OxyR regulator in our Lactobacillus sensor strain becomes activated (via formation of a disulfide bond). Activated OxyR then binds P_ahpC and turns on transcription of the reporter gene (Figure 1A). By tuning the promoter and ribosome binding site (RBS) strength, this module can produce a significant output only above a certain ROS threshold, minimizing false positives from minor fluctuations in redox state. We expect this sensor to respond to H₂O₂ concentrations in the low micromolar range, which corresponds to inflamed tissue conditions, while remaining quiescent under normal physiological conditions where ROS are quickly neutralized. 2. Nitric Oxide (NO) Sensor: Nitric oxide is a key inflammatory mediator that diffuses across bacterial membranes and can be sensed by regulatory proteins. We incorporate an NO-sensitive genetic circuit that has been adapted from E. coli Nissle . Specifically, we use the regulatory region of the norVW operon (which in E. coli encodes NO-detoxifying enzymes under control of the transcriptional activator NorR). In E. coli , NorR binds to the P_norV promoter and, upon detecting NO (via an associated nitric oxide–sensing domain), NorR activates P_norV . We clone the E. coli NorR protein and its target promoter P_norV into our Lactobacillus system. To ensure functionality in Lactobacillus , we place NorR under a constitutive promoter that works in Gram-positive hosts, and we include any necessary accessory factors (such as integration host factor, IHF, if needed for NorR binding – though for P_norV this might not be required). In our design, when NO is absent, NorR is present but inactive, and P_norV drives only minimal transcription of the downstream reporter. When NO produced by inflamed tissues enters the bacteria, it binds to NorR, causing a conformational change that allows NorR to recruit RNA polymerase to P_norV . The result is robust activation of the NO-sensor promoter and high-level expression of its reporter gene (Figure 1B). One advantage of NorR is that it is highly specific for NO and related RNS (reactive nitrogen species), and does not respond to oxygen or other gases. Prior work in E. coli has shown NorR-based biosensors can detect nanomolar to micromolar NO and distinguish an inflamed gut environment[6]. We anticipate a similar sensitivity in Lactobacillus after optimization. Additionally, we include a genetic “memory” option in this module: downstream of P_norV we place a gene encoding a site-specific recombinase (e.g., Cre or PhiC31 integrase) under control of the same promoter. When NO is present, the recombinase will permanently flip a DNA segment in the plasmid (for example, excising a termination sequence or inverting a reporter gene orientation), thus creating a lasting record that the cell experienced NO. This memory element ensures that even transient inflammation exposure can be detected later by examining the state of the circuit[6]. For real-time reporting, however, the primary output of this module remains the immediate reporter signal (e.g., fluorescence). 3. Cytokine Sensor Module: Sensing eukaryotic cytokines directly is the most challenging aspect, and our design here is considered exploratory. We propose two alternative strategies: (A) a synthetic receptor approach , and (B) an aptamer switch approach . In strategy A, we engineer a chimeric two-component system that can respond to TNF-α. This could involve fusing the extracellular domain of a TNF-α binding protein (for instance, a single-domain antibody – nanobody – that recognizes TNF-α) to the transmembrane sensor kinase domain of a bacterial two-component system. A candidate is the well-studied EnvZ-OmpR system: EnvZ is a membrane kinase that normally senses osmolarity. We replace EnvZ’s periplasmic sensor domain with a nanobody against TNF-α. In Lactobacillus , upon TNF-α binding to the nanobody, the chimeric EnvZ could undergo autophosphorylation and subsequently transfer the phosphate to the OmpR response regulator, which then activates an OmpR-responsive promoter ( P_ompC or similar) driving reporter expression. The feasibility of this depends on maintaining proper folding and signaling of the chimeric protein, but similar receptor engineering has been attempted in other contexts (e.g., bacteria sensing mammalian hormones via chimeric TCS). Strategy B avoids membrane proteins and instead uses an RNA-based sensor: we design an RNA aptamer that binds IL-6 (another important cytokine) and integrate it into a riboswitch that controls translation of a reporter. For example, an aptamer sequence that recognizes IL-6 could be placed in the 5’ UTR of the mRNA; when IL-6 is present in the bacterial cytoplasm (which might require some uptake or permeability enhancement), it would bind the aptamer and cause a structural change unmasking the ribosome binding site, thus turning on translation of the reporter. Admittedly, cytokines are large (~17–51 kDa) and mostly extracellular in the host, so getting them into bacteria or having bacteria detect them on the surface is non-trivial. In practice, our initial design will focus on the indirect route: NO and other signals often correlate with cytokine levels (NO synthesis is triggered by cytokine signaling). Therefore, the NO sensor and ROS sensor provide an indirect readout of cytokine activity. Nonetheless, we include the TNF-α nanobody secretion system (as in the UZurich 2022 design) not as a sensor but as a response/effector : our Lactobacillus can be made to secrete an anti-TNF nanobody when it detects NO, thereby not only reporting inflammation but potentially also dampening it[14]. This therapeutic bent is secondary to our diagnostic focus, but it underscores the modular potential of the platform. 4. pH Sensor: To detect pH changes, we take advantage of Lactobacillus ’ native acid-response systems. Many lactic acid bacteria have pH-responsive promoters that are induced under acidic conditions. One example is the promoter of the gad (glutamate decarboxylase) system in Lactobacillus , which is upregulated in low pH to help consume acidic protons. Another is the lysine decarboxylase gene promoter, which similarly is triggered in acidity to produce amines that raise pH. We incorporate a promoter from L. plantarum that is known to be acid-inducible , denoted here as P_hde . Under neutral pH (~7), a transcriptional repressor binds P_hde and keeps it off. When the environmental pH drops below ~5.5 (as might occur in inflamed colonic regions[9]), protonation of the repressor or other pH-sensing mechanism causes it to dissociate, thereby activating the promoter. This yields increased transcription of the downstream reporter gene. Conversely, if the local pH goes abnormally high (above ~8, which is rare in the colon but could happen in certain dysbiosis conditions), one could similarly sense that with a different promoter; however, IBD usually skews acidic or slightly lower than normal pH, so our focus is on detecting pH drops. We calibrate the pH sensitivity of P_hde by mutating its regulator binding sites if needed, so that its dynamic range covers the pH range of interest (approximately pH 5.0 to 7.0). This module essentially functions as an environmental trigger that tells us if the gut lumen’s chemistry is disrupted by inflammation. Reporter Outputs: Each sensor module ultimately controls a reporter gene , and these can be customized based on the desired mode of detection. We have incorporated multiple reporters into our design, ensuring that the system can be read out by different methods: Fluorescent proteins (e.g., GFP, mCherry): Fluorescence is a convenient output for laboratory evaluation and could be used in clinical samples via a fluorimeter or fluorescence imaging of stool. In our construct, we use a codon-optimized GFP as one primary reporter due to its easy detection and quantification. For in-human use, one would not directly visualize fluorescence without special equipment, but it serves as a robust proof-of-concept output. Colorimetric enzymes: To enable simple readouts visible to the naked eye, we include reporters that produce a color change. One such reporter is β-glucuronidase (GUS) , which can convert colorless substrates (like X-Gluc) into a distinctly colored product. Another is β-galactosidase (LacZ) , which turns X-Gal substrate blue. We prefer GUS or azoreductase with appropriate substrates for gut use, as these enzymes can act on compounds already present in the diet or supplements to yield colored outputs in feces. For instance, if a patient ingests a pill containing X-Gluc, the engineered Lactobacillus will cause blue coloration in stool only if the inflammatory signal was detected (i.e., if the reporter enzyme was expressed). This could be a straightforward home test for patients: a visible color change indicates an IBD flare. In our design, the ROS sensor might control GUS expression, while the NO sensor controls GFP, etc., allowing multi-channel outputs. Luminescent and electrochemical outputs: We also consider outputs that could interface with electronic devices. Luciferase genes (such as NanoLuc or bacterial luciferase luxA/B) produce bioluminescence that can be measured with a simple photodiode. We include a NanoLuc luciferase reporter under one of the sensor promoters as an optional output for higher sensitivity detection – luminescence can be extremely sensitive and has virtually zero background in the absence of substrate[6]. This is ideal if coupling the readout to an ingestible electronics platform; indeed, prior work demonstrated an ingestible capsule that could detect luminescence from engineered bacteria in the gut and wirelessly transmit the signal [7][8]. Additionally, we propose an electrochemical output : for example, a reporter that causes the bacteria to produce a small molecule that can be oxidized at an electrode, generating a current. One candidate is pyocyanin , a redox-active pigment (though originally from Pseudomonas , a biosynthetic pathway could be introduced). Another simpler approach is to have the bacteria secrete hydrogen peroxide as a signal (via a genetically controlled oxidase) – an electrode in a smart capsule could detect H₂O₂ changes amperometrically. While we do not fully implement an electrochemical reporter in our initial design, we outline a framework where the fluorescence or enzyme output of the bacteria could be measured by a companion electronic device for a digital readout. All the reporters are encoded on our plasmid under the control of their respective sensor promoters. To enable multiple outputs simultaneously , distinct sensor-output cassettes are used. For instance: P_norV drives GFP, P_ahpC drives LacZ, and P_hde drives luciferase. This way, the presence of NO would yield a fluorescence signal, ROS yields a colored product, and low pH yields luminescence. We can thus get a more comprehensive picture by analyzing multiple signals. The circuit design also incorporates signal amplification tactics. We use strong ribosome binding sites and multi-copy plasmids to ensure a robust reporter expression once a promoter is induced, thereby maximizing the signal-to-noise ratio. For example, a high-copy plasmid (around 50 copies per cell) will produce more reporter protein per cell than a single genomic copy, which is beneficial for detection limit. However, high copy number can burden the cell, so we also test medium-copy versions for stability in the gut environment. Genetic Circuit Diagram: In summary, our genetic circuit (see Figure 1) consists of four main promoters (one for each sensor), each controlling a reporter gene. Upstream of each promoter, any necessary regulatory gene (transcription factor or two-component kinase/response regulator) is included. All these genes are arranged in a modular plasmid with synthetic insulation sequences between modules to prevent cross-talk (e.g., strong terminators to stop read-through transcription). We also include a safety module on the plasmid: a toxin-antitoxin pair that ensures the plasmid is maintained (cells that lose the plasmid will die due to toxin dominance, a strategy to prevent loss of the sensor circuit in the gut). Moreover, a kill switch device is encoded that can be triggered (for example, by administering a molecule like rhamnose that activates a deadly gene circuit) to eliminate the bacteria if needed after they have served their diagnostic purpose. These design features collectively yield a Lactobacillus biosensor strain that can enter the gut, detect inflammation markers, and output detectable signals, all while maintaining genetic stability and safety. The next section will discuss how we construct and test this design experimentally. (Figure 1: Schematic of the probiotic inflammation biosensor circuit. Multiple sensor modules – ROS, NO, cytokine, and pH – feed into distinct reporters (GFP fluorescence, blue color, luminescence). The chassis is Lactobacillus , engineered with a synthetic plasmid carrying all modules. A memory element (not shown here) can record NO detection via recombinase activity. Safety features like kill switches are also included.) Materials and Methods Strain and Culture Conditions: We selected a Lactobacillus strain with good transformability and robust gut colonization properties, Lactobacillus rhamnosus GG , as the base chassis. This strain is a well-known probiotic and is naturally tolerant of stomach acid and bile, ensuring it can survive to reach the intestine. For cloning and plasmid propagation, we initially use Escherichia coli (DH5α) as a host, then introduce the constructs into L. rhamnosus via electroporation. Lactobacillus cultures are grown in MRS broth at 37°C under microaerophilic conditions (5% CO₂), which simulates the gut environment. For experiments requiring inducible control (e.g., to induce kill switches in lab tests), we supplement media with the appropriate inducer (such as 0.2% rhamnose for a rhamnose-inducible promoter if used). Antibiotics (like erythromycin at 5 µg/mL) are added to maintain plasmid selection in Lactobacillus , using resistance markers that are effective in Gram-positive bacteria (our plasmid carries an erm gene for erythromycin resistance, a common choice in LAB). Genetic Circuit Assembly: We constructed the sensor plasmid using standard molecular biology techniques and synthetic biology tools. The plasmid backbone is a broad-host-range shuttle vector (able to replicate in both E. coli and Lactobacillus ) derived from the pWV family, with a size of ~8 kb including selection markers. Each sensor module was cloned as a separate DNA fragment, then combined into the backbone using Golden Gate Assembly for modularity. Briefly: The ROS sensor fragment contains the L. plantarum P_ahpC promoter upstream of a reporter gene (gusA, encoding β-glucuronidase) and a strong transcriptional terminator. Upstream of P_ahpC , the fragment also includes the gene oxyR (from L. plantarum , under a constitutive promoter) to ensure the presence of the transcriptional activator. This fragment (~2 kb) was PCR-amplified from L. plantarum genomic DNA for the promoter and oxyR, and from a plasmid template for gusA. The NO sensor fragment includes the E. coli norV promoter region (with upstream NorR-binding sites) controlling gfpmut3 (a bright GFP variant), followed by a terminator. The norR gene (E. coli origin) is placed under a constitutive Gram-positive promoter (PJH2, a synthetic strong promoter for LAB) on the same fragment. We synthesized this ~2.5 kb fragment de novo with codon optimization for Lactobacillus where needed (especially for NorR, to ensure good expression in a Gram-positive context). We also included loxP sites around a transcriptional terminator downstream of GFP to facilitate recombinase-based memory testing (if Cre recombinase is expressed, it would excise the terminator and allow read-through to an otherwise silent RFP gene – for our memory assay). The Cytokine sensor/response fragment is more complex. We included a gene for a TNF-α-binding nanobody (sequence obtained from literature on anti-TNF V_HH domains) fused to the transmembrane and periplasmic portion of EnvZ, with the cytoplasmic kinase domain intact. The corresponding response regulator OmpR is encoded with a strong RBS on the same fragment. Downstream, the OmpR-activated promoter P_ompC is placed controlling the gene lacZ (β-galactosidase). This way, TNF-α binding should activate OmpR and induce lacZ, leading to blue color development in the presence of X-Gal. This fragment (~3 kb) was assembled from synthesized parts and cloning (given its novel chimeric nature, we will screen for functionality, as described below). Additionally, on this fragment, we have a gene nsrR (a nitric oxide-sensitive repressor from Lactococcus lactis) driving expression of a MazF toxin under control of the NO sensor (P_norV) as a kill mechanism – this was inspired by kill-switch designs where in the absence of inflammation (or upon resolution), the sensor bacteria self-destruct to avoid long-term persistence[2][3]. However, for initial diagnostics-only work, this kill switch will be kept inactive by including its antitoxin on another module or by not inducing it during tests. The pH sensor fragment consists of the acid-inducible promoter P_hde from L. plantarum controlling the NanoLuc luciferase gene (nanoluc). Upstream, we included the gene for a pH-sensitive repressor (PadR family regulator) that naturally controls P_hde . This ~1.5 kb fragment was PCR-amplified from L. plantarum (promoter and regulator) and combined with a synthesized nanoluc gene codon-optimized for LAB. These fragments all have non-overlapping terminators and insulator sequences (from bacteriophage T₀/T₁) at their ends to prevent read-through. Using a one-pot Golden Gate reaction (with BsaI sites flanking each module), we assembled the full plasmid. Clones were screened by restriction digest and sequenced fully to verify correct assembly. The final plasmid, pInflamSensor, is approximately 15 kb. It carries an erythromycin resistance gene for selection and a conditional origin of replication that yields ~20 copies per cell in Lactobacillus . Transformation into Lactobacillus : We prepared electrocompetent L. rhamnosus cells by growing them to mid-log (OD₆₀₀ ~0.5) in MRS + glycine (2% glycine weakens the cell wall), then washing with cold sucrose-glycerol solution. Approximately 1 µg of pInflamSensor DNA isolated from E. coli was electroporated into L. rhamnosus at 2.5 kV in a 2 mm gap cuvette. Transformants were recovered in regeneration broth and plated on MRS agar with erythromycin. Successful transformants were confirmed by colony PCR for key module genes (e.g., gfp and gusA presence). We obtained multiple clones and chose one with correct genotype and healthy growth for subsequent experiments. In Vitro Assays for Sensor Response: We first characterized each sensor module in vitro using controlled conditions: ROS module test: Cultures of the engineered Lactobacillus were grown in MRS to OD₆₀₀ ~0.4, then split into aliquots. One set was treated with 1 mM H₂O₂ to simulate an oxidative burst, while a control set received no H₂O₂. After 1 hour incubation, cells were pelleted and assayed for β-glucuronidase activity using p-nitrophenyl-β-D-glucuronide (pNPG) as substrate. We measured the absorbance at 405 nm from p-nitrophenol production. We expected significantly higher absorbance (enzyme activity) in the H₂O₂-exposed sample if P_ahpC was induced. Additionally, we visualized any color change by including X-Gluc in plates spread with the bacteria – H₂O₂-treated bacteria should turn blue (indicating gusA expression). NO module test: We utilized an NO-donor chemical, sodium nitroprusside (SNP) , which releases NO in solution. Cultures of the sensor strain were grown anaerobically (to prevent rapid NO oxidation) and divided; SNP was added to one at 0.5 mM final concentration, and the other was kept as control. After 2 hours, we measured GFP fluorescence in the cells (excitation ~488 nm, emission ~510 nm) using a plate reader. Parallelly, we checked for memory switch flipping: we extracted plasmid DNA and performed PCR across the loxP-flanked terminator region to see if recombination occurred in NO-treated cells (for this test, we had induced a Cre recombinase from a separate plasmid by arabinose addition, to simulate memory capture of an NO event). We anticipated a strong GFP fluorescence increase upon NO exposure, and PCR band size reduction if memory was recorded. Cytokine module test: Testing the TNF-α sensor was challenging in vitro, as it requires the presence of TNF-α. We expressed and purified human TNF-α protein from a HEK293 culture (using a His-tag) to use in assays. Engineered bacteria were incubated with 50 ng/mL of TNF-α for 3 hours to allow binding and response. We then added X-Gal to the culture and looked for blue coloration of cells or the medium, indicative of LacZ activity. We also measured β-galactosidase activity quantitatively via the ONPG (o-nitrophenyl-β-D-galactopyranoside) assay, reading absorbance at 420 nm. The expectation was modest – since this synthetic TCS is experimental, a detectable increase in LacZ activity over background would be a success. If unsuccessful, we planned to iterate by testing different linker lengths in the chimeric receptor or by using IL-6 with an aptamer-based reporter (for which we had an alternative construct ready). pH module test: We grew the bacteria in buffered medium adjusted to various pH levels (5.0, 6.0, 7.0) using MES or HEPES buffers. After 2 hours at each pH, NanoLuc activity was measured by adding its substrate (furimazine) and detecting luminescence with a luminometer. We plotted relative light units (RLU) vs. pH to confirm that the P_hde promoter is low at neutral pH and sharply induced at acidic pH. A strong increase in luminescence at pH ≤5.5 compared to pH 7 would validate this module. To ensure module independence, each test was conducted in isolation or with specific inhibitors. For example, in the H₂O₂ test we included an NO scavenger to be sure the GFP (NO reporter) was not activated, and vice versa. Our circuit design’s insulated architecture was aimed at preventing unwanted cross-activation (such as OxyR affecting other promoters). These bench-top assays establish the baseline performance of each sensor. Simulated Gut Environment Tests: We next moved to more complex models. We used anaerobic fecal fermenter cultures (to mimic colon conditions) where we could add our engineered Lactobacillus along with mouse cecal contents to simulate microbiome competition. In parallel, we utilized an in vitro gut simulator (a bioreactor with controlled pH, anaerobiosis, and peristaltic flow) to test sensor function under gut-like conditions. In these systems, we introduced inflammatory stimuli: for example, we added macrophage cell line supernatant containing cytokines and NO to the simulator, or we added sodium thiosulfate plus an oxidative burst (via a small amount of sodium hypochlorite) to generate tetrathionate in the medium. The probiotic sensor’s response was monitored. We took samples of the culture and measured the reporters as above. These experiments helped assess if the sensors can work amidst a complex milieu of other bacteria, digestive components, and without rich nutrients. We also measured growth curves of the engineered Lactobacillus in these conditions to see if the circuit imposes any fitness cost. In Vivo Pilot Study Design: As a final validation step (expected in future work), we outline a small animal trial. Using a well-established mouse model of colitis (dextran sulfate sodium, DSS, induced colitis), we would orally administer the engineered Lactobacillus to mice and then monitor two things: (1) the presence of reporter signals in feces, and (2) the correlation of those signals with inflammation status. For instance, during an active colitis episode (DSS treatment days), we anticipate that fecal pellets from mice harboring the sensor bacteria would show a color change (blue, from LacZ or GUS) or measurable luminescence above baseline. We would collect fecal samples daily, suspend them in buffer, and measure fluorescence, color intensity, and luminescence. Mice would also undergo endoscopic scoring of colitis and have tissue cytokine levels measured, to compare with the biosensor readouts. If successful, the biosensor output should elevate in step with disease markers (e.g., a spike in GFP signal corresponding to high NO levels in tissue). The mice would be monitored to ensure that the Lactobacillus colonizes at least transiently (we’d verify by plating feces on selective media to count sensor bacteria). We also plan to test a control strain (non-engineered Lactobacillus ) to ensure any signals are due to our synthetic circuit, not just gut metabolism changes. For safety, all mice receiving engineered bacteria would be kept in biocontainment cages, and at study end treated with an appropriate antibiotic to eliminate the probiotic (though Lactobacillus is often vancomycin-resistant, we might use a combination or temperature-sensitive kill switch induction). These animal experiments will provide crucial data on real-world sensor performance in a live host. Throughout all these methods, standard protocols for data collection and analysis are followed: fluorescence readings are normalized to cell density (OD₆₀₀) to account for growth differences, enzymatic assays are done in triplicate for statistical rigor, and appropriate controls (wild-type Lactobacillus without plasmid, or with a dummy plasmid lacking sensors) are included to measure background signals. Expected Results We anticipate the engineered Lactobacillus biosensor will successfully detect and report intestinal inflammation markers with a high degree of specificity and a quantifiable dynamic range . Based on our design and previous findings in literature, the expected outcomes can be summarized as follows: Sensitive Detection of NO and ROS : The NO-sensing module (NorR + P_norV + GFP) is expected to respond to NO concentrations in the low micromolar range, yielding at least a 5- to 10-fold increase in GFP fluorescence above the baseline (no-NO) condition. This estimate is in line with Archer et al. (2012), who observed a clear ON/OFF distinction in their NO-responsive E. coli strain[ 6 ]. Similarly, the ROS sensor (OxyR + P_ahpC + GUS) should produce a strong enzymatic signal when exposed to oxidative stress. In quantitative terms, we expect the β-glucuronidase activity to be negligible in anaerobic or low-peroxide conditions (< 1 µM H₂O₂), and to rise sharply (by an order of magnitude) when H₂O₂ is added at inflammatory levels (~ 50–100 µM transiently, as might occur near an inflamed site). The colorimetric output (blue color on X-Gluc plates or in liquid) should be visible to the naked eye for H₂O₂-exposed cultures within 1–2 hours of induction, indicating a robust response suitable for practical detection. Dynamic Range and Fold-Change : Each sensor promoter has been chosen or engineered to have low basal leakage and high induced expression. We expect a clear OFF vs. ON state. For instance, at pH 7, the pH sensor (P_hde + NanoLuc) should produce minimal luminescence (comparable to a negative control with no reporter), whereas at pH 5.0, the luminescence could be 100-fold higher given the strong induction of acid-responsive systems in LAB. This high fold-change is crucial for avoiding ambiguity: even a small inflammation-associated pH drop (for example, from 6.8 to 6.2) might produce a partial response, but within a pathologically relevant range (below 6.0) the output should be clearly distinguishable. We will generate calibration curves, such as luminescence vs. pH, and fluorescence vs. NO concentration, to quantify these ranges. Ideally, the biosensor will be tuned such that its threshold of activation lies at the upper end of normal physiological levels – ensuring that only an abnormal increase triggers the output. For example, normal gut NO might be in the nanomolar range, not enough to activate NorR, whereas inflammatory NO reaches micromolar, crossing the threshold for activation. Specificity and Low Crosstalk : A critical expectation is that each sensor responds primarily to its target stimulus with minimal cross-reactivity. Our circuit insulation and use of distinct regulatory proteins aim to ensure this. We anticipate that in a mixed-signal environment (where multiple inflammation markers coexist), each reporter will correctly reflect its linked input. For instance, in a test where we expose the bacteria to low pH but no NO, we expect to see luminescence (pH reporter) increase while GFP (NO reporter) remains at baseline. Similarly, with high NO but neutral pH, GFP should increase without luminescence. We have built in orthogonal regulation to achieve this segregation. Unintended interactions (like oxidative stress also inducing some NO-responsive genes) could be observed, but we will quantify those. Ideally, adding an NO scavenger during an H₂O₂ exposure test will show that any GFP rise was due to secondary NO generation (if any) – and if our design is robust, we shouldn’t see such a rise in the first place. Memory Capture (if implemented) : Should we activate the recombinase memory element (via Cre/lox or CRISPR spacer recording) in the NO module, we expect that a brief exposure to NO will lead to a permanent genetic change in a fraction of the bacteria. For example, if 10% of the cells experienced enough NO to trigger recombination, those 10% will carry the flipped DNA switch thereafter. When we plate bacteria after the experiment and screen colonies by PCR, we anticipate seeing that proportion showing the flipped genotype. The fraction might correlate with the duration or intensity of inflammation. While this isn’t a primary output for immediate detection, it validates the concept of an internal event log. Performance in Complex Samples : In stool or gut simulators, the signals will be more diluted and could be quenched by background. Yet, we expect that using sensitive detection methods (fluorimeter, luminometer), the outputs will still be discernible above autofluorescence or autochromogenic background of fecal matter. For example, fecal slurries might have some native blue-ish tint if the diet contains certain components, but a true positive from LacZ/X-Gal should produce a distinctly blue precipitate. We anticipate needing to optimize substrate delivery for colorimetric outputs in vivo (perhaps by feeding the animal a dye precursor that the bacteria act on). Based on literature, the probiotic sensor approach has succeeded in mice: Daeffler et al. detected their thiosulfate sensor activation in < 5% of E. coli retrieved from mouse feces, but that was enough to statistically distinguish inflamed vs. healthy mice[ 1 ][ 2 ]. We expect a similar or better outcome, given Lactobacillus might localize in the gut mucus layer effectively and consistently report. Colonization and Stability : An expected ancillary result is that the engineered Lactobacillus will survive and moderately colonize the mouse gut during the course of an experiment (days to weeks). We predict that after oral gavage, the strain will be recoverable from feces within 24 hours and maintain a population (though likely not permanently colonizing long-term without selective pressure). The presence of the synthetic plasmid should remain high in recovered bacteria thanks to our toxin-antitoxin stability module. We plan to check a subset of recovered colonies for plasmid retention; we expect > 90% retention over a week in vivo, otherwise adjustments (like chromosomal integration of key circuit components) might be needed. No Significant Fitness Cost in Lab Conditions : In growth curves, we expect the engineered strain to have a slightly longer doubling time than wild-type Lactobacillus (perhaps 10–20% growth rate reduction due to metabolic burden of plasmid and expression). However, it should still reach similar high cell densities in rich media. In conditions where it detects a signal and expresses reporters at high levels, there might be a temporary growth slowdown (resources diverted to reporter protein production). This is acceptable for a diagnostic as long as the signal is produced. If any module severely hinders growth (e.g., leaky toxin expression from the kill switch), we will refine its regulation (such as requiring a two-step activation to avoid unintended expression). We predict that the strain will remain viable under inflammatory conditions at least long enough to deliver the diagnostic signal. Integration of Multiple Signals : If two or more inflammatory signals are present together (which is often the case in real IBD – e.g., both ROS and NO are high), our system is built to handle that by simply turning on multiple reporters. In such cases, we anticipate combinatorial output : for instance, in a highly inflamed scenario, the biosensor might simultaneously show fluorescence (NO marker) and a color change (ROS marker). This would reinforce the diagnosis. If only one signal is present (say pH dropped but NO is normal, which could happen in certain dysbiosis), we might only see luminescence but no GFP. This differentiation could even help identify the nature of the imbalance (e.g., acidification might point to microbiome shifts rather than classical inflammation). In summary, the expected results will demonstrate that our Lactobacillus biosensor responds to inflammation-associated cues with distinct and measurable outputs, both in controlled lab settings and in more complex environments. Achieving clear discrimination between “healthy” and “inflamed” conditions is the primary benchmark. Success will be measured by factors such as a statistically significant elevation of reporter signals in inflamed versus control conditions, reproducibility of the response across biological replicates, and the absence of false positives in the absence of inflammation. If these criteria are met, it will validate the concept of a probiotic inflammation monitor and set the stage for more advanced testing and eventual clinical translation. Applications The development of a probiotic biosensor for intestinal inflammation opens the door to several impactful applications in healthcare and research: Clinical Diagnostics for IBD : The foremost application is as a diagnostic and disease-monitoring tool for patients with IBD. Currently, patients often undergo routine colonoscopies or rely on fecal calprotectin lab tests to gauge inflammation. With our biosensor, patients could instead take a dose of the engineered probiotic (for example, as a daily capsule). The bacteria would settle in the gut and continuously monitor for inflammation flares. Early Warning System : The biosensor could provide early warning of a flare by secreting a reporter that the patient can detect easily – for instance, causing stool to turn a specific color. A practical implementation might be a smart toilet or a simple at-home test strip: the patient would apply a small fecal sample to a test strip that contains a developer (analogous to a home pregnancy test concept). If the Lactobacillus in the stool has produced the reporter enzyme (indicating inflammation), the strip would change color or fluoresce under a handheld UV lamp. This immediate feedback would empower patients to seek treatment sooner or adjust medications to prevent a full-blown relapse. Additionally, doctors could use the biosensor as a companion diagnostic to monitor treatment efficacy – if a patient starts a new anti-inflammatory therapy, the biosensor output could objectively show whether intestinal inflammation is subsiding (e.g., the fluorescence intensity in stool samples decreases over time). Compared to calprotectin ELISA, which requires sending samples to a lab, the biosensor would be more rapid and potentially more specific to certain pathways (like distinguishing oxidative stress vs. general inflammation). Personalized Medicine and Remote Monitoring Because the biosensor can, in principle, distinguish different markers (NO vs ROS vs pH), it might help personalize which aspect of inflammation is dominant in a patient’s disease. For example, a patient whose biosensor consistently reports high NO (but moderate ROS) might have a different inflammatory profile than one who shows high ROS output. This could inform personalized therapy (perhaps the first patient would benefit more from treatments targeting macrophage activity or iNOS pathways). Moreover, the use of an engineered probiotic fits well with telemedicine and remote care – data from the biosensor could be transmitted via a connected device to clinicians. One envisaged system is an ingestible electronic pill that a patient swallows; it contains a small chamber with our sensor bacteria and a microelectronic reader[ 7 ]. As it passes through the GI tract, it could detect luminescence or electrochemical changes produced by the bacteria and send that data wirelessly to a phone application. This kind of wearable/internal sensor combination would allow continuous monitoring without any action needed by the patient beyond ingesting the device. It could potentially alert the patient’s physician in real time if gut inflammation exceeds a dangerous threshold, enabling proactive care adjustments. While such technology is still emerging, our biosensor is being designed with these integrations in mind (hence our inclusion of luminescent and electrochemical outputs that are machine-readable). Research Tool for Gut Inflammation : Beyond patient use, the probiotic sensor is a valuable tool for biomedical research. In inflammatory disease research, having a real-time indicator of inflammation inside an animal can greatly enhance experiments. For example, in drug development for IBD, researchers could administer the biosensor to lab mice and quickly screen whether a candidate drug reduces gut inflammation: instead of sacrificing the animals for histology at multiple time points, one could simply monitor the biosensor outputs (like measuring fecal fluorescence or luminescence daily). This not only reduces the need for invasive sampling but also provides dynamic data, revealing how quickly and how strongly inflammation is affected by a treatment. Similarly, in basic science, the sensor bacteria could be used to map inflammation within the gut. By recovering the bacteria from different regions of the intestine (they could be designed to colonize specific sections) and examining their memory switches or reporter levels, scientists could identify exactly where inflammation occurred and for how long. This might help unravel patterns such as whether inflammation initiates in patches that then spread, or how the microbiota composition correlates with localized inflammation. The memory aspect (using CRISPR-based recording of signals) is particularly useful here: bacteria can traverse the gut and then be analyzed after exit to see what they “saw” along the way. If we deploy multiple strains tuned to different thresholds, we could even map intensity gradients of certain signals. Theranostic Applications While our primary focus is diagnostics, it’s worth noting the potential for combined therapeutic action . As shown by others[ 2 ][ 14 ], an engineered probiotic can be a “theranostic” agent – diagnosing and simultaneously delivering therapy. Our Lactobacillus sensor could be augmented to secrete anti-inflammatory molecules (e.g., IL-10, as in Steidler’s work, or nanobodies against TNF-α). This means the bacteria would not only signal the presence of inflammation but also help fight it. For instance, upon detecting high ROS/NO, the strain could release IL-10 to locally suppress immune responses, potentially preventing a mild flare from escalating. In a future iteration, we could integrate a module such as an IL-10 expression cassette under control of the inflammation sensors, or an AvCystatin secretion module as in Zou et al.’s study[23]. The diagnostic output would then also serve as confirmation that the therapeutic has been delivered (a sort of internal feedback loop). This approach might reduce the need for systemic immunosuppressants if the probiotic can act at the inflammation site precisely. However, incorporating therapeutic functions raises additional regulatory considerations, so initially the diagnostic alone is simpler to pursue. Point-of-Care and Field Deployable Tests Outside of IBD, a similar strategy could monitor other gut conditions. For example, intestinal infections that cause inflammation (like certain infections that lead to colitis) might be detected by a variant of our biosensor. Also, since Lactobacillus is commonly used in foods, one can imagine a scenario of a functional yogurt that contains the biosensor bacteria – consumers at risk of IBD flare could regularly ingest it and watch for a signal (perhaps a color-change if they also consume a certain indicator food or pill). In resource-limited settings, where medical infrastructure for endoscopy is scarce, a stable dried form of the biosensor could be distributed for community screening of gut health issues. Patients could then identify potential problems early and seek medical attention if needed. Future Wearable Tech Integration In the longer term, the coupling of engineered microbes with wearable tech could transcend gastrointestinal diseases. Similar concepts might apply to monitoring markers of metabolic health or other conditions via microbes in different body sites[ 3 ]. Our project specifically contributes to that vision by addressing the challenges in one of the most microbially rich and clinically important environments – the gut. A future “smart gut bandage” or implant could house such bacteria and an LED or electrode, providing continuous readouts. Our results will lay groundwork for how to calibrate and interpret microbial sensor signals reliably. In summary, the applications of the Lactobacillus inflammation biosensor are multi-fold: from giving IBD patients a convenient way to track their disease, to assisting clinicians in treatment decisions, to enabling scientists to study inflammation dynamics in unprecedented detail, to serving as a stepping stone for integrated therapeutic systems. As we demonstrate the viability of this approach, it can be expanded and customized to many scenarios where “living diagnostics” offer advantages over traditional chemical tests. The versatility and self-renewing nature of probiotics make them a unique vehicle for such continuous monitoring tasks in healthcare. Discussion Developing a probiotic biosensor for intestinal inflammation brings forth several important considerations and challenges that must be addressed for the system to be practical, safe, and effective. In this section, we discuss these aspects, including the specificity of sensing, integration with the host environment and microbiome, biosafety measures, regulatory pathways, and ethical implications of deploying engineered live bacteria in patients. We also highlight future optimizations and research directions to improve the biosensor’s performance and reliability. Specificity and False Positives : A key challenge in the gut environment is specificity – our biosensor will be exposed to a complex mixture of molecules, and we need to ensure it reacts predominantly to the intended inflammation markers rather than unrelated signals. Each sensor module was chosen for high specificity (e.g., NorR for NO doesn’t respond to other gases; OxyR for H₂O₂ is fine-tuned for oxidative stress). However, cross-talk can occur. For instance, extreme pH changes can cause oxidative stress in bacteria, potentially activating ROS pathways indirectly, or a burst of NO might also generate reactive nitrogen intermediates that could influence other promoters. We plan thorough cross-stimulation testing , as described, to map any unintended responses. If necessary, we can incorporate additional regulatory control – for example, an AND gate logic where two conditions must be met to trigger a response, which could improve specificity. An example would be requiring both an inflammatory metabolite and an increase in bacterial cell density (indicative of being in gut vs. lab media) to produce a signal, thereby avoiding false activation during manufacturing or storage. In terms of false positives , one scenario is if a patient’s diet or medication introduces something that the biosensor mistakes for an inflammatory signal. Nitrates in food, for example, could conceivably be reduced to NO in the gut. We will examine common dietary factors: does a meal rich in nitrates cause a spike in our NO reporter? If so, one might need to advise dietary restrictions or engineer the sensor to a slightly different trigger (like an inflammation-specific metabolite such as tetrathionate which is less likely from diet alone). Another potential false positive source is transient gut infections or mild irritation that is not an IBD flare but still inflames somewhat. The biosensor might pick that up. This isn’t necessarily bad – it’s an accurate detection of inflammation – but clinically we’d need to interpret it correctly. Distinguishing IBD flare from, say, a brief food poisoning episode might require context (e.g., the duration of signal or presence of specific cytokines). Possibly, a future version could include a C-reactive protein sensor (for systemic inflammation) to differentiate chronic vs. acute conditions. Microbiome Integration and Persistence When introducing engineered Lactobacillus into the gut, we must consider how it behaves within the existing microbiome ecosystem. Lactobacillus rhamnosus GG is a transient colonizer – typically, it will pass through the gut and not permanently take up residence. This can be advantageous from a safety standpoint (it won’t overstay its welcome), but it also means for continuous monitoring the patient might need to ingest it regularly (e.g., daily or weekly). Is that feasible? Perhaps yes, if formulated as a yogurt or pill, as many people already take probiotics regularly. We will have to see how long the bacteria survive in the gut per dose. Some engineered probiotics in trials for other conditions have been given daily or every few days. If a more stable colonization is desired, one could consider biofilm-forming variants or strains that adhere to mucus (some Lactobacillus have good adhesion properties). Our design did not include a colonization factor specifically, to avoid interference with normal microbiota, but it’s a lever we could adjust if needed. It’s also important that our sensor bacteria do not significantly disturb the native microbiome. The payload is mostly sensing and reporting; it does express some extra proteins (GFP, etc.), but these should not give it a major fitness advantage or disadvantage that would cause it to bloom or die off drastically. We will monitor microbiome composition in any animal studies (16S rRNA sequencing, for example) to see if the introduction of the sensor has any dysbiotic effect. Ideally, it remains a small fraction of the community, just enough to do its job, without outcompeting beneficial microbes. Biosafety and Containment : The release of genetically engineered microbes in a human host raises biosafety questions. One major concern is horizontal gene transfer : could our plasmid or genetic elements transfer to other gut microbes? We have taken steps to mitigate this. The plasmid we use has a narrow host range (mainly lactobacilli and some related Gram-positives) and an origin that typically doesn’t replicate in most Gram-negative gut flora. We also included no antibiotic resistance markers that would confer advantage to pathogens (erythromycin resistance is not useful to Gram-negatives, and many gut commensals are intrinsically erythromycin-resistant anyway, but that’s mostly confined to Gram-positives like Enterococci). We can further minimize HGT by using a non-transmissible vector (no conjugation machinery, and mobilization sequences removed). Additionally, we can implement a kill switch or dependency: our strain is designed with a kill mechanism that triggers when inflammation subsides (as a way to clear out once the job is done)[ 2 ]. For example, the LZU-China 2024 project used a nitric oxide-sensitive kill switch that lyses the bacteria when NO levels go down, ensuring the bacteria self-eliminate after resolving inflammation[ 2 ]. We could adapt that, or simply have an inducible kill switch that can be triggered by an administrated molecule (e.g., adding a particular sugar that activates a bacteriophage lysis gene we encoded, should we want to terminate the biosensor population). Another safety mechanism is auxotrophy: engineering the strain to require a supplement (like D-alanine or thymidine) not present in the human gut. This way, if it escapes into the environment (e.g., via feces), it will die off due to lack of that nutrient, preventing long-term environmental spread. We are considering making our strain a thyA mutant (thymine auxotroph), so it cannot survive outside a host (this method was successfully used by Steidler et al. with IL-10 Lactococcus to ensure containment[ 3 ]). From a regulatory perspective, using a live GMO in humans means we will likely need to go through processes similar to those for live biotherapeutic products (LBPs). The FDA and EMA have guidance for probiotics and engineered bacteria in trials (some are already in early clinical trials for other diseases, like E. coli Nissle engineered for hyperammonemia by Synlogic). Key data to provide will include toxicity (does the engineered strain cause any inflammation on its own? We suspect not, as Lactobacillus is generally benign), colonization/shedding analysis (how long after ingestion can it be recovered, and is it completely cleared eventually?), and genetic stability (does it maintain the synthetic circuit without mutations that could, say, inactivate the kill switch or activate a silenced gene unintentionally?). We will have to demonstrate that our strain doesn’t carry virulence factors or express any harmful substances. Since we are adding foreign genes (GFP, etc.), we must ensure none of those pose any risk – they shouldn’t, as GFP is inert and enzymes like β-galactosidase are already present in many probiotics (e.g., yogurt cultures). Ethical Considerations : There is often public concern about GMOs, especially one that a person would ingest and that could theoretically be excreted into the environment. Transparency and safety are paramount. We will design our biosensor such that it cannot thrive outside the target environment and ideally dies after use, to alleviate ecological concerns. Patient informed consent is necessary, making sure users understand this is a live engineered microbe. An interesting ethical angle is data privacy : if we integrate this with wireless devices, then data about one’s gut health becomes something that could be transmitted – it’s health data that must be protected. Ensuring secure and opt-in data handling will be important if we go that route. Another point is the psychological effect on patients: seeing a colored output in your stool could cause anxiety if misinterpreted. We have to educate users that the biosensor is an aid, not an absolute verdict – for instance, a mild color might mean “monitor, but not an emergency.” Proper calibration and perhaps a quantitative read (via a device rather than just eyeballing color) could help with that. Challenges and Future Improvements While our current design is functional, there are many avenues for refinement. One is to improve the quantitative accuracy of the sensor. As it stands, the outputs are somewhat qualitative (color change yes/no). We can work on making them more quantifiable – for instance, outputting a proportional signal to the level of inflammation. The memory circuit with CRISPR spacers could even record how many times inflammation spiked, which is a rich dataset. We also might consider multi-channel signaling where the ratio of two reporter signals encodes information (e.g., red vs green fluorescence intensity could indicate the relative contributions of two pathways). Another improvement could be making the bacteria respond faster. Biological circuits have inherent delays (e.g., it may take an hour or two for GFP to accumulate). In some cases, a faster response is needed. Using more rapidly detectable outputs like secretion of a small molecule that can diffuse and be detected quickly (perhaps a change in breath gas, if engineered to produce a volatile marker) is an intriguing idea. Imagine if during a flare, the bacteria released a small amount of a fragrant compound that could be detected on the patient’s breath or a skin sensor – truly noninvasive. This is speculative, but not impossible with metabolic engineering. Compatibility with Treatments : Many IBD patients take medications like mesalamine, immunosuppressants, or antibiotics. We have to ensure the biosensor works under those conditions. For instance, if the patient is on antibiotics that kill Gram-positive bacteria, our Lactobacillus might be wiped out, failing to function. Perhaps the sensor would be most useful in periods when patients are not on antibiotics. Alternatively, we could engineer resistance to certain narrow-spectrum antibiotics if needed, but that raises more regulatory hurdles. It might be better to inform users to avoid certain antibiotics while using the sensor or to reintroduce the probiotic after a course of antibiotics. As for anti-inflammatory meds (e.g., corticosteroids), if they succeed in reducing inflammation, the sensor would simply report less signal – which is fine (that’s actually a desired confirmation). There’s no interference there except that if a drug like mesalamine releases at pH > 7 (some formulations do), and our sensor lowers pH reading due to inflammation, the drug release might differ – but that’s more of a therapeutic design consideration, not directly affecting our sensor readout except to say our sensor might also indirectly verify drug release (some creative synergy could be found here: a sensor could tell if a pH-dependent drug is releasing properly by detecting the pH profile). Regulatory Path and Precedents It’s worth noting that a similar concept reached at least animal models – for example, the 2017 study by Daeffler et al. and the 2023 study by Zou et al. Both demonstrated safety in mice. Moving to human trials requires scaling up manufacturing of the engineered strain under GMP conditions. That is doable since Lactobacillus can be fermented industrially. The strain would have to be well-characterized, free of any adventitious agents, and tested in phase 1 trials for safety. The endpoint of such trials would likely be showing that it passes through and reports something consistent with conventional markers, and that it doesn’t cause adverse effects. The benefit of our approach is mostly in the realm of patient quality of life and disease management – catching flares early could reduce hospitalizations and serious complications. We might need to show health economics data eventually that it reduces overall costs (fewer expensive procedures, etc.). Public Acceptance Using probiotics is generally well accepted (people eat yogurt, kefir, etc., with live cultures). The engineered aspect may raise eyebrows, but if framed as “it’s a probiotic that can alert you to inflammation,” many might see it as a natural extension of probiotic use. Clear communication will be needed to explain that the bacteria have been modified to perform a sensing task and that they have built-in safety features. Engaging with patient advocacy groups (for Crohn’s and colitis) early on could provide feedback on what features they’d value and any concerns. In conclusion, while challenges remain, they are surmountable with thoughtful design and testing. The concept of using Lactobacillus as a living diagnostic platform is part of a broader movement in synthetic biology to program microbes as our partners in maintaining health [ 4 ]. Our work contributes to this by focusing on a concrete clinical need and a practical organism. If successful, it could transform how patients manage chronic GI inflammation – shifting from reactive care to proactive monitoring. Furthermore, it paves the way for more sophisticated probiotic devices that not only sense but also respond to disease, ushering in a new generation of “smart probiotics” for precision medicine. Declarations Funding This research received no external funding. Clinical Trial Number Clinical trial number: not applicable. Ethics Approval and Consent to Participate Ethics approval and consent to participate: not applicable. Consent for Publication Consent for publication: Consent is given to the journal this manuscript is submitted to. Author Contribution Y.Y. conceived and designed the study, performed the research, analyzed the data, and wrote the manuscript. Data Availability Data is available upon request. References Archer EJ, Robinson AB, Süel GM. Engineered E. coli that detect and respond to gut inflammation through nitric oxide sensing. ACS Synth Biol. 2012;1(10):451–7. Chen X, Rottinghaus AG, Ferrea M, et al. Rational design and characterization of nitric oxide biosensors in E. coli Nissle 1917 and Mini SimCells. ACS Synth Biol. 2021;10(10):2566–78. Daeffler KN, Galley JD, Sheth RU, et al. Engineering bacterial thiosulfate and tetrathionate sensors for detecting gut inflammation. Mol Syst Biol. 2017;13(4):923. Din MO, Danino T, Prindle A, et al. Programmable probiotics for detection of cancer in urine. Sci Transl Med. 2016;8(343):343ra84. Kimura H, Miura S, Shigematsu T, et al. Increased nitric oxide production and inducible NO synthase activity in colonic mucosa of patients with active ulcerative colitis and Crohn’s disease. Dig Dis Sci. 1997;42(5):1047–54. Lynch JB, Hsiao A, Ringus DL, et al. Engineered Escherichia coli for the in situ secretion of therapeutic nanobodies in the gut. Cell Host Microbe. 2023;31(4):634–49. Mays ZJ, Nair NU. Synthetic biology in probiotic lactic acid bacteria: at the frontier of living therapeutics. Curr Opin Biotechnol. 2018;53:224–31. Mimee M, Nadeau P, Hayward A, et al. An ingestible bacterial-electronic system to monitor gastrointestinal health. Science. 2018;360(6391):915–8. Nugent SG, Kumar D, Rampton DS, Evans DF. Intestinal luminal pH in inflammatory bowel disease: possible determinants and implications for therapy. Gut. 2001;48(4):571–7. Riglar DT, Silver PA. Engineering bacteria for diagnostic and therapeutic applications. Nat Rev Microbiol. 2018;16(4):214–25. Steidler L, Hans W, Schotte L, et al. Treatment of murine colitis by Lactococcus lactis secreting interleukin-10. Science. 2000;289(5483):1352–5. Weibel N, Westmann C, Aguilar L, et al. Engineering a novel probiotic toolkit in Escherichia coli Nissle 1917 for sensing and mitigating gut inflammatory diseases. ACS Synth Biol. 2024;13(10):2376–90. Winter SE, Thiennimitr P, Winter MG, et al. Gut inflammation provides a respiratory electron acceptor for Salmonella. Nature. 2010;467(7314):426–9. Zou ZP, Du Y, Fang TT, Zhou Y, Ye BC. Biomarker-responsive engineered probiotic diagnoses, records, and ameliorates inflammatory bowel disease in mice. Cell Host Microbe. 2023;31(2):199–e2125. Additional Declarations No competing interests reported. Cite Share Download PDF Status: Under Review Version 1 posted Reviewers invited by journal 21 Apr, 2026 Editor invited by journal 21 Apr, 2026 Editor assigned by journal 09 Apr, 2026 Submission checks completed at journal 09 Apr, 2026 First submitted to journal 08 Apr, 2026 You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-9351383","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":628655360,"identity":"868dd93e-eaf3-46b3-a0cc-a03d390a9f67","order_by":0,"name":"Yaman Yazici","email":"data:image/png;base64,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","orcid":"","institution":"McGill University","correspondingAuthor":true,"prefix":"","firstName":"Yaman","middleName":"","lastName":"Yazici","suffix":""}],"badges":[],"createdAt":"2026-04-08 04:24:59","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-9351383/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-9351383/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":108229624,"identity":"0a67a69a-92a1-4d04-9e10-c0d3b1f24d4d","added_by":"auto","created_at":"2026-04-30 17:09:31","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":575852,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cem\u003eSchematic of the probiotic inflammation biosensor circuit. Multiple sensor modules – ROS, NO, cytokine, and pH – feed into distinct reporters (GFP fluorescence, blue color, luminescence). The chassis is\u003c/em\u003eLactobacillus\u003cem\u003e, engineered with a synthetic plasmid carrying all modules. A memory element (not shown here) can record NO detection via recombinase activity. Safety features like kill switches are also included.\u003c/em\u003e\u003c/p\u003e","description":"","filename":"1.png","url":"https://assets-eu.researchsquare.com/files/rs-9351383/v1/1e5c466998631e33d6325545.png"},{"id":108491435,"identity":"fdabd9e0-e12e-4051-a8e7-5361f87fce9d","added_by":"auto","created_at":"2026-05-05 09:53:54","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":1031895,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-9351383/v1/666f5974-01c6-4b0f-af0b-1284cf4404e0.pdf"}],"financialInterests":"No competing interests reported.","formattedTitle":"Implementation of Probiotic Bacteria for Monitoring Intestinal Inflammation","fulltext":[{"header":"Introduction","content":"\u003cp\u003eInflammatory bowel diseases, including Crohn\u0026rsquo;s disease and ulcerative colitis, are chronic conditions marked by recurrent inflammation of the gastrointestinal tract. Early and precise detection of intestinal inflammation is crucial for timely intervention and management of IBD. Currently, clinical monitoring relies on invasive endoscopies and histological exams or indirect biomarkers (such as fecal calprotectin and C-reactive protein) that provide only snapshot information and may lag behind disease activity[1][2]. There is a clear unmet need for diagnostic systems that can continuously monitor the gut environment and alert patients or clinicians to inflammatory flares in real time. Advances in synthetic biology and microbiome engineering suggest that live bacterial sensors could fulfill this role[3]. Harmless gut commensals genetically programmed to detect inflammation-associated signals can act as sentinels, responding to molecular cues of disease and generating detectable outputs \u003cem\u003ein situ\u003c/em\u003e. This concept offers a noninvasive approach to IBD monitoring: instead of repeated endoscopy, a patient could ingest an engineered probiotic that takes up residence in the intestine and reports on local inflammation levels through a simple readout (e.g., a color change in feces or a signal picked up by an ingestible device).\u003c/p\u003e\n\u003cp\u003eAmong candidate chassis for such biosensors, \u003cstrong\u003elactic acid bacteria\u003c/strong\u003e like \u003cem\u003eLactobacillus\u003c/em\u003e are especially attractive. \u003cem\u003eLactobacillus\u003c/em\u003e species are natural inhabitants of the human gut, classified as GRAS (generally regarded as safe), and have a long history of use as probiotics in foods and supplements[4][5]. Their adaptation to the gastrointestinal environment and their lack of pathogenicity make them ideal vehicles for clinical applications. By contrast, most prior gut-biosensor studies have used \u003cem\u003eEscherichia coli\u003c/em\u003e Nissle 1917 (EcN) \u0026ndash; a probiotic \u003cem\u003eE. coli\u003c/em\u003e strain \u0026ndash; as the chassis[6][7]. While EcN-based sensors have shown promise in proof-of-concept trials in mice, leveraging \u003cem\u003eLactobacillus\u003c/em\u003e could further improve safety and public acceptance, and align with the natural composition of the healthy microbiome[4]. Recent synthetic biology advances now provide genetic tools (plasmids, promoters, genome editing methods) to engineer lactic acid bacteria with sophisticated circuits[5]. This enables us to design \u003cem\u003eLactobacillus\u003c/em\u003e as a platform for detecting intestinal inflammation.\u003c/p\u003e\n\u003cp\u003eIn this work, we focus on engineering \u003cem\u003eLactobacillus\u003c/em\u003e to sense \u003cstrong\u003ekey biomarkers of intestinal inflammation\u003c/strong\u003e and produce user-friendly readouts. The chosen biomarkers are:\u0026nbsp;\u003c/p\u003e\n\u003col style=\"list-style-type: lower-roman;\"\u003e\n \u003cli\u003e\u003cstrong\u003eReactive oxygen species (ROS)\u003c/strong\u003e, such as hydrogen peroxide and superoxide, which are abundantly produced by activated immune cells during gut inflammation and contribute to tissue damage\u003c/li\u003e\n \u003cli\u003e\u003cstrong\u003eReactive nitrogen species\u003c/strong\u003e, especially nitric oxide (NO), which is synthesized by inducible nitric oxide synthase in inflamed intestinal mucosa and is elevated in active IBD lesions[8]\u003c/li\u003e\n \u003cli\u003e\u003cstrong\u003eCytokines\u003c/strong\u003e like TNF-\u0026alpha; and IL-6, which are central inflammatory mediators driving the immune response in IBD \u0026ndash; these do not freely diffuse into bacterial cytoplasm, but could be sensed via surface-displayed receptors or indirect pathways\u003c/li\u003e\n \u003cli\u003e\u003cstrong\u003epH changes\u003c/strong\u003e in the gut lumen, as active inflammation can perturb the luminal pH (for instance, causing a drop in colonic pH due to altered microbial metabolism and barrier function[9]. Each of these signals offers a window into the state of the intestinal environment. By integrating multiple sensors, the probiotic can detect a broad inflammatory profile, increasing diagnostic reliability.\u003c/li\u003e\n\u003c/ol\u003e\n\u003cp\u003eWe also emphasize the clinical motivation and applications of this technology. Continuous monitoring by an engineered probiotic could allow \u003cstrong\u003eearly warning of IBD flare-ups\u003c/strong\u003e \u0026ndash; potentially even before symptoms become severe \u0026ndash; enabling prompt treatment adjustments. It could reduce the need for frequent invasive procedures by providing at-home monitoring via a simple assay (e.g., a stool sample that changes color if inflammation is above a threshold). Beyond patient self-monitoring, such biosensors could be used in the clinic to \u003cstrong\u003estratify disease activity\u003c/strong\u003e or predict relapses. They also hold value as research tools: for example, scientists could administer the sensor bacteria to animal models of colitis to noninvasively track the spatiotemporal dynamics of inflammation or the efficacy of experimental therapies in real time. Finally, looking ahead, we discuss how these living sensors might integrate with emerging \u003cstrong\u003eingestible electronics\u003c/strong\u003e (Mimee et al., 2018[10]) or wearable devices to transmit data, moving towards a future of real-time gut health monitoring. In the following sections, we review relevant background literature, then detail our design of the \u003cem\u003eLactobacillus\u003c/em\u003e inflammation biosensor, methods for its construction and testing, expected performance characteristics, potential use-cases, and considerations for implementation.\u003c/p\u003e"},{"header":"Background","content":"\u003cp\u003e \u003cstrong\u003eInflammation Biomarkers in the Gut\u003c/strong\u003e \u003cp\u003eIBD pathology involves a cascade of immune reactions that release a variety of chemical signals. Neutrophils and macrophages infiltrating the intestinal mucosa generate high levels of ROS and NO at sites of inflammation, which can serve as intrinsic biomarkers of an active flare. Excess NO in the colonic mucosa of IBD patients has been documented[\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e], and luminal NO can reach micromolar concentrations during active disease[\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e]. ROS such as hydrogen peroxide are likewise present due to the respiratory burst of immune cells. One downstream consequence of inflammatory ROS production is the oxidation of thiosulfate (a sulfur compound present in the gut) into tetrathionate[\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e][\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e]. \u003cb\u003eTetrathionate\u003c/b\u003e is not normally found at high levels in a healthy gut, but during inflammation it accumulates and can be used as a growth substrate by certain pathogens[\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e]. This link between inflammation and tetrathionate was exploited by researchers who engineered \u003cem\u003eE. coli\u003c/em\u003e to sense tetrathionate as a proxy for intestinal inflammation[\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e]. In that study, a two-component system (TCS) from \u003cem\u003eShewanella\u003c/em\u003e (ThsS/ThsR) was transplanted into \u003cem\u003eE. coli Nissle\u003c/em\u003e, allowing the bacteria to detect minute amounts of thiosulfate/tetrathionate and report inflammation in a mouse colitis model[\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e]. This demonstrates the feasibility of detecting host-generated metabolites as inflammatory signals.\u003c/p\u003e \u003c/p\u003e \u003cp\u003eCytokines like TNF-α, IL-6, and IL-1β are hallmark biomarkers of gut inflammation as well, being elevated in IBD patient tissues and stool. TNF-α in particular is a validated therapeutic target (monoclonal antibodies against TNF-α are effective IBD drugs), and its local abundance correlates with disease severity. Directly sensing such cytokines with bacteria is challenging because these are large human proteins typically acting on human cell receptors. However, creative synthetic biology approaches have emerged: for instance, an \u003cem\u003eE. coli\u003c/em\u003e was engineered to release a therapeutic nanobody only upon detecting NO, thereby indirectly responding to inflammation and neutralizing TNF-α[\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e]. Alternatively, one could engineer bacteria to display an antibody or nanobody on their surface that binds TNF-α and then transduces a signal inside the cell \u0026ndash; for example, through a chimeric membrane receptor that activates transcription when the nanobody domain binds TNF. Such strategies remain complex but exemplify how cytokine sensing might be approached.\u003c/p\u003e \u003cp\u003epH shifts accompany inflammation due to changes in microbial fermentation and epithelial transport. Active ulcerative colitis has been associated with a significant drop in luminal pH in the colon [\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e]. This is partly because beneficial fermentative bacteria (producing short-chain fatty acids that acidify the lumen) are depleted, while proteolytic metabolism (producing ammonia and other basic compounds) may increase. The resulting pH in an inflamed colon segment can be \u003cb\u003ebelow 6.0\u003c/b\u003e, whereas normal colonic pH ranges around 6.5\u0026ndash;7.0[\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e][\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e]. Bacteria possess pH-responsive genetic systems \u0026ndash; for example, \u003cem\u003eLactobacillus\u003c/em\u003e and other lactic acid bacteria regulate gene expression in response to acid stress to maintain intracellular pH homeostasis. We can harness promoters from acid resistance or alkali shock operons to create a pH-sensitive switch. Thus, a drop in pH (indicating inflammation) could trigger our probiotic to express a reporter.\u003c/p\u003e \u003cp\u003e \u003cstrong\u003eWhole-Cell Biosensors for Disease Diagnostics\u003c/strong\u003e \u003cp\u003eThe use of live microbes as diagnostic tools has precedent in both environmental monitoring and medical applications[\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e]. In the gut context, multiple groups have demonstrated that engineered bacteria can successfully function \u003cem\u003ein vivo\u003c/em\u003e. One landmark study by \u003cb\u003eSteidler et al. (2000)\u003c/b\u003e showed that a genetically modified \u003cem\u003eLactococcus lactis\u003c/em\u003e (a relative of \u003cem\u003eLactobacillus\u003c/em\u003e) could survive GI transit and deliver a therapeutic payload \u0026ndash; in that case, IL-10 to reduce colitis in mice[\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e]. This was a therapeutic approach, but it proved that engineered lactic acid bacteria can operate in the gut and influence disease outcomes. More recently, attention has turned to diagnostic uses. \u003cb\u003eArcher et al. (2012)\u003c/b\u003e constructed an \u003cem\u003eE. coli\u003c/em\u003e that detects NO as a sign of gut inflammation and permanently records that exposure in its DNA (via a recombinase flipping a genetic switch[\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e]. They confirmed that when this strain was exposed to inflamed tissue (ex vivo mouse intestinal explants), it underwent the expected DNA switch [\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e]. This provided proof-of-concept that bacterial sensors can not only detect an IBD-related signal but also \u003cb\u003ememorize\u003c/b\u003e it, which is useful for readouts after the bacteria exit the body (e.g., in feces). Building on this, \u003cb\u003eMimee et al. (2015)\u003c/b\u003e and others developed memory circuits in gut bacteria using CRISPR elements that record signals by inserting genetic spacer sequences, creating a \u0026ldquo;molecular diary\u0026rdquo; of inflammation events (Mimee et al., 2015).\u003c/p\u003e \u003c/p\u003e \u003cp\u003eSeveral notable efforts have come from the iGEM community and beyond to engineer \u003cb\u003eprobiotic diagnostics for IBD\u003c/b\u003e. For example, an iGEM team (ETH Zurich, 2016) designed \u0026ldquo;Pavlov\u0026rsquo;s Coli,\u0026rdquo; a system that detects an inflammatory marker (NO) \u003cem\u003eand\u003c/em\u003e a secondary metabolite (to gather context on microbiome state), and then locks a memory element that can be read out from recovered bacteria[\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e][\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e]. Another project (UZurich, 2022) engineered \u003cem\u003eE. coli Nissle\u003c/em\u003e to sense NO and, in response, secrete an anti-TNF nanobody to simultaneously diagnose and treat inflammation[\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e]. These works underscore the modularity of the approach: one can mix-and-match sensor inputs and functional outputs for tailored applications. Indeed, a 2023 study by Zou \u003cem\u003eet al.\u003c/em\u003e created an \u003cb\u003eintegrated diagnostic and therapeutic probiotic\u003c/b\u003e named \u0026ldquo;\u003cb\u003ei-ROBOT\u003c/b\u003e\u0026rdquo; that records inflammatory signals and releases immunomodulatory therapy in a mouse model[\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e][\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e]. In their design, a thiosulfate-responsive sensor triggered a CRISPR-based \u003cb\u003ebase editor\u003c/b\u003e in the bacteria, permanently changing a DNA sequence (which served as a memory of inflammation) and concurrently inducing production of a protective protein (AvCystatin) to help quell the inflammation[\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e][\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e]. The success of i-ROBOT in mice \u0026ndash; effectively reporting on disease activity via fecal sample analysis and improving disease outcomes \u0026ndash; provides a strong foundation for our proposed Lactobacillus sensor strategy.\u003c/p\u003e \u003cp\u003e \u003cstrong\u003eSynthetic Biology Tools for \u003cem\u003eLactobacillus\u003c/em\u003e Chassis\u003c/strong\u003e \u003cp\u003eHistorically, engineering Gram-positive probiotic bacteria like \u003cem\u003eLactobacillus\u003c/em\u003e was less straightforward than engineering lab strains of \u003cem\u003eE. coli\u003c/em\u003e, due to a relative paucity of genetic tools. This has changed in recent years, with the development of well-characterized promoters, inducible systems, and plasmid vectors that function in lactic acid bacteria[\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e][\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e]. Broad-host-range plasmids (such as derivatives of pWV or pSH vectors) can replicate in \u003cem\u003eLactobacillus\u003c/em\u003e, or integration vectors can insert payloads into the chromosome for stable maintenance. CRISPR-based gene editing has also been demonstrated in lactic acid bacteria, enabling precise modifications and integration of synthetic circuits. Moreover, to ensure \u003cb\u003econtainment and safety\u003c/b\u003e, kill-switches and auxotrophic dependencies have been implemented in probiotics (e.g., strains requiring a nutrient not present outside the host, to prevent environmental survival)[\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e][\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e]. All these advances mean that \u003cem\u003eLactobacillus\u003c/em\u003e is now a viable and practical chassis for sophisticated biosensing functions. We can take advantage of native genetic elements \u0026ndash; for example, acid-tolerance promoters or two-component systems responding to gut-specific signals \u0026ndash; and combine them with heterologous elements (like an E. coli NO sensor) in the same cell. The result is a \u003cb\u003esynthetic gene network\u003c/b\u003e operating within \u003cem\u003eLactobacillus\u003c/em\u003e, connecting inputs (inflammation cues) to outputs (reporter signals). In the next section, we describe the design of such a network in detail, outlining how each sensor module works and how the outputs are generated in an easy-to-detect manner.\u003c/p\u003e "},{"header":"Biosensor Design","content":"\u003cp\u003eOur probiotic inflammation monitor is conceived as a \u003cstrong\u003emodular genetic circuit\u003c/strong\u003e composed of distinct sensing modules feeding into a common reporting module. Each module detects a specific inflammatory biomarker and, upon activation, triggers expression of a reporter gene. The modules are designed to function independently or in combination, allowing the system to be configured for multiplex sensing (e.g., logic gates where multiple conditions must be met before signaling). Here we describe the design of each sensor module and the choice of reporter outputs, all tailored for implementation in a \u003cem\u003eLactobacillus\u003c/em\u003e chassis. Figure 1 (A and B) (schematic overview) illustrates how these modules are organized in the engineered cell[6][7].\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e1. Reactive Oxygen Species (ROS) Sensor:\u003c/strong\u003e For detecting ROS such as hydrogen peroxide (H₂O₂), we utilize a promoter regulated by the bacterial oxidative stress response. One well-characterized example is the \u003cstrong\u003eOxyR regulon\u003c/strong\u003e in \u003cem\u003eE. coli\u003c/em\u003e: OxyR is a transcription factor that, when oxidized by H₂O₂, activates promoters like \u003cem\u003eP_katG\u003c/em\u003e (catalase gene promoter) to induce expression of antioxidant genes. \u003cem\u003eLactobacillus\u003c/em\u003e species have analogous systems; for instance, \u003cem\u003eLactobacillus plantarum\u003c/em\u003e encodes an OxyR-like regulator and peroxide-inducible genes. We have chosen the promoter of the \u003cem\u003eahpC\u003c/em\u003e gene (alkyl hydroperoxide reductase) from \u003cem\u003eL. plantarum\u003c/em\u003e as our ROS-responsive element, since it is strongly upregulated in the presence of peroxides. This promoter, \u003cem\u003eP_ahpC\u003c/em\u003e, is placed upstream of a reporter gene on our plasmid. In the absence of oxidative stress, a repressor protein (or an inactive activator) keeps \u003cem\u003eP_ahpC\u003c/em\u003e largely off. When ROS levels rise in the environment (e.g., during inflammation, neutrophils release H₂O₂ into the gut lumen), the OxyR regulator in our \u003cem\u003eLactobacillus\u003c/em\u003e sensor strain becomes activated (via formation of a disulfide bond). Activated OxyR then binds \u003cem\u003eP_ahpC\u003c/em\u003e and \u003cstrong\u003eturns on transcription\u003c/strong\u003e of the reporter gene (Figure 1A). By tuning the promoter and ribosome binding site (RBS) strength, this module can produce a significant output only above a certain ROS threshold, minimizing false positives from minor fluctuations in redox state. We expect this sensor to respond to H₂O₂ concentrations in the low micromolar range, which corresponds to inflamed tissue conditions, while remaining quiescent under normal physiological conditions where ROS are quickly neutralized.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e2. Nitric Oxide (NO) Sensor:\u003c/strong\u003e Nitric oxide is a key inflammatory mediator that diffuses across bacterial membranes and can be sensed by regulatory proteins. We incorporate an \u003cstrong\u003eNO-sensitive genetic circuit\u003c/strong\u003e that has been adapted from \u003cem\u003eE. coli Nissle\u003c/em\u003e. Specifically, we use the regulatory region of the \u003cem\u003enorVW\u003c/em\u003e operon (which in \u003cem\u003eE. coli\u003c/em\u003e encodes NO-detoxifying enzymes under control of the transcriptional activator NorR). In \u003cem\u003eE. coli\u003c/em\u003e, \u003cstrong\u003eNorR\u003c/strong\u003e binds to the \u003cem\u003eP_norV\u003c/em\u003e promoter and, upon detecting NO (via an associated nitric oxide\u0026ndash;sensing domain), NorR activates \u003cem\u003eP_norV\u003c/em\u003e. We clone the \u003cem\u003eE. coli\u003c/em\u003e NorR protein and its target promoter \u003cem\u003eP_norV\u003c/em\u003e into our \u003cem\u003eLactobacillus\u003c/em\u003e system. To ensure functionality in \u003cem\u003eLactobacillus\u003c/em\u003e, we place NorR under a constitutive promoter that works in Gram-positive hosts, and we include any necessary accessory factors (such as integration host factor, IHF, if needed for NorR binding \u0026ndash; though for \u003cem\u003eP_norV\u003c/em\u003e this might not be required). In our design, when NO is absent, NorR is present but inactive, and \u003cem\u003eP_norV\u003c/em\u003e drives only minimal transcription of the downstream reporter. When NO produced by inflamed tissues enters the bacteria, it binds to NorR, causing a conformational change that allows NorR to recruit RNA polymerase to \u003cem\u003eP_norV\u003c/em\u003e. The result is robust \u003cstrong\u003eactivation of the NO-sensor promoter\u003c/strong\u003e and high-level expression of its reporter gene (Figure 1B). One advantage of NorR is that it is highly specific for NO and related RNS (reactive nitrogen species), and does not respond to oxygen or other gases. Prior work in \u003cem\u003eE. coli\u003c/em\u003e has shown NorR-based biosensors can detect nanomolar to micromolar NO and distinguish an inflamed gut environment[6]. We anticipate a similar sensitivity in \u003cem\u003eLactobacillus\u003c/em\u003e after optimization. Additionally, we include a genetic \u0026ldquo;memory\u0026rdquo; option in this module: downstream of \u003cem\u003eP_norV\u003c/em\u003e we place a gene encoding a site-specific recombinase (e.g., Cre or PhiC31 integrase) under control of the same promoter. When NO is present, the recombinase will permanently flip a DNA segment in the plasmid (for example, excising a termination sequence or inverting a reporter gene orientation), thus creating a lasting record that the cell experienced NO. This memory element ensures that even transient inflammation exposure can be detected later by examining the state of the circuit[6]. For real-time reporting, however, the primary output of this module remains the immediate reporter signal (e.g., fluorescence).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e3. Cytokine Sensor Module:\u003c/strong\u003e Sensing eukaryotic cytokines directly is the most challenging aspect, and our design here is considered exploratory. We propose two alternative strategies: (A) a \u003cstrong\u003esynthetic receptor approach\u003c/strong\u003e, and (B) an \u003cstrong\u003eaptamer switch approach\u003c/strong\u003e. In strategy A, we engineer a chimeric two-component system that can respond to TNF-\u0026alpha;. This could involve fusing the extracellular domain of a TNF-\u0026alpha; binding protein (for instance, a single-domain antibody \u0026ndash; nanobody \u0026ndash; that recognizes TNF-\u0026alpha;) to the transmembrane sensor kinase domain of a bacterial two-component system. A candidate is the well-studied EnvZ-OmpR system: EnvZ is a membrane kinase that normally senses osmolarity. We replace EnvZ\u0026rsquo;s periplasmic sensor domain with a nanobody against TNF-\u0026alpha;. In \u003cem\u003eLactobacillus\u003c/em\u003e, upon TNF-\u0026alpha; binding to the nanobody, the chimeric EnvZ could undergo autophosphorylation and subsequently transfer the phosphate to the OmpR response regulator, which then activates an OmpR-responsive promoter (\u003cem\u003eP_ompC\u003c/em\u003e or similar) driving reporter expression. The feasibility of this depends on maintaining proper folding and signaling of the chimeric protein, but similar receptor engineering has been attempted in other contexts (e.g., bacteria sensing mammalian hormones via chimeric TCS). Strategy B avoids membrane proteins and instead uses an RNA-based sensor: we design an \u003cstrong\u003eRNA aptamer\u003c/strong\u003e that binds IL-6 (another important cytokine) and integrate it into a riboswitch that controls translation of a reporter. For example, an aptamer sequence that recognizes IL-6 could be placed in the 5\u0026rsquo; UTR of the mRNA; when IL-6 is present in the bacterial cytoplasm (which might require some uptake or permeability enhancement), it would bind the aptamer and cause a structural change unmasking the ribosome binding site, thus turning on translation of the reporter. Admittedly, cytokines are large (~17\u0026ndash;51 kDa) and mostly extracellular in the host, so getting them into bacteria or having bacteria detect them on the surface is non-trivial. In practice, our initial design will focus on the indirect route: NO and other signals often correlate with cytokine levels (NO synthesis is triggered by cytokine signaling). Therefore, the NO sensor and ROS sensor provide an indirect readout of cytokine activity. Nonetheless, we include the TNF-\u0026alpha; nanobody secretion system (as in the UZurich 2022 design) not as a sensor but as a \u003cstrong\u003eresponse/effector\u003c/strong\u003e: our \u003cem\u003eLactobacillus\u003c/em\u003e can be made to secrete an anti-TNF nanobody when it detects NO, thereby not only reporting inflammation but potentially also dampening it[14]. This therapeutic bent is secondary to our diagnostic focus, but it underscores the modular potential of the platform.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e4. pH Sensor:\u003c/strong\u003e To detect pH changes, we take advantage of \u003cem\u003eLactobacillus\u003c/em\u003e\u0026rsquo; native acid-response systems. Many lactic acid bacteria have pH-responsive promoters that are induced under acidic conditions. One example is the promoter of the \u003cem\u003egad\u003c/em\u003e (glutamate decarboxylase) system in \u003cem\u003eLactobacillus\u003c/em\u003e, which is upregulated in low pH to help consume acidic protons. Another is the \u003cem\u003elysine decarboxylase\u003c/em\u003e gene promoter, which similarly is triggered in acidity to produce amines that raise pH. We incorporate a promoter from \u003cem\u003eL. plantarum\u003c/em\u003e that is known to be \u003cstrong\u003eacid-inducible\u003c/strong\u003e, denoted here as \u003cem\u003eP_hde\u003c/em\u003e. Under neutral pH (~7), a transcriptional repressor binds \u003cem\u003eP_hde\u003c/em\u003e and keeps it off. When the environmental pH drops below ~5.5 (as might occur in inflamed colonic regions[9]), protonation of the repressor or other pH-sensing mechanism causes it to dissociate, thereby activating the promoter. This yields increased transcription of the downstream reporter gene. Conversely, if the local pH goes abnormally high (above ~8, which is rare in the colon but could happen in certain dysbiosis conditions), one could similarly sense that with a different promoter; however, IBD usually skews acidic or slightly lower than normal pH, so our focus is on detecting pH drops. We calibrate the pH sensitivity of \u003cem\u003eP_hde\u003c/em\u003e by mutating its regulator binding sites if needed, so that its dynamic range covers the pH range of interest (approximately pH 5.0 to 7.0). This module essentially functions as an environmental trigger that tells us if the gut lumen\u0026rsquo;s chemistry is disrupted by inflammation.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eReporter Outputs:\u003c/strong\u003e Each sensor module ultimately controls a \u003cstrong\u003ereporter gene\u003c/strong\u003e, and these can be customized based on the desired mode of detection. We have incorporated multiple reporters into our design, ensuring that the system can be read out by different methods:\u003c/p\u003e\n\u003cul type=\"disc\"\u003e\n \u003cli\u003e\u003cstrong\u003eFluorescent proteins\u003c/strong\u003e (e.g., GFP, mCherry): Fluorescence is a convenient output for laboratory evaluation and could be used in clinical samples via a fluorimeter or fluorescence imaging of stool. In our construct, we use a \u003cstrong\u003ecodon-optimized GFP\u003c/strong\u003e as one primary reporter due to its easy detection and quantification. For in-human use, one would not directly visualize fluorescence without special equipment, but it serves as a robust proof-of-concept output.\u003c/li\u003e\n \u003cli\u003e\u003cstrong\u003eColorimetric enzymes:\u003c/strong\u003e To enable simple readouts visible to the naked eye, we include reporters that produce a color change. One such reporter is \u003cstrong\u003e\u0026beta;-glucuronidase (GUS)\u003c/strong\u003e, which can convert colorless substrates (like X-Gluc) into a distinctly colored product. Another is \u003cstrong\u003e\u0026beta;-galactosidase (LacZ)\u003c/strong\u003e, which turns X-Gal substrate blue. We prefer GUS or \u003cstrong\u003eazoreductase\u003c/strong\u003e with appropriate substrates for gut use, as these enzymes can act on compounds already present in the diet or supplements to yield colored outputs in feces. For instance, if a patient ingests a pill containing X-Gluc, the engineered \u003cem\u003eLactobacillus\u003c/em\u003e will cause blue coloration in stool only if the inflammatory signal was detected (i.e., if the reporter enzyme was expressed). This could be a straightforward home test for patients: a visible color change indicates an IBD flare. In our design, the ROS sensor might control GUS expression, while the NO sensor controls GFP, etc., allowing multi-channel outputs.\u003c/li\u003e\n \u003cli\u003e\u003cstrong\u003eLuminescent and electrochemical outputs:\u003c/strong\u003e We also consider outputs that could interface with electronic devices. \u003cstrong\u003eLuciferase\u003c/strong\u003e genes (such as NanoLuc or bacterial luciferase luxA/B) produce bioluminescence that can be measured with a simple photodiode. We include a NanoLuc luciferase reporter under one of the sensor promoters as an optional output for higher sensitivity detection \u0026ndash; luminescence can be extremely sensitive and has virtually zero background in the absence of substrate[6]. This is ideal if coupling the readout to an ingestible electronics platform; indeed, prior work demonstrated an ingestible capsule that could detect luminescence from engineered bacteria in the gut and wirelessly transmit the signal [7][8]. Additionally, we propose an \u003cstrong\u003eelectrochemical output\u003c/strong\u003e: for example, a reporter that causes the bacteria to produce a small molecule that can be oxidized at an electrode, generating a current. One candidate is \u003cstrong\u003epyocyanin\u003c/strong\u003e, a redox-active pigment (though originally from \u003cem\u003ePseudomonas\u003c/em\u003e, a biosynthetic pathway could be introduced). Another simpler approach is to have the bacteria secrete \u003cstrong\u003ehydrogen peroxide as a signal\u003c/strong\u003e (via a genetically controlled oxidase) \u0026ndash; an electrode in a smart capsule could detect H₂O₂ changes amperometrically. While we do not fully implement an electrochemical reporter in our initial design, we outline a framework where the fluorescence or enzyme output of the bacteria could be measured by a companion electronic device for a digital readout.\u003c/li\u003e\n\u003c/ul\u003e\n\u003cp\u003eAll the reporters are encoded on our plasmid under the control of their respective sensor promoters. To enable \u003cstrong\u003emultiple outputs simultaneously\u003c/strong\u003e, distinct sensor-output cassettes are used. For instance: \u003cem\u003eP_norV\u003c/em\u003e drives GFP, \u003cem\u003eP_ahpC\u003c/em\u003e drives LacZ, and \u003cem\u003eP_hde\u003c/em\u003e drives luciferase. This way, the presence of NO would yield a fluorescence signal, ROS yields a colored product, and low pH yields luminescence. We can thus get a more comprehensive picture by analyzing multiple signals. The circuit design also incorporates \u003cstrong\u003esignal amplification\u003c/strong\u003e tactics. We use strong ribosome binding sites and multi-copy plasmids to ensure a robust reporter expression once a promoter is induced, thereby maximizing the signal-to-noise ratio. For example, a high-copy plasmid (around 50 copies per cell) will produce more reporter protein per cell than a single genomic copy, which is beneficial for detection limit. However, high copy number can burden the cell, so we also test medium-copy versions for stability in the gut environment.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eGenetic Circuit Diagram:\u003c/strong\u003e In summary, our genetic circuit (see Figure 1) consists of four main promoters (one for each sensor), each controlling a reporter gene. Upstream of each promoter, any necessary regulatory gene (transcription factor or two-component kinase/response regulator) is included. All these genes are arranged in a modular plasmid with synthetic insulation sequences between modules to prevent cross-talk (e.g., strong terminators to stop read-through transcription). We also include a \u003cstrong\u003esafety module\u003c/strong\u003e on the plasmid: a toxin-antitoxin pair that ensures the plasmid is maintained (cells that lose the plasmid will die due to toxin dominance, a strategy to prevent loss of the sensor circuit in the gut). Moreover, a \u003cstrong\u003ekill switch\u003c/strong\u003e device is encoded that can be triggered (for example, by administering a molecule like rhamnose that activates a deadly gene circuit) to eliminate the bacteria if needed after they have served their diagnostic purpose. These design features collectively yield a \u003cem\u003eLactobacillus\u003c/em\u003e biosensor strain that can enter the gut, detect inflammation markers, and output detectable signals, all while maintaining genetic stability and safety. The next section will discuss how we construct and test this design experimentally.\u003c/p\u003e\n\u003cp\u003e\u003cem\u003e(Figure 1: Schematic of the probiotic inflammation biosensor circuit. Multiple sensor modules \u0026ndash; ROS, NO, cytokine, and pH \u0026ndash; feed into distinct reporters (GFP fluorescence, blue color, luminescence). The chassis is\u003c/em\u003e Lactobacillus\u003cem\u003e, engineered with a synthetic plasmid carrying all modules. A memory element (not shown here) can record NO detection via recombinase activity. Safety features like kill switches are also included.)\u003c/em\u003e\u003c/p\u003e"},{"header":"Materials and Methods","content":"\u003cp\u003e\u003cstrong\u003eStrain and Culture Conditions:\u003c/strong\u003e We selected a \u003cem\u003eLactobacillus\u003c/em\u003e strain with good transformability and robust gut colonization properties, \u003cem\u003eLactobacillus rhamnosus GG\u003c/em\u003e, as the base chassis. This strain is a well-known probiotic and is naturally tolerant of stomach acid and bile, ensuring it can survive to reach the intestine. For cloning and plasmid propagation, we initially use \u003cem\u003eEscherichia coli\u003c/em\u003e (DH5\u0026alpha;) as a host, then introduce the constructs into \u003cem\u003eL. rhamnosus\u003c/em\u003e via electroporation. \u003cem\u003eLactobacillus\u003c/em\u003e cultures are grown in MRS broth at 37\u0026deg;C under microaerophilic conditions (5% CO₂), which simulates the gut environment. For experiments requiring inducible control (e.g., to induce kill switches in lab tests), we supplement media with the appropriate inducer (such as 0.2% rhamnose for a rhamnose-inducible promoter if used). Antibiotics (like erythromycin at 5 \u0026micro;g/mL) are added to maintain plasmid selection in \u003cem\u003eLactobacillus\u003c/em\u003e, using resistance markers that are effective in Gram-positive bacteria (our plasmid carries an erm gene for erythromycin resistance, a common choice in LAB).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eGenetic Circuit Assembly:\u003c/strong\u003e We constructed the sensor plasmid using standard molecular biology techniques and synthetic biology tools. The plasmid backbone is a broad-host-range shuttle vector (able to replicate in both \u003cem\u003eE. coli\u003c/em\u003e and \u003cem\u003eLactobacillus\u003c/em\u003e) derived from the pWV family, with a size of ~8 kb including selection markers. Each sensor module was cloned as a separate DNA fragment, then combined into the backbone using Golden Gate Assembly for modularity. Briefly:\u003c/p\u003e\n\u003cul type=\"disc\"\u003e\n \u003cli\u003eThe \u003cstrong\u003eROS sensor fragment\u003c/strong\u003e contains the \u003cem\u003eL. plantarum\u003c/em\u003e \u003cem\u003eP_ahpC\u003c/em\u003e promoter upstream of a reporter gene (gusA, encoding \u0026beta;-glucuronidase) and a strong transcriptional terminator. Upstream of \u003cem\u003eP_ahpC\u003c/em\u003e, the fragment also includes the gene \u003cem\u003eoxyR\u003c/em\u003e (from \u003cem\u003eL. plantarum\u003c/em\u003e, under a constitutive promoter) to ensure the presence of the transcriptional activator. This fragment (~2 kb) was PCR-amplified from \u003cem\u003eL. plantarum\u003c/em\u003e genomic DNA for the promoter and oxyR, and from a plasmid template for gusA.\u003c/li\u003e\n \u003cli\u003eThe \u003cstrong\u003eNO sensor fragment\u003c/strong\u003e includes the \u003cem\u003eE. coli\u003c/em\u003e \u003cem\u003enorV\u003c/em\u003e promoter region (with upstream NorR-binding sites) controlling gfpmut3 (a bright GFP variant), followed by a terminator. The \u003cem\u003enorR\u003c/em\u003e gene (E. coli origin) is placed under a constitutive Gram-positive promoter (PJH2, a synthetic strong promoter for LAB) on the same fragment. We synthesized this ~2.5 kb fragment de novo with codon optimization for \u003cem\u003eLactobacillus\u003c/em\u003e where needed (especially for NorR, to ensure good expression in a Gram-positive context). We also included loxP sites around a transcriptional terminator downstream of GFP to facilitate recombinase-based memory testing (if Cre recombinase is expressed, it would excise the terminator and allow read-through to an otherwise silent RFP gene \u0026ndash; for our memory assay).\u003c/li\u003e\n \u003cli\u003eThe \u003cstrong\u003eCytokine sensor/response fragment\u003c/strong\u003e is more complex. We included a gene for a \u003cstrong\u003eTNF-\u0026alpha;-binding nanobody\u003c/strong\u003e (sequence obtained from literature on anti-TNF V_HH domains) fused to the transmembrane and periplasmic portion of EnvZ, with the cytoplasmic kinase domain intact. The corresponding response regulator OmpR is encoded with a strong RBS on the same fragment. Downstream, the OmpR-activated promoter \u003cem\u003eP_ompC\u003c/em\u003e is placed controlling the gene \u003cem\u003elacZ\u003c/em\u003e (\u0026beta;-galactosidase). This way, TNF-\u0026alpha; binding should activate OmpR and induce lacZ, leading to blue color development in the presence of X-Gal. This fragment (~3 kb) was assembled from synthesized parts and cloning (given its novel chimeric nature, we will screen for functionality, as described below). Additionally, on this fragment, we have a gene \u003cem\u003ensrR\u003c/em\u003e (a nitric oxide-sensitive repressor from \u003cem\u003eLactococcus\u003c/em\u003e lactis) driving expression of a \u003cstrong\u003eMazF toxin\u003c/strong\u003e under control of the NO sensor (P_norV) as a kill mechanism \u0026ndash; this was inspired by kill-switch designs where in the absence of inflammation (or upon resolution), the sensor bacteria self-destruct to avoid long-term persistence[2][3]. However, for initial diagnostics-only work, this kill switch will be kept inactive by including its antitoxin on another module or by not inducing it during tests.\u003c/li\u003e\n \u003cli\u003eThe \u003cstrong\u003epH sensor fragment\u003c/strong\u003e consists of the acid-inducible promoter \u003cem\u003eP_hde\u003c/em\u003e from \u003cem\u003eL. plantarum\u003c/em\u003e controlling the NanoLuc luciferase gene (nanoluc). Upstream, we included the gene for a pH-sensitive repressor (PadR family regulator) that naturally controls \u003cem\u003eP_hde\u003c/em\u003e. This ~1.5 kb fragment was PCR-amplified from \u003cem\u003eL. plantarum\u003c/em\u003e (promoter and regulator) and combined with a synthesized nanoluc gene codon-optimized for LAB.\u003c/li\u003e\n\u003c/ul\u003e\n\u003cp\u003eThese fragments all have non-overlapping terminators and insulator sequences (from bacteriophage T₀/T₁) at their ends to prevent read-through. Using a one-pot Golden Gate reaction (with BsaI sites flanking each module), we assembled the full plasmid. Clones were screened by restriction digest and sequenced fully to verify correct assembly. The final plasmid, pInflamSensor, is approximately 15 kb. It carries an erythromycin resistance gene for selection and a conditional origin of replication that yields ~20 copies per cell in \u003cem\u003eLactobacillus\u003c/em\u003e.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eTransformation into \u003cem\u003eLactobacillus\u003c/em\u003e:\u003c/strong\u003e We prepared electrocompetent \u003cem\u003eL. rhamnosus\u003c/em\u003e cells by growing them to mid-log (OD₆₀₀ ~0.5) in MRS + glycine (2% glycine weakens the cell wall), then washing with cold sucrose-glycerol solution. Approximately 1 \u0026micro;g of pInflamSensor DNA isolated from \u003cem\u003eE. coli\u003c/em\u003e was electroporated into \u003cem\u003eL. rhamnosus\u003c/em\u003e at 2.5 kV in a 2 mm gap cuvette. Transformants were recovered in regeneration broth and plated on MRS agar with erythromycin. Successful transformants were confirmed by colony PCR for key module genes (e.g., gfp and gusA presence). We obtained multiple clones and chose one with correct genotype and healthy growth for subsequent experiments.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eIn Vitro Assays for Sensor Response:\u003c/strong\u003e We first characterized each sensor module in vitro using controlled conditions:\u003c/p\u003e\n\u003cul type=\"disc\"\u003e\n \u003cli\u003e\u003cem\u003eROS module test:\u003c/em\u003e Cultures of the engineered \u003cem\u003eLactobacillus\u003c/em\u003e were grown in MRS to OD₆₀₀ ~0.4, then split into aliquots. One set was treated with 1 mM H₂O₂ to simulate an oxidative burst, while a control set received no H₂O₂. After 1 hour incubation, cells were pelleted and assayed for \u0026beta;-glucuronidase activity using p-nitrophenyl-\u0026beta;-D-glucuronide (pNPG) as substrate. We measured the absorbance at 405 nm from p-nitrophenol production. We expected significantly higher absorbance (enzyme activity) in the H₂O₂-exposed sample if \u003cem\u003eP_ahpC\u003c/em\u003e was induced. Additionally, we visualized any color change by including X-Gluc in plates spread with the bacteria \u0026ndash; H₂O₂-treated bacteria should turn blue (indicating gusA expression).\u003c/li\u003e\n \u003cli\u003e\u003cem\u003eNO module test:\u003c/em\u003e We utilized an NO-donor chemical, \u003cstrong\u003esodium nitroprusside (SNP)\u003c/strong\u003e, which releases NO in solution. Cultures of the sensor strain were grown anaerobically (to prevent rapid NO oxidation) and divided; SNP was added to one at 0.5 mM final concentration, and the other was kept as control. After 2 hours, we measured GFP fluorescence in the cells (excitation ~488 nm, emission ~510 nm) using a plate reader. Parallelly, we checked for memory switch flipping: we extracted plasmid DNA and performed PCR across the loxP-flanked terminator region to see if recombination occurred in NO-treated cells (for this test, we had induced a Cre recombinase from a separate plasmid by arabinose addition, to simulate memory capture of an NO event). We anticipated a strong GFP fluorescence increase upon NO exposure, and PCR band size reduction if memory was recorded.\u003c/li\u003e\n \u003cli\u003e\u003cem\u003eCytokine module test:\u003c/em\u003e Testing the TNF-\u0026alpha; sensor was challenging in vitro, as it requires the presence of TNF-\u0026alpha;. We expressed and purified human TNF-\u0026alpha; protein from a HEK293 culture (using a His-tag) to use in assays. Engineered bacteria were incubated with 50 ng/mL of TNF-\u0026alpha; for 3 hours to allow binding and response. We then added X-Gal to the culture and looked for blue coloration of cells or the medium, indicative of LacZ activity. We also measured \u0026beta;-galactosidase activity quantitatively via the ONPG (o-nitrophenyl-\u0026beta;-D-galactopyranoside) assay, reading absorbance at 420 nm. The expectation was modest \u0026ndash; since this synthetic TCS is experimental, a detectable increase in LacZ activity over background would be a success. If unsuccessful, we planned to iterate by testing different linker lengths in the chimeric receptor or by using IL-6 with an aptamer-based reporter (for which we had an alternative construct ready).\u003c/li\u003e\n \u003cli\u003e\u003cem\u003epH module test:\u003c/em\u003e We grew the bacteria in buffered medium adjusted to various pH levels (5.0, 6.0, 7.0) using MES or HEPES buffers. After 2 hours at each pH, NanoLuc activity was measured by adding its substrate (furimazine) and detecting luminescence with a luminometer. We plotted relative light units (RLU) vs. pH to confirm that the \u003cem\u003eP_hde\u003c/em\u003e promoter is low at neutral pH and sharply induced at acidic pH. A strong increase in luminescence at pH \u0026le;5.5 compared to pH 7 would validate this module.\u003c/li\u003e\n\u003c/ul\u003e\n\u003cp\u003eTo ensure module independence, each test was conducted in isolation or with specific inhibitors. For example, in the H₂O₂ test we included an NO scavenger to be sure the GFP (NO reporter) was not activated, and vice versa. Our circuit design\u0026rsquo;s insulated architecture was aimed at preventing unwanted cross-activation (such as OxyR affecting other promoters). These bench-top assays establish the baseline performance of each sensor.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eSimulated Gut Environment Tests:\u003c/strong\u003e We next moved to more complex models. We used anaerobic fecal fermenter cultures (to mimic colon conditions) where we could add our engineered \u003cem\u003eLactobacillus\u003c/em\u003e along with mouse cecal contents to simulate microbiome competition. In parallel, we utilized an \u003cem\u003ein vitro\u003c/em\u003e gut simulator (a bioreactor with controlled pH, anaerobiosis, and peristaltic flow) to test sensor function under gut-like conditions. In these systems, we introduced inflammatory stimuli: for example, we added \u003cem\u003emacrophage cell line supernatant\u003c/em\u003e containing cytokines and NO to the simulator, or we added sodium thiosulfate plus an oxidative burst (via a small amount of sodium hypochlorite) to generate tetrathionate in the medium. The probiotic sensor\u0026rsquo;s response was monitored. We took samples of the culture and measured the reporters as above. These experiments helped assess if the sensors can work amidst a complex milieu of other bacteria, digestive components, and without rich nutrients. We also measured growth curves of the engineered \u003cem\u003eLactobacillus\u003c/em\u003e in these conditions to see if the circuit imposes any fitness cost.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eIn Vivo Pilot Study Design:\u003c/strong\u003e As a final validation step (expected in future work), we outline a small animal trial. Using a well-established mouse model of colitis (dextran sulfate sodium, DSS, induced colitis), we would orally administer the engineered \u003cem\u003eLactobacillus\u003c/em\u003e to mice and then monitor two things: (1) the presence of reporter signals in feces, and (2) the correlation of those signals with inflammation status. For instance, during an active colitis episode (DSS treatment days), we anticipate that fecal pellets from mice harboring the sensor bacteria would show a color change (blue, from LacZ or GUS) or measurable luminescence above baseline. We would collect fecal samples daily, suspend them in buffer, and measure fluorescence, color intensity, and luminescence. Mice would also undergo endoscopic scoring of colitis and have tissue cytokine levels measured, to compare with the biosensor readouts. If successful, the biosensor output should elevate in step with disease markers (e.g., a spike in GFP signal corresponding to high NO levels in tissue). The mice would be monitored to ensure that the \u003cem\u003eLactobacillus\u003c/em\u003e colonizes at least transiently (we\u0026rsquo;d verify by plating feces on selective media to count sensor bacteria). We also plan to test a control strain (non-engineered \u003cem\u003eLactobacillus\u003c/em\u003e) to ensure any signals are due to our synthetic circuit, not just gut metabolism changes. For safety, all mice receiving engineered bacteria would be kept in biocontainment cages, and at study end treated with an appropriate antibiotic to eliminate the probiotic (though \u003cem\u003eLactobacillus\u003c/em\u003e is often vancomycin-resistant, we might use a combination or temperature-sensitive kill switch induction). These animal experiments will provide crucial data on real-world sensor performance in a live host.\u003c/p\u003e\n\u003cp\u003eThroughout all these methods, standard protocols for data collection and analysis are followed: fluorescence readings are normalized to cell density (OD₆₀₀) to account for growth differences, enzymatic assays are done in triplicate for statistical rigor, and appropriate controls (wild-type \u003cem\u003eLactobacillus\u003c/em\u003e without plasmid, or with a dummy plasmid lacking sensors) are included to measure background signals.\u003c/p\u003e"},{"header":"Expected Results","content":"\u003cp\u003eWe anticipate the engineered \u003cem\u003eLactobacillus\u003c/em\u003e biosensor will successfully detect and report intestinal inflammation markers with a \u003cb\u003ehigh degree of specificity and a quantifiable dynamic range\u003c/b\u003e. Based on our design and previous findings in literature, the expected outcomes can be summarized as follows:\u003c/p\u003e \u003cp\u003e \u003cul\u003e \u003cli\u003e \u003cp\u003e \u003cb\u003eSensitive Detection of NO and ROS\u003c/b\u003e: The NO-sensing module (NorR\u0026thinsp;+\u0026thinsp;P_norV\u0026thinsp;+\u0026thinsp;GFP) is expected to respond to NO concentrations in the low micromolar range, yielding at least a \u003cb\u003e5- to 10-fold increase\u003c/b\u003e in GFP fluorescence above the baseline (no-NO) condition. This estimate is in line with Archer et al. (2012), who observed a clear ON/OFF distinction in their NO-responsive \u003cem\u003eE. coli\u003c/em\u003e strain[\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e]. Similarly, the ROS sensor (OxyR\u0026thinsp;+\u0026thinsp;P_ahpC\u0026thinsp;+\u0026thinsp;GUS) should produce a strong enzymatic signal when exposed to oxidative stress. In quantitative terms, we expect the β-glucuronidase activity to be negligible in anaerobic or low-peroxide conditions (\u0026lt;\u0026thinsp;1 \u0026micro;M H₂O₂), and to rise sharply (by an order of magnitude) when H₂O₂ is added at inflammatory levels (~\u0026thinsp;50\u0026ndash;100 \u0026micro;M transiently, as might occur near an inflamed site). The colorimetric output (blue color on X-Gluc plates or in liquid) should be \u003cb\u003evisible to the naked eye\u003c/b\u003e for H₂O₂-exposed cultures within 1\u0026ndash;2 hours of induction, indicating a robust response suitable for practical detection.\u003c/p\u003e \u003c/li\u003e \u003cli\u003e \u003cp\u003e \u003cb\u003eDynamic Range and Fold-Change\u003c/b\u003e: Each sensor promoter has been chosen or engineered to have low basal leakage and high induced expression. We expect a clear OFF vs. ON state. For instance, at pH 7, the pH sensor (P_hde\u0026thinsp;+\u0026thinsp;NanoLuc) should produce minimal luminescence (comparable to a negative control with no reporter), whereas at pH 5.0, the luminescence could be \u003cb\u003e100-fold higher\u003c/b\u003e given the strong induction of acid-responsive systems in LAB. This high fold-change is crucial for avoiding ambiguity: even a small inflammation-associated pH drop (for example, from 6.8 to 6.2) might produce a partial response, but within a pathologically relevant range (below 6.0) the output should be clearly distinguishable. We will generate calibration curves, such as luminescence vs. pH, and fluorescence vs. NO concentration, to quantify these ranges. Ideally, the biosensor will be tuned such that its threshold of activation lies at the upper end of normal physiological levels \u0026ndash; ensuring that only an abnormal increase triggers the output. For example, normal gut NO might be in the nanomolar range, not enough to activate NorR, whereas inflammatory NO reaches micromolar, crossing the threshold for activation.\u003c/p\u003e \u003c/li\u003e \u003cli\u003e \u003cp\u003e \u003cb\u003eSpecificity and Low Crosstalk\u003c/b\u003e: A critical expectation is that each sensor responds primarily to its target stimulus with minimal cross-reactivity. Our circuit insulation and use of distinct regulatory proteins aim to ensure this. We anticipate that in a mixed-signal environment (where multiple inflammation markers coexist), each reporter will correctly reflect its linked input. For instance, in a test where we expose the bacteria to low pH but no NO, we expect to see luminescence (pH reporter) increase while GFP (NO reporter) remains at baseline. Similarly, with high NO but neutral pH, GFP should increase without luminescence. We have built in orthogonal regulation to achieve this segregation. Unintended interactions (like oxidative stress also inducing some NO-responsive genes) could be observed, but we will quantify those. Ideally, adding an NO scavenger during an H₂O₂ exposure test will show that any GFP rise was due to secondary NO generation (if any) \u0026ndash; and if our design is robust, we shouldn\u0026rsquo;t see such a rise in the first place.\u003c/p\u003e \u003c/li\u003e \u003cli\u003e \u003cp\u003e \u003cb\u003eMemory Capture (if implemented)\u003c/b\u003e: Should we activate the recombinase memory element (via Cre/lox or CRISPR spacer recording) in the NO module, we expect that a brief exposure to NO will lead to a permanent genetic change in a fraction of the bacteria. For example, if 10% of the cells experienced enough NO to trigger recombination, those 10% will carry the flipped DNA switch thereafter. When we plate bacteria after the experiment and screen colonies by PCR, we anticipate seeing that proportion showing the flipped genotype. The fraction might correlate with the duration or intensity of inflammation. While this isn\u0026rsquo;t a primary output for immediate detection, it validates the concept of an internal event log.\u003c/p\u003e \u003c/li\u003e \u003cli\u003e \u003cp\u003e \u003cb\u003ePerformance in Complex Samples\u003c/b\u003e: In stool or gut simulators, the signals will be more diluted and could be quenched by background. Yet, we expect that using sensitive detection methods (fluorimeter, luminometer), the outputs will still be discernible above autofluorescence or autochromogenic background of fecal matter. For example, fecal slurries might have some native blue-ish tint if the diet contains certain components, but a true positive from LacZ/X-Gal should produce a distinctly blue precipitate. We anticipate needing to optimize substrate delivery for colorimetric outputs in vivo (perhaps by feeding the animal a dye precursor that the bacteria act on). Based on literature, the probiotic sensor approach has succeeded in mice: Daeffler et al. detected their thiosulfate sensor activation in \u0026lt;\u0026thinsp;5% of \u003cem\u003eE. coli\u003c/em\u003e retrieved from mouse feces, but that was enough to statistically distinguish inflamed vs. healthy mice[\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e][\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e]. We expect a similar or better outcome, given \u003cem\u003eLactobacillus\u003c/em\u003e might localize in the gut mucus layer effectively and consistently report.\u003c/p\u003e \u003c/li\u003e \u003cli\u003e \u003cp\u003e \u003cb\u003eColonization and Stability\u003c/b\u003e: An expected ancillary result is that the engineered \u003cem\u003eLactobacillus\u003c/em\u003e will survive and moderately colonize the mouse gut during the course of an experiment (days to weeks). We predict that after oral gavage, the strain will be recoverable from feces within 24 hours and maintain a population (though likely not permanently colonizing long-term without selective pressure). The presence of the synthetic plasmid should remain high in recovered bacteria thanks to our toxin-antitoxin stability module. We plan to check a subset of recovered colonies for plasmid retention; we expect\u0026thinsp;\u0026gt;\u0026thinsp;90% retention over a week in vivo, otherwise adjustments (like chromosomal integration of key circuit components) might be needed.\u003c/p\u003e \u003c/li\u003e \u003cli\u003e \u003cp\u003e \u003cb\u003eNo Significant Fitness Cost in Lab Conditions\u003c/b\u003e: In growth curves, we expect the engineered strain to have a slightly longer doubling time than wild-type \u003cem\u003eLactobacillus\u003c/em\u003e (perhaps 10\u0026ndash;20% growth rate reduction due to metabolic burden of plasmid and expression). However, it should still reach similar high cell densities in rich media. In conditions where it detects a signal and expresses reporters at high levels, there might be a temporary growth slowdown (resources diverted to reporter protein production). This is acceptable for a diagnostic as long as the signal is produced. If any module severely hinders growth (e.g., leaky toxin expression from the kill switch), we will refine its regulation (such as requiring a two-step activation to avoid unintended expression). We predict that the strain will remain viable under inflammatory conditions at least long enough to deliver the diagnostic signal.\u003c/p\u003e \u003c/li\u003e \u003cli\u003e \u003cp\u003e \u003cb\u003eIntegration of Multiple Signals\u003c/b\u003e: If two or more inflammatory signals are present together (which is often the case in real IBD \u0026ndash; e.g., both ROS and NO are high), our system is built to handle that by simply turning on multiple reporters. In such cases, we anticipate \u003cb\u003ecombinatorial output\u003c/b\u003e: for instance, in a highly inflamed scenario, the biosensor might simultaneously show fluorescence (NO marker) and a color change (ROS marker). This would reinforce the diagnosis. If only one signal is present (say pH dropped but NO is normal, which could happen in certain dysbiosis), we might only see luminescence but no GFP. This differentiation could even help identify the nature of the imbalance (e.g., acidification might point to microbiome shifts rather than classical inflammation).\u003c/p\u003e \u003c/li\u003e \u003c/ul\u003e \u003c/p\u003e \u003cp\u003eIn summary, the expected results will demonstrate that our \u003cem\u003eLactobacillus\u003c/em\u003e biosensor responds to inflammation-associated cues with distinct and measurable outputs, both in controlled lab settings and in more complex environments. Achieving clear discrimination between \u0026ldquo;healthy\u0026rdquo; and \u0026ldquo;inflamed\u0026rdquo; conditions is the primary benchmark. Success will be measured by factors such as a statistically significant elevation of reporter signals in inflamed versus control conditions, reproducibility of the response across biological replicates, and the absence of false positives in the absence of inflammation. If these criteria are met, it will validate the concept of a probiotic inflammation monitor and set the stage for more advanced testing and eventual clinical translation.\u003c/p\u003e "},{"header":"Applications","content":"\u003cp\u003eThe development of a probiotic biosensor for intestinal inflammation opens the door to several impactful applications in healthcare and research:\u003c/p\u003e \u003cp\u003e \u003cb\u003eClinical Diagnostics for IBD\u003c/b\u003e: The foremost application is as a diagnostic and disease-monitoring tool for patients with IBD. Currently, patients often undergo routine colonoscopies or rely on fecal calprotectin lab tests to gauge inflammation. With our biosensor, patients could instead take a dose of the engineered probiotic (for example, as a daily capsule). The bacteria would settle in the gut and continuously monitor for inflammation flares. \u003cb\u003eEarly Warning System\u003c/b\u003e: The biosensor could provide early warning of a flare by secreting a reporter that the patient can detect easily \u0026ndash; for instance, causing stool to turn a specific color. A practical implementation might be a smart toilet or a simple at-home test strip: the patient would apply a small fecal sample to a test strip that contains a developer (analogous to a home pregnancy test concept). If the \u003cem\u003eLactobacillus\u003c/em\u003e in the stool has produced the reporter enzyme (indicating inflammation), the strip would change color or fluoresce under a handheld UV lamp. This immediate feedback would empower patients to seek treatment sooner or adjust medications to prevent a full-blown relapse. Additionally, doctors could use the biosensor as a companion diagnostic to monitor treatment efficacy \u0026ndash; if a patient starts a new anti-inflammatory therapy, the biosensor output could objectively show whether intestinal inflammation is subsiding (e.g., the fluorescence intensity in stool samples decreases over time). Compared to calprotectin ELISA, which requires sending samples to a lab, the biosensor would be more rapid and potentially more specific to certain pathways (like distinguishing oxidative stress vs. general inflammation).\u003c/p\u003e \u003cp\u003e \u003cstrong\u003ePersonalized Medicine and Remote Monitoring\u003c/strong\u003e \u003cp\u003eBecause the biosensor can, in principle, distinguish different markers (NO vs ROS vs pH), it might help personalize which aspect of inflammation is dominant in a patient\u0026rsquo;s disease. For example, a patient whose biosensor consistently reports high NO (but moderate ROS) might have a different inflammatory profile than one who shows high ROS output. This could inform personalized therapy (perhaps the first patient would benefit more from treatments targeting macrophage activity or iNOS pathways). Moreover, the use of an engineered probiotic fits well with telemedicine and remote care \u0026ndash; data from the biosensor could be transmitted via a connected device to clinicians. One envisaged system is an \u003cb\u003eingestible electronic pill\u003c/b\u003e that a patient swallows; it contains a small chamber with our sensor bacteria and a microelectronic reader[\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e]. As it passes through the GI tract, it could detect luminescence or electrochemical changes produced by the bacteria and send that data wirelessly to a phone application. This kind of \u003cb\u003ewearable/internal sensor\u003c/b\u003e combination would allow continuous monitoring without any action needed by the patient beyond ingesting the device. It could potentially alert the patient\u0026rsquo;s physician in real time if gut inflammation exceeds a dangerous threshold, enabling proactive care adjustments. While such technology is still emerging, our biosensor is being designed with these integrations in mind (hence our inclusion of luminescent and electrochemical outputs that are machine-readable).\u003c/p\u003e \u003c/p\u003e \u003cp\u003e \u003cb\u003eResearch Tool for Gut Inflammation\u003c/b\u003e: Beyond patient use, the probiotic sensor is a valuable tool for biomedical research. In inflammatory disease research, having a real-time indicator of inflammation inside an animal can greatly enhance experiments. For example, in drug development for IBD, researchers could administer the biosensor to lab mice and quickly screen whether a candidate drug reduces gut inflammation: instead of sacrificing the animals for histology at multiple time points, one could simply monitor the biosensor outputs (like measuring fecal fluorescence or luminescence daily). This not only reduces the need for invasive sampling but also provides dynamic data, revealing how quickly and how strongly inflammation is affected by a treatment. Similarly, in basic science, the sensor bacteria could be used to map inflammation within the gut. By recovering the bacteria from different regions of the intestine (they could be designed to colonize specific sections) and examining their memory switches or reporter levels, scientists could identify exactly where inflammation occurred and for how long. This might help unravel patterns such as whether inflammation initiates in patches that then spread, or how the microbiota composition correlates with localized inflammation. The \u003cb\u003ememory aspect\u003c/b\u003e (using CRISPR-based recording of signals) is particularly useful here: bacteria can traverse the gut and then be analyzed after exit to see what they \u0026ldquo;saw\u0026rdquo; along the way. If we deploy multiple strains tuned to different thresholds, we could even map intensity gradients of certain signals.\u003c/p\u003e \u003cp\u003e \u003cstrong\u003eTheranostic Applications\u003c/strong\u003e \u003cp\u003eWhile our primary focus is diagnostics, it\u0026rsquo;s worth noting the potential for \u003cb\u003ecombined therapeutic action\u003c/b\u003e. As shown by others[\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e][\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e], an engineered probiotic can be a \u0026ldquo;theranostic\u0026rdquo; agent \u0026ndash; diagnosing and simultaneously delivering therapy. Our \u003cem\u003eLactobacillus\u003c/em\u003e sensor could be augmented to secrete anti-inflammatory molecules (e.g., IL-10, as in Steidler\u0026rsquo;s work, or nanobodies against TNF-α). This means the bacteria would not only signal the presence of inflammation but also help fight it. For instance, upon detecting high ROS/NO, the strain could release IL-10 to locally suppress immune responses, potentially preventing a mild flare from escalating. In a future iteration, we could integrate a module such as an \u003cb\u003eIL-10 expression cassette\u003c/b\u003e under control of the inflammation sensors, or an \u003cem\u003eAvCystatin\u003c/em\u003e secretion module as in Zou et al.\u0026rsquo;s study[23]. The diagnostic output would then also serve as confirmation that the therapeutic has been delivered (a sort of internal feedback loop). This approach might reduce the need for systemic immunosuppressants if the probiotic can act at the inflammation site precisely. However, incorporating therapeutic functions raises additional regulatory considerations, so initially the diagnostic alone is simpler to pursue.\u003c/p\u003e \u003c/p\u003e \u003cp\u003e \u003cstrong\u003ePoint-of-Care and Field Deployable Tests\u003c/strong\u003e \u003cp\u003eOutside of IBD, a similar strategy could monitor other gut conditions. For example, intestinal infections that cause inflammation (like certain infections that lead to colitis) might be detected by a variant of our biosensor. Also, since \u003cem\u003eLactobacillus\u003c/em\u003e is commonly used in foods, one can imagine a scenario of a functional yogurt that contains the biosensor bacteria \u0026ndash; consumers at risk of IBD flare could regularly ingest it and watch for a signal (perhaps a color-change if they also consume a certain indicator food or pill). In resource-limited settings, where medical infrastructure for endoscopy is scarce, a stable dried form of the biosensor could be distributed for community screening of gut health issues. Patients could then identify potential problems early and seek medical attention if needed.\u003c/p\u003e \u003c/p\u003e \u003cp\u003e \u003cstrong\u003eFuture Wearable Tech Integration\u003c/strong\u003e \u003cp\u003eIn the longer term, the coupling of engineered microbes with wearable tech could transcend gastrointestinal diseases. Similar concepts might apply to monitoring markers of metabolic health or other conditions via microbes in different body sites[\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e]. Our project specifically contributes to that vision by addressing the challenges in one of the most microbially rich and clinically important environments \u0026ndash; the gut. A future \u0026ldquo;smart gut bandage\u0026rdquo; or implant could house such bacteria and an LED or electrode, providing continuous readouts. Our results will lay groundwork for how to calibrate and interpret microbial sensor signals reliably.\u003c/p\u003e \u003c/p\u003e \u003cp\u003eIn summary, the applications of the \u003cem\u003eLactobacillus\u003c/em\u003e inflammation biosensor are multi-fold: from giving IBD patients a convenient way to track their disease, to assisting clinicians in treatment decisions, to enabling scientists to study inflammation dynamics in unprecedented detail, to serving as a stepping stone for integrated therapeutic systems. As we demonstrate the viability of this approach, it can be expanded and customized to many scenarios where \u003cb\u003e\u0026ldquo;living diagnostics\u0026rdquo;\u003c/b\u003e offer advantages over traditional chemical tests. The versatility and self-renewing nature of probiotics make them a unique vehicle for such continuous monitoring tasks in healthcare.\u003c/p\u003e"},{"header":"Discussion","content":"\u003cp\u003eDeveloping a probiotic biosensor for intestinal inflammation brings forth several important considerations and challenges that must be addressed for the system to be practical, safe, and effective. In this section, we discuss these aspects, including the specificity of sensing, integration with the host environment and microbiome, biosafety measures, regulatory pathways, and ethical implications of deploying engineered live bacteria in patients. We also highlight future optimizations and research directions to improve the biosensor\u0026rsquo;s performance and reliability.\u003c/p\u003e \u003cp\u003e \u003cb\u003eSpecificity and False Positives\u003c/b\u003e: A key challenge in the gut environment is specificity \u0026ndash; our biosensor will be exposed to a complex mixture of molecules, and we need to ensure it reacts predominantly to the intended inflammation markers rather than unrelated signals. Each sensor module was chosen for high specificity (e.g., NorR for NO doesn\u0026rsquo;t respond to other gases; OxyR for H₂O₂ is fine-tuned for oxidative stress). However, cross-talk can occur. For instance, extreme pH changes can cause oxidative stress in bacteria, potentially activating ROS pathways indirectly, or a burst of NO might also generate reactive nitrogen intermediates that could influence other promoters. We plan thorough \u003cb\u003ecross-stimulation testing\u003c/b\u003e, as described, to map any unintended responses. If necessary, we can incorporate additional regulatory control \u0026ndash; for example, an \u003cb\u003eAND gate logic\u003c/b\u003e where two conditions must be met to trigger a response, which could improve specificity. An example would be requiring both an inflammatory metabolite and an increase in bacterial cell density (indicative of being in gut vs. lab media) to produce a signal, thereby avoiding false activation during manufacturing or storage. In terms of \u003cb\u003efalse positives\u003c/b\u003e, one scenario is if a patient\u0026rsquo;s diet or medication introduces something that the biosensor mistakes for an inflammatory signal. Nitrates in food, for example, could conceivably be reduced to NO in the gut. We will examine common dietary factors: does a meal rich in nitrates cause a spike in our NO reporter? If so, one might need to advise dietary restrictions or engineer the sensor to a slightly different trigger (like an inflammation-specific metabolite such as tetrathionate which is less likely from diet alone). Another potential false positive source is transient gut infections or mild irritation that is not an IBD flare but still inflames somewhat. The biosensor might pick that up. This isn\u0026rsquo;t necessarily bad \u0026ndash; it\u0026rsquo;s an accurate detection of inflammation \u0026ndash; but clinically we\u0026rsquo;d need to interpret it correctly. Distinguishing IBD flare from, say, a brief food poisoning episode might require context (e.g., the duration of signal or presence of specific cytokines). Possibly, a future version could include a \u003cb\u003eC-reactive protein sensor\u003c/b\u003e (for systemic inflammation) to differentiate chronic vs. acute conditions.\u003c/p\u003e \u003cp\u003e \u003cstrong\u003eMicrobiome Integration and Persistence\u003c/strong\u003e \u003cp\u003eWhen introducing engineered \u003cem\u003eLactobacillus\u003c/em\u003e into the gut, we must consider how it behaves within the existing microbiome ecosystem. \u003cem\u003eLactobacillus rhamnosus\u003c/em\u003e GG is a transient colonizer \u0026ndash; typically, it will pass through the gut and not permanently take up residence. This can be advantageous from a safety standpoint (it won\u0026rsquo;t overstay its welcome), but it also means for continuous monitoring the patient might need to ingest it regularly (e.g., daily or weekly). Is that feasible? Perhaps yes, if formulated as a yogurt or pill, as many people already take probiotics regularly. We will have to see how long the bacteria survive in the gut per dose. Some engineered probiotics in trials for other conditions have been given daily or every few days. If a more stable colonization is desired, one could consider \u003cb\u003ebiofilm-forming\u003c/b\u003e variants or strains that adhere to mucus (some \u003cem\u003eLactobacillus\u003c/em\u003e have good adhesion properties). Our design did not include a colonization factor specifically, to avoid interference with normal microbiota, but it\u0026rsquo;s a lever we could adjust if needed. It\u0026rsquo;s also important that our sensor bacteria do not significantly disturb the native microbiome. The payload is mostly sensing and reporting; it does express some extra proteins (GFP, etc.), but these should not give it a major fitness advantage or disadvantage that would cause it to bloom or die off drastically. We will monitor microbiome composition in any animal studies (16S rRNA sequencing, for example) to see if the introduction of the sensor has any dysbiotic effect. Ideally, it remains a small fraction of the community, just enough to do its job, without outcompeting beneficial microbes.\u003c/p\u003e \u003c/p\u003e \u003cp\u003e \u003cb\u003eBiosafety and Containment\u003c/b\u003e: The release of genetically engineered microbes in a human host raises biosafety questions. One major concern is \u003cb\u003ehorizontal gene transfer\u003c/b\u003e: could our plasmid or genetic elements transfer to other gut microbes? We have taken steps to mitigate this. The plasmid we use has a narrow host range (mainly lactobacilli and some related Gram-positives) and an origin that typically doesn\u0026rsquo;t replicate in most Gram-negative gut flora. We also included no antibiotic resistance markers that would confer advantage to pathogens (erythromycin resistance is not useful to Gram-negatives, and many gut commensals are intrinsically erythromycin-resistant anyway, but that\u0026rsquo;s mostly confined to Gram-positives like Enterococci). We can further minimize HGT by using a \u003cb\u003enon-transmissible vector\u003c/b\u003e (no conjugation machinery, and mobilization sequences removed). Additionally, we can implement a \u003cb\u003ekill switch\u003c/b\u003e or dependency: our strain is designed with a kill mechanism that triggers when inflammation subsides (as a way to clear out once the job is done)[\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e]. For example, the LZU-China 2024 project used a nitric oxide-sensitive kill switch that lyses the bacteria when NO levels go down, ensuring the bacteria self-eliminate after resolving inflammation[\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e]. We could adapt that, or simply have an inducible kill switch that can be triggered by an administrated molecule (e.g., adding a particular sugar that activates a bacteriophage lysis gene we encoded, should we want to terminate the biosensor population). Another safety mechanism is auxotrophy: engineering the strain to require a supplement (like D-alanine or thymidine) not present in the human gut. This way, if it escapes into the environment (e.g., via feces), it will die off due to lack of that nutrient, preventing long-term environmental spread. We are considering making our strain a \u003cb\u003ethyA mutant\u003c/b\u003e (thymine auxotroph), so it cannot survive outside a host (this method was successfully used by Steidler et al. with IL-10 \u003cem\u003eLactococcus\u003c/em\u003e to ensure containment[\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e]).\u003c/p\u003e \u003cp\u003eFrom a \u003cb\u003eregulatory\u003c/b\u003e perspective, using a live GMO in humans means we will likely need to go through processes similar to those for live biotherapeutic products (LBPs). The FDA and EMA have guidance for probiotics and engineered bacteria in trials (some are already in early clinical trials for other diseases, like \u003cem\u003eE. coli\u003c/em\u003e Nissle engineered for hyperammonemia by Synlogic). Key data to provide will include toxicity (does the engineered strain cause any inflammation on its own? We suspect not, as \u003cem\u003eLactobacillus\u003c/em\u003e is generally benign), colonization/shedding analysis (how long after ingestion can it be recovered, and is it completely cleared eventually?), and genetic stability (does it maintain the synthetic circuit without mutations that could, say, inactivate the kill switch or activate a silenced gene unintentionally?). We will have to demonstrate that our strain doesn\u0026rsquo;t carry virulence factors or express any harmful substances. Since we are adding foreign genes (GFP, etc.), we must ensure none of those pose any risk \u0026ndash; they shouldn\u0026rsquo;t, as GFP is inert and enzymes like β-galactosidase are already present in many probiotics (e.g., yogurt cultures).\u003c/p\u003e \u003cp\u003e \u003cb\u003eEthical Considerations\u003c/b\u003e: There is often public concern about GMOs, especially one that a person would ingest and that could theoretically be excreted into the environment. Transparency and safety are paramount. We will design our biosensor such that it cannot thrive outside the target environment and ideally dies after use, to alleviate ecological concerns. Patient informed consent is necessary, making sure users understand this is a live engineered microbe. An interesting ethical angle is \u003cb\u003edata privacy\u003c/b\u003e: if we integrate this with wireless devices, then data about one\u0026rsquo;s gut health becomes something that could be transmitted \u0026ndash; it\u0026rsquo;s health data that must be protected. Ensuring secure and opt-in data handling will be important if we go that route. Another point is the psychological effect on patients: seeing a colored output in your stool could cause anxiety if misinterpreted. We have to educate users that the biosensor is an aid, not an absolute verdict \u0026ndash; for instance, a mild color might mean \u0026ldquo;monitor, but not an emergency.\u0026rdquo; Proper calibration and perhaps a quantitative read (via a device rather than just eyeballing color) could help with that.\u003c/p\u003e \u003cp\u003e \u003cstrong\u003eChallenges and Future Improvements\u003c/strong\u003e \u003cp\u003eWhile our current design is functional, there are many avenues for refinement. One is to improve the \u003cb\u003equantitative accuracy\u003c/b\u003e of the sensor. As it stands, the outputs are somewhat qualitative (color change yes/no). We can work on making them more quantifiable \u0026ndash; for instance, outputting a proportional signal to the level of inflammation. The memory circuit with CRISPR spacers could even record how many times inflammation spiked, which is a rich dataset. We also might consider \u003cb\u003emulti-channel signaling\u003c/b\u003e where the ratio of two reporter signals encodes information (e.g., red vs green fluorescence intensity could indicate the relative contributions of two pathways). Another improvement could be making the bacteria respond faster. Biological circuits have inherent delays (e.g., it may take an hour or two for GFP to accumulate). In some cases, a faster response is needed. Using more rapidly detectable outputs like secretion of a small molecule that can diffuse and be detected quickly (perhaps a change in breath gas, if engineered to produce a volatile marker) is an intriguing idea. Imagine if during a flare, the bacteria released a small amount of a fragrant compound that could be detected on the patient\u0026rsquo;s breath or a skin sensor \u0026ndash; truly noninvasive. This is speculative, but not impossible with metabolic engineering.\u003c/p\u003e \u003c/p\u003e \u003cp\u003e \u003cb\u003eCompatibility with Treatments\u003c/b\u003e: Many IBD patients take medications like mesalamine, immunosuppressants, or antibiotics. We have to ensure the biosensor works under those conditions. For instance, if the patient is on antibiotics that kill Gram-positive bacteria, our \u003cem\u003eLactobacillus\u003c/em\u003e might be wiped out, failing to function. Perhaps the sensor would be most useful in periods when patients are not on antibiotics. Alternatively, we could engineer resistance to certain narrow-spectrum antibiotics if needed, but that raises more regulatory hurdles. It might be better to inform users to avoid certain antibiotics while using the sensor or to reintroduce the probiotic after a course of antibiotics. As for anti-inflammatory meds (e.g., corticosteroids), if they succeed in reducing inflammation, the sensor would simply report less signal \u0026ndash; which is fine (that\u0026rsquo;s actually a desired confirmation). There\u0026rsquo;s no interference there except that if a drug like mesalamine releases at pH\u0026thinsp;\u0026gt;\u0026thinsp;7 (some formulations do), and our sensor lowers pH reading due to inflammation, the drug release might differ \u0026ndash; but that\u0026rsquo;s more of a therapeutic design consideration, not directly affecting our sensor readout except to say our sensor might also indirectly verify drug release (some creative synergy could be found here: a sensor could tell if a pH-dependent drug is releasing properly by detecting the pH profile).\u003c/p\u003e \u003cp\u003e \u003cstrong\u003eRegulatory Path and Precedents\u003c/strong\u003e \u003cp\u003eIt\u0026rsquo;s worth noting that a similar concept reached at least animal models \u0026ndash; for example, the 2017 study by Daeffler et al. and the 2023 study by Zou et al. Both demonstrated safety in mice. Moving to human trials requires scaling up manufacturing of the engineered strain under GMP conditions. That is doable since \u003cem\u003eLactobacillus\u003c/em\u003e can be fermented industrially. The strain would have to be well-characterized, free of any adventitious agents, and tested in phase 1 trials for safety. The \u003cb\u003eendpoint\u003c/b\u003e of such trials would likely be showing that it passes through and reports something consistent with conventional markers, and that it doesn\u0026rsquo;t cause adverse effects. The \u003cb\u003ebenefit\u003c/b\u003e of our approach is mostly in the realm of patient quality of life and disease management \u0026ndash; catching flares early could reduce hospitalizations and serious complications. We might need to show health economics data eventually that it reduces overall costs (fewer expensive procedures, etc.).\u003c/p\u003e \u003c/p\u003e \u003cp\u003e \u003cstrong\u003ePublic Acceptance\u003c/strong\u003e \u003cp\u003eUsing probiotics is generally well accepted (people eat yogurt, kefir, etc., with live cultures). The engineered aspect may raise eyebrows, but if framed as \u0026ldquo;it\u0026rsquo;s a probiotic that can alert you to inflammation,\u0026rdquo; many might see it as a natural extension of probiotic use. Clear communication will be needed to explain that the bacteria have been modified to perform a sensing task and that they have built-in safety features. Engaging with patient advocacy groups (for Crohn\u0026rsquo;s and colitis) early on could provide feedback on what features they\u0026rsquo;d value and any concerns.\u003c/p\u003e \u003c/p\u003e \u003cp\u003eIn conclusion, while challenges remain, they are surmountable with thoughtful design and testing. The concept of using \u003cem\u003eLactobacillus\u003c/em\u003e as a living diagnostic platform is part of a broader movement in synthetic biology to \u003cb\u003eprogram microbes as our partners in maintaining health\u003c/b\u003e[\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e]. Our work contributes to this by focusing on a concrete clinical need and a practical organism. If successful, it could transform how patients manage chronic GI inflammation \u0026ndash; shifting from reactive care to proactive monitoring. Furthermore, it paves the way for more sophisticated probiotic devices that not only sense but also respond to disease, ushering in a new generation of \u0026ldquo;smart probiotics\u0026rdquo; for precision medicine.\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eFunding\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis research received no external funding.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eClinical Trial Number\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eClinical trial number: not applicable.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eEthics Approval and Consent to Participate\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eEthics approval and consent to participate: not applicable.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eConsent for Publication\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eConsent for publication: Consent is given to the journal this manuscript is submitted to.\u003c/p\u003e\u003ch2\u003eAuthor Contribution\u003c/h2\u003e\u003cp\u003eY.Y. conceived and designed the study, performed the research, analyzed the data, and wrote the manuscript.\u003c/p\u003e\u003ch2\u003eData Availability\u003c/h2\u003e\u003cp\u003eData is available upon request.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eArcher EJ, Robinson AB, S\u0026uuml;el GM. Engineered E. coli that detect and respond to gut inflammation through nitric oxide sensing. ACS Synth Biol. 2012;1(10):451\u0026ndash;7.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eChen X, Rottinghaus AG, Ferrea M, et al. Rational design and characterization of nitric oxide biosensors in E. coli Nissle 1917 and Mini SimCells. ACS Synth Biol. 2021;10(10):2566\u0026ndash;78.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDaeffler KN, Galley JD, Sheth RU, et al. Engineering bacterial thiosulfate and tetrathionate sensors for detecting gut inflammation. Mol Syst Biol. 2017;13(4):923.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDin MO, Danino T, Prindle A, et al. Programmable probiotics for detection of cancer in urine. Sci Transl Med. 2016;8(343):343ra84.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKimura H, Miura S, Shigematsu T, et al. Increased nitric oxide production and inducible NO synthase activity in colonic mucosa of patients with active ulcerative colitis and Crohn\u0026rsquo;s disease. Dig Dis Sci. 1997;42(5):1047\u0026ndash;54.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLynch JB, Hsiao A, Ringus DL, et al. Engineered Escherichia coli for the in situ secretion of therapeutic nanobodies in the gut. Cell Host Microbe. 2023;31(4):634\u0026ndash;49.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMays ZJ, Nair NU. Synthetic biology in probiotic lactic acid bacteria: at the frontier of living therapeutics. Curr Opin Biotechnol. 2018;53:224\u0026ndash;31.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMimee M, Nadeau P, Hayward A, et al. An ingestible bacterial-electronic system to monitor gastrointestinal health. Science. 2018;360(6391):915\u0026ndash;8.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNugent SG, Kumar D, Rampton DS, Evans DF. Intestinal luminal pH in inflammatory bowel disease: possible determinants and implications for therapy. Gut. 2001;48(4):571\u0026ndash;7.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRiglar DT, Silver PA. Engineering bacteria for diagnostic and therapeutic applications. Nat Rev Microbiol. 2018;16(4):214\u0026ndash;25.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSteidler L, Hans W, Schotte L, et al. Treatment of murine colitis by Lactococcus lactis secreting interleukin-10. Science. 2000;289(5483):1352\u0026ndash;5.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWeibel N, Westmann C, Aguilar L, et al. Engineering a novel probiotic toolkit in Escherichia coli Nissle 1917 for sensing and mitigating gut inflammatory diseases. ACS Synth Biol. 2024;13(10):2376\u0026ndash;90.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWinter SE, Thiennimitr P, Winter MG, et al. Gut inflammation provides a respiratory electron acceptor for Salmonella. Nature. 2010;467(7314):426\u0026ndash;9.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZou ZP, Du Y, Fang TT, Zhou Y, Ye BC. Biomarker-responsive engineered probiotic diagnoses, records, and ameliorates inflammatory bowel disease in mice. Cell Host Microbe. 2023;31(2):199\u0026ndash;e2125.\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"discover-bacteria","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"","sideBox":"Learn more about [Discover Bacteria](https://link.springer.com/journal/44351)","snPcode":"44351","submissionUrl":"https://submission.springernature.com/new-submission/44351/3","title":"Discover Bacteria","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"Discover Series","inReviewEnabled":true,"inReviewRevisionsEnabled":true},"keywords":"probiotic biosensor, Lactobacillus, gut inflammation, inflammatory bowel disease, synthetic biology, intestinal diagnostics","lastPublishedDoi":"10.21203/rs.3.rs-9351383/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-9351383/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eInflammatory bowel diseases (IBD) are characterized by chronic intestinal inflammation, driving a need for real-time, noninvasive monitoring of inflammatory biomarkers. Here we propose a \u003cb\u003eLactobacillus-based probiotic biosensor\u003c/b\u003e engineered to detect molecular signals of gut inflammation and report them via easily measurable outputs. The chassis organism, \u003cem\u003eLactobacillus\u003c/em\u003e (a commensal lactic acid bacterium), is genetically programmed with synthetic circuits that respond to \u003cb\u003ereactive oxygen species (ROS)\u003c/b\u003e, \u003cb\u003ereactive nitrogen species (e.g., nitric oxide, NO)\u003c/b\u003e, \u003cb\u003epro-inflammatory cytokines\u003c/b\u003e (such as tumor necrosis factor-α, TNF-α, or interleukin-6, IL-6), and \u003cb\u003epH shifts\u003c/b\u003e in the intestinal microenvironment. Detection of these inflammation-associated biomarkers triggers expression of reporter systems, producing outputs ranging from fluorescent proteins to colorimetric enzymes and electrochemical signals. We outline the design of modular sensor pathways \u0026ndash; for example, a hydrogen peroxide-responsive promoter for ROS, an NO-responsive genetic circuit, and a pH-sensitive two-component system \u0026ndash; all optimized for \u003cem\u003ein vivo\u003c/em\u003e function in the gut. A broad-host-range plasmid system is used to deploy these circuits in \u003cem\u003eLactobacillus\u003c/em\u003e, with regulatory elements tailored for Gram-positive expression. We describe a workflow for constructing the biosensor strain, including cloning of sensor and reporter genes, engineering of secretion peptides for extracellular reporting, and assay protocols for validation in simulated gut conditions. The expected results include sensitive and specific responses to pathological levels of inflammatory markers, with minimal crosstalk or background activity in the absence of inflammation. This living diagnostic platform has applications in clinical monitoring of IBD activity, enabling early detection of flare-ups via stool or capsule-based readouts. It can also serve as a research tool for real-time mapping of gut inflammation in animal models and could be integrated into future \u003cb\u003ewearable or ingestible devices\u003c/b\u003e for continuous gastrointestinal health monitoring. In the Discussion, we examine challenges such as maintaining sensor specificity in the complex gut milieu, ensuring the engineered \u003cem\u003eLactobacillus\u003c/em\u003e remains contained and stable in the microbiome, and addressing biosafety and regulatory hurdles for therapeutic use. This work demonstrates a path toward \u003cb\u003eprobiotic diagnostics\u003c/b\u003e, harnessing synthetic biology to create intestinal sentinels that detect and report on inflammation from within the gastrointestinal tract.\u003c/p\u003e","manuscriptTitle":"Implementation of Probiotic Bacteria for Monitoring Intestinal Inflammation","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2026-04-30 17:09:10","doi":"10.21203/rs.3.rs-9351383/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"reviewersInvited","content":"","date":"2026-04-22T01:53:33+00:00","index":"","fulltext":""},{"type":"editorInvited","content":"","date":"2026-04-21T12:46:59+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2026-04-09T11:10:33+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2026-04-09T11:09:56+00:00","index":"","fulltext":""},{"type":"submitted","content":"Discover Bacteria","date":"2026-04-08T04:08:29+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"discover-bacteria","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"","sideBox":"Learn more about [Discover Bacteria](https://link.springer.com/journal/44351)","snPcode":"44351","submissionUrl":"https://submission.springernature.com/new-submission/44351/3","title":"Discover Bacteria","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"Discover Series","inReviewEnabled":true,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"d75cdc7f-2c3e-4255-8e2e-bd22d5e048fa","owner":[],"postedDate":"April 30th, 2026","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"under-review","subjectAreas":[],"tags":[],"updatedAt":"2026-04-30T17:09:11+00:00","versionOfRecord":[],"versionCreatedAt":"2026-04-30 17:09:10","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-9351383","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-9351383","identity":"rs-9351383","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}