Autonomic Integration in Nested Protocell Communities

The self-driven organization of model protocells into higher-order nested cytomimetic systems with coordinated structural and functional relationships offers a step toward the autonomic implementation of artificial multicellularity. Here, we describe an endosymbiotic-like pathway in which proteinosomes are captured within membranized alginate/silk fibroin coacervate vesicles by guest-mediated reconfiguration of the host protocells. We demonstrate that interchange of coacervate vesicle and droplet morphologies through proteinosome-mediated urease/glucose oxidase activity produces discrete nested communities capable of integrated catalytic activity and selective disintegration. The self-driving capacity is modulated by an internalized fuel-driven process using starch hydrolases sequestered within the host coacervate phase, and structural stabilization of the integrated protocell populations can be achieved by on-site enzyme-mediated matrix reinforcement involving dipeptide supramolecular assembly or tyramine–alginate covalent cross-linking. Our work highlights a semi-autonomous mechanism for constructing symbiotic cell-like nested communities and provides opportunities for the development of reconfigurable cytomimetic materials with structural, functional, and organizational complexity.

measurements were used to characterize the charge of the silk-based polymers (CSF and SF, 1 mg/ml) using a zeta potentiometer analyser (Malvern Instruments, UK).
The primary amino content in different silk-based polymers was determined using TNBSA (2,4,6-trinitrobenzene sulfonic acid), which reacts with primary amino group (-NH 2 ) to generate a coloured product. [2] Typically, CSF and SF were separately diluted to 0.2 mg/ml in Na 2 CO 3 /NaHCO 3 buffer (0.1 M, pH 8.5), followed by additions of 0.25 ml TNBSA (0.05%, w/v, buffer 8.5) to each group (0.5 ml) and incubation at 37 o C for 2 h. 0.25 ml HCl (0.5 M) and 0.5 ml SDS were then added to each sample to stop and stabilize the reaction before analysis by UV/VIS spectroscopy (250-700 nm, Lambda 35, PerkinElmer, USA). Using the above protocol, BSA (0.05, 0.1, 0.15, 0.2 mg/ml) was utilized as a standard reagent to produce a calibration curve (absorption at 335 nm vs -NH 2 content) to determine the primary amino content in the CSF or SF polymers.

Preparation and characterization of alginate/CSF coacervates
Silk-based coacervates were produced via an electrostatically mediated associative liquid-liquid phase separation in mixtures of positively charged CSF and negatively charged alginate (Mw:140-160 kDa). The mixtures were prepared at room temperature at different charge ratios. The alginate [COOH] : CSF [NH 2 ] charge ratio was estimated as the ratio of alginate monomer concentration (x1 COOH per monomer) to the primary amino concentration of CSF determined by TNBSA analysis. The charge ratio was controlled by compositional changes of [COOH] (alginate) and [NH 2 ] (CSF) or via pH manipulation through additions of NaOH or HCl.
Typically, positively charged coacervate vesicles (CV), positively charged multicompartmentalized coacervate droplets (MCV) and neutral coacervate droplets (CD), were prepared by increasing the volume of a stock alginate solution ([COOH] 100 mM, 0.5-20 μl) added to a stock CSF solution ([NH 2 ] 8.98 mM, 11-67 μl) under stirring, followed by additions of different amounts of H 2 O to make up the total volume to 100 μl for each sample. The final [COOH] (alginate) concentrations were 0.25-20 mM and the [COOH] (alginate) : [NH 2 ] (CSF) ratios were between 0.05-3. Alternatively, different microstructures were produced by increasing the volume of added alginate ([COOH] 100 mM, 2-8 μl) to stock CSF solutions ([NH 2 ] 8.98 mM, 22-88 μl) followed by additions of different amounts of H 2 O to make up the total volume to 100 μl for each sample to give final [COOH] (alginate) and [NH 2 ] (CSF) concentrations of 2/2, 4/4, 6/6 or 8/8 mM. Stoichiometric aliquots of NaOH or HCl were then added to each group to control the pH from pH 3 to 9. The silk-based coacervate microstructures were imaged using a Leica DMI3000 B fluorescence microscope (Leica, Germany) and/or a SP5-II confocal laser scanning microscope (LSCM, Leica, Germany) and analysed using Image J software. Zeta potential distributions were used to characterize the surface charge of the different alginate/CSF microstructures (Malvern Instruments, UK).

Preparation of proteinosomes
Synthesis of PNIPAAm polymer, preparation of BSA/PNIPAAm nanoconjugates and assembly of proteinosomes were achieved by following a reported procedure [2] . Typically, an aqueous mixture (40 μl) comprising BSA/PNIPAAm nanoconjugates (30 mg/ml, 15 μl), Na 2 CO 3 /NaHCO 3 (100/100 mM, pH 8.5, 15 μl) and dissolved PEG-NHS (cross-linker, 2.5 mg, 10 μl) were mixed with 650 μl 2-ethyl-1hexanol (oil phase) under mechanical disturbance to produce a dispersed water-in-oil Pickering emulsion droplets. The emulsion was left at 4 o C for 2 days to stabilize and precipitate the emulsion droplets. The precipitated emulsion droplets were then transferred into 1 ml of 70 % ethanol after removal of the oil supernatant, followed by centrifugation at 10000 rpm for 5 min to precipitate the proteinosomes and removed the supernatant. Redispersion and centrifugation of the proteinosomes were then repeated twice using 1 ml 40 % ethanol and then DI water. The proteinosome aggregate was then dispersed into 300 μl DI water and stored at 4 o C. If required, proteinosomes with smaller size could be produced by sonicating the emulsion droplets in an ultrasonic bath at 300 W for 1 min.
The number densities present in single populations of the coacervate vesicles and proteinosomes were determined using LSCM and Image J software. The number of protocells in a unit volume were counted. Typically, when mixed, the proteinosome and coacervate vesicle final number densities were approximately 1 x 10 7 /ml and 2 x 10 7 /ml, respectively. Thus, experiments involving the capture of proteinosomes in positively charged coacervate vesicles were undertaken at an approximate proteinosome : vesicle number ratio of 1 : 2.
Changes in pH were monitored using a pH meter (METTLER TOLEDO, Switzerland) controlled by Lab X software. LSCM was used to record the formation and reconfiguration of the hybrid protocells protocell at different pH values. FACS analyses were undertaken on single populations of positively charged coacervate vesicles ([NH 2 ] 4 mM, [COOH] 4 mM, pH ca. 6.0), membrane-less coacervate droplets ([NH 2 ] 4 mM, [COOH] 4 mM, pH ca. 9.0) or enzyme-containing proteinosomes. FACS analyses were also undertaken on binary populations of positively charged vesicles and proteinosomes before and after the stepwise addition of urea (30 mM) and glucose (60 mM). All investigations used a FACS Canto II flow cytometer operating at a low pressure with a 100 μm sorting nozzle. 2D pseudo-color plots of the FSC-A and SSC-A light were determined for a total of 20,000 particles in the single or binary populations. Data analysis was performed on FlowJo 7.6 software.
Glucose production by enzymatic hydrolysis of soluble starch An integrated enzyme system comprising α-amylase and amyloglucosidase was used to efficiently hydrolyze soluble potato starch to glucose. Amylase was used to hydrolyse starch polymers to maltose, maltotriose, dextrin and minimal amounts of glucose, while amyloglucosidase was included to hydrolyse the maltose/maltotriose/dextrin products to glucose.
Test experiments to assess the efficiency of starch hydrolysis were undertaken as follows. A concentrated starch solution (500 mM, monomer concentration) was prepared by dissolving potato starch in DI water under stirring and leaving the solution at 80 o C for 2 h. The starch solution was then diluted to 0-90 mM by adding aliquots of the solution (500 mM, 0-18 μl) to different volumes of DI water to make up the total volume to 90 μl. An amylase/amyloglucosidase solution (10 μl, 200/200 IU/ml) was then added to each sample and the mixtures incubated at 25 o C for 0-24 h to enable starch hydrolysis and glucose production. The reaction was stopped at various times by denaturing the integrated enzymes with 50 μl NaOH (1 M), followed by sonication for 10 min and then addition of 50 μl HCl (1 M). The supernatants were collected by centrifugation at 10,000 rpm for 5 min and the precipitates discarded. The supernatants were then diluted 10-times with DI water and if required stored at room temperature. A mixture comprising 100 μl oPD (100 mM), 100 μl GOx (10,000 IU/ml), 100 μl HRP (200 IU/ml) and 400 μl PBS buffer (pH 6.0) was added to the supernatants (100 μl) and the reaction mixtures incubated at 37 o C for 30 min. The production of 2,3-DAP from the GOx/HRP enzyme cascade was stopped at different times by addition of 200 μl HCl solution (3 M), followed by 5-times dilution. Glucose concentrations were quantitively measured by UV/VIS absorption spectroscopy using the 2,3-DAP absorbance at 490 nm. Standard plots obtained from known glucose concentrations (0-0.2 mM) were used to calibrate the GOx/HRP assay.

Determination of partitioning for starch and starch hydrolases
Soluble starch (500 mM, 1 μl), amylase (10 mg/ml, 1 μl) or amyloglucosidase (10 mg/ml, 1 μl), were separately mixed with alginate/CSF coacervate droplets (200 μl, [NH 2 ] 4 mM, [COOH] 4 mM, pH 4). Partition constants were obtained by calculating the ratio of red fluorescence intensity (grey value) inside and outside individual coacervate droplets imaged by LSCM images and analysed using Image J software. Transformation of the loaded hybrid droplets to positively charged coacervate vesicles was undertaken by addition of aqueous NaOH (0.05 M) to increase the pH to ca. 9.0. The distribution of starch and the starch hydrolases in the coacervate vesicles was then characterized by LSCM.
Proteinosome-in-coacervate droplets (100 μl, RITC-CSF, [NH 2 ] 8 mM, [COOH] 8 mM, pH 9.0, proteinosome number density, 2 x 10 7 /ml) were prepared in the presence (1.5 mg/ml) or absence (0 mg/ml) of Fmoc-AA-OH. Glucose was then added (0-24 μl, 500 mM) to the suspension, followed by different amounts of DI water to make the total volume up to 200 μl in each group of experiments. Samples were then incubated at room temperature for up to 12 hours and the pH monitored at different time points (0-12 h) and images recorded by LCSM and SEM (Jeol IT300 SEM, Japan). SEM samples were prepared by mounting a drop (10 μl) of the suspensions (10 times diluted) onto the surface of PEG functionalized cover slips and then drying in a N 2 flow for 15 min, followed by a coating with silver prior to SEM imaging.
Changes in the fluidity of the membrane domain in proteinosome-in-coacervate vesicles containing Fmoc-AA-OH filaments were determined by FRAP experiments. Briefly, the dynamic fluorescence recovery of the coacervate matrix was measured using time-sequence LSCM images (over 25 min) of the samples after photo-bleaching localized regions of the host coacervate phase. Corresponding LSCM images were analyzed using Image J software. Similar experiments were done using the unmodified coacervate vesicles containing proteinosomes as well as proteinosome-incoacervate droplets with or without sequestered Fmoc-AA-OH.
The structural stability of the proteinosome-in-coacervate vesicles after infiltration with Fmoc-AA-OH filaments was investigated by exposing the hybrid protocells to a NaCl diffusion gradient (4 M). The NaCl gradient was generated by adding aqueous NaCl (10 μl, 4 M) to one end of a circular glass channel containing a suspension of the proteinosomes-in-coacervate vesicles (50 μl). Timedependent LSCM images were recorded and Image J software used to evaluate the time-dependent changes in fluorescence intensities (grey value) of the host coacervate phase and surrounding waterfilled environment. Similar experiments were done using the unmodified coacervate vesicles containing proteinosomes as well as proteinosome-in-coacervate droplets with or without sequestered Fmoc-AA-OH.
Retention of enzyme activity in the Fmoc-AA-OH modified proteinosome-in-coacervate vesicles under high ionic strength was investigated by monitoring the lipase-mediated hydrolysis of non-fluorescent calcein-AM to produce a calcein green fluorescence output. Fmoc-AA-OH-containing proteinosome-in coacervate vesicle suspensions (200 μl, [NH 2 ] 4 mM, [COOH] 4 mM) were prepared as above but with the inclusion of an acid-resistant lipase (0 or 1 μl, > 100 LCLU/g). The samples were placed in a circular glass channel and a drop (30 μl) of an aqueous solution containing calcein-AM (ca. 5.0 x 10 -2 mM) and NaCl (2.3 M) added at one end to generate a chemical diffusion gradient. Timedependent images of the hybrid protocells were recorded over 0-8 h by LSCM.

Host-mediated matrix reinforcement in proteinosome-in-coacervate vesicles
Tyramine-functionalized alginate (alginate-Tyr) was synthesized by an EDC/NHS-activated amination reaction. Briefly, 200 mg of sodium alginate and 500 mg of Tyramine · HCl were blended and then dissolved in 10 ml MES buffer (0.2 M, pH 4.7). The reaction was then started by addition of 160/120 mg of EDC/NHS. The reaction mixture was left at room temperature and under stirring for 24 hours and the alginate-Tyr product purified by dialysis against DI water for 3-4 days using a cellulose dialysis tube (Sigma, MWCO: 12-14 kDa) and then lyophilized. The product was characterized using FI-TR spectroscopy (4000-550 cm -1 , PerkinElmer, USA) and UV/VIS spectroscopy (250-700 nm, Lambda 750, PerkinElmer, USA).
As a test experiment, a suspension of GOx-containing proteinosomes (4 mg/ml, 4 x 10 7 /ml, 100 μl) and HRP solution (200 IU/ml, 0 or 15 μl) were added to an alginate-Tyr solution (10 mg/ml, 200 ul), followed by addition of different amounts of DI water to give a final volume of 400 μl. Gelation was then achieved by addition of 12 μl glucose solution (1 M) to the mixture and an incubation at room temperature for 0-6 hours. UV/VIS absorption spectra of the reaction mixture was collected at different time points (0-6 h). Gelation was estimated by inversion of a glass vial containing this reaction mixture.
Host-mediated matrix reinforcement of proteinosome-in-coacervate vesicles was undertaken as follows. Positively charged coacervate vesicles were prepared by mixing aqueous solutions of RITC-CSF ([NH 2 ] 8 mM) and alginate-Tyr) ([COOH] 10 mM) at pH 4.0 (total volume, 100 μl). Urease/GOx/gelatin-containing proteinosomes (GOx 4 mg/ml, 4 x 10 7 /ml, 50 µl) were then added at an approximate proteinosome : vesicle number ratio of 1 : 2, followed by addition of different volumes of RITC-labelled HRP solution (0-20 µl, 200 IU/ml) and DI water to give a final volume of 200 μl and the pH raised to 9.0 to capture the proteinosomes. Glucose was then added (1.5 M, 4 μl) and the samples incubated at room temperature for 12h. GOx-mediated acidification transformed the coacervate droplets into vesicles while GOx-induced production of hydrogen peroxide gave rise to crosslinking of alginate-Tyr chains in the coacervate matrix. LSCM was used to image the samples before and after addition of glucose. Images were analysed by using Image J software.
The structural stability of the proteinosome-in-coacervate vesicles after alginate-Tyr crosslinking was monitored by LSCM using the following procedures; (i) incubation of the samples for 3 h at room temperature at pH 9 (NaOH) or 8.7 (urease/urea), and (ii) placing the samples in a circular glass tube and exposing the protocells to a diffusion gradient of aqueous NaOH (25 mM) or aqueous sodium chloride (3 M) at pH 4 as described above.

Fuel-driven three-enzyme network in proteinosome-in-coacervate vesicles
Alginate/CSF coacervate droplets ([NH 2 ] 8 mM, [COOH] 8 mM, RITC-CSF, 100 μl) containing a sequestered amylase/amyloglucosidase mixture (40/40 IU/ml) along with captured GOx-containing proteinosomes ([GOx] = 4 mg/ml, number density = 1 x 10 7 /ml) and HRP-containing proteinosomes ([HRP], 4 mg/ml, number density = 1 x 10 7 /ml) were prepared at pH 9. Soluble starch (500 mM, 0-24 μl) and oPD (300 mM, 20 μl) were then added followed by different amounts of DI water to make up the total volume to 200 μl in each group. The samples were then incubated at room temperature for 12 h and the pH recorded at various time intervals (METTLER TOLEDO, Switzerland). 10 μl of the reaction mixture was taken at different time points (1, 3, 6 h) and then diluted with 890 μl DI water, followed by addition of 100 μl HCl (3 M) to quench oPD oxidation. The stabilized reaction mixtures were then centrifugated at 10,000 rpm for 5 min to remove any precipitates prior to the UV/VIS absorption measurement. The 2,3-DAP output from the starch hydrolase/GOx/HRP enzyme cascade was measured over time by monitoring the intensity of the absorption band spectroscopy at 490 nm. Time-dependent LSCM images and line profiles were recorded on individual guest/host protocells undergoing 2,3-DAP production.
Control experiments involving the addition of soluble starch (40 mM) to coacervate droplets with single or double populations of GOx-and HRP-containing proteinosomes prepared in the absence of amylase/amyloglucosidase were also undertaken. Test experiments were initially undertaken on coacervate-free suspensions of binary proteinosome populations as follows. A mixture of amylase/amyloglucosidase (200/200 IU/ml, 20 μl) and oPD (300 mM, 20 μl) was added to a binary population of GOx-and HRP-containing proteinosomes in aqueous solution (100 μl; number ratio, 1:1; number density, 2 x 10 7 /ml). Aliquots of soluble starch were added (500 mM, 0-24 μl) followed by different amounts of DI water to make up the total volume to 200 μl in each group an incubation at room temperature for 6 hours. Control experiments involving the addition of soluble starch (16 μl, 500 mM, total volume, 200 μl) to single or double populations of GOx-and HRP-containing proteinosomes prepared in the absence of amylase/amyloglucosidase were also undertaken.
The level of proteinosome disassembly was determined by addition of different amounts of glucose (200 mM, 0-20 μl), HRP (300 IU/ml, 0 or 10 μl) and TCEP (400 mM, 2.5 μl) to proteinosome suspensions (100 μl, number density, 3 x 10 4 /ml), followed by addition of different volumes of DI water to make up the total volume to 200 μl, and incubation at room temperature for 6 hours. The incubated mixtures were then diluted with 200 μl NaCl solution (1 M, coacervate disassembly) and centrifuged at 10,000 rpm for 10 min to discard the precipitates prior to the detection of fluorescence emission in the supernatant. Three parallel samples were used in each group for statistical analysis. Intrinsic background fluorescence (I b ) was determined by monitoring the incubated proteinosome population in the absence of TCEP as a control group.
The percentage of proteinosome disassembly was determined by using the followed equation: where I i and I 0 are the intensity of fluorescence emission (430 nm) of samples in the presence and absence of glucose addition.

TCEP-induced disassembly of guest proteinosomes in hybrid coacervate droplets and vesicles
A suspension of GOx/gelatin/proteinosomes-in-coacervate droplets ([NH 2 ] 4 mM, [COOH] 4 mM; GOx, 4 mg/ml; Dylight 405 labelled-proteinosomes, number density, 1 x 10 7 /ml; pH ~ 9.0) was prepared. HRP (0 or 10 μl, 300 IU/ml) was then sequestered into the coacervate phase, and the samples incubated at room temperature for 15 min. Disassembly of the guest proteinosomes was started by addition of TCEP (2.5 μl, 400 mM, pH ~ 9.0) in the absence or presence of added volumes of glucose (500 mM, 0-8 μl), followed by different additions of DI water to make up the total volume to 200 μl in each group and incubation at room temperature for 6 h. LSCM used to record images of the coacervate droplet to vesicles reconfiguration and disintegration of the guest proteinosomes 6 h after addition of glucose. Guest proteinosome disassembly was also quantitively characterized by measuring the changes in Dylight 405 blue fluorescence emission in the aqueous continuous phase. Typically, the supernatants of the above samples were collected by centrifugation at 10,000 rpm for 5 min after a dilution with 200 μl NaCl solution (1 M) for coacervate disassociation. Supernatants were examined using a FluoroMax-4 Spectrofluorometer (HORIBA Scientific, Japan; excitation 400 nm, emission 410-650 nm) and the percentage of percentage of proteinosome disassembly calculated from equation 1. Scheme showing supramolecular assembly of Fmoc-AA-OH nanofilaments in semi-permeable GOx-containing proteinosomes exposed to glucose and the dipeptide. Addition of glucose and Fmoc-AA-OH results in acidification of the proteinosome interior, which in turn induces filament assembly initially inside and on the surface of the protocells, and then in the external environment leading to hydrogelation of the medium. (B) Time-series of fluorescence microscopy images showing populations of GOx-containing proteinosomes (blue; initial conditions, t 0 , pH 0 ~ 9.0, Na 2 CO 3 /NaHCO 3 10/10 mM) after addition of glucose (40 mM) and Fmoc-AA-OH (2 mM, 0.75 mg/ml). Self-assembled Fmoc-AA-OH filaments are observed as the pH decreases to 7, and extensively hydrogelled proteinosomes are observed at pH 6. Lower pH values give rise to filament formation in the external environment due to acidification of the bulk solution. Fmoc-AA-OH filaments are stained with Nile red. Scale bars are 20 μm. Figure 14. UV spectrum (black plot) showing characteristic peaks for Fmoc-AA-OH in GOx/proteinosome-in-coacervate vesicles after supramolecular assembly ([NH 2 ] 4 mM, [COOH] 4 mM, pH ~ 4.0, Fmoc-AA-OH 0.75 mg/ml). The modified guest/host protocells were collected by centrifugation (10,000 rpm, 5 min, 200 ml) and the supernatant removed, followed by treatment with 2 ml Na 2 CO 3 /NaHCO 3 buffer (100/100 mM) at pH ca. 9.0 to release Fmoc-AA-OH if present. The solutions were then analysed by UV spectroscopy. The spectrum associated with corresponding experiments undertaken without Fmoc-AA-OH is also shown (grey plot). Disassembly of Fmoc-AA-OH nanofilaments in the coacervate membrane domain has no effect on reconfiguration to the proteinosome-in-coacervate droplet. The captured proteinosomes (blue) are not observed in the images recorded after 10 min due to their movement out of the focal plane. Scale bars are 20 μm. Live yeast cells (ca. 2.5 x 10 7 /ml) are initially incorporated at pH 9.0 into membrane-less alginate/CSF coacervate droplets (CD, light orange circle, [NH 2 ] 4 mM, [COOH] 4.5 mM) containing sequestered starch hydrolases (purple hexagons, amylase/amyloglucosidase). Addition of soluble starch induces endogenous production of glucose which subsequently initiates fermentation by the captured yeast cells to produce CO 2 and acidification (pH 4) of the protocell environment, resulting in membranization/reconfiguration of the coacervate droplets into coacervate vesicles (CV, dark orange circle) with captured living yeast cells. Compression of the coacervate phase during osmotically induced expansion results in expulsion of the yeast cells into the aqueous-filled lumen of CV where they proliferate by budding division to produce a densely packed viable population. (B) LSCM images of individual yeast@CD (upper row) and yeast/PCV (bottom row); the latter were recorded 36 h after starch (60 mM) addition. Yeast cells were stained with SYTO 9 (green, live cells) and PI (red, dead cells).