In situ Tracking of Exoenzyme Activity Using Droplet Luminescence Concentrators for Ratiometric Detection of Bacteria

We demonstrate a novel, rapid, and cost-effective biosensing paradigm that is based on an in situ visualization of bacterial exoenzyme activity using biphasic Janus emulsion droplets. Sensitization of the droplets toward dominant extracellular enzymes of bacterial pathogens is realized via selective functionalization of one hemisphere of Janus droplets with enzyme-cleavable surfactants. Surfactant cleavage results in an interfacial tension increase at the respective droplet interface, which readily transduces into a microscopically detectable change of the internal droplet morphologies. A macroscopic fluorescence read-out of such morphological transitions is obtained via ratiometrically recording the angle-dependent anisotropic emission signatures of perylene-containing droplets from two different angles. The optical read-out method facilitates detection of marginal morphological responses of polydisperse droplet samples that can be easily produced in any environment. The performance of Janus droplets as powerful optical transducers and signal amplifiers is highlighted by rapid (<4 h) and cost-effective antibody and DNA-free identification of three major foodborne pathogens, with detection thresholds of below 10 CFU mL–1 for Salmonella and <102 to 103 CFU mL–1 for Listeria and Escherichia coli.


Instruments:
Nuclear magnetic resonance spectra were carried out using a Bruker Advance (400MHz) spectrometer.The angle-dependent droplet emission was recorded using an Avantes (model: StarLine AvaSpec-ULS2048CL-EVO-RS) spectrometer.Interfacial tension measurements were carried out using a drop shape ananalyzer tensiometer (DSA10-MK2, Krüss) in the pendant drop setting and images were recorded using a CCD camera.Bacteria samples were incubated using incubator from Binder (model BD 23) set to 37 o C.

Droplet preparation:
Janus droplets comprising a 1:1 volume ratio of diethylbenzene (containing 2.5mM perylene) and HFE7500 were prepared by batch one-step emulsification via an established thermal phaseseparation approach. [3]Emulsion droplets prepared by batch-scale vortex mixing were polydisperse in size displaying an average diameter of 55.7 µm ± 51.9 µm.In brief, the oil mixture and the surfactant-containing aqueous continuous phase were separately heated above the critical solution temperature of the two oils.Subsequently, 10 vol.% of the homogenous oil mixture were transferred to the aqueous solution and vortexed at 2100 rppm for 10 sec Vortex Genie 2, Scientific Industries).After emulsification, single phase droplets were allowed to settle and cool back to room temperature, which induced phase separation of the two oils inside the droplets.The morphology of droplets is exclusively determined by the force balance of interfacial tensions acting at the individual interfaces that can be fine-tuned and controlled by adjusting the surfactant composition in the continuous phase.Emulsion droplets are kinetically stabilized and are therefore prone to aging and mechanical agitation.In our experiments we therefore employed droplets freshly prepared on the same day.However, we did not observe any noticeable changes in droplet morphology, composition, or size distribution when stored on the benchtop for up to 7 days.

Imaging and microscopy:
Side-view micrographs of Janus droplets were recorded with a BRESSER MicroCam SP 3.1 microscope camera (and MicroCamLabII software), using a custom-made side-view microscopy setup, comprising of a 200mm tube lens (Thorlabs) and planar optical microscopy lens (Olympus).Emission studies were carried out using a custom-built tilting fluorescence microscope with a planar objective (10x Olympus), a tube lens (200mm, Thorlabs), camera (Allied vision) and a fluorescence cube (Thorlabs) containing a UV filter (Thorlabs MD416) and MDF-BFP dichroic filter.The setup is designed for tracking perylene emission between 425-475 nm, via filtering out the excitation light using the fluorescence cube.

Contact angle analysis
In order to analyze contact angles of gravity-aligned Janus droplets, sideview micrographs were recorded.From the obtained micrographs, the triple phase contact line was used to determine droplet's contact angle, as a quantitative description of droplet morphology.To this end, the distance between the droplet radius and the radius of the internal curvature was recorded and used to determine the contact angle, as well as the surface area by employing Neumann construction along with the law of cosines. [4]Additionally, a correction factor Rreal = nmedium/nouterRimage was implemented due to the refractive index contrast of the individual emulsion phases.All the contact angles were determined using imaging software Fiji.

Ratiometric angle-dependent emission detection setup
Prepared droplets were placed on the sample holder (Thermo-Fisher Scientific Invitrogen Attofluor Cell Chamber) by first deposing 1 mL of surfactant followed by a deposition of a droplet monolayer (20 µL) in the center of the sample holder.The sample holder was then placed on a RPS-SMA sample stage.Optical fibers were fixed on top of the sample holder (13mm above the stage) using adjustable fiber optic probe stand (Thorlabs, RPA-SMA).One of the fibers (Thorlabs, 400µm) was placed at 45 o to the sample holder and used to record sideway emission.A second bifurcated fiber (Thorlabs, 400µm) was placed vertically (at 0 o ) above the sample holder and attached to both the spectrometer and the light source (395nm LED light from Thorlabs).Light output was recorded using a spectrometer from Avantes (model: StarLine AvaSpec-ULS2048CL-EVO-RS). Morphology-dependent changes in perylene emission were recorded using both fibers 0 o and 45 o and recording changes in the 472nm emission peak of perylene.Subsequently, the ratio of emission intensity recorded by the fiber placed at 0 o over emission intensity from the fiber placed at 45 o was calculated and then normalized, where value of 1 corresponded to fully encapsulated H/F/W morphology obtained by preparing droplets in a pure Zonyl FS-300 solution.

Bacteria culture:
Salmonella enterica: S. enterica DSM 554 was purchased from DSMZ.The culture was prepared in tryptic soy broth (TBS), where 5ml of TSB was inoculated with S. enterica from a frozen glycerol stock and incubated at 37 o C overnight.Late-logarithmic phase was obtained, with an OD600 value reaching ca.0.8.The sample was diluted to an OD600 of 0.4 and incubated for ca. 3 hours until the desired OD of 0.65 was reached, which corresponds to app. 10 9 CFU/ml (calculated from serial agar plate dilution assays).Subsequently, serial dilutions in TBS were carried out, giving concentrations between 1-10 8 CFU/ml.From that, 100 µL of each diluted bacteria solution was transferred to 900 µL of surfactant 1 and zonyl-FS300 solution in PBS containing 300 µL of 1wt% of surfactant 1and 500 µL of 0.2wt% Zonyl and 100 µL of PBS.Samples were then placed in the incubator at 37 o C. Subsequently, samples were placed on a heating stage along with separate vial containing 1:1 volume mixture of DEB with 2.5mM perylene and HFE7500.Vials were heated above the Tc of the oil mixture and 100 µL of oil mix was transferred into surfactant samples.Samples were then vortexed at 2100rpm for 10s ale left to cool down to room temperature.Changes in droplet morphology were compared with blank samples, where 100 µL of bacterial sample was replaced with 100 µL of PBS.Each sample was measured after different time points: 1h, 2h, 4h and 6h.The time required for surfactant cleavage by commercial enzyme sensing cannot be directly translated into time needed for bacteria to cleave the surfactant, because bacterial enzymes are continuously produced -meaning that the amounts of enzymes will vary between samples.Our method relies on continuous enzyme production by live bacteria.Thus, the longer incubation time, the more bacteria in the sample, the more enzymes are produced leading to larger extent of surfactant cleavage.Morphology changes were then recorded using a portable phone microscope and optical setup.All data points were measured in triplicates.

Listeria monocytogenes:
L. monocytogenes DSM 20600 was purchased with DSMZ.The cultures were prepared in brain heart infusion (BHI) broth from a frozen glycerol stock.Culture was grown at 37 o C for 20 hours until desired OD600 of 0.65 was reached, corresponding to 8*10 8 CFU/ml. [5]bsequently serial dilutions ranging from 10 2 CFU/mL to 10 8 CFU/mL were prepared by placing broth containing bacteria to appropriate amounts BHI broth.From that, 100 µL of each diluted bacteria solution was transferred to 900 µL of surfactant 2 and zonyl-FS300 solution in PBS containing 300 µL of 1wt% of surfactant 2and 500 µL of 0.2wt% Zonyl and 100 µL of PBS.Samples were then placed in the incubator at 37 o C. Subsequently, samples were placed on a heating stage along with separate vial containing 1:1 volume mixture of DEB with 2.5mM perylene and HFE7500.Vials were heated above the Tc of the oil mixture and 100 µL of oil mix was transferred into surfactant samples.Samples were then vortexed at 2100rpm for 10s ale left to cool down to room temperature.Changes in droplet morphology were compared with blank samples, where 100 µL of bacterial sample was replaced with 100 µL of PBS.Each sample was measured after different time points: 1h, 2h, 4h and 6h.The time required for surfactant cleavage by commercial enzyme sensing cannot be directly translated into time needed for bacteria to cleave the surfactant, because bacterial enzymes are continuously produced -meaning that the amounts of enzymes will vary between samples.Our method relies on continuous enzyme production by live bacteria.Thus, the longer incubation time, the more bacteria in the sample, the more enzymes are produced leading to larger extent of surfactant cleavage.Morphology changes were then recorded using a portable phone microscope and optical setup.All data points were measured in triplicates.

Escherichia coli
E.coli DH5α was purchased from New England Biolabs.The culture was grown from a frozen glycerol stock in lysogeny broth (LB).The culture was incubated at 37 o C overnight.The next day the culture was diluted to an OD600 of 0.6 and allowed to grow to an OD600 of 1.0, which corresponds to app. 10 9 CFU/ml. [6]Subsequently serial dilutions ranging from 10 2 CFU/mL to 10 8 CFU/mL were prepared by placing broth containing bacteria to appropriate amounts LB broth.From that, 100 µL of each diluted bacteria solution was transferred to 900 µL of surfactant 3 and zonyl-FS300 solution in PBS containing 400 µL of 1wt% Surfactant 3and 500 µL of 0.2wt% Zonyl.Samples were then placed in the incubator at 37 o C. Subsequently, samples were placed on a heating stage along with separate vial containing 1:1 volume mixture of DEB with 2.5mM perylene and HFE7500.Vials were heated above the Tc of the oil mixture and 100 µL of oil mix was transferred into surfactant samples.Samples were then vortexed at 2100rpm for 10s and left to cool down to room temperature.Changes in droplet morphology were compared with blank samples, where 100 µL of bacterial sample was replaced with 100 µL of PBS.Each sample was measured after different time points: 1h, 2h, 4h and 6h.The time required for surfactant cleavage by commercial enzyme sensing cannot be directly translated into time needed for bacteria to cleave the surfactant, because bacterial enzymes are continuously produced -meaning that the amounts of enzymes will vary between samples.Our method relies on continuous enzyme production by live bacteria.Thus, the longer incubation time, the more bacteria in the sample, the more enzymes are produced leading to larger extent of surfactant cleavage.Morphology changes were then recorded using a portable phone microscope and optical setup.All data points were measured in triplicates.  of one hemisphere of Janus droplets and that variations on the order in 0.5 mN m -1 suffice to induce morphological changes on the order of 20° (that suffice for a full optical response), 7 the theoretical limits of detection of the latter can become very low, particularly when operated below the cmc of the target surfactants.

1 H NMR cleavage studiesenzymatic cross tests:
Surfactants kept at concentration of 0.1wt% and enzymes at 1U/ml.Progress of the cross cleavage was measured after different time points.

Detection of enzymatic activity using Janus double emulsions
For all cleavage studies, enzymes were added and incubated with an appropriate surfactant mixture prior to emulsification to avoid any additional interactions between droplet oils and enzymes.The same approach was employed throughout bacteria detection studies.

Porcine liver esterase (PLE):
experiments were conducted at room temperature in PBS at pH 7.4.All solutions, including surfactant stocks and enzyme stocks were prepared using PBS.
Experiments have been carried out with a constant surfactant concentration (300 µL of 1wt% of surfactant 1and 500 µL of 0.2wt% Zonyl and 100 µL of PBS) against varying PLE concentrations (100 µL) after 2 hours or different reaction times using 0.025 U/mL of final PLE concentration.After enzyme addition, samples were placed on a shaker to ensure correct stirring and even enzyme distribution throughout the sample.Subsequently, samples were placed on a heating stage along with separate vial containing 1:1 volume mixture of DEB with 2.5mM perylene and HFE7500.Vials were heated above the Tc of the oil mixture and 100 µL of oil mix was transferred into surfactant samples.Samples were then vortexed at 2100rpm for 10s ale left to cool down to room temperature.Changes in droplet morphology were compared with blank samples, where 100 µL of enzyme solution was replaced with 100 µL of PBS.
Morphology changes were then recorded using a sideview microscope and optical setup.All data points were measured in triplicates.β-glucosidase: experiments were conducted at room temperature in PBS at pH 7.4.All solutions, including surfactant stocks and enzyme stocks were prepared using PBS.
Experiments have been carried out with a constant surfactant concentration (300 µL of 1wt% of surfactant 2and 500 µL of 0.2wt% Zonyl and 100 µL of PBS) against varying βglucosidase concentrations after 2 hours or different reaction times using 1 U/mL of final βglucosidase concentration.After enzyme addition, samples were placed on a shaker to ensure correct stirring and even enzyme distribution throughout the sample.Subsequently, samples were placed on a heating stage along with separate vial containing 1:1 volume mixture of DEB with 2.5mM perylene and HFE7500.Vials were heated above the Tc of the oil mixture and 20 100 µL of oil mix was transferred into surfactant samples.Samples were then vortexed at 2100rpm for 10s ale left to cool down to room temperature.Changes in droplet morphology were compared with blank samples, where 100 µL of enzyme solution was replaced with 100 µL of PBS.Morphology changes were then recorded using a sideview microscope and optical setup.All data points were measured in triplicates.
β-galactosidase: experiments were conducted at room temperature in PBS at pH 7.4.All solutions, including surfactant stocks and enzyme stocks were prepared using PBS.
Experiments have been carried out with a constant surfactant concentration (400 µL of 1wt% Surfactant 3and 500 µL of 0.2wt% Zonyl) against varying β-galactosidase concentrations after 2 hours or different reaction times using 2.5 U/mL of final β-galactosidase concentration.
After enzyme addition, samples were placed on a shaker to ensure correct stirring and even enzyme distribution throughout the sample.Subsequently, samples were placed on a heating stage along with separate vial containing 1:1 volume mixture of DEB with 2.5mM perylene and HFE7500.Vials were heated above the Tc of the oil mixture and 100 µL of oil mix was transferred into surfactant samples.Samples were then vortexed at 2100rpm for 10s ale left to cool down to room temperature.Changes in droplet morphology were compared with blank samples, where 100 µL of enzyme solution was replaced with 100 µL of PBS.Morphology changes were then recorded using a sideview microscope and optical setup.All data points were measured in triplicates.µL of surfactant solution 330 µL of 1wt% of surfactant 1and 570 µL of 0.2wt% Zonyl.The same procedure was carried out for surfactant 2: 2.5U/mL of enzyme solution in PBS solution (100 µL) was added to 900 µL of surfactant solution containing 420 µL of 1wt% of surfactant 2and 480 µL of 0.2wt% Zonyl.Samples were then placed on a shaker for 2 hours.
Subsequently, samples were placed on a heating stage along with separate vial containing 1:1 volume mixture of DEB with 2.5mM perylene and HFE7500.Vials were heated above the Tc of the oil mixture and 100 µL of oil mix was transferred into surfactant samples.Samples were then vortexed at 2100rpm for 10s and left to cool down to room temperature.Changes in droplet morphology were compared with blank samples, where 100 µL of enzyme solution was replaced with 100 µL of PBS.Morphology changes were then recorded using a sideview microscope and optical setup.All data points were measured in triplicates.

Enzyme kinetics:
To calculate concentration of surfactant within the emulsion system upon cleavage, droplet contact angle was plotted against HC surfactant concentration used.Plot for each surfactant shows linear behavior, therefore y=mx+b equation can be used, where x is a surfactant concentration assigned to a given droplet morphology.Those values were then applied to calculate enzyme kinetics and rate constants.
Given that enzymes exhibit first order kinetics, rate constant of the enzymatic cleavage, k, can be calculated by following equation: Where -k is the slope, t is the investigated reaction time, [A]0 (µmol/mL) is a starting surfactant concentration and [A]t (µmol/mL) is surfactant concentration after given incubation time.
Linear fit was placed and rate constants after 2 H of cleavage were compared.Concentration "spilled" (CFU/mL)

Fig. S27 1 4 . 2 4 . 3 1 Fig
Fig. S27 Concentration dependence study showing a linear behavior between the decrease in the contact angle of droplets stabilized with surfactant 3 and β-galactosidase concentration.

Fig. S32
Fig. S32 Swab test results employing surfactant 3 to detect L. monocytogenes, where 0 corresponds to blank sample where only 100 µL of growth medium (broth) was added to the solution.