Programmable receptors enable bacterial biosensors to detect pathological biomarkers in clinical samples

The study addresses the need for simple, field-deployable diagnostic tools for early detection and monitoring of chronic diseases, focusing on liver dysfunction. Conventional in vitro diagnostics often require centralized facilities, costly equipment, and trained personnel. Whole-cell biosensors (genetically engineered living cells) offer a promising alternative due to their innate ability to sense diverse chemical cues and their recently improved robustness and sensitivity in complex media. Liver diseases are prevalent and deadly, and current biomarkers and methods can lack specificity, are hospital-bound, and may miss early disease. Bile salts in serum are a highly specific, dynamic biomarker of liver function and are gold-standard for pregnancy cholestasis, with relevance to early cirrhosis, hepatitis, and drug-induced injury. However, current bile salt assays are centralized and lack species discrimination. The research goal is to engineer modular, programmable bacterial receptors (EMeRALD platform) in Escherichia coli to detect clinically relevant bile salts, improve sensitivity and limit of detection via directed evolution, and develop a practical colorimetric biosensor operable in clinical samples.

Background work highlights whole-cell biosensors’ advancements in robustness and signal processing enabling use in complex matrices like wastewater and clinical samples. Liver disease diagnostics typically involve biomarker panels; serum bile salts have emerged as sensitive, dynamic indicators for various liver conditions and are the standard for diagnosing intrahepatic cholestasis of pregnancy. Enteropathogenic Vibrio species naturally sense bile salts to regulate virulence via TcpP/TcpH (Vibrio cholerae) and VtrA/VtrC (Vibrio parahaemolyticus) signaling. Prior studies implicate bile salts and specific residues (e.g., disulfide-forming cysteines) in TcpP function and virulence regulation. The authors build upon modular receptor platforms (EMeRALD) enabling swappable ligand-binding domains connected to a CadC-based transcriptional scaffold, and leverage deep mutational scanning and NGS to map sequence–function relationships for improving receptor performance.

- Receptor design and construction: The EMeRALD platform uses a synthetic receptor scaffold comprising a transmembrane region, a juxtamembrane linker, and the CadC DNA-binding domain (DBD). Vibrio bile salt-sensing modules were repurposed by fusing CadC DBD to the periplasmic and transmembrane domains of TcpP (V. cholerae) or to VtrA (V. parahaemolyticus) with co-expression of their cofactors TcpH or VtrC, respectively. Reporters: sfGFP under the CadC target promoter (PcCadB) for fluorescence readouts; LacZ with CPRG substrate for colorimetric output. - Expression tuning and characterization: Constitutive promoters (e.g., P14, P10, P9 for CadC-TcpP; P5 for TcpH) were used to tune receptor/cofactor expression. Transfer functions were measured in E. coli grown to exponential phase, incubated with bile salts for 4 h, and analyzed by flow cytometry (Attune NxT) to quantify GFP in reference promoter units. - Specificity profiling: The TcpP/TcpH-EMeRALD and VtrA/VtrC-EMeRALD systems were tested against a panel of 12 bile salts (primary and secondary, conjugated and unconjugated), including TCA, GCDCA, GDCA, TDCA, and others, to determine specificity and dynamic range. - Directed evolution of TcpP loop: Guided by multiple sequence alignments and structural modeling (PSIPRED secondary structure prediction and Rosetta ab initio modeling), the flexible loop between conserved cysteines Cys207 and Cys218 in TcpP was targeted. A comprehensive NNK mutagenesis library was constructed at residues A210–G213 (NYEQ region). Libraries were cloned into CadC-TcpP plasmids driving GFP. - Functional screening and selection: Libraries were induced with taurocholic acid (TCA). Fluorescent responding cells were enriched by FACS across three rounds (TCA at 200 μM for rounds 1–2; 20 μM for round 3), progressively gating for increased sensitivity (20–80 μM range). Individual clones were isolated post-sorting for detailed transfer function analysis. - Sequencing and analysis: The pool of selected variants was subjected to NGS (Illumina PE250). Sequence counts were bias-corrected; sequence logos were generated to identify enriched/depleted residues at targeted positions, revealing sequence–function determinants. - Kinetic analysis: Time-course measurements compared variant response dynamics and inferred changes in apparent affinity relative to wild-type TcpP. - Colorimetric assay development: The best-performing variant (TcpP18) was fused into a LacZ reporter system (TcpP18-LacZ). Using CPRG substrate, color change (yellow to purple via chlorophenol red) was quantified as ΔA580. Cell density and incubation time were tuned to adjust LOD and linear range; linear response to GCDCA from 0 to 40 μM was established after optimization. - Clinical sample testing: The colorimetric bactosensor (TcpP18-LacZ) was tested on 21 serum samples from liver transplant patients (followed at the Montpellier hospital). Samples were diluted 10-fold and incubated for 2 h. Results were compared to a standard enzymatic bile salt assay kit to assess concordance and the ability to detect pathological elevations. - Additional methods: Flow cytometry settings, cell sorting parameters (Bio-Rad S3), NGS data processing scripts (GitHub), and growth/measurement conditions (e.g., media, incubation times, microplate reader settings) are provided in Methods and Supplementary Information.

- Modular receptors in E. coli: The EMeRALD platform successfully rewired Vibrio bile salt-sensing modules into E. coli. Both TcpP/TcpH-EMeRALD and VtrA/VtrC-EMeRALD constructs were functional, showing robust ligand-induced reporter activation. - Specificity differences: TcpP/TcpH-EMeRALD was highly specific to primary conjugated bile salts, especially taurocholic acid (TCA) and glycochenodeoxycholic acid (GCDCA), and showed negligible response to secondary bile salts. VtrA/VtrC-EMeRALD exhibited broader specificity, primarily responding to secondary conjugated bile salts (e.g., GDCA, TDCA). - Sensitivity improvement by directed evolution: Targeted mutagenesis of the TcpP loop (A210–G213) followed by FACS selection yielded variants with improved sensitivity and lower EC50. Variant V18 (TcpP18) achieved an EC50 of approximately 28.3 μM for TCA versus ~89.4 μM for wild-type, indicating markedly increased sensitivity. Kinetic analyses showed TcpP18 had roughly a 13-fold increase in apparent ligand affinity and faster response at low ligand concentrations compared to wild type. - Sequence–function insights: NGS of functional variants revealed depletion of negatively charged (Asp, Glu) and polar (Asn, Gln) residues at key loop positions (e.g., 211), and enrichment of hydrophobic and aromatic residues (Phe, Tyr, Leu), uncovering previously unrecognized determinants for TcpP function relevant to virulence signaling. - Colorimetric biosensor performance: The TcpP18-LacZ colorimetric assay, using CPRG, produced a naked-eye-detectable readout. After optimization of cell density and incubation time, it showed a linear response to GCDCA from 0 to 40 μM, with adjustable thresholding to cover disease-relevant ranges. - Clinical sample detection: In 21 serum samples from liver transplant patients, the bactosensor detected elevated bile salt levels consistent with measurements from a standard enzymatic bile salt assay kit, demonstrating feasibility for clinical monitoring of post-transplant complications (e.g., biliary stenosis, acute cellular rejection).

The findings demonstrate that modular synthetic receptors can transplant natural pathogen-sensing capabilities into safe surrogate hosts, enabling practical diagnostics. By connecting TcpP/TcpH and VtrA/VtrC bile salt sensors to the EMeRALD scaffold, the authors showed that essential sensing components alone suffice for functional bile detection, validating the platform's modularity and scalability. Directed evolution of TcpP yielded substantial gains in sensitivity and reduced LOD, and provided new mechanistic insights into loop-sequence features governing receptor function—information not easily inferred from natural sequence alignments. The colorimetric TcpP18-LacZ sensor translates molecular detection into a simple, naked-eye readout suited to low-resource settings. Application to transplant patient sera underscores clinical relevance for early detection and monitoring of liver dysfunction. The work suggests that EMeRALD receptors can be integrated with genetic circuits for multiplexing, logic, memory, and amplification, expanding capabilities for diagnostics, environmental monitoring, and control of therapeutic microbes. Future translations could include inhibitor discovery for virulence pathways and gut-delivery control of engineered probiotics.

This study introduces a programmable, modular receptor platform (EMeRALD) to repurpose natural bacterial sensing modules in E. coli, enabling precise detection of clinically relevant bile salts. The authors engineered and characterized two bile salt sensors with complementary specificity (TcpP/TcpH and VtrA/VtrC), enhanced TcpP sensitivity via directed evolution (TcpP18), and developed a practical colorimetric assay. The bactosensor detected pathological bile salt levels in clinical serum samples, positioning the technology for point-of-care diagnostics and longitudinal patient monitoring. Future work should focus on full clinical validation in larger cohorts, integration with amplification and logic circuits, device miniaturization (e.g., microfluidics), stabilization (e.g., lyophilization), and expansion to additional biomarkers and sample types.

- Clinical validation: The clinical cohort was limited (21 serum samples) and heterogeneous; larger, diverse cohorts are required to establish diagnostic performance metrics (sensitivity, specificity, predictive values). - Operational complexity: The current assay still involves several laboratory steps (cell culture, incubation, centrifugation/lysis for some formats), limiting immediate point-of-care deployment. - Sample matrix constraints: Serum handling requires additional processing; while urine was tested as an alternative non-invasive matrix, endogenous bile salt detection in urine was unreliable due to prevalent sulfation. - Platform maturity: While sensitivity and LOD were improved, further optimization and integration with user-friendly hardware (e.g., lab-on-chip) and stable formulations (e.g., lyophilized cells) are needed for at-home or field use.