Overview¶

We assessed vesicle functionality in hydrogels through two complementary readouts: (i) quantification of T7-driven GFP expression yield in giant unilamellar vesicles (GUVs), and (ii) release of encapsulated dye, Chlorophenol red-β-D-galactopyranoside (CPRG) from large unilamellar vesicles (LUVs), monitored in a panel of chemically and mechanically distinct hydrogels alongside a solution control. Comparing across gel types allowed us to separate the effects of network chemistry, stiffness, and mesh size on vesicle behavior.

[fig-schematic-hydrogel-overview, file:general/schematic-hydrogel-overview.png, caption: (Schematic representation of unilamellar vesicles contained in a hydrogel matrix.)]

Methods¶

Vesicle preparation¶

LUVs

Hydration

1. Hydrate a dried POPC lipid film (~2 mg) in buffer containing glucose, 50 mM HEPES, and 15 v/v% Optiprep, at a final osmolarity of 1200 mOsm. The buffer contains 15 mg/mL CPRG. The lipid film was dried under a nitrogen atmosphere and then dried under vacuum for ~2 hours before hydration.

2. Vortex, then sonicate for 15 min at room temperature.

3. Apply 5 freeze-thaw cycles, liquid nitrogen, then thaw in a room-temperature water bath.

Clean-up by centrifugation

1. Split the mixture into 5 tubes, each with 200 µL vesicles and 1 mL outer solution, glucose and 50 mM HEPES, at a final osmolarity of 1200 mOsm.

2. Spin at 15 000 g for 15 min at room temperature.

3. Resuspend the pellet in outer solution and wash again by pelleting at the same settings.

4. Repeat until the outer solution of the pellet looks clear, 2 to 4 washes.

[fig-photo-luv-pellet, file:general/photo-luv-pellet.png, caption: (Snapshot of the LUV pellet after 4 washes.)]

Hydrogel preparation¶

Three hydrogels were prepared and plated using inputs shared across the London and Chicago nodes.

ULGA (1 w/v%)

LGA (2.8 w/v%)

PEG4Nb (80 mM)

1. Prepare ULGA at 1 w/v% in the outer solution used during vesicle preparation, that is, glucose and 50 mM HEPES, at a final osmolarity of ~1200 mOsm.

2. Microwave in 10 s bursts, 5 to 7 times, until fully dissolved.

3. Keep on the 55 °C bead bath until use.

1. Dissolve LGA at 2.8 w/v% in the outer solution used for vesicle preparation, glucose and 50 mM HEPES, adjusting to a final osmolarity of ~1200 mOsm.

2. Heat the solution in a 95 °C heat block until fully dissolved, vortexing occasionally to facilitate dissolution.

3. Transfer an aliquot to an Eppendorf tube and maintain at 45 °C until use. Since LGA can undergo thermogelation at room temperature, keeping it warm prevents premature gelation during mixing.

4. Mix the LGA solution with the vesicles at 45 °C, then dispense the mixture into a well plate for gel formation.

1. Prepare a precursor solution using the concentrations in Table 2. Briefly, dissolve 80 mM PEG4Nb, 20 mM PEG-4SH, a 1:1 stoichiometric ratio with PEG4Nb, and 16.9 mM LAP in 1 mL of PBS. You can even use DI water or any buffer. This is a photosensitive reaction, so prepare the precursor solution in a dark room and wrap the vial with foil for storage.

2. Vortex the solution at 2000 rpm for a minute and then keep at RT for 3 to 5 min.

3. Vortex again for 2 to 4 min and ensure all the chemicals are dissolved well. Repeat this part until a clear solution is obtained.

4. Put the desired volume in a 96-well plate and photopattern for exactly 60 s at 405 nm using DLP projectors. If needed, you can also do a 1:1 dilution in the well, for example 50 µL of precursor gel solution and 50 µL PBS in 1 well.

[fig-photo-photopatterning-peg4nb, file:general/photo-photopatterning-peg4nb.jpg, caption: (Photopatterning PEG4Nb hydrogels in a 96-well plate.)]

Reagent	Product Name	Conc.	Molecular weight	Storage Conditions	Supplier
PEG4Nb	4-Arm-PEG-Norbornene	80 mM	5000 Da	freezer at -20 °C in the dark	Creative Pegworks
PEG-4SH crosslinker	4-Arm PEG-Thiol	20 mM	2000 Da	freezer at -20 °C in the dark	Creative Pegworks
LAP photoinitiator	Lithium phenyl-2,4,6-trimethylbenzoylphosphinate	16.9 mM	294.21 g/mol	2 to 8 °C, in the dark, photosensitive	Sigma-Aldrich
glucose in 50 mM HEPES buffer (~1200 Osm)	—	1 mL	—	4 °C	Sigma-Aldrich
DLP projectors	—	—	—	—	Zwants Supplies Engineering (PRO4500-92-405)

Experimental design¶

Two vesicle conditions were tested, each in solution and in three hydrogels:

1. CPRG-loaded LUVs and β-galactosidase enzyme

[fig-diagram-cprg-cpr-reaction, file:general/diagram-cprg-cpr-reaction.png, caption: (CPRG is cleaved by beta-galactosidase to CPR and galactose.)]

Plate map: 15-QRtbOExMxpLzzK5mwe1lvsUzQgJlNAuqf9PWi7KRs.

Plating conditions

• Vesicles in outer solution only, no gel, serve as the reference control.

• Mixed 15 µL vesicle solution with 15 µL gel stock, then plated 30 µL per well. Plated all samples in duplicate.

Results¶

CPRG-loaded LUVs across hydrogels¶

[fig-luv-stability-hydrogels, notebook:luv_stability_hydrogels_matlabcode.txt, platemap:none, data source:20261009-113443-synergy2-OJMB-gfp-pla-guv-luv-gels-16UmL-LacZ.txt, caption: (Stability of CPRG-loaded LUVs across hydrogels, end-point A570/A412 ratio in four formats.)]

CPRG-loaded LUVs were embedded in each hydrogel, and the CPRG-to-CPR conversion was read as the endpoint absorbance ratio A570/A412 (n = 2 wells per condition, blank-corrected). Since CPRG is hydrolysed to the coloured product CPR only when the substrate meets β-galactosidase, a higher ratio means more of the dye has been exposed to the enzyme and converted. This reports loss of LUV integrity or leakage of the encapsulated CPRG. The PEG4Nb hydrogel gave the lowest ratio of all formats (about 0.4), so CPRG-loaded LUVs were most stable, with the least leakage and colour change, in the PEG4Nb hydrogel. The conversion was highest in LGA (about 1.5), above the free-in-solution LUV control (about 1.2) and ULGA (about 1.0). LGA therefore showed the strongest colour change, consistent with the greatest release of CPRG and its conversion to CPR. These are endpoint measurements on duplicate wells, so the differences are descriptive and not statistically tested.

GFP-expressing GUVs¶

Plate map 2: 15F2gLuDLFGxJRSbk73k1AJpaMZUq_iTpFFEllDdRHug.

Wells, as the source lists them:

• E12: GFP-GUVs+LUVs in solution and G12 is control

• E15: GFP-GUVs+LUVs in ULGA and G15 is control

• E14 precursor PEG4Nb hydrogel and E21 PEG4Nb hydrogel conditions with G14 and G21 as controls

• E16 GFP-GUVs+LUVs in LGA and G16 as control

All four GFP-GUV gel conditions are shown as interactive viewers, deep-linked by well, on Interactive viewers. So are the solution condition and its control: E12, G12, E14, E15, E16 and E21.

Archival figures for the four gel conditions, one field of view each, are below, stitched into one image with each well’s label as its panel header. Each panel is a single z-plane, a centre crop covering 1/9 of the field’s area. Brightfield is blended in at low weight for the gel matrix as context.

GFP contrast is grouped where the two wells’ signal levels allow it, and kept per-well otherwise. The PEG4Nb pre-gel and gel wells (E14, E21) share one stretch, 11864 to 20155 counts, since their GFP medians are close. ULGA and LGA (E15, E16) do not share a window. E15’s whole GFP distribution sits close to E16’s ceiling. Pooling the two, even with the high percentile raised to 99.99, left E15 a near-uniform wash and E16 almost dark. E15 is stretched at 8405 to 11247 counts, and E16 at 4158 to 7020 counts, each its own 50th-to-99.9th percentile. See general/render-microscopy-fovs.py and figures/fov-render-params.json for the full method.

[fig-fov-composite, file:figures/fov-composite.png, zarr-viewer, source:https://data.nucleus.engineering/microscopy/nucleus-bnext-01/OP-20261006_endpoint-vescile-gel-MB-zstack-E15_2026-10-06_12-58-55.990402.zarr, caption: (GFP-GUVs+LUVs across the four gel conditions: E14, E21, E15, E16. Contrast limits and grouping in the figure caption above.)]

GFP expression was detected in GUVs encapsulated in all three hydrogel formats: the photocrosslinked PEG4Nb hydrogel and the physical (thermogelling) hydrogels ULGA and LGA. Cell-free expression was therefore supported inside each gel network. GUVs have previously been shown to remain functional in thermogelling hydrogels. To our knowledge, this is the first demonstration of GFP expression, and therefore retained biological function, in a photocrosslinked hydrogel, which indicates that GUV-based expression systems are compatible with photocrosslinking. The PEG4Nb hydrogel showed considerably higher autofluorescence than ULGA and LGA, which gave lower background fluorescence. This reduces the signal-to-background ratio for GFP in PEG4Nb relative to the thermogels.

Discussion¶

This work shows that cell-free GFP expression can be carried out inside GUVs encapsulated in all three hydrogel formats tested: the photocrosslinked PEG4Nb hydrogel and the physical gels ULGA and LGA. We examined two vesicle types, GUVs and LUVs, so the platform is not tied to a single vesicle architecture. Earlier studies of vesicle-hydrogel interactions were largely limited to physical, thermogelling networks. By extending the work to a photocrosslinked gel, we widen the range of materials that can be combined with synthetic cells and make material fabrication and processing more reliable across gel types. Because expression was detected after gel formation, including exposure to photoinitiator and 405 nm light during photopolymerization, the vesicles appear to tolerate hydrogel formation and subsequent processing well enough to remain functional. This is a functional readout, and we did not separately quantify vesicle yield. How the crosslinking dynamics, such as light dose, crosslinking rate and network density, affect vesicle functionality is an open question, and a systematic study of it would be a worthwhile next step.

The formats also differ in how easily they can be handled. PEG4Nb and ULGA were easy to pipette. PEG4Nb has a practical advantage because it gels only when irradiated at 405 nm with a projector, not in response to temperature. The precursor therefore stays liquid until crosslinking is triggered, so the sample can be loaded and checked before gelation and expression begin. This is simpler than handling thermogels, whose viscosity depends on temperature during preparation.

Measurement remains the main limitation. Compared with solution-state assays, gels introduce refractive-index mismatch and autofluorescence, which complicate imaging and plate-reader quantification. PEG4Nb in particular showed substantial background fluorescence. Metrology suited to hydrogels will need further exploration, including alternative instruments, imaging modes, and data-processing approaches, to expand the platform. Future work should also quantify expression yield and GFP fluorescence, so that the conclusions rest on numerical comparisons across gels and not on detection alone.

Finally, this platform is a practical starting point for groups without a synthetic-biology background, because many of the control experiments needed to establish baseline behaviour have already been carried out.

Closed by the two revisions of 2026-10-11 and 2026-10-12:

• “[Paste the graph here]” and “[link to dataset URL here...]” are both filled.

• PEG4Nb has numbered preparation steps, as ULGA and LGA already did.

• The LGA transfer step is stated.

• ULGA’s formulation reference, previously “[insert link]”, is given.

• Figure 2 previously carried the placeholder caption “[Caption]”. It now has the caption the source gives it.

• The Discussion was five sentence fragments. It is now three written paragraphs.

• Figure 5 was an empty heading. It is now microscopy, carried as two viewers.

• The design list said four and listed two. It now says two and lists two. See gap-design-narrowed for what that cost.

• The PEG4Nb concentrations are settled. See gap-peg4nb-units.

Still open, and each has its own flag above:

• The Conclusion heading has no text.

• The plasmid question is open in words: is T7-EsaO-mNG linear or a plasmid, is T7-EsaO-PLA1 the same, and does either carry a tetO operator. Two sequence files were linked on the source on 2026-10-10, and they answer identity, not this. See The two DNA constructs could not be verified.

• Rows I and J, in gap-rows-ij-conflict, and whether that conflict carries into the two new plate maps.

• The PEG4Nb solvent and the plate format, in gap-peg4nb-solvent.

• The microscopy has no archival figure, in gap-microscopy-no-static.

• Three GFP claims need their evidence named, in gap-gfp-claims.

• The pre-gel condition cannot be measured by this assay, in gap-prepeg-dropped.

<!-- # Open questions

:::{admonition}

1. Are LUVs leaky?
2. Can PLA lyse?
3. Do only the gels lyse?

::: -->

<!-- Moved to the end of the page per an @claude comment left in this
file: "maybe open questions come at the end?" -->

<!-- # Resources

The full asset chain, the review history, and every flagged
disagreement between sources are in `main-supplement.md`, in this same
directory. It is not in `curvenote.yml`'s `toc`, so it does not publish
as a page of this DevNote. It is a plain file in the repository. -->