Development and Characterization of Indole-Responsive Whole-Cell Biosensor Based on the Inducible Gene Expression System from Pseudomonas putida KT2440

Indole is a biologically active compound naturally occurring in plants and some bacteria. It is an important specialty chemical that is used as a precursor by the pharmaceutical and chemical industries, as well as in agriculture. Recently, indole has been identified as an important signaling molecule for bacteria in the mammalian gut. The regulation of indole biosynthesis has been studied in several bacterial species. However, this has been limited by the lack of in vivo tools suitable for indole-producing species identification and monitoring. The genetically encoded biosensors have been shown to be useful for real-time quantitative metabolite analysis. This paper describes the identification and characterization of the indole-inducible system PpTrpI/PPP_RS00425 from Pseudomonas putida KT2440. Indole whole-cell biosensors based on Escherichia coli and Cupriavidus necator strains are developed and validated. The specificity and dynamics of biosensors in response to indole and its structurally similar derivatives are investigated. The gene expression system PpTrpI/PPP_RS00425 is shown to be specifically induced up to 639.6-fold by indole, exhibiting a linear response in the concentration range from approximately 0.4 to 5 mM. The results of this study form the basis for the use of whole-cell biosensors in indole metabolism-relevant bacterial species screening and characterization.


Introduction
Indole is an N-heterocyclic aromatic compound that belongs to a large group of naturally occurring biologically active molecules. It plays a signaling role in organisms between various species and kingdoms [1,2]. Importantly, indole has been identified as a significant antimicrobial agent, which can impact the bacterial communication network, decreasing the bacterial resistance, as shown in the example of enteric Escherichia coli and Citrobacter rodentium infection treatment [3]. Indole-based derivatives also have potential as antibacterial agents against methicillin-resistant Staphylococcus aureus [4]. The range of indole applications is wide and versatile, as it can be used either directly or as a precursor to producing pharmaceuticals, amino acid L-tryptophan, plant growth regulators, and dyes [5][6][7][8]. For example, indole-3-butyric acid (3-IBA) is synthesized from indole and is commonly applied as a rooting agent [9]. Notably, biologically active compounds that can be derived from indole, including tryptophan, melatonin, and serotonin, are of high pharmaceutical interest. For instance, indole derivatives are used for the treatment of migraines, depression, various mental disorders, cancer, viral infections, and others [10][11][12][13].
Currently, the main industrial production method of indole and its derivatives is based on Fischer indole synthesis [14]. While the Fischer method is most commonly used, other The trpIAB operon, encoding transcription factor TrpI (locus tag PP_RS00430), and tryptophan synthase subunits alpha (TrpA, PP_RS00420) and beta (TrpB, PP_RS00425) in P. putida KT2440. The gene arrangement of (B) indole-inducible gene expression system PpTrpI/PPP_RS00425 with rfp reporter gene assembled in the construct pPM0081 that was used for developing the whole-cell biosensor, (C) intergenic region 'only' containing promoter PPP_RS00425 followed by rfp gene (pPM0082), and (D) inducible system PpTrpI/PPP_RS00425 with rfp reporter gene followed by tnaA gene encoding tryptophanase (pPM0083). AatII, NdeI, BglII, and BamHI represent positions of restriction sites, which were used for cloning and restriction analysis.
Tryptophan synthases are found in most bacteria, and they are present in all known proteobacteria [22], including E. coli and C. necator. Moreover, the tryptophan synthase α subunit (TrpA) catalyzes reversible conversion between indole and I3GP. Based on this, we reasoned that the putative gene expression system PpTrpI/PPP_RS00425 can be activated in the presence of indole and therefore be used for developing an indole-specific biosensor. To this end, the putative gene expression system PpTrpI/PPP_RS00425, harboring trpI and the adjacent promoter region PPP_RS00425, was assembled into the vector pBRC1 as described in see Section 4. To demonstrate the wider applicability of the indole-inducible system, the resulting plasmid construct pPM0081 ( Figure 1B) was introduced into the gammaproteobacterium E. coli, a commonly used bacterial chassis in synthetic biology, and the evolutionary distant betaproteobacterium C. necator. The obtained whole-cell biosensors were validated by confirming their response to indole ( Figure 2). The trpIAB operon, encoding transcription factor TrpI (locus tag PP_RS00430), and tryptophan synthase subunits alpha (TrpA, PP_RS00420) and beta (TrpB, PP_RS00425) in P. putida KT2440. The gene arrangement of (B) indole-inducible gene expression system PpTrpI/P PP_RS00425 with rfp reporter gene assembled in the construct pPM0081 that was used for developing the wholecell biosensor, (C) intergenic region 'only' containing promoter P PP_RS00425 followed by rfp gene (pPM0082), and (D) inducible system PpTrpI/P PP_RS00425 with rfp reporter gene followed by tnaA gene encoding tryptophanase (pPM0083). AatII, NdeI, BglII, and BamHI represent positions of restriction sites, which were used for cloning and restriction analysis.
Tryptophan synthases are found in most bacteria, and they are present in all known proteobacteria [22], including E. coli and C. necator. Moreover, the tryptophan synthase α subunit (TrpA) catalyzes reversible conversion between indole and I3GP. Based on this, we reasoned that the putative gene expression system PpTrpI/P PP_RS00425 can be activated in the presence of indole and therefore be used for developing an indole-specific biosensor. To this end, the putative gene expression system PpTrpI/P PP_RS00425 , harboring trpI and the adjacent promoter region P PP_RS00425 , was assembled into the vector pBRC1 as described in see Section 4. To demonstrate the wider applicability of the indole-inducible system, the resulting plasmid construct pPM0081 ( Figure 1B) was introduced into the gammaproteobacterium E. coli, a commonly used bacterial chassis in synthetic biology, and the evolutionary distant betaproteobacterium C. necator. The obtained whole-cell biosensors were validated by confirming their response to indole ( Figure 2). Furthermore, to confirm that the transcription factor TrpI regulates the PPP_RS00425 promoter region, the fluorescence outputs of biosensors containing inducible system PpTrpI/PPP_RS00425 (pPM0081) were compared to those of cells harboring construct pPM0082 ( Figure 1C) with the PpPPP_RS00425 promoter only. Figure 2 shows that a response to 1 mM indole was observed only in the presence of the trpI gene. These results demonstrate that transcription factor TrpI is essential and acts as a transcriptional activator for the inducible gene expression system PpTrpI/PPP_RS00425. Interestingly, the observed 10% induction using the C. necator H16/PpPPP_RS00425 biosensor indicates that some bacteria outside the Pseudomonas genus, such as C. necator, may possess TR, contributing to the activation of the PpPPP_RS00425 promoter.

Characterization of Indole-Inducible Biosensors
The indole-inducible gene expression system PpTrpI/PPP_RS00425 (pPM0081) was subsequently characterized by evaluating its response to different concentrations of indole (dose-response) and dynamic range of induction as described previously [36,37]. The fluorescence and absorbance outputs were measured over time using both E. coli and C. necator-based biosensors containing the inducible system PpTrpI/PPP_RS00425 in Luria-Bertani (LB) medium and minimal medium (MM) supplemented with different concentrations of indole. For both biosensors, the inducible system's activation, represented by increased fluorescence output compared to the uninduced state, starts 60 min after the addition of indole (Supplementary Figure S1). Notably, the induction kinetics profiles differ between E. coli and C. necator-based biosensors, most likely due to variability in their growth rates and responses to the indole. E. coli growth was less affected by the indole, with no significant growth inhibition up to 1 mM, which is a commonly occurring concentration of this compound in the intestinal tract [4]. However, C. necator was more susceptible to the indole, with significant growth inhibition observed at concentrations of 0.125 mM and above. Figure 3 shows the dose-response curves representing the correlation between extracellular indole concentration and relative normalized fluorescence 4 h after inducer Furthermore, to confirm that the transcription factor TrpI regulates the P PP_RS00425 promoter region, the fluorescence outputs of biosensors containing inducible system PpTrpI/P PP_RS00425 (pPM0081) were compared to those of cells harboring construct pPM0082 ( Figure 1C) with the PpP PP_RS00425 promoter only. Figure 2 shows that a response to 1 mM indole was observed only in the presence of the trpI gene. These results demonstrate that transcription factor TrpI is essential and acts as a transcriptional activator for the inducible gene expression system PpTrpI/P PP_RS00425 . Interestingly, the observed 10% induction using the C. necator H16/PpP PP_RS00425 biosensor indicates that some bacteria outside the Pseudomonas genus, such as C. necator, may possess TR, contributing to the activation of the PpP PP_RS00425 promoter.

Characterization of Indole-Inducible Biosensors
The indole-inducible gene expression system PpTrpI/P PP_RS00425 (pPM0081) was subsequently characterized by evaluating its response to different concentrations of indole (dose-response) and dynamic range of induction as described previously [36,37]. The fluorescence and absorbance outputs were measured over time using both E. coli and C. necator-based biosensors containing the inducible system PpTrpI/P PP_RS00425 in Luria-Bertani (LB) medium and minimal medium (MM) supplemented with different concentrations of indole. For both biosensors, the inducible system's activation, represented by increased fluorescence output compared to the uninduced state, starts 60 min after the addition of indole (Supplementary Figure S1). Notably, the induction kinetics profiles differ between E. coli and C. necator-based biosensors, most likely due to variability in their growth rates and responses to the indole. E. coli growth was less affected by the indole, with no significant growth inhibition up to 1 mM, which is a commonly occurring concentration of this compound in the intestinal tract [4]. However, C. necator was more susceptible to the indole, with significant growth inhibition observed at concentrations of 0.125 mM and above. Figure 3 shows the dose-response curves representing the correlation between extracellular indole concentration and relative normalized fluorescence 4 h after inducer supplementation. Data indicate that biosensors can be tuned in the range of approximately 0.4 to 5 mM for a linear fluorescence output. supplementation. Data indicate that biosensors can be tuned in the range of approximately 0.4 to 5 mM for a linear fluorescence output. E. coli-and C. necator-based biosensors exhibited a dynamic range of 639.6-and 11.9fold in MM or 373.5-and 101.4-fold in a rich LB medium, respectively (Table 1). Notably, the observed variability of the dynamic range between E. coli and C. necator biosensors in a rich and minimal medium can be attributed to the indole's signaling and antimicrobial properties, which may have a differential effect on the cell growth and/or indole transport of diverse bacterial species. A lower dynamic range of E. coli/PpTrpI/PPP_RS00425 in a rich medium than that in MM can be caused by traces of indole or structurally similar compounds that might be present in LB and interact with the transcription factor TrpI. Whereas the significantly lower dynamic range of C. necator/PpTrpI/PPP_RS00425 in LB and MM than that of E. coli/PpTrpI/PPP_RS00425 is likely associated with the indole degradation observed in several bacterial species [4]. Indeed, a decrease in fluorescence output was observed at the later stage of the time course experiment (between 8 and 12 h after the addition of indole) compared to the E. coli biosensor (Supplementary Figure S1A,B), indicating consumption of indole by C. necator. A similar response attributable to the inducer's consumption has been previously reported for acrylate-inducible systems [38]. Moreover, a protein sequence similarity search using BLAST (www.ncbi.nlm.nih.gov; accessed on 1 March 2022) showed that C. necator H16 contains the gene cluster H16_RS17385-H16_RS17400 encoding enzyme homologues responsible for indole conversion into anthranilate as reported for Acinetobacter sp. O153 and Cupriavidus sp. SHE [39,40]. The low dynamic range of C. necator/PpTrpI/PPP_RS00425 in MM could also be associated with the catabolic repression proposed previously for aromatic compounds [37]. Notwithstanding the E. coliand C. necator-based biosensors exhibited a dynamic range of 639.6-and 11.9fold in MM or 373.5-and 101.4-fold in a rich LB medium, respectively (Table 1). Notably, the observed variability of the dynamic range between E. coli and C. necator biosensors in a rich and minimal medium can be attributed to the indole's signaling and antimicrobial properties, which may have a differential effect on the cell growth and/or indole transport of diverse bacterial species. A lower dynamic range of E. coli/PpTrpI/P PP_RS00425 in a rich medium than that in MM can be caused by traces of indole or structurally similar compounds that might be present in LB and interact with the transcription factor TrpI. Whereas the significantly lower dynamic range of C. necator/PpTrpI/P PP_RS00425 in LB and MM than that of E. coli/PpTrpI/P PP_RS00425 is likely associated with the indole degradation observed in several bacterial species [4]. Indeed, a decrease in fluorescence output was observed at the later stage of the time course experiment (between 8 and 12 h after the addition of indole) compared to the E. coli biosensor (Supplementary Figure S1A,B), indicating consumption of indole by C. necator. A similar response attributable to the inducer's consumption has been previously reported for acrylate-inducible systems [38]. Moreover, a protein sequence similarity search using BLAST (www.ncbi.nlm.nih.gov; accessed on 1 March 2022) showed that C. necator H16 contains the gene cluster H16_RS17385-H16_RS17400 encoding enzyme homologues responsible for indole conversion into anthranilate as reported for Acinetobacter sp. O153 and Cupriavidus sp. SHE [39,40]. The low dynamic range of C. necator/PpTrpI/P PP_RS00425 in MM could also be associated with the catabolic repression proposed previously for aromatic compounds [37]. Notwithstanding the above, similar indole concentrations are required to mediate the gene expression from inducible system PpTrpI/P PP_RS00425 in both types of biosensors and media with K m values ranging from 0.9 to 1.8 mM. Data are mean of three biological replicas. K m represents the indole concentration required to achieve the half-maximal activation of inducible system.

Specificity of Indole-Inducible Biosensors
To evaluate the specificity, the inducible system PpTrpI/P PP_RS00425 -based whole-cell biosensors were tested for induction with compounds that are structurally similar to indole. These included the indole metabolism product L-tryptophan (2) and phytohormones such as indole-3-acetic acid (3-IAA) (3), indole-3-propionic acid (3-IPA) (4), and indole-3-butyric acid (3-IBA) (5) ( Figure 4A). RFP fluorescence assays using these compounds were carried out. To this end, E. coliand C. necator-based biosensors harboring the inducible system PpTrpI/P PP_RS00425 (pPM0081) were cultivated in MM. The fluorescence and absorbance outputs were monitored 10 h after supplementation with each compound to a final concentration of 1 mM, and relative normalized fluorescence values were calculated as % of the normalized fluorescence obtained by using 1 mM indole. Results revealed that neither L-tryptophan (2), 3-IAA (3), 3-IPA (4), nor 3-IBA (5) induces the system PpTrpI/P PP_RS00425 with a confidence level of more than 99.0 and 98.5% for E. coliand C. necator-based biosensors, respectively ( Figure 4B,C). Therefore, it can be concluded that the developed biosensors exhibit a strong affinity for their primary ligand indole and are strictly specific.

Specificity of Indole-Inducible Biosensors
To evaluate the specificity, the inducible system PpTrpI/PPP_RS00425 -based whole-cell biosensors were tested for induction with compounds that are structurally similar to indole. These included the indole metabolism product L-tryptophan (2) and phytohormones such as indole-3-acetic acid (3-IAA) (3), indole-3-propionic acid (3-IPA) (4), and indole-3-butyric acid (3-IBA) (5) ( Figure 4A). RFP fluorescence assays using these compounds were carried out. To this end, E. coli-and C. necator-based biosensors harboring the inducible system PpTrpI/PPP_RS00425 (pPM0081) were cultivated in MM. The fluorescence and absorbance outputs were monitored 10 h after supplementation with each compound to a final concentration of 1 mM, and relative normalized fluorescence values were calculated as % of the normalized fluorescence obtained by using 1 mM indole. Results revealed that neither L-tryptophan (2), 3-IAA (3), 3-IPA (4), nor 3-IBA (5) induces the system PpTrpI/PPP_RS00425 with a confidence level of more than 99.0 and 98.5% for E. coli-and C. necator-based biosensors, respectively ( Figure 4B,C). Therefore, it can be concluded that the developed biosensors exhibit a strong affinity for their primary ligand indole and are strictly specific.  Furthermore, to evaluate if L-tryptophan (2), 3-IAA (3), 3-IPA (4), or 3-IBA (5) can competitively repress indole-mediated activation of the inducible system PpTrpI/P PP_RS00425 , the E. coli-based biosensor was subjected to the ligand competition assays in MM (Figure 4D). Fluorescence measurements using equimolar mixtures of ligands showed that none of the structurally similar compounds tested had a significant effect on the biosensor's response to indole, suggesting that they do not directly interfere with the binding of indole to the transcription factor TrpI and induction of PpTrpI/P PP_RS00425 .

Application of Whole-Cell Biosensor
Indole is produced from the aromatic amino acid L-tryptophan by tryptophanase (TnaA) [41]. Deletion of the tnaA gene has previously been shown to improve L-tryptophan biosynthesis in E. coli [42]. A protein sequence similarity search using BLAST revealed no homologue in C. necator H16. Besides, its genome contains the gene cluster H16_RS17385-H16_RS17400 encoding enzyme homologues responsible for indole consumption, as reported previously for Acinetobacter sp. O153 and Cupriavidus sp. SHE [39,40]. Therefore, we reasoned that the heterologous expression of the tnaA gene in this bacterium could allow us to monitor the transient indole synthesis and its accumulation in the cell.
The biosensor developed on the basis of C. necator was applied to screen bacterial strains possessing the functional tnaA gene. Cells were transformed with the plasmid pPM0083 ( Figure 1D). This construct contained the tnaA gene from E. coli (locus tag b3708), which was cloned downstream and transcriptionally coupled to the inducible gene expression system PpTrpI/P PP_RS00425 , enabling simultaneous synthesis and detection of indole. Subsequently, C. necator H16 cells harboring the plasmid pPM0083 were inoculated into the MM and were grown overnight. The resulting cultures were then subjected to the measurement of absorbance and fluorescence (Supplementary Figure S2). Figure 5 shows elevated fluorescence levels for strain A3, indicating the biosynthesis and accumulation of indole intracellularly and, therefore, the presence of functional gene tnaA, whereas low fluorescence levels in the case of strain A2 show either the absence of tryptophanase activity or rapid consumption of indole in this strain background. Further research will be required to identify the genetic differences between strains A2 and A3 and if A2 accumulated spontaneous mutations, resulting in an indole-negative phenotype. similar compounds with 1 + 2, 1 + 3, 1 + 4, and 1 + 5 representing equimolar mixtures of indole and structurally similar compounds (each at a concentration of 1 mM). Measurements were carried out in MM. Data are mean ± SD, n = 3. Asterisks indicate a statistically significant increase in fluorescence in the samples supplemented with inducer comparing to samples without the ligand (** p < 0.01, * p < 0.015, unpaired two-tailed t-test).
Furthermore, to evaluate if L-tryptophan (2), 3-IAA (3), 3-IPA (4), or 3-IBA (5) can competitively repress indole-mediated activation of the inducible system PpTrpI/PPP_RS00425, the E. coli-based biosensor was subjected to the ligand competition assays in MM ( Figure 4D). Fluorescence measurements using equimolar mixtures of ligands showed that none of the structurally similar compounds tested had a significant effect on the biosensor's response to indole, suggesting that they do not directly interfere with the binding of indole to the transcription factor TrpI and induction of PpTrpI/PPP_RS00425.

Application of Whole-Cell Biosensor
Indole is produced from the aromatic amino acid L-tryptophan by tryptophanase (TnaA) [41]. Deletion of the tnaA gene has previously been shown to improve L-tryptophan biosynthesis in E. coli [42]. A protein sequence similarity search using BLAST revealed no homologue in C. necator H16. Besides, its genome contains the gene cluster H16_RS17385-H16_RS17400 encoding enzyme homologues responsible for indole consumption, as reported previously for Acinetobacter sp. O153 and Cupriavidus sp. SHE [39,40]. Therefore, we reasoned that the heterologous expression of the tnaA gene in this bacterium could allow us to monitor the transient indole synthesis and its accumulation in the cell.
The biosensor developed on the basis of C. necator was applied to screen bacterial strains possessing the functional tnaA gene. Cells were transformed with the plasmid pPM0083 ( Figure 1D). This construct contained the tnaA gene from E. coli (locus tag b3708), which was cloned downstream and transcriptionally coupled to the inducible gene expression system PpTrpI/PPP_RS00425, enabling simultaneous synthesis and detection of indole. Subsequently, C. necator H16 cells harboring the plasmid pPM0083 were inoculated into the MM and were grown overnight. The resulting cultures were then subjected to the measurement of absorbance and fluorescence (Supplementary Figure S2). Figure 5 shows elevated fluorescence levels for strain A3, indicating the biosynthesis and accumulation of indole intracellularly and, therefore, the presence of functional gene tnaA, whereas low fluorescence levels in the case of strain A2 show either the absence of tryptophanase activity or rapid consumption of indole in this strain background. Further research will be required to identify the genetic differences between strains A2 and A3 and if A2 accumulated spontaneous mutations, resulting in an indole-negative phenotype.  Overall, these results show that the developed biosensor is applicable for the screening of microbial strains possessing tryptophanase activity. Moreover, it can potentially assist with monitoring the expression of indole metabolism-relevant genes, screening enzyme variants, and evaluating of their catalytic properties.

Discussion
Indole has attracted much attention due to its widespread presence in the natural environment and various wastes, including byproducts of petroleum, tobacco, farming, and mining industries [43]. Importantly, it is widely used in pharmaceuticals, polymers, dyes, agro, and other chemicals. TR-based whole-cell-bacterial biosensors have found application in environmental microbiology, synthetic biology, and biotechnology [30][31][32][33]. However, very little research has been performed on developing such tools for indole detection and monitoring. In this study, we developed and characterized biosensors that respond to indole in a concentration-dependent manner. To achieve this, we identified the inducible system PpTrpI/P PP_RS00425 in Pseudomonas putida KT2440. The construct containing the TR gene (trpI), promoter region, and rfp reporter gene was successfully assembled and introduced into E. coli and C. necator bacterial cells, resulting in whole-cell biosensors based on two different classes of bacteria, gammaproteobacteria and betaproteobacteria, respectively. Using these biosensors, we confirmed that TrpI is essential for the regulation of the P PP_RS00425 -dependent inducible gene expression system. It needs to be mentioned that the TrpI protein belongs to the LysR-type family of prokaryotic transcriptional regulators [17] and has previously been shown to bind I3GP, resulting in gene expression activation [18].
Furthermore, developed indole-inducible biosensors were thoroughly characterized by assessing their dynamic range and dose-dependence. A very high dynamic range of 639.6-fold was observed using an E. coli-based biosensor in MM. Despite the moderate level of sensitivity with K m values in the mM concentration range, both biosensors showed a high degree of specificity to indole when compared to L-tryptophan (2), 3-IAA (3), 3-IPA (4), and 3-IBA (5). The indole biosensor's utility was further exemplified by its application for tryptophanase-positive strain screening.
Recently, Herud-Sikimić and colleagues have shown that a genetically encoded biosensor for 3-IAA enables real-time monitoring of the uptake and disposal of auxin by individual cells and within cell sections in planta [44]. However, with few reports focused on indole derivatives, no indole-specific whole-cell biosensor has been described in the literature to date [44,45]. This study shows that our developed biosensor is specific to indole and can be used to measure the intracellular and extracellular concentrations of this compound. Future work can be focused on the application of this tool for the monitoring of indole concentration in the environment, industrial wastewater, or sewage [46,47]. In addition, it has the potential to be used for (a) screening enzymes and microbial strains exhibiting different L-tryptophan and indole specific catalytic properties, (b) evaluating flux in indole producing or degrading bacteria, (c) measuring indole transport and cross-species communication, and (d) indole metabolism-relevant metabolic engineering and biotechnology applications.

Bacterial Strains and Media
All strains employed in this study are listed in Supplementary Table S1. For cloning and plasmid propagation, E. coli Top10 (Thermo Fisher Scientific, Waltham, MA, USA) was used. To characterize indole-inducible systems and determine dose-response, dynamic range, and specificity of biosensors, RFP fluorescence assays were performed using E. coli With some modifications to standard protocol, E. coli transformations were performed by mixing 50 µL of formerly prepared chemically competent E. coli Top10 with 50 ng plasmid DNA, incubating on ice for 5 min, carrying out the heat shock at 42 • C for 1.5 min and subsequently incubating on ice for 5 min [48]. The recovery of transformed cells was achieved by incubation in 1 mL of LB medium (Thermo Fisher Scientific, Waltham, MA, USA) at 37 • C for 1 h. The transformants were plated on LB agar containing the respective antibiotic and incubated overnight at 37 • C.
Transformations of C. necator were carried out by adding 100 ng plasmid DNA to 100 µL of electrocompetent C. necator H16 in a pre-chilled electroporation cuvette (0.2 cm gap width, Bio-Rad). The mixture was incubated on ice for 5 min [49], following the electroporation at 2.5 kV using a Micropulser (Bio-Rad, Hercules, CA, USA). The recovery of the transformed cells was performed by incubation in 1 mL of SOC medium at 30 • C for 2 h. After recovery, the transformants were plated on LB agar containing the respective antibiotic and incubated at 30 • C for 2 days.
Plasmid pPM0083 was assembled to contain both indole-inducible system PpTrpI/ P PP_RS00425 and tnaA gene encoding the tryptophanase ( Figure 1D and Supplementary Figure S3C). The tnaA was PCR amplified using oligonucleotide primers P065 and P066 from E. coli MG1655 genomic DNA. PCR fragment was inserted into pPM0081 using BamHI restriction site and placing tnaA in direct orientation downstream to the rfp gene.

Fluorescence and Absorbance Measurements
For determination of normalized fluorescence, the RFP fluorescence and cell culture absorbance were measured at multiple time points. In assay preparation, freshly grown bacterial cells were used to inoculate 2 mL of LB medium or MM containing appropriate antibiotics in 50-mL conical centrifuge tubes, and cell cultures were grown overnight at 200 rpm and 30 • C. To obtain logarithmically growing cells, E. coli and C. necator overnight cultures were diluted to 0.05 and 0.1 OD 600 , respectively, in 5 mL of fresh medium with antibiotic. After the cells were grown to an OD 600 of 0.15-0.2, culture aliquots of 142.5 µL were transferred into 96-well plate (flat and clear bottom, black; Corning Incorporated, Corning, NY, USA) and 7.5 µL of indole was added to each mini culture to the required final concentration. Absorbance and fluorescence of mini cultures were measured using an Infinite ® M Nano+ microplate reader (Tecan, Grödig, Austria). Multiple time point measurements were taken every 10 min over the course of 12 h with a plate reader measuring RFP fluorescence at excitation and emission wavelengths set to 585 nm and 620 nm, respectively, and absorbance at 600 nm. The fluorescence gain factor was assigned to 120%. Obtained fluorescence and absorbance values were corrected by subtracting the autofluorescence and absorbance of the culture medium. The absolute normalized fluorescence was determined dividing the fluorescence value by corresponding absorbance value.
The dose-response of biosensors was determined by using single time-point values of fluorescence and absorbance, which were obtained from the multiple time-point measurements 4 h after supplementation of indole at different concentrations ranging from 0.02 to 2.5 mM. Similarly, the fluorescence and absorbance data for biosensor specificity analysis or tryptophanase-positive strain screening were obtained from the multiple time-point measurements as a single time-point 10 h after addition of ligand at 1 mM or unknown concentration, respectively. For all assays, the cultures of E. coli and C. necator were prepared as described above, and 142.5 µL of logarithmically growing cells, supplemented with 7.5 µL of ligand, were cultured in 96-well plate. The fluorescence and absorbance were measured using the same conditions and plate reader as indicated above.

Non-Linear Least-Squares Fitting and Dynamic Range Calculation
To perform a non-linear least-squares fit using the Hill function, the absolute normalized fluorescence values were analyzed and plotted as a function of ligand concentration using software GraphPad Prism 9 as follows: where RFP(I)-absolute normalized fluorescence value at given ligand concentration I; b max and b min -the maximum and minimum levels of reporter output in absolute normalized fluorescence units, respectively; h-the Hill coefficient; K m -the inducer concentration, corresponding to the half-maximal reporter's output. The dynamic range µ was calculated using Formula (2) as follows: