Contributed by Sidney Altman, January 7, 2017 (sent for review September 15, 2016; reviewed by Robert S. Phillips and Israel Silman)
Biocompatible microfluidic double water-in-oil-in-water emulsion (MDE) enables in-droplet cultivation of different living species. The combination of droplet-generating machinery with FACS followed by next-generation sequencing and liquid chromatography-mass spectrometry analysis of the secretomes of encapsulated organisms yielded detailed genotype/phenotype descriptions. The MDE–FACS platform we developed enabled highly sensitive single-cell selection of predesigned activity and exploration of pairwise interactions between target and effector cells without interference from other microbiota species.
Ultrahigh-throughput screening (uHTS) techniques can identify unique functionality from millions of variants. To mimic the natural selection mechanisms that occur by compartmentalization in vivo, we developed a technique based on single-cell encapsulation in droplets of a monodisperse microfluidic double water-in-oil-in-water emulsion (MDE). Biocompatible MDE enables in-droplet cultivation of different living species. The combination of droplet-generating machinery with FACS followed by next-generation sequencing and liquid chromatography-mass spectrometry analysis of the secretomes of encapsulated organisms yielded detailed genotype/phenotype descriptions. This platform was probed with uHTS for biocatalysts anchored to yeast with enrichment close to the theoretically calculated limit and cell-to-cell interactions. MDE–FACS allowed the identification of human butyrylcholinesterase mutants that undergo self-reactivation after inhibition by the organophosphorus agent paraoxon. The versatility of the platform allowed the identification of bacteria, including slow-growing oral microbiota species that suppress the growth of a common pathogen, Staphylococcus aureus, and predicted which genera were associated with inhibitory activity.
The ultrahigh-throughput (1, 2) technique of screening (uHTS) in a double emulsion was applied in directed enzyme evolution (3, 4) to investigate the idea that a universal genotype–phenotype linkage was provided by compartmentalization. Artificial compartments of double emulsions were produced with high polydispersity by shear stress (5, 6), which significantly decreased the portion of uniform droplets and thereby reduced the sensitivity and the maximal sorting rate. By contrast, sophisticated custom sorters demonstrated the screening of precise monodisperse droplets of water-in-oil emulsions generated by microfluidic technology (1, 7). However, it is not always convenient to use custom devices, and the use of oil as a continuous phase limits the sorting rate. Alternatively, compartmentalization in microfluidic double emulsion (MDE) enables uHTS of >10,000 events/s using commercially available cell sorters (8, 9). Furthermore, biocompatible oil and water phases provide viability and proliferation of Escherichia coli cells (10) inside the microenvironment of a double emulsion.
Here, we propose an MDE–FACS platform that combines the benefits of previously reported systems based on compartmentalization in MDE and FACS selection together with modern omics (Fig. 1). We succeeded in assembling this platform using commercially available parts, which included straightforward microfluidics (Fig. S1) for MDE generation, multiparametric FACS for uHTS, next-generation sequencing (NGS) for bioinformatic predictions, and mass spectrometry for proteome and secretome analysis. We demonstrated this idea with several single-cell methods (Fig. S2), including the selection of different biocatalytic activities, screening enzymes with different levels of the same activity, de novo creation of enzymes with artificial activity (Fig. 2), and investigation of bacterial cell-to-cell interactions (Fig. 3).
Principal scheme of the MDE–FACS technique. Compartmentalization of a library of species with specific fluorescence-generating machinery in MDE using emulsification in microfluidic chips enabled single-cell probing of a targeted function. Only specific phenotypes (indicated by the yellow star) fit and activated the machinery in droplets, thus leading to the production of fluorescent substances. Because the droplet volume was uniform, similar concentrations and conditions resulted in a narrow fluorescence distribution that corresponded to the same activity. The fluorescent signal was registered by a conventional cell sorter that separated the variants that activated the machinery (activators) from those that did not (trash). Unculturable species were analyzed by the direct sequencing of MDE. If the selected activators could be cultivated in vitro, they were regenerated from the droplets, grown, and analyzed through the combination of classical methods (kinetics, LC-MS, sequencing, and others). W/O/W, double water-in-oil-in-water emulsion.
Screening of biocatalysts anchored to the yeast surface using MDE–FACS. (A) Compartmentalization of active (RFP-positive) and inactive (nonfluorescent) yeast cells with fluorogenic substrate. After the mixture of active and inactive cells was encapsulated, the fluorescent product accumulated solely inside the droplets with active cells, which were selected using FACS. (B) Visualization of biochemical reaction by merging the signals of green fluorescence (reaction product), red fluorescence (reporter protein), and the visible light image. (Scale bar, 100 μm.) (C) The plot represents enrichment efficiency depending on the dilution of active cells with inactive cells after one round of MDE–FACS. (D) Enrichment efficiency of MDE–FACS for active cells displaying different biocatalysts (DNase I, EK, and BChE) and inactive cells (Fab) mixed in a 1:1:1:100 ratio. (E) Selection of BChE mutants with different levels of activity from the BChE library before selection (Lib) and after the selection using gates G1–G3. (Inset) Overlay of the MDE–FACS spectra of droplets with Fab, BChE library, and WT BChE. The gates used for sorting are indicated. Median ± interquartile ranges are shown for each group. P < 0.0001 for gates G1–G3 vs. library (Mann–Whitney–Wilcoxon test). (F) Distribution of droplets with encapsulated BChE mutants and Fab by fluorescence of the reaction product. The upper panel shows activity of the BChE mutants in comparison with WT BChE. Gates have distinct fluorescence levels: high (the median of cl.3), medium (cl.8), and low (cl.13). (G) Enrichment efficiency of BChE mutants mixed with Fab in a 1:1:1:1,000 ratio. The analysis was performed using high, medium, and low gates as defined in F. Asterisks indicate an undetectable level of enrichment. (H) Interaction of BChE (E) with OP. The cl.14 mutant rather than WT BChE displays catalytic hydrolysis of POX-R. Ki, inhibition constant; k1, phosphylation constant; k2, self-reactivation constant. The BChE concentration was 0.45 μM. All data points plotted are derived from at least two measurements; error bars denote SD (n = 3 technical replicates). RFU, relative fluorescence units.
Screening of bacteria inhibiting S. aureus growth using MDE–FACS. (A–D) The concept for screening of antibiotic producers in MDE. (A) Target S. aureus cells with a GFP reporter were coencapsulated with either killer S. venezuelae cells producing red fluorescent metabolites or mate E. coli cells with a far-red fluorescent reporter. (B) Coculturing of S. aureus with S. venezuelae in contrast to E. coli inhibited S. aureus growth and yielded distinct combinations of fluorescent signals. (C and D) Selection of droplets with the lowest level of green fluorescence resulted in the enrichment of killers (C) rather than mates (D). (E) Visualization of droplets containing S. aureus cocultured with S. venezuelae (Left) or E. coli (Right). (Scale bar, 100 μm.) (F) Selection scheme for bacteria inhibiting S. aureus growth from human oral microbiota. The target S. aureus cells were labeled with the red fluorescent dye sCy5 and were further encapsulated with oral microbiota in MDE. Coculturing the target cells with the effectors provided four different scenarios. (1) From left to right: (i) the effector kills the target; (ii) both the target and the effector are dead; (iii) both the target and the effector are alive; and (iv) the target kills the effector. (2) The addition of Calcein Violet AM to the emulsion after coculturing resulted in the accumulation of blue fluorescent product only inside the droplets with live bacteria. (3) Gating of droplets with sCy5high, GFPlow, and Calcein Violethigh fluorescence resulted in desirable droplets with proliferated effector and inhibited target cells. (G) Prediction of bacterial genera from human oral microbiota inhibiting S. aureus growth using 16S-ribosome sequencing. (Inset) Estimated enrichment of bacterial genera as indicated; subpopulations indicated by line differ with P < 0.01. (H) Comparative analysis of the inhibition activity of bacteria selected by conventional plate analysis (plates) and MDE–FACS (droplets). Data represent mean ± SD (n = 3 biological replicates from one experiment).
Microfluidic cell encapsulation in MDE droplets. (A) Sequential emulsification in single-emulsion chips. (B and C) Channel geometries of microfluidic chips used to produce MDE. (B) The 20-μm chip used for yeast encapsulation. (C) The 60-μm chip used for bacteria and yeast encapsulation. W/O, single water-in-oil emulsion; W/O/W, double water-in-oil-in-water emulsion.
Single-cell compartmentalization in droplets followed Poisson statistics. (A) The probability (P) of finding x cells in the volume of each individual droplet was strictly defined by the λ parameter, which is the average number of cells in droplets. (B) Schematic visualization of droplet occupancy at λ = 0.5 (the most frequently used value in our screenings). (C) Relationships between maximal theoretical purity of sorting (purity) and λ and percentage of droplets with single cells (single cell) and λ.
Enzyme screening was performed with two types of yeast (Fig. 2A): “active” cells displaying a membrane-anchored target enzyme and producing the RFP reporter mCherry (Fig. S3 A and B) and control cells with an “inactive” protein [a fragment of an antibody (Fab) lacking catalytic activity] without any fluorescent markers. A mixture of active and inactive cells was encapsulated with a specific green fluorogenic substrate in MDE droplets. The incubation of active cells with the substrate inside droplet compartments resulted in the generation of a fluorescent product. Active and inactive cells were distinguished by green fluorescence using FACS. Droplets with inactive cells had low green fluorescence and low red fluorescence, whereas the active cells had high red and green fluorescence. Fig. 2B shows an example of an enzymatic reaction in droplets with the encapsulated mixture of active and inactive cells (1:10). The efficiency of biocatalyst selection was quantified by mixing active and inactive cells in ratios varying from 1:10–1:105 within one screening round (Fig. 2C). For low dilutions (1:10 and 1:100), we managed to achieve the theoretical enrichment maximum. Moreover, enrichment efficiency was more than 3.5 × 104, corresponding to 35% of active cells, even at an extremely high dilution (1:105). Thus this approach may be useful for the specific selection of exceptionally rare events.
Expression of enzymes anchored on the yeast surface. (A) Genetic constructs for expression of enzymes exposed on the surface of yeast cells. (B) Active BChE, EK, and DNase I are displayed on the yeast cell surface. Visualization was performed using antibodies against the HA-epitope conjugated with Alexa 488 (catalog no. 26183-A488; Thermo Fisher Scientific). Colorimetric determination of anchored BChE activity was measured using Ellman’s assay. The activity of the anchored enteropeptidase light chain was measured using Gly-(Asp)4-Lys-naphthylamide (Sigma-Aldrich), DNase I activity was measured using the fluorogenic substrate, oligonucleotide 5′FAM-AAAAAAACCCCCCCATATAGCGCGTTTTTTT-3′RTQ1 (Syntol). (C) Reaction schemes illustrating different principles of fluorescent signal generation in droplets used for different enzyme–substrate pairs. (1) Oligonucleotide with FRET pair for DNase I. (2) Nonfluorescent amide for EK. (3) Coupled reaction for BChE. (D) Enzymatic properties of biocatalysts and the respective properties of enzymes anchored on yeasts measured directly in yeast cell suspensions. Data are shown as mean values ± SD. n = 3 technical replicates. Ab, antibody; AOX, alcohol oxidase 1 promoter; FAM, fluorescein amidite; RFU, relative fluorescence units; RTQ, real-time quencher; SAG1, α-agglutinin anchor subunit Aga1p from Saccharomyces cerevisiae.
To test whether the MDE–FACS platform could be used to select different enzymatic activities from the biocatalyst bouillon, we mixed separate individual clones displaying phosphodiesterase (DNase I), protease (bovine enteropeptidase, EK), and esterase (human butyrylcholinesterase, BChE) activities, and a Fab fragment in a 1:1:1:100 ratio. A set of fluorescence-generating systems (Fig. S3 C and D) for sensitive discrimination of different substrate specificities was designed. The results obtained after one round of MDE–FACS (Fig. 2D) indicated that selection was both efficient and specific for all enzymes. The experimental enrichment efficiency (74 ± 5) was close to the theoretical maximum (78.7), and nonspecific enrichment was not observed.
To focus on another biotechnologically relevant issue, the isolation of clones with different levels of the same activity, we created a BChE library (Fig. S4) mutated in the acyl-binding loop (284–TPLSV–288). This loop is highly variable among mammalian BChEs and represents an important region for mutagenesis aimed at altering the enzyme catalytic properties. Mutations dramatically reduced the average activity of the library to ∼1% of WT BChE. Three different gates with decreasing fluorescence levels (G1 > G2 > G3) were used for selection in the MDE–FACS mode (Fig. 2E). Individual yeast clones were regenerated from droplets before and after selection using G1–G3 to illustrate qualitatively that the gate fluorescence correlated with the activity of the selected BChE mutants. To determine quantitatively the selection efficiency of enzymes with different levels of the same activity (Fig. 2F), we mixed representative BChE mutants (cl.3 with high activity, cl.8 with medium activity, and cl.13 with low activity) selected using G1–G3 with the control yeasts in ratios of 1:1:1:1 and 1:1:1:1,000. The fraction of BChE mutants with different levels of BChE activity after the gate selection process with high, medium, and low fluorescence levels indicated efficient, specific enrichment of each BChE mutant in the 1:1:1:1 mixture (Fig. S5). Cl.3 and cl.8 also were selected efficiently from the more diluted 1:1:1:1,000 mixture (Fig. 2G); specific enrichment in cl.13 was lower, indicating limited selection efficiency of mutants with activity close to background. In both mixtures, cl.3 was not selected in gates with medium or low fluorescence levels, suggesting that the selection of droplets with less fluorescence resulted in efficient exclusion of the most active mutant. As a part of the negative selection, this strategy enabled us to identify mutations that abolished enzymatic activity.
Library of human BChE mutants. (A) Human BChE mutants have multiple amino acid substitutions in the acyl-binding loop 284–TPLSV–288 (yellow) close to the active site. The catalytic triad is shown in gray. (B) Sequences of selected BChE mutants with substitutions in the 284–TPLSV–288 positions of WT human BChE.
The MDE–FACS platform enabled the selection of enzymes with different levels of the same activity. (A and B) The portion of BChE mutants with high (cl.3), medium (cl.8), and low (cl.13) activity were selected using G1–G3, and control yeasts after selection using high, medium, and low gates (Fig. 2F) were determined using real-time PCR for 1:1:1:1 (A) and 1:1:1:1,000 (B) mixtures. (C and D) The respective fold enrichment was calculated using initial 1:1:1:1 (C) and 1:1:1:1,000 (D) ratios. Asterisks indicate that the level of cl.3 was too low to be determined. Bars are colored as shown in A. Data represent the mean ± SD (n = 3 technical replicates).
The MDE–FACS platform allowed us to identify BChE mutants resistant to irreversible inhibition by organophosphate (OP) nerve agents using a BChE library preinhibited with OPs (Fig. S6A): paraoxon (POX) or a coumarinyl analog of soman [GDC, 3-chloro-4-methyl-2-oxo-2H-chromen-7-yl (3,3-dimethylbutan-2-yl) methylphosphonate]. BChE mutants displaying resistance to POX (cl.14 and cl.15) and GDC (cl.14 and cl.19) were isolated after one round of screening. The interaction between WT BChE and OP leads to fast (high bimolecular reaction constant k1/Ki) and irreversible (k2 ∼0) inactivation of WT BChE by phosphylation with OP (Fig. 2H) (11). However, several BChE mutants (cl.14 and cl.15) showed residual BChE activity even after prolonged incubation with POX (Fig. S6B). We attribute this property to self-reactivation (k2 > 0), i.e., hydrolysis of the enzyme–OP adduct, which was demonstrated using the resorufin analog POX-R (Fig. 2H). The k1/Ki and k2 values (Fig. S6C) indicated that the selection of BChE mutants resistant to OP inactivation occurred via two distinct pathways: by decreasing the reactivity with OP (low k1/Ki values obtained for cl.19) and the emergence of self-reactivation (e.g., paraoxonase activity observed for cl.14). A combination of both pathways also was observed for cl.15. The selected BChE mutants were of high interest because they revealed a previously unreported target (the BChE acyl-binding loop) for the creation of catalytic bioscavengers based on human BChE, which are used extensively for protection against nerve agent poisoning and postexposure treatment (12, 13).
Screening of the BChE library using MDE–FACS enabled the selection of mutants resistant to OP inhibition by self-reactivation reactivity or decreased affinity to OP. (A) The reaction scheme illustrating the formation of a reversible BChE–OP complex, irreversible BChE–OP adduct formation, and BChE self-reactivation. Ki, inhibition constant; k1, phosphylation rate constant; k2, self-reactivation rate constant. Structures of OP used for selection of OP-resistant mutants (GDC, coumarin analogue of soman; POX, paraoxon; POX-R, resorufin analogue of POX). (B) Pseudo first-order inhibition plots showing the residual activity as a function of time after incubation of 2 nM BChE with 100 nM POX for cl.14 and cl.15 mutants selected for resistance against POX inhibition, control WT BChE, and cl.19 selected for resistance against GDC inhibition. All data points plotted are derived from at three independent measurements. (C) Kinetic constants (mean values ± SD, n = 3 technical replicates) from the reaction scheme in A estimated for WT BChE and mutants. Cells indicating constants that were altered in selected mutants relative to WT BChE are highlighted with orange for POX and blue for GDC. ND, no detectable activity, i.e., an activity level <10−6/s.
The scope of the MDE–FACS platform may be expanded to highlight the peculiarities of bacterial coexistence, cell-to-cell interaction, and signaling, as demonstrated using a simple model (Fig. 3 A–D). The concept was proved by the coencapsulation of cell pairs (target/killer and target/mate) with a well-studied coexistence (14) into MDE. A common pathogen, Staphylococcus aureus, was used as a target, and an antibiotic producer, Streptomyces venezuelae, which inhibits the growth of S. aureus, was the model killer (Fig. 3A); E. coli was used as the mates, because it does not influence the growth of S. aureus. All the bacteria had distinct fluorescent reporters for monitoring their presence in droplets: S. aureus produced GFP, S. venezuelae produced red fluorescent prodiginines (15, 16), and E. coli produced the far-red fluorescent protein Katushka2S (17). The inhibition of S. aureus and the growth of S. venezuelae corresponded to droplets with low green and high red fluorescence in cocultivation (Fig. 3 B and E). Simultaneously, cocultivation of S. aureus and E. coli led to the growth of both bacteria in droplets with high green and far-red fluorescence (Fig. 3 B and E). Thus killers and mates could be distinguished by the fluorescence of their compartments, and selection of the droplets with the least green fluorescence resulted in the enrichment of killers (Fig. 3C) rather than mates (Fig. 3D). We observed that the enrichment efficiency was limited and was dramatically dependent on the S. venezuelae dilution (Fig. S7). These limitations result from similarity between “desirable” droplets with low green fluorescence from S. venezuelae inhibition and “empty” droplets during negative selection using a single GFP reporter.
Enrichment efficiency was limited and depended dramatically on killer S. venezuelae dilution during negative selection using a single GFP reporter. (A) The droplets with the least green fluorescence could have low green fluorescence because of coencapsulation with S. venezuelae or the absence of rapidly dividing S. aureus (empty droplets). The portion of empty droplets was predetermined by the λ parameter of stochastic S. aureus encapsulation; the portion of S. venezuelae-positive droplets decreased with increasing S. venezuelae dilution. (B) Negative selection of droplets (Fig. 3 C and D) resulted in enrichment of killer S. venezuelae rather than mate E. coli cells. The theoretical values were calculated from Poisson statistics (Fig. S2) at λ = 1. The experimental data points plotted are derived from three measurements; error bars denote SD (n = 3 technical replicates).
Despite the high pathogenic potential of S. aureus (18), it is rarely associated with dentoalveolar infection (19). We hypothesized that the suppression of S. aureus infection relates to a natural inhibitory activity of some unknown effectors from human oral microbiota. We tested this possibility by identifying S. aureus killers from the oral microbiota at the level of bacterial clones and their genomes and secretomes. Accordingly, the initial scheme of bacterial screening was modified to use a combination of fluorescent signals for both positive and negative selection to enable the efficient isolation of S. aureus killers. The following additional fluorescent reporters were used: sCy5 to evaluate the initial S. aureus load and Calcein Violet to estimate the total number of viable cells in each individual droplet after cocultivation (Fig. 3F). The gating of droplets with simultaneous sCy5high, GFPlow, and Calcein Violethigh fluorescence (Fig. 3 F, 3) yielded desirable droplets with a high initial S. aureus load, a low number of viable S. aureus cells, and a high number of unknown effector cells after cocultivation. The selected droplets were used for (i) NGS and bioinformatic analysis of the bacteria, especially unculturable bacteria, and (ii) proteomic and secretomic analysis of culturable bacteria regenerated by agar plating.
The MDE–FACS platform provided an essential link between the genotype and its functionality for a diverse population at the single-cell level, a challenging problem in microbiome analysis. Isolation of droplets containing an inhibited target with subsequent NGS and bioinformatic analysis allows the prediction of both culturable and, more critically, unculturable killers, thus revealing potential effectors that are not identifiable using classical microbiology approaches. The prediction of S. aureus inhibitors by 16S rRNA sequencing uncovered distinct populations of bacterial genera characterized by different enrichment efficiencies (Fig. 3G). Propionibacterium, Stenotrophomonas, Sphingomonas, Pseudomonas, and Escherichia were specifically and efficiently enriched in selected droplets, and Corynebacterium, Janthinobacterium, Serratia, Enterobacter, and Streptococcus were enriched with lower efficiency. Thus, the technology we developed allows the prediction of potential cell coexistence and allows the microbiome to be subdivided into discrete functional subpopulations. Whole-genome sequencing confirmed the dramatic enrichment of slow-growing Propionibacterium acnes. The Streptococcus mitis group (S. pneumoniae, S. mitis, S. oralis, and S. pseudopneumoniae), Prevotella dentalis (slow-growing bacteria), Staphylococcus epidermidis (the well-known S. aureus effector), and Pseudomonas aeruginosa were considerably amplified after selection.
This platform allowed us to verify predictions for culturable bacterial species. Fig. 3H shows substantial improvement of the selection procedure for inhibitory clones compared with a conventional agar plating screening assay, allowing us to select bacterial clones with enhanced S. aureus inhibition. Mass spectrometry enabled us to perform taxonomical classification and secretome analysis of culturable S. aureus killers. Most (>90%) belonged to the Streptococcus genus, and 64% were classified as Streptococcus oralis. Growth media obtained from S. oralis strains were characterized by a high (up to 16-fold) inhibiting dilution, although S. oralis did not abolish S. aureus growth in diluted liquid coculture. P. aeruginosa, an uncommon bacteria in the analyzed microbiota samples (Fig. 3G), was selected exclusively using MDE–FACS but not with the conventional screening assay. P. aeruginosa displayed outstanding eradication of S. aureus at a dilution of >1:106 in coculture (Fig. S8). Liquid chromatography-mass spectrometry (LC-MS) was used to analyze the secretome of P. aeruginosa for metabolites inhibiting S. aureus. Pyocyanin, phenazine-1-carboxylic acid, and heptyl-4-hydroxyquinoline were the principle active compounds in P. aeruginosa, displaying pronounced synergetic inhibition of S. aureus growth (Figs. S8 and S9). It is noteworthy that Pseudomonas and Streptococcus predicted by 16S rRNA sequencing (Fig. 3G) were selected using MDE–FACS platform as culturable bacteria that inhibited S. aureus growth.
The P. aeruginosa strain selected using the improved screening strategy for the isolation of S. aureus killers displayed outstanding inhibition of S. aureus growth. (A) P. aeruginosa cells selected after droplet screening were seeded in double dilutions together with S. aureus (∼106 cells per well) in a 96-well plate. S. aureus growth was registered after 8 h of coculturing using the fluorescence of the GFP reporter. Pronounced growth inhibition was observed for S. aureus even at a single P. aeruginosa cell per well. At high P. aeruginosa dilutions (>1:104), there was a considerable increase in blue fluorescence (associated with the induction of P. aeruginosa pyoverdine production). (B) Growth media obtained after the coculturing of S. aureus (SA) and P. aeruginosa (PA) in flasks at different SA:PA dilutions (measured in colony-forming units per milliliter) inhibited S. aureus growth with different efficiencies. The medium with the highest inhibitory activity was obtained from the highest SA:PA ratio. A positive correlation between the inhibition of S. aureus growth (decreased OD800) and pyoverdine production (blue RFU) was observed. (C) Medium obtained after SA and PA coculturing was extracted with chloroform. After evaporation, the chloroform solid residual was resolubilized, and all three phases (medium before extraction, medium after extraction, and solubilized extract) were tested for inhibition of S. aureus growth (SA RFU) and pyoverdine fluorescence (blue RFU). We observed that the chloroform extract rather than the water phase represented the majority of inhibition activity not associated with pyoverdine. All data points plotted are derived from at least two measurements. (D) A chloroform extract of medium obtained after SA and PA cocultivation was fractionated by HPLC in an ACN gradient using a C18 column. AU, arbitrary units. (E and F) The most active fractions 1–7 that displayed prominent inhibition of S. aureus growth after 10 (E) or 48 (F) h were used for further analysis. Fractions 1 and 6 produced by P. aeruginosa displayed a synergetic inhibition of S. aureus growth. Fractions 1–7 were diluted eightfold and were mixed pairwise with each other. The obtained mixtures were used for inhibition of S. aureus growth. (G and H) Green fluorescence of S. aureus was measured after 10 (G) and 72 (H) h. All data points plotted are derived from at least two measurements. RFU, relative fluorescence units.
Mass spectrometric identification of metabolites produced by P. aeruginosa inhibiting S. aureus growth. Mass spectra of pyocyanin (A), phenazine-1-carboxylic acid (B), and 2-heptyl-4-hydroxyquinoline (C). Peaks were selected for fragmentation in Auto MS/MS mode by the following parameters: mass range 50–2,000 m/z, spectra rate 2 Hz, number of precursors 50, precursors selected by maximum intensity, precursor excluded after three spectra and released after 5 min, isolation width 5 Da, collision energy 10, 20, or 30 eV (SI Materials and Methods).
The platform used for the selection of biocatalysts and identification of cell-to-cell interactions allowed the screening of enormous (>108) repertoires of DNA-encoded information. Theoretically, the rate-limiting step for the application of this platform is the FACS-assisted droplet-screening technology (20), which is limited to several tens of thousands of events per second. The characteristics of the MDE–FACS platform—specifically, sensitive detection combined with specific separation—allow its use in practical research on drug design, including the selection of novel desired activity, direct isolation of probiotics, antibiotic producers, and prediction of unculturable effectors. However, the analysis of metabolites or secretomes seems to be the main limiting factor in MDE–FACS automation. The MDE–FACS platform enabled highly sensitive single-cell selection of predesigned activity and the exploration of pairwise interactions between target and effector cells without interference from other microbiota species. The unique ability this technique to accomplish the separation of living species in biocompatible conditions has great potential applications in the field of synthetic biology.
Microfluidic chips with flow-focusing geometry (Fig. S1 B and C) were produced using standard soft lithographic techniques (21, 22), i.e., SU-8/Si masters were used to make polydimethylsiloxane (PDMS) slides with 60 × 70 µm and 20 × 22 µm (width × height) orifices. Inlets and outlets were made using a 1.2-mm Harris Uni-Core biopsy punch. Slides were bound to PDMS or glass slides using an Atto plasma cleaner (Diener) after a 1-min treatment [maximum power, p(O2) = 0.7 mbar]. Immediately after bonding, chips were treated with 1% (wt/vol) polyvinyl alcohol (23) Mowiol 23–88 (Kuraray Specialties Europe) for hydrophilic chips or 0.5% (wt/vol) trichloro(octadecyl)silane (Sigma-Aldrich) in mineral oil for hydrophobic chips [Aquapel (PPG Industries) was used for fluorocarbon oil]. After a 1-min treatment, surface modifiers were removed thoroughly by vacuum, and the chips were heated for 3 min in a drying oven at 120 °C. After modification, hydrophobic chips could be stored for more than 3 y, and hydrophilic chips retained their properties for at least 1 y of storage at room temperature.
Our fluidics were based on an OB1-MkII piezoelectric pressure controller (Elveflow) that independently operated with four liquid streams: inner water 1 (IW1) for cell suspension; inner water 2 (IW2) for substrate solution; oil (O); and other water (OW). Actual flow rates were measured using microfluidic flow sensors (Elveflow) automatically operated in flow-control mode using an inner feedback loop. The IW1 reservoir and the corresponding flow sensor were arranged vertically to minimize cell sedimentation. The inner water streams were mixed before using the hydrophobic chip with MicroTee (P-890; IDEX) and tubing with a dead volume of ∼150 nL. Subsequently, the joint inner water stream was sequentially emulsified in hydrophobic and hydrophilic chips using O [light mineral oil (Sigma-Aldrich) with 3% (wt/vol) Abil EM 180 (Evonic) or fluorocarbon oil (Novec 7500) with 2% (wt/vol) Pico-Surf 2 (Dolomite)] and the OW phase [2% (wt/vol) Pluronic F127 (Sigma-Aldrich), 0.1% Mowiol 23-88, and 50 mM potassium phosphate buffer, pH 7.4]. The characteristic flow rates were 6:6:4:200 µL/min for the IW1:IW2:O:OW phase for the 60-µm chips and 4:4:2:50 µL/min for the 20-µm chips.
Recombinant BChE, human EK, and human DNase I were produced in Pichia pastoris GS115 (Invitrogen) using the modified expression vector pPICZ-mCherry-F2A-HSAss-AfeI/PvuI-HA-SAG1 based on pPICZαA (Invitrogen). This vector contained the fluorescent reporter protein mCherry, self-processing F2A peptide, the human serum albumin (HSA) signal sequence, AfeI and PvuI cloning sites, a-agglutinin anchor subunit Aga1p from Saccharomyces cerevisiae, and an HA epitope. The fragments encoding BChE, EK, and DNase I sequences were PCR-amplified from pFUSE PRAD-F2A-BChE (24), pHENm/l-HEP/C122S (25), and p6E-rhDNaseI, which was kindly provided by Pharmsynthez PJSC, and were cloned into vector pPICZ-mCherry-F2A-HSAss-AfeI/PvuI-HA-SAG1 digested with AfeI and PvuI. Primers used for amplification are presented in SI Materials and Methods.
Yeasts producing anchored enzymes or a Fab fragment were grown overnight in liquid culture using yeast extract/peptone/dextrose (YPD) medium and were induced by growth in buffered methanol-complex medium (BMMY). Subsequently, the yeasts were washed twice with 50 mM potassium phosphate buffer (pH 7.4) and resuspended in appropriate buffer [1 mM MnCl2, 0.1 mM CaCl2, 20 mM Tris⋅HCl (pH 7.4) for DNase I; 0.2 mM CaCl2, 20 mM Tris⋅HCl (pH 7.4) for EK; and 50 mM potassium phosphate buffer (pH 7.4) for BChE]. Next, the yeasts were filtered using a 20-µm solvent filter A-313 (IDEX), diluted with buffer to reach λ = 0.5 (OD600 = 0.3 and OD600 = 3 for the 60-μm and 20-μm chips, respectively), and mixed in an 1:10–1:105 enzyme:antibody ratio. Finally, they were compartmentalized in drops of double emulsions with an appropriate substrate in a corresponding buffer [1 µM FAM: AAAAAAACCCCCCCATATAGCGCGTTTTTTT-RTQ1 (Syntol) for DNase I; 10 µM SensoLyte Rh110 substrate (AnaSpec) for EK, and 30 µM butyrylthiocholine iodide (Sigma-Aldrich) with 30 µM 3-(7-Hydroxy-2-oxo-2H-chromen-3-ylcarbamoyl)acrylic acid methyl ester (Millipore) for BChE].
We used S. aureus constitutively producing GFP that was kindly provided by Andrei Shkoporov of the Department of Microbiology and Virology, Russian National Research Medical University, Moscow. The S. venezuelae Ehrlich Aс-505 strain was obtained from the Russian Collection of Microorganisms. The E. coli JW5503 strain (kindly provided by Hironori Niki, National Institute of Genetics, Mishima, Shizuoka 411-8540, Japan) was transformed with a pKatushka2S-B vector (Evrogen) that enabled inducible expression of the far-red fluorescent protein Katushka2S. Bacteria were cultured using 2YT medium (16 g/L tryptone, 10 g/L yeast extract, 5.0 g/L NaCl) supplemented with 2% glucose (for Streptomyces) or 100 μg/mL ampicillin (for E. coli) in shaking flasks at 37 °C (S. aureus and E. coli) or 28 °C (S. venezuelae) at 250 rpm. S. aureus and E. coli were grown for 4–6 h until they reached a logarithmic phase of growth; S. venezuelae was grown for 2 d. Subsequently, liquid cultures were washed three times with cultivation medium [0.16% tryptone, 0.1% yeast extract, 0.05% NaCl, 0.5% glycerol, 0.67% yeast nitrogen base (YNB) and 1 mM isopropyl β-d-1-thiogalactopyranoside (IPTG)]. After washing, cultures were filtered using 40-μm cell strainers (Greiner Bio-One) and were diluted to reach (i) OD600 = 0.1 (λ = 10) for S. aureus; (ii) OD600 = 0.03 (λ = 2) for E. coli; and (iii) OD600 = 0.3 (λ = 1) for S. venezuelae. S. venezuelae and E. coli cells were encapsulated pairwise with S. aureus in double-emulsion droplets using 60-µm chips. Similarly, effector cells (S. venezuelae and E. coli) were diluted by target cells in a 1:10 or 1:100 effector:target ratio and were encapsulated with S. aureus in double-emulsion droplets. After encapsulation, the droplets were incubated overnight for 2 d at 25 °C.
Microfluidic double-emulsion droplets with encapsulated yeast or bacterial cells were loaded into a hemocytometer and were visualized using an Eclipse Ti inverted fluorescence microscope (Nikon) with standard FITC and Texas Red filters.
Droplets were sorted using FACSAria III (BD). Initial gating of intact double-emulsion droplets was performed using scattering and background fluorescence of the substrate or culture medium. Sorting of positive events was performed using a 530/30 nm emission filter (or 450/50 nm for BChE). Yeast and bacteria clones were regenerated by agar plating after sorting in three replicates. Individual yeast colonies with displayed enzymes were analyzed by fluorescence from the mCherry reporter using VersaDoc (Bio-Rad). The number of E. coli colonies was calculated by measurement of the far-red fluorescence of reporter Katushka2S. The number of S. venezuelae colonies was calculated manually based on specific morphology. The degree of enrichment was defined as the ratio of the percentage of positive colonies after and before sorting in three replicates. Maximal theoretical enrichment was calculated as a dilution multiplied by the maximal theoretical purity of sorting estimated using a Poisson distribution (Fig. S2).
Recombinant BChE and its library were produced in the methylotrophic yeast Pichia pastoris GS115 (Invitrogen) using the modified expression vector pPic9k-α-SfiI-FLAG-anchor based on pPIC9k (Invitrogen). The degenerated primers (SI Materials and Methods) were used to amplify regions flanking the mutated 284–TPLSV–288 loop of WT human BChE and were cloned into expression vector pPic9k-α-Lib_BChE-FLAG-anchor using SfiI. Vector was linearized by PmeI and transformed into P. pastoris GS115 as previously described (26). The diversity of the library was ∼9 × 105.
Yeasts producing a library of anchored BChE mutants were grown, induced, and washed as previously described. Next, a cell suspension was incubated with OPs [0.2 mM POX (Sigma-Aldrich) or 1 mM GDC] for 0.5 h at 25 °C, washed with 50 mM potassium phosphate buffer (pH 7.4) supplemented with 2 μM of the respective OP, filtered as previously described, and encapsulated with BChE substrate, including 100 µM butyrylthiocholine and 100 µM 3-(7-Hydroxy-2-oxo-2H-chromen-3-ylcarbamoyl)acrylic acid methyl ester. After 3 h, the most fluorescent droplets were screened using FACS. The regenerated yeast colonies were analyzed for BChE activity in a 384-well plate. The most active clones after OP inactivation (cl.14 and cl.15 for POX and cl.19 for GDC) were selected and sequenced to determine substitutions in the 284–TPLSV–288 sequence of WT human BChE (Fig. S4). The selection of BChE mutants with altered activity is presented in SI Materials and Methods.
Recombinant human BChE was produced in FreeStyle 293-F cells (Thermo Fisher Scientific) using the modified expression vector pFUSE-BChE-6xHis based on pFUSE-hIgG1-Fc (Invivogen) containing truncated BChE W541H6Δ (27). Genomic DNA was extracted from clones with improved resistance against irreversible OP inhibition (28) and was used as a PCR template with primer pair i/iv (SI Materials and Methods). This PCR product was used to replace the insert between the NheI and NcoI sites in pFUSE-BuChE-6xHis. Transfection of FreeStyle 293-F cells was performed using 293fectin Transfection Reagent (Thermo Fisher Scientific). WT BChE and its mutants were purified using TALON Metal Affinity Resin (Clontech). Purified BChE mutants were used to determine the bimolecular inhibition constant (k1/Ki) according to the Kitz–Wilson method (29). The residual BChE activity was determined according to Ellman’s method (SI Materials and Methods) (30). The activity was measured using a Varioskan Flash multimode reader (Thermo Fisher Scientific) at λEx/λEm = 570/585 nm (for POX-R) and λEx/λEm = 360/450 nm (for GDC). All data were derived from three technical replicates.
Samples of oral microbiota were obtained by scraping the gingivae and sublingual fold of three independent healthy donors. The study was approved by the Local Ethics Committee of the Scientific Research Institute of Physical-Chemical Medicine (SRI-PCM). All donors provided written informed consent. Scraped bacteria were resuspended in medium for coculturing. Target S. aureus cells producing the GFP reporter at a logarithmic phase were stained with 5 μM sulfo-Cyanine5 NHS (Lumiprobe) for 1 h in PBS and were washed three times with medium for coculturing. S. aureus were coencapsulated with an oral microbiota suspension using 60 µm chips. After overnight incubation at 35 °C, Calcein Violet AM (Thermo Fisher Scientific) was added to the droplet emulsion to a final concentration of 10 µM. After 30 min of incubation, droplets with simultaneous sCy5high, GFPlow and Calcein Violethigh fluorescence (using 660/20-, 530/30-, and 450/50-nm filters) were sorted. For detailed information SI Materials and Methods. Bacterial colonies that regenerated after agar plating and demonstrated inhibition of S. aureus growth were analyzed by mass spectrometry. Bacterial cells were spotted on a sample spot of a MALDI target plate (MSP 96 target, ground steel; Bruker Daltonics) and were overlaid with 2 μL of matrix solution HCCA (saturated solution of α-4-cyano-hydroxycinnamic acid; Bruker Daltonics) in 50% acetonitrile (Sigma-Aldrich) and 2.5% trifluoroacetic acid solution (Sigma-Aldrich). Mass spectra profiles were acquired using a Microflex spectrometer (Bruker Daltonics). The molecular ions were measured automatically in linear positive ion mode with instrument parameters optimized for a range of 2,000–20,000 m/z. The software packages flexControl 3.0 (Bruker Daltonics) and flexAnalysis 3.0 (Bruker Daltonics) were used for mass spectra recording and processing. Spectra identification and analysis were carried out using the MALDI Biotyper 3.0 (Bruker Daltonics). Identification was performed by comparing the obtained spectra with the MALDI Biotyper 3.0 library (version 3.2.1.1). All additional methods are described in SI Materials and Methods.
The following primers were used:
(i) GCTTACTCGGAAGATGACATCATAATTGCAACAA
(ii) CTCCTTCGATCGAGACTTTTGGAAAAAATGATGTCCAG
(iii) GCTTACTCGATTGTTGGTGGCTCTAACGC
(iv) CTCCTTCGATCGACGCATGCAGAAAGCTCTGG
(v) GCTTACTCGCTGAAGATCGCAGCCTTCAAC
(vi) CTCCTTCGATCGACTTCAGCATCACCTCCACTG
The following primers were used:
(i) GCGCTAGCTGCGGCCCAGCCGGCCGAAGATGACATCATAATTGCAACAAAGAAT
(ii) CTTTGTAGTCCATCAGGCCCCCGAGGCCGAGACCCACACAACTTTCTTTCTTG
(iii) CCCATAGGGGACAACAAATGCTTC
(iv) AACTTTGGTCCGACCGTGGATG
(v) GAAGCATTTGTTGTCCCCTATGGGMVWMVWMDWMRWVDWAACTTTGGTCCGACCGTGGATG
The BChE library was screened using three different gates (Fig. 2E) with increasing levels of droplet fluorescence: G3 < G2 < G1. Representative clones with different levels of BChE activity relative to WT BChE (highly active cl.3, cl.8 with medium activity, and cl.13 with low activity) were selected using G1–G3 and were sequenced. Yeast clones with different levels of WT BChE activity (cl.3, cl.8, and cl.13) were mixed with inactive yeast producing anchored antibody Fab fragment in 1:1:1:1 and 1:1:1:1,000 ratios. Next, the clones were screened for BChE activity in droplets using the fluorescence level gates (high, medium, and low) corresponding to the median fluorescence of droplets with encapsulated cl.3, cl.8, and cl.13 (Fig. 2F). The regenerated yeast colonies were analyzed by quantitative PCR using primers specific to BChE mutants and Fab.
Highly diluted (∼2 nM) samples of BChE mutants were incubated with varying concentrations of OP (1.6–0.012 μM for POX and 125–1 μM for GDC) for different times ranging from 1 min to 2 h, and the residual BChE activity was determined according to Ellman’s method using 1 mM butyrylthiocholine iodide (Sigma-Aldrich) and 0.5 mM 5,5-dithio-bis-(2-nitrobenzoic acid) (Sigma-Aldrich), in PBS (pH 7.4) at 25 °C. The self-reactivation constant k2 was determined using 0.2–1 μM BChE concentrations incubated with 0.7–6 μM POX-R (Annova Chem Inc.) or 1–4 μM GDC for 14 h. The activity was measured using a Varioskan Flash Multimode Reader (Thermo Fisher Scientific) at λex/λem = 570/585 nm (for POX-R) and λex/λem = 360/450 nm (for GDC). All data were derived from three technical replicates.
Conventional selection of S. aureus inhibitors was provided by a soft agar overlay assay based on the plating of oral microbiota samples of three independent healthy donors on LB agar plates. Bacterial colonies appeared after 2 d of incubation at 35 °C and were overlaid with 0.7% LB agar containing S. aureus with OD600 = 0.03. After overnight incubation, zones of clearance appeared around bacterial colonies producing inhibitors of S. aureus growth. Representative clones of inhibitors selected using conventional agar plating were used as a control to evaluate the efficiency of selection using droplets.
Inhibition of S. aureus growth was measured separately for cell suspensions obtained after 2 d of growth in liquid culture at 35 °C and growth medium sterilized by filtering using 0.22-µm syringe filters. Bacterial cells and growth medium were diluted by a doubling dilution in 96-well plates with a S. aureus cell suspension OD600 = 0.03 in 2YT medium. After overnight incubation at 35 °C, S. aureus growth was analyzed by GFP fluorescence (λex/λem = 488/513 nm) and OD600 using a Varioskan Flash Multimode Reader (Thermo Fisher Scientific).
Growth medium obtained after coculturing the selected P. aeruginosa strain and S. aureus was extracted with chloroform and fractionated by HPLC in an acetonitrile (ACN) gradient using a C18 column µBondapak (10 µm × 3.9 mm × 300 mm) (Waters). The resulting fractions were analyzed for S. aureus growth inhibition in culture (Fig. S7). The most active fractions 1–7 that displayed prominent inhibition of S. aureus growth were used for LC-MS analysis. The mass spectrometry experiments were conducted on an ultrahigh-resolution (UHR) quadrupole time-of-flight (QTOF) maXis mass spectrometer with an upgraded high-definition collision (HDC) cell (Bruker Daltonics GmbH) equipped with an electrospray ionization (ESI) source used in positive ion mode. A Windows 7 Professional (Microsoft) workstation running otofControl (version 3.4; Bruker Daltonics) and Compass Data Analysis (version 4.2, Compass for otofSeries version 1. 7; Bruker Daltonics) was used for data acquisition and processing. Bruker ESI source parameters at the flow rate of 3 μL/min were as follows: end plate offset 500 V, capillary voltage 4,500 V, dry gas current 4 L/min, dry gas temperature 200 °C, nebulizer gas 0.3 bar. Samples and calibration mix were input directly into the electrospray source using a syringe (Hamilton) installed in the syringe injection pump (KD Scientific). Before the experiment, the mass spectrometer TOF was calibrated using ESI-L low-concentration tuning mix (Agilent Technologies); the calibration mode was enhanced quadratic, and the summarized SD was 0.2 ppm. Peaks were selected for fragmentation in Auto MS/MS mode by following parameters: mass range 50–2,000 m/z, spectra rate 2 Hz, number of precursors 50, precursors selected by maximum intensity, precursor excluded after three spectra and released after 5 min, isolation width 5 Da, collision energy 10, 20, 30, or 50 eV.
Total DNA was isolated using a QIAamp DNA Investigator Kit (Qiagen). Amplicon libraries were prepared using primers 341F (TCGTCGGCAGCGTCAGATGTGTATAAGAGACAGCCTACGGGNGGCWGCAG) and 801R (GTCTCGTGGGCTCGGAGATGTGTATAAGAGACAGGACTACHVGGGTATCTAATCC) according to the 16S Metagenomic Sequencing Library Preparation manual (Part 15044223 Rev. B). Sequencing of libraries and data analysis were performed using the genetic analyzer MiSeq (Illumina) and MiSeq Reagent Kit v2 (300 cycles) according to the manufacturer’s instructions. The data obtained were processed using software build in MiSeq (Illumina). The threshold of 300 (the total number of reads in the blank experiment) was used in the reads of each genus. The number of reads obtained was normalized to the total number of reads in each sample. Specific enrichment was defined as the number of normalized reads after selection divided by the number of normalized reads before selection.
Total DNA was isolated using a QIAamp DNA Investigator Kit (Qiagen). Whole-genome amplification was performed using the REPLI-g Single Cell Kit (Qiagen). Fragment libraries were prepared using the Ion Xpress Plus Fragment Library Kit and the Ion Xpress Barcode Adapters 1–16 Kit (Life Technologies) according to the manufacturer’s instructions. Sequencing of libraries was performed using the genetic analyzer Ion Proton, the Ion PI HI-Q Sequencing 200 Kit, and the Ion PI Chip Kit v2 (Life Technologies) according to the manufacturer’s instructions. The shotgun metagenomic reads were processed using two taxonomic classification algorithms: Metaphlan2 and CLARK. The number of genera reads obtained was normalized to the total number of reads in each sample. The threshold of 0.005% was used to normalize each read. Specific enrichment was defined as the number of reads after selection divided the number of reads before selection.
This work was supported by Grant RFMEFI60414X0069 from the Ministry of Education and Science of Russia.
↵1S.S.T. and I.V.S. contributed equally to this work.
↵2Deceased December 16, 2016.
Author contributions: S.S.T., I.V.S., N.A.P., M.V.D., A.G.G., and S.A. designed research; S.S.T., I.V.S., A.V.S., T.V.B., Y.A.M., O.V.K., V.V.B., M.T.V., M.V.M., M.A.K., and A.A.V. performed research; Y.A.M., M.P.R., M.O.G., A.A.M., A.S.B., M.V.D., and E.N.I. contributed new reagents/analytic tools; S.S.T., I.V.S., A.V.S., N.A.P., A.A.B., E.S.K., A.I.M., A.V.T., A.A.V., E.N.I., P.M., A.G.G., and S.A. analyzed data; and S.S.T., I.V.S., A.G.G., and S.A. wrote the paper.
Reviewers: R.S.P., University of Georgia; and I.S., Weizmann Institute of Science.
The authors declare no conflict of interest.
This article contains supporting information online at www.pnas.org/lookup/suppl/doi:10.1073/pnas.1621226114/-/DCSupplemental.
Freely available online through the PNAS open access option.
