Overview¶

We present a synthetic-cell based hydrogel device which can sense bacterial contamination through quorum sensor induced cell-free protein expression. POPC GUVs encapsulating the E. coli S30 lysate and the LuxR-mNeonGreen plasmid become fluorescent in the presence of purified 3-oxo-C6-HSL quorum sensing molecule, verified using fluorescence microscopy. Additionally, GUVs embedded within ultra low gelling agarose still sense and respond to the HSL, offering a material with improved structural and mechanical properties. Finally, the LuxI plasmid, which encodes for bacterial production of 3-oxo-C6-HSL, was transformed into E. coli, with supernatant derived from overnight cultures used to induce protein expression within the synthetic cells. A mixture of fluorescence microscopy, Z-stack fluorescence microscopy and fluorescence well plate assays were used to verify HSL induced cell-free protein expression in bulk lysate and vesicles.

Constructs¶

Results¶

Result 1 — LuxR-mNeonGreen sensor plasmid in bulk lysate¶

To verify whether E. coli lysate and the LuxR-mNeonGreen plasmid can detect 3-oxo-C6-HSL in bulk, a 12.5 µL bulk lysate fluorescence assay was conducted in 384 well plates. The LuxR-mNeonGreen plasmid constitutively expresses the LuxR transcription factor in E. coli lysate, which is activated by the HSL signal. Upon activation, LuxR dimerizes, enabling it to bind to the Lux box situated adjacent to the RNA polymerase promoter sequence. The activated LuxR recruits the σ70 subunit of RNA polymerase, activating transcription. The constitutive mNeonGreen plasmid does not contain the Lux box, so RNA polymerase does not require activated LuxR for protein expression.

Similar cell-free biosensors for bacterial detection using quorum sensing have been developed in lysate-based systems with the LasR transcription factor and 3-oxo-C12-HSL quorum sensor ().

It is clear there is a significant difference in fluorescent signal in the presence and absence of 10 µM 3-oxo-C6-HSL, confirming that the sensor construct can successfully detect low concentrations of HSL in E. coli lysate with minimal leaky expression (Figure 2). Additionally, the constitutive control has significantly higher mNeonGreen expression than the HSL induced reaction. Future experiments will focus on lowering the HSL concentration to find the limit of detection for the combined lysate and sensor construct.

Result 2 — detecting purified HSL using lysate-based synthetic cells¶

To demonstrate the ability of synthetic cells to detect HSL, lysate-based GUVs encapsulating the LuxR-mNeonGreen plasmid were resuspended in outer solution containing HSL and visualized by brightfield and fluorescence microscopy over time. Additionally, to improve vesicle stability and robustness, the synthetic cells were embedded in 1 wt% (w/v) ultra-low gelling temperature agarose (ULGA), with purified HSL added to the top of the gel to induce mNeonGreen expression. Time point and Z-stack fluorescence microscopy were used throughout this section.

3-oxo-C6-HSL has been shown elsewhere to passively diffuse across artificial lipid bilayer membranes and induce protein expression in E. coli lysate-based systems (). It would therefore be expected that HSL can permeate the GUV lipid membrane into the synthetic cell. As HSL is non-fluorescent, this could not be verified directly; it is instead verified indirectly through LuxR-induced expression of a fluorescent protein.

Detecting purified HSL using lysate-based synthetic cells in solution¶

Three experimental conditions were used for HSL sensing synthetic cells in solution, shown in Table 3. Methods for lysate-based synthetic cell formation are in Methods, under Lysate GUVs preparation.

Microscopy images were acquired every 20 minutes across 17 time points, in the store Lysate_LuxR_Cytosol_EsaR_2026-10-05_18-59-02.423661.zarr. Plate map: https://docs.google.com/spreadsheets/d/17V5zdl5ebfRpTSc5XnVHenCmLCfFspMTcQ0KqSPfDu0/edit?gid=1906581583#gid=1906581583.

HSL sensing GUVs

Negative control

Constitutive control

[fig-solution-j4, platemap: none, data source:Lysate_LuxR_Cytosol_EsaR_2026-10-05_18-59-02.423661.zarr/J/4/0, caption: (Well J4, HSL sensing GUVs, T = 17, GFP channel only, center crop.)]

[fig-solution-j7, platemap: none, data source:Lysate_LuxR_Cytosol_EsaR_2026-10-05_18-59-02.423661.zarr/J/7/0, caption: (Well J7, negative control, T = 17, GFP channel only, center crop.)]

[fig-solution-j10, platemap: none, data source:Lysate_LuxR_Cytosol_EsaR_2026-10-05_18-59-02.423661.zarr/J/10/0, caption: (Well J10, constitutive control, T = 17, GFP channel only, center crop.)]

From the endpoint microscopy data, the HSL sensing GUVs have increased fluorescence output compared to the negative control without HSL. This shows that the combined lysate and DNA template used for bulk assays in Figure 2 can be successfully encapsulated within GUVs, and that HSL can permeate into the vesicles and induce protein expression. Although not quantified, the difference in signal between the negative control and HSL induced expression appears significantly lower than the bulk lysate reactions. This suggests that either the HSL concentration is significantly lower, or there is reduced concentration of essential components for CFPS within the vesicles, such as the plasmid, which can occur during vesicle encapsulation.

Similarly to the bulk lysate plate assays, the constitutive control has significantly higher and quicker mNeonGreen expression than the HSL induced system in GUVs. This is expected, due to only requiring the production of a single protein and non-reliance on HSL membrane transport and binding for mNeonGreen expression. Future work for this section will focus on reducing the HSL concentration to quantify the limit of detection of the system within synthetic cells.

Detecting purified HSL using lysate-based synthetic cells in gels¶

Embedding synthetic cells within hydrogels allows a medium to distribute the vesicle population evenly throughout the material, and lets it be placed on solid surfaces for potential biosensing applications. Additionally, gel materials offer mechanical support and should lead to increased stability of GUVs compared with solution, with greater robustness broadening the scope for potential applications. Throughout this section, HSL sensing GUVs are embedded within 0.5% ULGA, following the same lysate-based GUV protocol as the solution subsection above.

Methods, under GUV embedding in gels for synthetic AHL experiment, covers the steps for embedding GUVs in 0.5% ULGA. In summary, 1 wt% ULGA is heated above its gelling temperature and mixed thoroughly with the GUV pellet at equal volumes, then added to a 384 well plate. Z-stack fluorescence microscopy is then used to assess fluorescence throughout the gel after 16 hours. After gelling, 10 µL of stock HSL is added to the top of the gel, corresponding to a final concentration of 100 µM; 10 µL of outer solution was added to the top of gels for experiments without HSL.

HSL sensing gels

Negative control gels

Constitutive control gels

[fig-gel-d6, platemap: none, data source:J_N_+100uM_AHL_D6_gels_2026-10-07_14-36-46.345035.zarr/D/6/0, caption: (Well D6, HSL sensing gels, z = 50, GFP channel only, center crop.)]

[fig-gel-d5, platemap: none, data source:J_N_-100uM_AHL_D5_gels_2026-10-07_14-44-55.273559.zarr/D/5/0, caption: (Well D5, negative control gels, z = 50, GFP channel only, center crop.)]

[fig-gel-d11, platemap: none, data source:J_N_Consitutive_AHL_D11_gels_2026-10-07_14-53-48.138799.zarr/D/11/0, caption: (Well D11, constitutive control gels, z = 50, GFP channel only, center crop.)]

Constitutive mNeonGreen expression with no HSL added (wells D9, D10, D11):

• J_N_Consitutive_AHL_D9_gels

• J_N_Consitutive_AHL_D10_gel_bigger_areas — a bigger area was measured because of a large air bubble

• J_N_Consitutive_AHL_D11_gels

LuxR-mNeonGreen with no HSL added (wells D3, D4, D5):

• J_N_-100uM_AHL_D3_gels

• J_N_-100uM_AHL_D4_gels

• J_N_-100uM_AHL_D5_gels

LuxR-mNeonGreen with 100 µM HSL added (wells D6, D7, D8):

• J_N_+100uM_AHL_D6_gels

• J_N_+100uM_AHL_D7_gels

• J_N_+100uM_AHL_D8_gels

Empty ULGA gels (wells D12, D13, D14):

• J_N_just_gel_1

• J_N_just_gel_2

• J_N_just_gel_3

Log directories: the GUVs-in-solution run (2nd/3rd October) is at https://drive.google.com/drive/folders/15EQ5kkf6LlnIJqQShonYNE3fKh8O3PgD, and the gel run is at https://drive.google.com/drive/folders/1j4vF36gjF_jpaGyI8fI-_uyUs-M6mC2d.

Result 3 — detecting bacterially produced HSL in solution¶

To model bacterial sensing using GUVs, the LuxI plasmid was transformed into NEB5α E. coli (NEB5α w/ LuxI) to encode production of 3-oxo-C6-HSL, the small-molecule quorum sensor detected in the previous sections. The same LuxR-mNeonGreen plasmid was used to verify whether lysate-based synthetic cells are capable of detecting bacterially produced HSL, implemented through resuspending the GUV pellet in bacterial supernatant. Control experiments were performed with the NEB5α strain without the LuxI plasmid (NEB5α ΔLuxI), which removes HSL from the supernatant.

Lysate GUVs were prepared following the standard protocol in Methods up to the resuspension step. The GUV pellets were resuspended in 100 µL of 1:10 supernatant to outer solution, and 30 µL was added to a 384 well plate for visualization. One GUV preparation was run per condition, split into three wells for microscopy. Brightfield and 488 nm fluorescence microscopy were both run on each sample over time, images acquired every 20 minutes for about 7 hours.

Supernatant from NEB5α carrying the LuxI plasmid induced mNeonGreen expression within lysate GUVs carrying the LuxR-mNeonGreen plasmid, visible by fluorescence microscopy at 488 nm (wells C3 to C5). After about 80 minutes, mNeonGreen expression becomes visible within some GUVs, with more GUVs expressing over time. Expression is clearly visible throughout the well, but overlapping brightfield and 488 nm fluorescence shows that many GUVs had no expression, likely because they did not encapsulate all the machinery lysate-based TXTL needs.

Lysate GUVs carrying the same plasmid, resuspended in supernatant from bacteria without the LuxI plasmid, had minimal expression over the 7 hour experiment (wells C6 to C8). This shows that HSL is required to induce expression within lysate GUVs, and that HSL made by E. coli can be detected by synthetic cells. Interestingly, the fluorescence intensity of LuxR-mNeonGreen lysate GUVs induced by bacterial supernatant was significantly higher than with 10 µM purified HSL, indicating that significantly more HSL is produced by overnight inoculation of NEB5α w/ LuxI than the 10 µM purified HSL used in Table 3. Future work should include supernatant in the bulk lysate assays (Figure 2), with supernatant extracted at various time points.

Synthetic cells carrying the constitutive mNeonGreen plasmid in NEB5α ΔLuxI supernatant expressed after about 40 minutes, with expression increasing over the experiment (wells C9 to C11). The faster expression compared to LuxR-mNeonGreen follows from synthesizing one protein only, with no dependency on HSL entering the cell and activating mNeonGreen expression.

The three conditions are shown as interactive viewers, one well per condition, on 2026-10-08 — Interactive viewer. That page carries the open questions on contrast and timepoint.

HSL sensing GUVs

Negative control

Constitutive control

[fig-fov-c4, platemap: none, data source:nc_lysate_cyt_c_d_2026-10-08_20-00-25.520782.zarr/C/4/0, caption: (Well C4, HSL sensing GUVs, T = 19, GFP channel only, cropped and zoomed.)]

[fig-fov-c7, platemap: none, data source:nc_lysate_cyt_c_d_2026-10-08_20-00-25.520782.zarr/C/7/0, caption: (Well C7, negative control, T = 19, GFP channel only, cropped and zoomed.)]

[fig-fov-c10, platemap: none, data source:nc_lysate_cyt_c_d_2026-10-08_20-00-25.520782.zarr/C/10/0, caption: (Well C10, constitutive control, T = 19, GFP channel only, cropped and zoomed.)]

Methods¶

Bulk lysate (12.5 µL reaction)¶

Mix amino acids using equal volumes from each provided tube (−Met, −Cys, −Lys). Assemble the following from the Promega circular lysate kit on ice:

Place 10 µL of the reaction in a 384 low volume plate and incubate in a plate reader at 37 °C for 4 hours.

Lysate GUVs preparation¶

Create lipid thin films in a glass vial using 160 µL of 25 mg/mL POPC chloroform stock and dry under nitrogen gas. Place the films in a vacuum desiccator overnight. Add 1 mL of mineral oil to each film.

Prepare the inner solution using the table below.

Prepare the outer solution with the concentrations in the table below.

Label two microcentrifuge tubes “inner” and “outer”. Pipette 300 µL outer solution into the outer tube and 200 µL lipid-in-oil into the inner tube. Add 20 µL of lysate solution to the lipid-in-oil and emulsify 20 to 25 times using a P1000 pipette. Immediately layer the emulsion onto the outer solution and let it sit at room temperature for 5 minutes, to let the monolayer form at the aqueous-oil interface. Centrifuge at 9000 ×g for 15 minutes. Clear the oil layer and the old outer solution. Resuspend the pellet with 100 µL outer solution. Incubate overnight at 37 °C in a plate reader or microscope.

GUV embedding in gels for synthetic AHL experiment¶

Make a 1% ULGA stock, with ULGA dissolved in the outer solution from Table 9. Heat in a microwave or on a hot plate until the solution is bubbling. Allow it to cool to a temperature that can be held by hand, and before it gels, pipette 25 µL of GUVs and 25 µL of ULGA into a 384 well plate and mix. Place on ice or in a fridge for 10 minutes. Add 10 µL of 600 µM AHL, giving a final gel concentration of 100 µM, and incubate at 37 °C in a plate reader or microscope overnight.

Supernatant preparation¶

Inoculate 5 mL LB (ampicillin) with the NEB5α strain carrying the LuxI producing plasmid. Incubate overnight at 37 °C. Centrifuge for 15 minutes. Pipette the supernatant into a separate tube. Sterile filter the supernatant with a 0.22 µm filter. Store at −20 °C until use.

Summary¶

1. E. coli lysate-based cell-free protein expression with the LuxR-mNeonGreen plasmid can successfully detect 3-oxo-C6-HSL in bulk.

2. The same system can be successfully encapsulated in synthetic cells, which is activated with 10 µM purified HSL in solution, visible using fluorescence microscopy.

3. Lysate-containing synthetic cells encapsulated within ULGA hydrogels can sense 100 µM HSL added to the top of the gels.

4. Bacterial supernatant collected from E. coli containing the LuxI plasmid induced GFP expression in lysate GUVs containing the LuxR-mNeonGreen plasmid.

What’s next¶

1. Test bacterial supernatant with lysate GUVs embedded in agarose hydrogel.

2. Test orthogonal sensing with 3-oxo-C6-HSL and 3-oxo-C12-HSL quorum sensors using LuxR and LasR to regulate transcriptional regulation.

3. Investigate whether the LuxR system can be used for visible colorimetric detection of bacterial contamination, through encapsulation of a substrate and expression of an enzyme.

4. Perform tests around the mechanical stability of GUVs embedded in hydrogel.