Catalytic processing in ruthenium-based polyoxometalate coacervate protocells

The development of programmable microscale materials with cell-like functions, dynamics and collective behaviour is an important milestone in systems chemistry, soft matter bioengineering and synthetic protobiology. Here, polymer/nucleotide coacervate micro-droplets are reconfigured into membrane-bounded polyoxometalate coacervate vesicles (PCVs) in the presence of a bio-inspired Ru-based polyoxometalate catalyst to produce synzyme protocells (Ru4PCVs) with catalase-like activity. We exploit the synthetic protocells for the implementation of multi-compartmentalized cell-like models capable of collective synzyme-mediated buoyancy, parallel catalytic processing in individual horseradish peroxidase-containing Ru4PCVs, and chemical signalling in distributed or encapsulated multi-catalytic protocell communities. Our results highlight a new type of catalytic micro-compartment with multi-functional activity and provide a step towards the development of protocell reaction networks.

Optical microscopy was performed on a Zeiss Axioskop transmitted light microscope at 4x, 10x or 20x magnification. Images were analysed with ImageJ software.
Raman analysis was performed on a Renishaw 2000 laser system, with samples acquired at an excitation wavelength 480 nm for 10 minutes with 60 scans. Samples of Ru4PCVs and PTACVs were lyophilized and sputtered on a microscope slide for measurements.
ICP-AES analysis was performed on an Agilent 710 ICP-OES fitted with an Agilent SPS 3 auto-sampler. The instrument was calibrated using standard solutions of 0.2 ppm, 0.4 ppm, 0.6 ppm, 0.8 ppm and 1 ppm prepared from solution of the single element ruthenium (in 7% wt. HCl, 1000 ppm) and tungsten (in nitric acid and hydrofluoric acid, 1000 ppm). For measurements, samples of Ru4PCVs (5.7 mg) and PTACVs (5.3 mg) were freshly prepared, lyophilized, and digested in H2SO4 (98 %, 1 mL) for five days. The resulting solutions were evaporated to almost dryness (heating ca. 4 h), and subsequently redissolved in nitric acid (1 %, 25 mL) and filtered (22 µDa particle size) before measurements. Because of the presence of the precipitate that needed to be filtered, the amounts of heavy metals might be underestimated.
X-ray photoelectron spectroscopy (XPS) was carried out using an Argus spectrometer working at a base pressure of 4.0 × 10 -11 mbar. Core-level photoemission spectra were acquired in grazing incidence, i.e. 45° between the sample surface and the normal of the electron analyzer, with a monochromatic Al Kα (1,486.7 eV) source. The pass energy was set to 20 eV, 50 eV and 100 eV for high-resolution, survey, and high-throughput regions, respectively. The measurements were acquired at room temperature. The electron charging was neutralized by a flood gun. The binding energy scale was referenced to the carbon C-C bond in the C 1s photoemission line at 284.8 eV. Ru4PCVs and PTACVs samples were measured as lyophilized powders.
Molecular oxygen concentration was measured using an OceanOptics NeoFox and FOSPOR-R probe with silicon overcoat. Before every measurement, the sensor was calibrated using a two-point calibration: 0 % oxygen was acquired under argon, and 20.9 % oxygen in the open air. Data was logged using NeoFox Viewer Ver. 2.40 averaging over 10 readings.

Kinetic studies of Ru4PCV-mediated decomposition of H2O2
In a 1.75 mL glass vial, different amounts of Ru4PCVs (0.38, 1.50 and 2.25 mg) were dispersed in PBS buffer (10 mM, pH 6.5) to obtain a final volume of 500 µL. Subsequently, the vial was sealed, the oxygen probe was inserted through the seal to be held above the liquid, and the PCV dispersion was stirred and purged with Ar gas. A pre-purged H2O2 solution (200 µL, 30 w/w %) was added using a micro-syringe and the evolution of oxygen gas was monitored for 60 min. Control experiments were conducted following the same general protocol, but either in the absence of H2O2 (by injecting pre-purged H2O), or in the absence of Ru4POM (using PTA-CVs).
The Ru4POM-mediated decomposition of H2O2 follows second order kinetics 1 : where R is the reaction rate, k is the second order rate constant, [H2O2] is the concentration of hydrogen peroxide (2.8 M), and [Ru4POM] is the concentration of inorganic synzyme (0.96 ± 0.02 nmol mg -1 of PCVs). By using this equation it was possible to calculate the catalytic response of Ru4PCVs; typically, k = 43.0 x 10 -3 M -1 s -1 (5.90 x 10 -8 mg -1 s -1 ) (see Supplementary Figure 7).

Oxygen bubble nucleation in Ru4PCV-containing aminoclay/DNA synthetic protocells
Statistical studies: All experiments were performed on freshly made aminoclay/DNA protocells containing different amounts of Ru4PCVs (1, 2 and 3 batches; see Methods Ru4PCV-containing aminoclay/DNA synthetic protocells). Controlled numbers of aminoclay/DNA multicompartmentalized synthetic protocells (typically 100 µL of concentrated dispersion) were transferred into well plates using a glass pipette, and 100 µL hydrogen peroxide (30 w/w%) were added. The wells were monitored using an optical microscope, and the statistics related to the number of protocells capable of nucleating oxygen bubbles within their lumen as a function of the amount of Ru4PCV proto-organelles was determined.
Bubble growth rate: Buoyant microcapsules were transferred using a glass pipette into a well plate containing a mixture of 100 µL of H2O2 and 100 µL of dilute dispersion of AMP clay (0.5 mg/mL). The growth of the microbubbles was monitored using an optical microscope and the image stacks were analysed using ImageJ software to quantify the oxygen bubble growth rate.
Protocell buoyancy experiments: Approximatively 300 µL of a dilute dispersion of aminoclay (0.5 mg/mL) were pipetted into a plastic cuvette followed by careful addition of 100 µL of hydrogen peroxide (30 w/w%) using a pipette to ensure that that the substrate and clay dispersion were preferentially located at the bottom of the cuvette. Ru4PCV-containing aminoclay/DNA synthetic protocells or PTA-CV-containing aminoclay/DNA microcapsules were then transferred into the cuvette using a glass pipette and allowed to sink under the influence of gravity. To capture the vertical (z direction) motion of the protocells, a Zeiss Axioskop microscope equipped with a Canon EOS 500D camera and a 1.25x objective lens was aligned side-on to the cuvette to switch the viewing plane from xy to xz.

Parallel catalytic processing in synzyme protocells.
An aqueous solution of ATP (100 μL, 50 mM,) was added to a mixture (1 mL) of PDDA (5 mM monomer, 100-200 kDa) and horseradish peroxidase (HRP, 0.01 mg/ml) in the presence of two orthogonal acoustic standing waves (6.76/6.78 MHz, 10 V). After 30 min, the supernatant was carefully removed and exchanged with Milli-Q water three times without disturbing the array of coacervate microdroplets. The 2D HRP-containing PCV array was prepared by injecting a mixture of PTA and Ru4POM (7.5 µL, PTA/Ru4POM 7:1 mol/mol; [PTA] = 17.6 mM; [Ru4POM] = 2.4 mM; pH 6.5) from one side of the acoustic trapping chamber. After 1 h of the injection, the supernatant in the acoustic trapping chamber was carefully removed and exchanged with Milli-Q water for three times. To avoid concentration gradients for the substrates, in an Eppendorf tube H2O2 (4 uL, 10 mM) and Amplex Red (1 uL, 10 mM) were diluted to 500 µL with MilliQ water. Subsequently 500 µL of supernatant were removed from the chamber, and replaced with the freshly made H2O2/Amplex Red dilute solution. The added solution was rendered homogeneous by gentle pipetting. The final concentrations of H2O2 and Amplex Red in the chamber were 40 µM and 10 µM, respectively. Fluorescence microscopy (λex = 515 -560 nm and λem = 580 nm) was used to monitor the HRP-mediated conversion of the non-fluorescence Amplex Red to the fluorescence resorufin over time.

Chemical communication within dispersed PTA-CV/Ru4PCV enzyme/synzyme consortia
FITC-tagged-GOx-filled PTA-CVs (25 μL of a stock dispersion in 500 uL of PBS buffer 0.01 M pH 6.5), RITCtagged-HRP-filled PTA-CVs (25 μL) and Ru4PCVs (25 μL) were prepared as described in Methods FITCtagged GOx-or RITC-tagged HRP-containing PTA-CVs and Preparation of Ru4PCVs, washed as described in Methods Preparation of Ru4PCVs, and homogeneously mixed in an Eppendorf tube with PBS buffer (115 µL, 10 mM, pH 6.5) before depositing them on a Petri dish. Substrates glucose (100 μL, 150 mM in PBS 10 mM pH 6.5) and o-PD (10 µL, 30 mM in PBS 10 mM pH 6.5) were added, and images were acquired every 15 s. for 1 h. Images were captured simultaneously on four different channels using a fluorescence confocal microscope, as described in Supplementary Table 1. Brightfield images were also acquired to visualize the non-fluorescent Ru4PCVs. All image stacks were then analysed with ImageJ. to give a triple concentrate of Ru4PCV dispersion of 6 mg mL -1 , instead of 1.5 mg mL -1 . The microplate was prepared by introducing in sequence GOx-filled PTA-CVs (25 μL), HRP-filled PTA-CVs (25 μL), variable volumes of Ru4PCVs (0, 25, 70 and 140 μL), and variable volumes of PBS buffer (140, 115, 70 and 0, 10 mM pH 6.5). The substrates o-PD (10 μL,30 mM in PBS 10 mM, pH 6.5) and glucose (100 μL, 150 mM in PBS 10 mM, pH 6.5) were automatically injected immediately before the measurements. The final solution volume in each well was 300 μL. The o-PD solution was prepared immediately before use to minimise degradation from atmospheric oxygen. The kinetics were tracked by fluorescence spectroscopy (λexc = 405-410 and λem = 550 nm) by acquiring data points every 2 s for 15 min. The 2,3-DAP concentration over time was then determined by allowing the enzymatic reaction to go to completion and determining the fluorescence intensity when all o-PD was converted to product.

Fabrication and enzyme/synzyme activity of PTA-CV/Ru4PCV-containing proteinosomes
FITC-tagged-GOx-filled PTA-CVs, RITC-tagged-HRP-filled PTA-CVs and Ru4PCVs were prepared separately as described in Methods FITC-tagged GOx-or RITC-tagged HRP-containing PTA-CVs and Preparation of Ru4PCVs, allowed to sediment to the bottom of an Eppendorf tube, and the supernatant was removed. Subsequently, 10 µL of each PCV population were mixed together in a separate Eppendorf tube. In another Eppendorf tube 15 µL of an aqueous solution of RITC-labelled BSA/PNIPAM nanoconjugates (8 mg/mL) were mixed with 5 µL of glucose solution (0.55 M in PBS buffer 10 mM, pH 6.5) and 5 µL of o-PD (11 mM in PBS buffer 10 mM, pH 6.5). This solution was quickly injected into the Eppendorf tube containing the ternary PCV mixture, the dispersion gently mixed, and then injected into a 1.75 mL glass vial containing 1 mL of 2-ethyl-1-hexanol. The vial was shaken for 30 s to obtain PCVcontaining proteinosomes as water-in-oil emulsion droplets. The sample was then loaded into a microscope channel slide, which was sealed to avoid evaporation. Images were acquired every 30 s for 2 h using a fluorescence confocal microscope. Images were captured simultaneously on four different channels, as described in Supplementary Table 2. Brightfield images were also acquired to visualize the non-fluorescent Ru4PCVs. All images were then analysed with ImageJ.  Table (bottom).