Liquid–Liquid Phase Separation‐Mediated Photocatalytic Subcellular Hybrid System for Highly Efficient Hydrogen Production

Abstract Plant chloroplasts have a highly compartmentalized interior, essential for executing photocatalytic functions. However, the construction of a photocatalytic reaction compartment similar to chloroplasts in inorganic–biological hybrid systems (IBS) has not been reported. Drawing inspiration from the compartmentalized chloroplast and the phenomenon of liquid–liquid phase separation, herein, a new strategy is first developed for constructing a photocatalytic subcellular hybrid system through liquid–liquid phase separation technology in living cells. Photosensitizers and in vivo expressed hydrogenases are designed to coassemble within the cell to create subcellular compartments for synergetic photocatalysis. This compartmentalization facilitates efficient electron transfer and light energy utilization, resulting in highly effective H2 production. The subcellular compartments hybrid system (HM/IBSCS) exhibits a nearly 87‐fold increase in H2 production compared to the bare bacteria/hybrid system. Furthermore, the intracellular compartments of the photocatalytic reactor enhance the system's stability obviously, with the bacteria maintaining approximately 81% of their H2 production activity even after undergoing five cycles of photocatalytic hydrogen production. The research brings forward visionary prospects for the field of semi‐artificial photosynthesis, offering new possibilities for advancements in areas such as renewable energy, biomanufacturing, and genetic engineering.


Figure S2 .
Figure S2.Prediction of protein disorder of (A) hydrogenase (HydA), (B) hydrogenase with his-tag (HH) using the IUPRed2 program.A score greater than 0.5 (above the horizontal line) indicates an unordered region, while a score less than 0.5 indicates an ordered region.The fusion of hydrogenase with MaSpI8 IDPs theoretically possesses the capability for LLPS.

Figure S4 .
Figure S4.(A) SDS-PAGE analysis of HM and its relevant maturases (HydEFG) under varying IPTG concentrations.The concentration of IPTG is 0.1, 0.5, and 1 mM.Under 0.1 mM IPTG conditions, the expression level of hydrogenase is highest.

Figure S5 .
Figure S5.(A) Hydrogen production is measured by detecting the hydrogen content in sealed bottle headspace after 24 hours of induction.The results reveal that the fusion of different fragments (2×His-tag and MaSpI8) at the C-terminus of hydrogenase do not affect its hydrogen production activity.Only a small amount of hydrogen gas is produced in the growth medium.(B) Comparing the cell growth of engineered strains that have undergone genetic modifications through monitoring optical density (OD600).Similar rates of cell growth indicate that the expressed protein has no adverse impact on cell growth.(C) At room temperature, HM protein appears cloudy when mixed with the macromolecular crowding agent Dextran-70.Rotate the mixture through centrifugation, and the bottom of the centrifuge tube shows the coexistence of two liquid phases, as shown by the red arrow in the photograph.

Figure S6 .
Figure S6.The bright-field and confocal laser scanning microscopy images of the E. coli expressing empty Vector (A) and HydA (B) stained with ThT reveal minimal green fluorescence.The fluorescence intensity line-cuts analysis showed no significant change in bacterial fluorescence.

Figure S7 .
Figure S7.CLSM images for the CdSe x S 1-x QDs mineralized in HydA/IBS.In this biohybrid systems, mineralization of CdSe x S 1-x QDs occurred simultaneously and exhibited consistent fluorescence trends within the cells.Scale bar = 1 μm.

Figure
Figure S8.(A) TEM images for the CdSe x S 1-x mineralized HM/IBSCS.Cd, Se, and S elements exhibit spatially oriented distribution, with aggregated distribution at the two poles of E. coli.(B) The result of EDS analysis of the precipitated CdSe x S 1-x nanoparticles.(C) Relative abundance of Cd, Se, and S elements within and outside the condensates in the HM-expressing cells.

Figure S9 .
Figure S9.HRTEM images for the QDs mineralized in HydA/IBS (A) and HH/IBS (B).The result of EDS analysis of the precipitated Se particle.The white arrows pointed to the large-sized Se precipitate.

Figure S11 .
Figure S11.The MTT reduction activity was used to represent cell activity after treatment with 1 mM CdCl 2 and Na 2 SeO 3 ,

Figure S13 .
Figure S13.TEM characterizes two different morphologies of E. coli present in the HM/IBSCS.(A) TEM images of HM E. coli mineralized by CdSexS1-xQDs.Two types of bacteria were observed to exhibit significant adhesion (blue border), with only a unipolar dark region at the extreme of the bacteria, which may be due to bacterial division.(B) The TEM image reveals mineralization of CdSexS1-x QDs at bipolar of the HM/IBSCS.(C) The TEM image demonstrates the unipolar mineralization of CdSexS1-x QDs at bacteria.(D) Cd, Se, and S elements exhibit spatially oriented distribution, with aggregated distribution at the single poles of E. coli.(E) Bio-TEM image of thin-sectioned HM/IBSCS, which is undergoing a division process, with a noticeable depression in the middle part of the cell, which is a characteristic of the division process.(F) Bio-TEM image of thin-sectioned HM/IBSCS shows unipolar.Thin-sectioned HM/IBSCS Bio-TEM images display unipolar subcellular compartments.

Figure S14 .
Figure S14.In vitro phase separation system based on HH protein.(A) Prediction of protein disorder of hydrogenase fusion with his-tag (HH) using the IUPRed2 program.A score greater than 0.5 (above the horizontal line) indicates an unordered region, while a score less than 0.5 indicates an ordered region.The HH theoretically does not have the ability for LLPS.(B) The particle size of the HH solution did not change before and after mixing (C) with the macromolecular crowding agent dextran 70.(D) The bright-field and confocal laser scanning microscopy images of the E. coli expressing empty HH stained with ThT reveal minimal green fluorescence.The fluorescence intensity line-cuts analysis showed no significant change in bacterial fluorescence.(E) The CLSM images for the CdSe x S 1-x QDs mineralized in HH/IBS.In these two biohybrid systems, mineralization of CdSe x S 1-x QDs occurred simultaneously and exhibited consistent fluorescence trends within the cells.Scale bar = 1 μm.

Figure S15 .
Figure S15.In vitro reconstitution of CdSe x S 1-x QDs biosynthesis in HM protein droplets.(A) Schematic diagram illustrating the in vitro reconstitution of CdSe x S 1-x QDs biosynthesis.In the LLPS protein condensate, the imidazole group of histidine residues in the histidine tag forms a strong coordination bond with CdSe x S 1-X QDs, effectively anchoring the QDs around the hydrogenase.Cys acts as a reducer.(B) In the protein condensate droplets formed in vitro, partial fusion of protein condensates (circular regions) is observed.Additionally, the formation of QDs within the protein condensates is observed (red arrows).(C) The HRTEM images of the CdSe x S 1-X QDs.The inset in the illustration shows an individual nanocrystal with lattice planes spaced at 0.349 nm, scale bar = 5 nm.(D) The related fast Fourier transform (FFT) patterns (inset of C) of the individual nanocrystal CdSe x S 1-x QDs.(E) Size distribution of the CdSe x S 1-x QDs synthesized in vitro.

Figure S16 .
Figure S16.Characterization of QD formation under anaerobic conditions incubation with 1 mM CdCl 2 and Na 2 SeO 3 .(A) UV-vis absorption.The broken E. coli cells exhibit a clear absorption peak at 400 nm, which can be attributed to the absorption peak of QDs.(B) The images from the QDs under a bright field or UV 365 nm light irradiation.(C) The fluorescence emission spectrum QDs under excitation at 400 nm shows a maximum emission peak at ~575 nm.(D) Optical image of QDs placed in a 96 well plate under bright field and UV 365 nm irradiation.(E) Measure the fluorescence intensity of lysate E. coli cells using a Microplate reader.

Figure S17 .
Figure S17.(A) UV-vis-DRS spectra of the QDs.(B) Tauc plot from UV-vis-DRS spectrum, (C) valence band XPS spectra of QDs.(D) a diagram of the band structures of QDs.

Figure S18 .
Figure S18.After IPTG induction of hydrogen enzyme expression, the HM strain was transferred to a freshly prepared sterile hydrogen production buffer under anaerobic conditions, and hydrogen production was observed under different temperature conditions (Buffer: 100 mM Tris HCl, 150 mM NaCl, 50 mM Glucose 24 h).

Figure S19 .
Figure S19.The viability of E. coli in the HM/IBSCS.(A) The CFU images and quantitative data (B) of HM/IBSCS after the visible light irradiation.Comparison of ROS content in the HM/IBSCS before and after light irradiation.Fluorescence images of ROS in bacteria before and after light irradiation (C) and the corresponding fluorescence quantification (D).scale bar = 20 μm.

Table S1 .
Plasmids used in this study

Table S2 .
Sequences of proteins in this work.

Table S3 .
Comparison of apparent quantum yields (AQY) of biological hybrid systems constructed under different construction strategies for photocatalytic hydrogen production.

Table S4 .
Comparison of the difference in multiple (fold) of the products obtained under the conditions of biological hybridization system and photocatalysis with naked bacteria or photosensitizer.