Bacterial strains
E. coli strain DH10β was used for all cloning and plasmid preparations. E. coli strain BL21 Star (DE3) ΔlacIZYA was created by lambda red recombination60 and used for in-house cell-free lysate preparation. Genomic DNA from E. coli strains DH10β and BL21 Star (DE3) were used as negative controls for target-specific amplification of Stx1 and Stx2 triggers (Supplementary Fig. 19). Genomic DNA from B. thetaiotaomicron (ATCC 29148D) and STEC O157:H7 (ATCC 51657GFP) were used for testing detection of pathogenic bacteria.
Genetic parts assembly and plasmid preparation
Sequences of all parts used in this study are provided in the Source Data, under file name Sequence Information. DNA oligonucleotides for cloning and sequencing were synthesized by Eurofins Genomics. Partial sequences for small molecule sensors and toehold switches were obtained from previously published sequences and were synthesized either as gene fragments or ssDNA-annealed oligonucleotides from Eurofins Genomics. Plasmids expressing regulators and reporter proteins were cloned using either Gibson Assembly61 or blunt end ligation into plasmid backbone pJL1. Assembled constructs were transformed into DH10β cells, and isolated colonies were grown overnight in LB with antibiotics. Plasmid DNA from overnight cultures was purified using EZNA miniprep columns (OMEGA Bio-Tek). Plasmid sequences were verified with Sanger DNA sequencing (Eurofins Genomics).
Plasmid DNA used for all cell-free and protocell reactions was purified from EZNA midiprep columns (OMEGA Bio-Tek), followed by isopropanol and ethanol precipitation. The purified DNA pellet was reconstituted in elution buffer, measured on a Nanodrop 2000 for concentration, and stored at −20 °C.
Genomic DNA for B. thetaiotaomicron (ATCC 29148D) used for pathogen detection was purchased from American Type Culture Collection (ATCC). Genomic DNA for STEC O157:H7 (ATCC 51657GFP) used for pathogen detection was obtained from an overnight culture grown in tryptic soy broth supplemented with 1% glucose at 37 °C. DNA was extracted using an Invitrogen PureLink Microbiome DNA Purification Kit (A29790). Genomic DNA for DH10β and BL21 Star (DE3) were obtained from a 5 mL overnight culture grown in LB medium and were extracted using Quick-DNA Plus Kit (Zymo Research) according to the manufacturer’s protocol.
Preparation of in-house cell-free lysate
Cellular lysate for all experiments was prepared as described by Sun et al.62 with a few protocol modifications. Briefly, BL21 Star (DE3) ΔlacIZYA cells were grown in 2×YTP medium at 37 °C and 220 rpm to an optical density (OD) between 1.5–2.0, corresponding to the mid-exponential growth phase. Lysate prepared for toehold switch expression had an additional IPTG (0.4 mM) induction step when the OD reached 0.4 to activate expression of T7 RNA polymerase, creating a T7 RNAP-enriched lysate. Cells were centrifuged at 2700 × g and washed via resuspension with S30A buffer (50 mM tris, 14 mM magnesium glutamate, 60 mM potassium glutamate, 2 mM dithiothreitol, and pH-corrected to 7.7 with acetic acid). These centrifugation and wash steps were repeated twice for a total of three S30A washes. After the final centrifugation, the wet cell mass was determined, and cells were resuspended in 1 mL of S30A buffer per 1 g of wet cell mass. The cellular resuspension was divided into 1 mL aliquots. Cells were lysed using a Q125 sonicator (Qsonica) at a frequency of 20 kHz and 50% of amplitude. Cells were sonicated on ice with cycles of 10 s on and 10 s off, delivering approximately 200–250 J, at which point the cells appeared visibly lysed. An additional 4 mM dithiothreitol was added to each tube, and the sonicated mixture was then centrifuged at 12,000 × g and 4 °C for 10 min. After centrifugation, the supernatant was removed, divided into 0.5 mL aliquots, and incubated at 37 °C and 220 rpm for 80 min. After this runoff reaction, the cellular lysate was centrifuged at 12,000 × g and 4 °C for 10 min. The supernatant was removed and loaded into a 10 kDa molecular weight cutoff dialysis cassette (Thermo Fisher). Lysate was dialyzed in 1 L of S30B buffer (14 mM magnesium glutamate, 60 mM potassium glutamate, 1 mM dithiothreitol, and pH-corrected to 8.2 with tris) at 4 °C for 3 h. Dialyzed lysate was removed and centrifuged at 12,000 × g and 4 °C for 10 min. The supernatant was removed, aliquoted, flash-frozen in liquid nitrogen, and stored at −80 °C for future use.
Cell-free reactions
Cell-free reactions were assembled as described by Kwon and Jewett63. Briefly, reaction mixtures were composed of 27 v/v% of in-house prepared lysate, 2 mM each proteinogenic amino acid, 1.2 mM ATP, 0.85 mM each of GTP, CTP, and UTP, 0.2 mg/mL tRNA, 0.27 mM CoA, 0.33 mM NAD, 0.068 mM folinic acid, 1.5 mM spermidine, 33 mM PEP, 130 mM potassium glutamate, 10 mM Ammonium glutamate, 12 mM magnesium glutamate, 4 mM sodium oxalate, and specified concentrations of plasmids (described in Supplementary Table 1), RNA triggers, and small molecules. For experiments with RNA triggers (Supplementary Fig. 6), Rnase Inhibitor Murine (New England Biolabs) was added to the bulk phase at 0.5 v/v%. Each assembled cell-free reaction was 10 µL in volume and placed in a black-bottomed 384-well plate (Greiner Bio-One) and incubated at 37 °C for 3 h for GFP expression. A clear adhesive film was used to cover the plate and prevent evaporation.
Protocell CFE reactions
Polymers used to establish ATPS-based membrane-less protocells were prepared by dissolving either 400k Ficoll, 500k dextran, or 35k PEG into nuclease-free water. The bulk phase at the time of preparation and before the addition of Ficoll or dextran protocells consisted of 5 v/v% of 35k PEG, 1× concentration of all reagents added for cell-free reactions (excluding lysate), specified concentrations of small molecules or nucleic acids, and nuclease-free water to a final volume of 200 µL for the 4-plex system or 100 µL for the 9-plex system. For experiments with RNA triggers (Fig. 4, Supplementary Figs. 4 and 6), Rnase Inhibitor Murine (New England Biolabs) was added to the bulk phase at 0.5 v/v%. For experiments in Fig. 5 and Supplementary Figs. 11–13, RNase Inhibitor Murine (New England Biolabs) was added to a concentration of 1.5 v/v% in the bulk phase to decrease serum Rnase activity. For colorimetric output in cell-free reactions, color substrates were added to the bulk phase to a final concentration of 0.6 mg/mL for CPRG or 0.2 mg/mL for X-gal.
Concentrations for individual plasmid sensors and reaction additives used in each figure are provided in Supplementary Table 1. Briefly, each protocell sensor consisted of 10 v/v% Ficoll or 5 v/v% dextran polymers, 27 v/v% cell-free lysate, 1× concentration of cell-free reagents, specific concentrations of plasmid DNA, and water to a final volume of 2 µL for the 4-plex system or 1 µL for the 9-plex system. The assembled protocell solution was then vortexed at a medium-high setting to ensure homogenous mixing. Protocells for detection of linear DNA also contained 2 µM of χDNA to protect linear DNA against endonucleases present in the CFE lysate, except for the Stx1 switch in Supplementary Fig. 10 which used 10 µM of the GamS protein (Arbor Bioscience). Reactions with χDNA had slightly higher fold-change compared to reactions with GamS for linear DNA protection (Supplementary Fig. 20).
To assemble the protocell arrays, a bulk phase solution containing specified concentrations of targets was first pipetted into the custom-made microwell plate (PHASIQ) to fill up each microwell. Protocell droplets were then pipetted into the bulk phase solution at their designated micro-basins. Unless otherwise noted, microwells containing assembled protocell arrays were incubated at 37 °C for 3 h before imaging on a ChemiDoc MP (Bio-Rad) imaging system. A clear adhesive film was used to cover the plate and prevent evaporation.
Data acquisition and analysis
For cell-free reactions, endpoint GFP readings were taken using a plate reader (Synergy4, BioTek) with its companion Gen 5 software. The excitation and emission wavelengths were 485 and 528 nm, with a gain setting of 70. For experiments in Supplementary Figs. 6–8, the gain was reduced to 60 due to signal overflow. Data acquired from Gen 5 software was exported to Microsoft Excel files for further analysis. For protocell array reactions with fluorescent reporters, the ChemiDoc MP imaging system (Bio-Rad) was used for fluorescent plate imaging. Image Lab software (Bio-Rad) was used for image collection with settings of 0.5 s exposure time, grayscale image color, 530/28 filter for GFP detection, and Blue Epi illumination as a light source. An image transform procedure was uniformly applied to all images in Image Lab (with high, low, and gamma values of 10000, 0, and 1, respectively) before exporting files for analysis. For image analysis, each image was first converted to an 8-bit grayscale image. An image processing software, Fiji, was used to manually extract signal intensities from each 1.5 mm diameter microwell (pixel size 50) for data analysis. The signal intensities were recorded and analyzed using Microsoft Excel. For colorimetric protocell array reactions, all pictures were taken with an iPhone X (Apple) in a light-controlled setting. Adobe Photoshop was used to crop individual wells for data presentation. A brightness adjustment was uniformly applied to all colorimetric ATPS photos to make them better resemble appearance to the naked eye. Untouched photos are provided with Source Data.
Trigger preparation
DNA encoding each trigger RNA used in experiments was either amplified from cloned plasmid or from genomic DNA of specified species of bacteria by PCR with Q5 DNA polymerase (New England Biolabs). Primers were designed to create linear DNA with a T7 promoter and additional protective sequences on the 5′ and 3′ ends of the linear template. Sequences for primers used to amplify triggers from DNA template or genomic DNA are provided in Supplementary Table 2. After PCR amplification, all products were run on a 2 w/v% agarose gel to verify successful amplification of targets and then purified using a PCR purification kit (Omega Bio-Tek). The prepared linear DNA was either directly used in cell-free and protocell reactions or used as a template for in vitro transcription.
RNA triggers were transcribed from linear DNA template using T7 polymerase according to the manufacturer’s protocol (New England Biolabs). Following RNA synthesis, Dnase I (Zymo Research) was added to degrade the linear DNA template. The RNA products were then purified using an RNA Clean and Concentrator kit (Zymo Research) according to the manufacturer’s protocol. Following purification, RNA concentration was measured on a Nanodrop 2000, sub-aliquoted to reduce freeze-thaw cycles, and stored −20 °C.
STEC toehold switch development
Toehold switches targeting gene sequences of Shiga toxin proteins (Stx1 and Stx2) in STEC O157:H7 were designed using NUPACK with series B toehold switch design29 and cloned into a pJL1 plasmid containing a GFP reporter. Trigger sequences were synthesized by Eurofins Genomics as gene fragments containing a T7 promoter and 20–35 bp of protective sequences before and after the actual trigger sequence, and were cloned onto the pJL1 plasmid backbone to facilitate sensor screening. Trigger/sensor pairs were then tested in cell-free reactions containing 2.5 nM toehold switch sensor-GFP reporter plasmid and 5 nM of trigger plasmid to verify successful sensor activation and orthogonality (Supplementary Fig. 21). Following trigger/sensor pair validation, primers used to amplify different trigger sequences from genomic DNA were verified for specificity toward their targets using PCR reactions with nontarget templates (Supplementary Fig. 19).
Serum processing
Pooled human serum was purchased from Corning (Corning, NY). Endogenous zinc was removed from serum through Chelex-100 treatment. In total, 1 g of Chelex-100 resin was added to 100 mL of serum, and the mixture was vigorously stirred for 2 h at room temperature. Resin was then isolated from the serum through centrifugation followed by syringe filtering. All serum samples were aliquoted to minimize freeze-thaw cycles and stored at −20 °C until use.
Measurement of successful zinc removal from serum was conducted at the University of Georgia Laboratory for Environmental Analysis (Supplementary Fig. 14). Samples were digested with concentrated acid and analyzed on ICP-MS according to EPA method 3052.
Lyophilization
Protocell sensors (1 µL) containing 55 v/v% lysate, specified concentrations of plasmid DNA (Supplementary Table 1), and 10 v/v% Ficoll were spotted into micro-basins, and the plate was stored at −80 °C to freeze for at least 5 h. Bulk phase solutions (360 µL per bulk phase condition, 120 µL per technical replicate) containing 5 v/v% PEG, 0.2 mg/mL X-gal, and 1× cell-free energy buffer were prepared in 2 mL Eppendorf tubes and stored at −80 °C to freeze for at least 5 h. Frozen plates and bulk phase solutions were transferred to a prechilled Labconco Fast-Freeze flask as quickly as possible to prevent reagent thawing. The flask was connected to a Labconco benchtop lyophilizer and lyophilized at −50 °C and 0.05 mbar overnight (>14 h).
Freeze-dried reactions were taken out of the lyophilizer the following day. For the bulk phase, water mixed with 50 nM of specified bacterial triggers was used to reconstitute the lyophilized bulk phase solution to 360 µL (120 µL per technical triplicate). One microliter of water was used to rehydrate the freeze-dried protocells, and the plate was placed at room temperature for 5 min for protocells to congeal, thereby reducing the potential for protocell sensors to float into neighboring wells during rehydration. Reconstituted bulk phase solutions were added to microwells before incubating at 37 °C. A clear adhesive film was used to cover the plate and prevent evaporation.
Reporting summary
Further information on research design is available in the Nature Research Reporting Summary linked to this article.
