Engineered microbial hydrogels with confined architecture and binary microbes for efficient hydrogen production

Abstract

Renewable energy sources, particularly hydrogen, offer a promising solution to address global energy crisis and carbon emissions. Microalgae-driven hydrogen production has attracted immense interest in both scientific and industrial fields. However, challenges such as high oxygen sensitivity, substantial water demand, and low hydrogen production efficiency limit their potential. Here, we develop a core–shell symbiotic hydrogel system for enhanced hydrogen production via leveraging coaxial 3D bioprinting to spatially separate microalgae (i.e., core component) and bacteria (i.e., shell component). These networks optimize light and nutrient utilization while providing a localized anaerobic microenvironment to facilitate hydrogen production from microalgal photosynthesis. The symbiotic system enables a high hydrogen yield (1763 ± 98 mL L^–1). The system not only provides a highly efficient, liquid-free strategy for biohydrogen generation, but also advances the understanding of symbiotic relationships and microorganism-material interactions for creating advanced living material systems.

Microalgae‑based hydrogen production faces challenges including oxygen sensitivity, high water demand, and low efficiency. The authors develop a symbiotic hydrogel system that enhances hydrogen production by using coaxial 3D bioprinting to spatially separate microalgae and bacteria.

Introduction

The increasing global energy demand and environmental concerns have catalyzed the development of sustainable and clean energy solutions^1,2. Hydrogen is considered a key element in the future of energy transition due to its high energy density, zero carbon emissions, and renewable characteristics³. Traditional hydrogen production methods, such as steam reforming of fossil fuels, are often accompanied by high energy consumption and greenhouse gas emissions. Water electrolysis is a promising approach for obtaining clean and renewable energy, but it requires relatively high energy consumption at present, raising potential concerns about the viability of hydrogen as a clean and renewable energy source^4,5. Biological hydrogen production has been widely studied as a sustainable technology including photo-fermentation production, microbial electrolysis production, and synthetic biology-enabled production^6–8. In particular, microalgae-driven hydrogen production via photosynthetic activity has attracted much attention, as it directly utilizes the abundant and clean solar energy. The microalgal [FeFe]-hydrogenase uses electrons obtained from the photosynthetic electron transport chain to reduce protons into hydrogen under anaerobic conditions⁹, representing an ideal method for sustainable hydrogen production. However, another intrinsic product of photosynthesis, oxygen, is a strong inhibitor of the activity and expression of [FeFe]-hydrogenase¹⁰. Therefore, depleting oxygen and creating an anaerobic environment are essential for microalgae hydrogen production.

Several strategies have been developed to overcome this challenge^11–13, including genetic engineering, cell-material interface engineering, microalgae immobilization, and co-culture method. For example, the lba gene from soybean was introduced into the green microalga Chlamydomonas reinhardtii, resulting in a significant increase of hydrogen production due to enhanced respiration rates and altered starch metabolism¹⁴. However, this approach involves complex transgenic biotechnology and may lead to potential ecological and biosafety concerns. In contrast, engineering the microalgae-material interface (e.g., modulating microbial metabolic pathways through the intervention of materials¹⁵) eliminates the gene risk (e.g., potential biosecurity concerns caused by transgenic technology) and provides an effective strategy for creating anaerobic microenvironments. While enzyme-regulated anaerobic encapsulation technique was introduced to encapsulate single Chlorella cell by using laccase and polydopamine for sustained H₂ production¹⁶, the application of such enzymes and materials increases the cost and limits the large-scale application. Additionally, immobilized microalgae maintained high photosynthetic activity while effectively reducing oxygen diffusion, thereby enhancing hydrogenase activity. For example, gel immobilisation increased the hydrogen yield of Chlamydomonas reinhardtii by approximately twofold compared with suspended counterparts. Nevertheless, conventional immobilized systems exhibit limited light penetration and low mass transfer efficiency, which constrain efficient hydrogen production¹⁷. Co-culturing algal-bacterial symbiosis is able to generate hydrogen, where the anaerobic conditions are maintained by oxygen-scavenging bacteria¹⁸. Such methods typically exhibit the maximum accumulation of hydrogen around 120 mL L⁻¹, primarily due to nutrient competition, toxic metabolite accumulation¹⁹, and particularly reduced light transmission efficiency²⁰ (limited light availability to microalgae), thus reducing hydrogen production efficiency. Given these limitations, optimizing the symbiotic system is crucial to improving hydrogen yield while maintaining its eco-friendly and cost-effective advantages.

Here, we show a core–shell binary hydrogel network system containing microalgae and bacteria for efficient hydrogen production, denoted as microbial symbiotic networks (MSNs), which enhances and prolongs the activity of hydrogenases by creating a localized anaerobic microenvironment to boost the hydrogen production capacity of microalgae (Fig. 1). Specifically, we prepare hydrogel networks composed of sodium alginate (SA), carrageenan, and poly (ethylene glycol) diacrylate (PEGDA), with Chlamydomonas reinhardtii (C. reinhardtii) and Bacillus subtilis (B. subtilis) spatially compartmentalized in the core and shell parts, respectively, through coaxial 3D printing technology (Supplementary Fig. 1). C. reinhardtii is a model species for photosynthetic hydrogen production due to its high hydrogenase activity, fast growth rate, and low culture cost^21,22, while B. subtilis is able to consume oxygen by respiration²⁰. The hydrogel nature reduces the need for a liquid medium, and its core–shell structure mitigates nutrient competition. Moreover, the customized hydrogel networks effectively distribute light, allowing the microalgae within the scaffold to adequately receive light for enhanced hydrogen production. The MSNs not only enhance the hydrogen production efficiency of microalgae in hydrogel but also shed light on the regulated symbiotic relationship in living material systems.

Fig. 1. 3D bioprinting of a core-shell structure containing algae and bacteria for hydrogen production.

a Schematic of coaxial printing of MSNs with algae in the core and bacteria in the shell. b Hydrogen production mechanism of MSNs in anaerobic conditions. c Digital image of MSNs demonstrating good light transmission and reduced self-shading effects. d Digital images of living MSNs after 14 days of cultivation.

Results and discussion

Preparation of microbial symbiotic networks (MSNs)

SA and PEGDA were the main components used in the coaxial bioprinting technology for preparing core–shell hydrogels due to their mechanical properties and biocompatibility^23,24. SA and carrageenan were mixed to prepare the bioink for the core part (i.e., microalgae region), while they were further combined with PEGDA for the shell part (i.e., bacteria region) to improve the mechanical properties and structural stability of the final hydrogel networks^25,26. Specifically, extrusion-based 3D printing requires materials with shear-thinning properties; hence, carrageenan was incorporated into the bioink to modulate its viscosity. The addition of an appropriate amount of carrageenan improved the quality of the printed filaments (Supplementary Fig. 2). Moreover, such composite bioink exhibited effective shear thinning behavior and good elastic recovery, and the addition of C. reinhardtii or B. subtilis did not change the shear thinning behavior (Supplementary Fig. 3). Introducing the corresponding culture medium (TAP medium for C. reinhardtii and LB medium for B. subtilis) to each bioink enables the precise core–shell spatial manipulation of C. reinhardtii and B. subtilis with suitable nutrient distribution for their mutual growth. The crosslinked MSNs with stable architecture were eventually obtained after calcium chloride treatment and UV curing process (Supplementary Fig. 4 and Supplementary Table 1).

The printing properties of core–shell hydrogels were investigated by adjusting different parameters, including filament diameter and structural patterns. We demonstrated that core–shell hydrogels can be printed in different sizes and heights, with C. reinhardtii (emitting red fluorescence due to chlorophyll) and B. subtilis (emitting green fluorescence due to green fluorescent protein (GFP) expression) uniformly distributed in the core and shell regions of the hydrogels (Fig. 2a and Supplementary Fig. 5). In addition, the MSNs can be fabricated into structures of various shapes with excellent flexibility (Fig. 2b, c). The obtained MSNs can withstand more than 50% compressive strain and 30% tensile strain, and the compressive strength can reach 94 kPa (Supplementary Fig. 6), showing excellent mechanical strength for potential large-scale applications.

Fig. 2. Construction of core–shell MSNs containing C. reinhardtii and B. subtilis.

a Fluorescence images of MSN filaments with different diameters and MSN constructs with different heights. Red: C. reinhardtii. Green: B. subtilis. b 3D printing of MSN hydrogel constructs in different shapes. c Digital images of flexible MSN hydrogel constructs. d Activity of C. reinhardtii and B. subtilis before and after 3D printing. Before: C. reinhardtii or B. subtilis from suspension culture. After: C. reinhardtii or B. subtilis recovered from hydrogels. e SEM images of C. reinhardtii and B. subtilis extracted from the MSN hydrogel constructs after 14 days of culture. f Digital photographs of MSN hydrogel constructs (5 cm in diameter) and corresponding microscope images and confocal images of MSNs captured from day 0 to 14. g Time-dependent chlorophyll content of microalgae during 14-day culture. Data were presented as the mean ± s.d. (n = 3). Each experimental group consisted of three biological replicates. h Time-dependent fluorescence intensity of C. reinhardtii and B. subtilis during 14-day culture. i Distribution of C. reinhardtii and B. subtilis in MSN hydrogel constructs after liquid-free culture for 14 days. CC: C. reinhardtii. BS: B. subtilis. Source data are provided as a Source Data file.

Flow cytometry revealed that C. reinhardtii and B. subtilis maintained high viability throughout the processing steps, including extrusion, cross-linking, and photocuring (Fig. 2d and Supplementary Fig. 7). In addition, glutathione S-transferase activity was monitored, as it is a critical indicator of the level of microbial oxidative stress. No significant difference was observed in glutathione S-transferase activity of C. reinhardtii and B. subtilis before and after photocuring (Supplementary Fig. 8a). Furthermore, when the UV-treated cells were inoculated at the same concentration as the pristine ones, no difference in the number of colonies was observed (Supplementary Fig. 8b,c), further confirming that the negligible effect of printing and photopolymerization on microbial activity. MSNs was placed in a sealed vial and an agar substrate was provided at the bottom to maintain the humidity for the MSN culture (Supplementary Fig. 9). The growth of C. reinhardtii and B. subtilis were monitored during 14-day culture, during which additional liquid medium was not required to suspend the MSNs. Compared to the scanning electron microscope (SEM) images and optical microscope of native C. reinhardtii and B. subtilis (Fig. 2e and Supplementary Fig. 10), this culture method had no adverse effects on the morphology of cells. Moreover, the MSN hydrogel constructs exhibited the increased chlorophyll content and transitioned from a light green color on day 0 to a dark green after 14 days of culture (Fig. 2f, g), confirming the growth of microalgae. The steady-state fluorescence spectrum further revealed that the chlorophyll (at 680 nm) corresponding to C. reinhardtii continued to increase over a period of two weeks, while the peak intensity of the green fluorescent protein (in the range of 400–550 nm) significantly enhanced in the first three days and then gradually decreased to a stable level, which might be due to nutrient deficiency (Fig. 2h and Supplementary Fig. 11). C. reinhardtii and B. subtilis were largely confined in their designated compartments after 14-day culture with clear boundaries for each microbial community (Fig. 2i). We ascribed this superior confinement to the physical properties of the hydrogel (e.g., porosity and microenvironment of the hydrogel), which restricted their movement and made C. reinhardtii and B. subtilis more inclined to colonize within their inoculated regions. Additionally, the composition of the core and shell media, which were tailored for C. reinhardtii and B. subtilis respectively, further facilitated their partitioned growth. Collectively, we created an engineered binary living material with complex structures, providing a favorable environment for the growth and interaction of C. reinhardtii and B. subtilis.

Photosynthetic hydrogen production in MSNs

We next investigated whether B. subtilis could deplete oxygen through respiration to create an anaerobic environment, which thereby activated hydrogenase to achieve hydrogen production. Tungsten oxide (WO₃), which can react with hydrogen to produce tungsten bronze and therefore changes color from pale yellow to blue, was used for perceivable hydrogen detection^27,28. In the typical experimental setup, the MSNs were sealed in a glass container with WO₃ powder. After a 48-hour illumination of the MSNs, the powder transformed from yellow-green to blue-gray color, indicating the presence of hydrogen. Whereas, the color change was negligible in the absence of light (Fig. 3a), indicating that light is essential for microalgae-driven hydrogen production. Oxygen availability was typically constrained by low dissolved oxygen levels in traditional liquid cultures, while MSNs under liquid-free conditions were exposed to sealed headspace air, which contains ~20% oxygen. Therefore, the effect of oxygen in the headspace of container on hydrogen production efficiency was investigated. Initially, the overall oxygen level decreased, as the bacteria within the shell consumed oxygen through respiration. After the first cycle of hydrogen production (72 hours), the overall oxygen generated by MSNs system was higher, leading the system to gradually shift toward an aerobic state (Supplementary Fig. 12). We inferred that the depletion of acetic acid in the medium may reduce microalgal respiratory oxygen consumption, as acetate serves as a primary substrate for mitochondrial respiration²⁹. Consequently, reduced oxygen uptake led to the accumulation of photosynthetically generated oxygen, driving the system from an anaerobic to an aerobic state and thereby inhibiting the oxygen-sensitive hydrogenase. In this scenario, the system exhibited a hydrogen production rate of 905 ± 47 mL H₂ L^–1 (Supplementary Fig. 13), representing only 51.3% of the purged system. Therefore, after introducing the MSNs into the setup, nitrogen purging is recommended to remove oxygen in the headspace and establish an anaerobic environment for initiating hydrogen production.

Fig. 3. MSNs-driven H₂ production under light.

a The MSNs were incubated in the presence of WO₃ powder (in the small vial on top) with or without light irradiation. b Time-dependent hydrogen concentration determined by gas chromatography under different ratios of C. reinhardtii to B. subtilis. c Time-dependent DO concentration of MSNs and hydrogels with C. reinhardtii or B. subtilis. d The PAM fluorescence images of C. reinhardtii cultured in different modes evaluated by Imaging-PAM fluorometry. e F_v/F_m values of C. reinhardtii cultured under different conditions. f Digital photos of hydrogels with different structures on day 4 of culture. g Chlorophyll content in MSNs and homogenous microbial hydrogels during cultivation. h Hydrogen yield of MSNs and homogenous hydrogels in liquid-free culture for 2 days. i Comparison of hydrogen production between MSN system and SM system. j Hydrogen yield after purging the MSN system with nitrogen. k Comparison of the hydrogen production performance of MSNs with other reported methods. Data were presented as the mean ± s.d. (n = 3). Each experimental group consisted of three biological replicates. CC: C. reinhardtii. BS: B. subtilis. Source data are provided as a Source Data file.

Hydrogenase activity governs the efficiency of electron transport and hydrogen production, while it is highly sensitive to oxygen and rapidly deactivated even at low concentrations of oxygen¹⁷. To maintain the hydrogenase activity, a sustained anaerobic environment is required. During this process, B. subtilis were used to continuously create the atmosphere of low oxygen level by consuming oxygen via respiration (Supplementary Fig. 14). We varied the proportions of C. reinhardtii and B. subtilis to investigate the optimal ratio for hydrogen production (Fig. 3b). The C. reinhardtii-containing hydrogel system (i.e., shell bioink without bacteria) produced hydrogen with a yield of 933 ± 64 mL H₂ L^–1, which is superior to most conventional suspension-based microalgal (SM) systems (Supplementary Table 2). We ascribed this enhanced hydrogen production to the improvement of cell density and light utilization^30,31. No hydrogen was detected in the monoculture system with B. subtilis alone for 48 h, while the addition of B. subtilis into C. reinhardtii increased hydrogen production (Supplementary Figs. 14 and 15), as B. subtilis helped maintain the oxygen concentration at a low level, which is essential for regulating the activity of hydrogenase. For example, when the C. reinhardtii and B. subtilis were employed at a cell number ratio of 1:15, the yield of hydrogen was 1763 ± 98 mL H₂ L^–1, approximately 3985 ± 327 μmol H₂ (mg Chl)⁻¹, which increased by 89.0% compared to the C. reinhardtii-containing hydrogel. In contrast, excessive use of B. subtilis (i.e., the ratio of C. reinhardtii to B. subtilis was 1:20) led to their obvious diffusion into the algal region in the MSNs (Supplementary Fig. 16), where it likely competed with C. reinhardtii for nutrients and interfered with light transmission²⁰. Moreover, the MSN with B. subtilis had a longer hydrogen production time than the group without B. subtilis (Fig. 3b). We refer that B. subtilis consumed oxygen through respiration to maintain a low-oxygen environment, thereby improving the hydrogenase activity and prolonging the time of hydrogen production at relatively high levels. To investigate whether MSNs with C. reinhardtii: B. subtilis ratio of 1:15 can establish and maintain a local anaerobic environment, time-dependent dissolved oxygen (DO) concentration of C. reinhardtii-containing hydrogels, B. subtilis-containing hydrogels, and MSNs were monitored (Fig. 3c). The results showed that the C. reinhardtii-containing hydrogels elevated DO levels through continuous production of photosynthetic oxygen. In contrast, B. subtilis-containing hydrogels caused a rapid decline in DO due to bacterial respiration. In the MSN system, the oxygen generated by C. reinhardtii in the core was consumed by the excessive B. subtilis in the shell, leading to a decrease in DO to a relatively low level. This confirmed that MSNs efficiently created an anaerobic environment conducive to hydrogen production.

Under light conditions, algal cells decompose water through the oxygen-evolving complex (OEC) in Photosystems II (PSII) to produce electrons, which are the main source of algal photosynthesis hydrogen³². Therefore, the activity of PSII can significantly affect the hydrogen production efficiency of microalgae. Pulse-amplitude-modulation chlorophyll fluorescence imaging (Imaging-PAM) was used to evaluate the PSII activity of C. reinhardtii under three different culture conditions, including the SM system, the homogenous algal-bacterial hydrogel (Homo system), and the core–shell MSNs. The results showed that the PSII activity of C. reinhardtii under the three culture conditions was similar at the beginning, indicating that the PSII activity of microalgae would not be restricted in the hydrogel. The PSII activity in the Homo system slightly decreased after 4 days of cultivation, whereas that in the MSNs system increased (Fig. 3d, e). These results suggested that the microalgae of MSNs maintained higher photosynthetic activity over time, indicating that the core–shell structure provided a favorable microenvironment for supporting efficient photosynthesis.

The influence of core–shell MSN architecture (e.g., filament diameter, projected area, filling interval, and construct height) on hydrogen production was evaluated (Supplementary Fig. 17). Varying filament diameter (0.8, 1.0, and 1.2 mm) or projected area (300, 600, and 900 mm²) at a constant height of 4 mm had minimal impact on hydrogen yield per unit volume (Supplementary Fig. 17b, c). In contrast, MSNs with no filling intervals produced less hydrogen than those with 2.5 or 5.0 mm intervals (Supplementary Fig. 17d). This reduction is attributed to restricted gas diffusion and high light attenuation, which suppress microalgal growth within the 1–2 mm thin layer beneath the cube periphery³³. However, height significantly affected production, with 4 and 8 mm constructs showing similar efficiency (1742 ± 47 and 1745 ± 9 mL L^–1, respectively), while the height of 12 mm reduced efficiency (1419 ± 91 mL L^–1) probably due to limited light penetration (Supplementary Fig. 17e). Moreover, F_v/F_m measurements after 48 hours confirmed that higher heights (e.g., 12 mm) led to relatively lower value (Supplementary Fig. 18).

To assess the advantage of the spatial compartment of the two microorganisms for hydrogen production, we compared the core–shell MSNs to a Homo system. The core–shell MSNs exhibited robust growth for 8 days, while the Homo system turned yellow on the fourth day, indicating the decreased chlorophyll content for photosynthetic activity (Fig. 3f, g, and Supplementary Fig. 19). It is noteworthy that C. reinhardtii-containing hydrogel (i.e., non-core–shell structure) also exhibited robust growth (Supplementary Fig. 19), suggesting that the death of algae in the Homo system was related to the bacteria. This highlights the advantage of MSNs system, which spatially separates microalgae and bacteria, effectively avoiding the potential inhibitory effect of bacteria on microalgae growth. On day 8, the chlorophyll content of MSNs reached 29.8 ± 1.8 mg L^–1, which was 2.2 times higher than that of Homo system (Fig. 3g). Moreover, the core–shell MSNs produced higher amounts of hydrogen (Fig. 3h). This enhancement is attributed to compartmentalizing C. reinhardtii and B. subtilis in distinct spatial domains, thereby providing tailored microenvironments for each microorganism, which are conducive to maintaining the ideal growth and metabolic state of each microorganism for hydrogen production^34,35. During the cultivation process of MSNs and Homo systems, the metabolic process of B. subtilis significantly reduced the mechanical properties of the hydrogel over time by microbial degradation of the hydrogel matrix (Supplementary Fig. 20)³⁶.

Particularly, the MSNs can not only conserve water resources and improve space utilization, but also outperform SM system in hydrogen production (Fig. 3i). TAP medium was used for hydrogen production in the SM system, as C. reinhardtii grew well in the TAP medium, whereas B. subtilis underwent limited proliferation due to insufficient necessary nutrients (Supplementary Fig. 21). When the initial cell numbers of C. reinhardtii and B. subtilis were consistent with those in MSNs, the yield of hydrogen in SM is 22.4 ± 3.2 mL H₂ L^–1. In contrast, the hydrogen yield of MSNs achieved 1763 ± 98 mL H₂ L^–1, 78 times higher than that of the conventional SM system. MSNs were able to create a more tailored microenvironment for C. reinhardtii and B. subtilis, which facilitated the accumulation of microbial population and, in turn, promoted the efficient generation of hydrogen. Several reports showed that in the symbiotic system of microalgae and bacteria, the hydrogen production commonly ceased after reaching a certain level due to nutrient deficiency and anaerobic environment disruption^37,38. In this MSN system (the ratio of C. reinhardtii to B. subtilis was 1:15), hydrogen production reached its maximum accumulation after 48 hours. Therefore, the resulting hydrogen was collected, and nitrogen was introduced into the reaction system to re-establish anaerobic conditions for another production cycle. The results indicated that MSNs were capable of producing hydrogen in consecutive experiments for 8 days, though hydrogen production showed an obvious decline with increased cycle numbers (Fig. 3j). We speculate that nutrients in the medium may be gradually depleted, limiting the growth and metabolic activity of the binary microbial systems. The hydrogel system containing only microalgae also showed a decreasing production trend in successive rounds, with hydrogen production in each cycle significantly lower than that of the MSNs system containing B. subtilis (Supplementary Fig. 22). These results indicated that the initial nitrogen purging cannot maintain a continuous anaerobic environment and therefore additional B. subtilis was necessary to ensure the activity of microalgal hydrogenase for enhanced hydrogen production.

Current microalgae-based photobiological hydrogen production mainly includes the use of oxygen-resistant strains³⁹, adoption of pulse illumination⁴⁰, depletion of generated oxygen from algae^41–43, co-culturing microalgae and bacteria^19,44, promoting algae cell aggregation²⁷, material-modified algae⁴⁵, immobilization system⁴⁶, and employing genetic engineered strains with high hydrogen production capabilities (Fig. 3k and Supplementary Table 2)^47,48. In contrast, the spatial separation strategy of C. reinhardtii and B. subtilis in MSNs significantly improves hydrogen production efficiency and extends operational duration by optimizing nutrient and light utilization, while also creating a localized anaerobic environment. Additionally, the spatial arrangement facilitates the establishment of localized anaerobic microenvironments essential for efficient biohydrogen synthesis. Consequently, this strategy enables the development of a sustainable biohydrogen production platform while reducing environmental footprint.

Variation of photosynthetic system and hydrogenase gene expression

During the photosynthesis of microalgae, the electron transport process involves Photosystems II (PSII) and I (PSI), where PSII and PSI work in tandem to drive the electron transport chain. PSII initiates the process by capturing light energy and generating high-energy electrons, which are then conveyed through the electron transport chain to plastoquinone (PQ), cytochrome b6/f complex (Cyt b6/f), and plastocyanin (PC) (Fig. 4a)⁴⁹. Following light-induced charge separation, PSI donates its light-generated electron to ferredoxin (Fd) and accepts the electron from PC. Fd acts as the final electron carrier in the photosynthetic linear electron transport chain and is responsible for distributing electrons to different metabolic pathways. Under anaerobic conditions, Fd translocates a portion of its electrons to [FeFe]-hydrogenase, wherein the enzymatic catalysis facilitates the combination of these electrons with protons for the biogenesis of hydrogen^50,51. The transfer of electrons from the primary quinone receptor (QA) to the secondary quinone receptor (QB) of PSII is a key step in the overall process. The compound 3-(3, 4-dichlorophenyl)-1,1-dimethylurea (DCMU) is a specific inhibitor of PSII that effectively blocks the electron transfer process from the QA to QB^52,53. When DCMU was added to the core ink, hydrogen production showed an obvious decrease with the increasing of DCMU concentration (Fig. 4b). This confirms the critical coupling of photosynthetic pathways and hydrogenase in MSN-driven hydrogen generation.

Fig. 4. Photosynthetic hydrogen production mechanism and differential genes of microalgae in MSNs system, conventional SM system, and Homo system.

a Photosynthetic hydrogen production mechanism of the biophotolysis in microalgae. b Effect of DCMU on hydrogen production from MSNs. Data were presented as the mean ± s.d. (n = 3). Each experimental group consisted of three biological replicates. c The volcano plot showing the DEG. d KEGG enrichment for DEGs. e GO enrichment for DEGs. f Gene expression of C. reinhardtii related to photosynthetic hydrogen production in different systems. g Schematic diagram of microalgae hydrogenase and its related gene expression in microalgae. Source data are provided as a Source Data file.

We further explored the mechanism of improved hydrogen production of MSNs by transcriptome analysis of C. reinhardtii in different coculture systems (i.e., the SM system or the Homo system served as the control group, while the MSNs system was the experimental group). The repeatability and correlation of each sample in transcriptome sequencing were reasonable, and the differential gene clustering effect was significant, which proved the reliability of transcriptome analysis data (Supplementary Fig. 23a). The expression patterns of C. reinhardtii in MSN system were significantly different from those in SM system or Homo system, as well as the data distribution and density of C. reinhardtii in MSNs (Supplementary Fig. 23b). Compared to the SM system, 9313 differentially expressed genes (DEGs) were identified, including 5056 upregulated and 4257 downregulated genes. While compared to Homo system, there were 2553 up-regulated genes and 1985 down-regulated genes (Fig. 4c and Supplementary Fig. 23c). The MSN system showed more gene expression differences compared to the SM system (Supplementary Fig. 23d). Compared to SM system, a certain number of DEGs in MSN systems were enriched in the pathways related to the expression of metabolic pathways (i.e., pyruvate metabolism, citrate cycle, etc.) according to the Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment (Fig. 4d). Moreover, the gene ontology (GO) enrichment analysis revealed that the pathways related to photosynthesis were significantly enriched (Fig. 4e and Supplementary Fig. 24). Compared to Homo system, some pathways related to metabolism and photosynthesis were also enriched (Supplementary Fig. 25). The microalgae photosynthesis pathway (e.g., photosynthesis, light reaction, light harvesting) was enriched, which was directly related to the improved light utilization efficiency. Enrichment of pathways related to nitrogen metabolism and ABC transporters indicated the enhanced metabolic activity and efficient nutrient utilization in the microalgae¹⁷. Collectively, the improved hydrogen production capacity of the MSNs system is likely linked to the activation of photosynthesis and energy metabolism.

Gene expression of the photosynthetic pathways in MSNs was significantly upregulated (Fig. 4f). In PS II, the expression of extrinsic proteins (psbO, psbP, psbQ, psbR) and structural proteins (psbY, psb27) is stimulated compared to SM system, and their expression levels significantly affected the efficiency of PSII. Similarly, the upregulation of genes encoding PSI led to an increased PSI activity. Moreover, the genes encoding Fd, cyt b6f complex, and ATPase in the MSNs system were generally upregulated compared to the SM system. The optimization of the electron transfer pathway enhanced the energy conversion efficiency, thereby activating the expression of hydrogenase. Furthermore, the expression levels of [Fe-Fe] hydrogenase and photosystem genes in different systems were compared by heat maps (Supplementary Fig. 26). The expression of most genes in the SM system were lower than that in the other two systems, indicating that the liquid-free cultivation environment had a positive effect on the photosynthesis and hydrogen production efficiency of microalgae. Moreover, the expression of C. reinhardtii hydrogenase-related genes in three systems were compared (Fig. 4g). HydEF and HydG are related genes encoding hydrogenase maturation proteins. These two genes in MSNs systems were up-regulated by 1.79 and 2.10 times compared to the SM system, and by 2.00 and 2.69 times compared to the Homo system. Compared to the SM system and Homo system, the HydA1 gene encoding the catalytic subunit of hydrogenase in the MSN system was up-regulated by 1.16 and 3.95 times, respectively. In contrast, the low expression of these genes in the Homo system highlighted that its microenvironment suppressed hydrogenase transcription, confirming that the microenvironment of the core–shell structure was more conducive to the growth of C. reinhardtii for improved hydrogen production.

In summary, this study demonstrates that the binary microbial hydrogels featuring a confined core–shell architecture represent a breakthrough approach for enhancing biohydrogen production. The spatial separation strategy achieved through coaxial 3D bioprinting creates optimal photosynthetic microenvironments for microalgae while establishing a localized anaerobic condition that facilitates hydrogen generation via bacteria respiration. The MSN system not only achieves high hydrogen production efficiencies but also reduces reliance on aqueous environments and maximizes space utilization, highlighting its great potential for sustainable and efficient green energy production. We expect that this innovative strategy provides a promising solution for microalgae-based biomanufacturing, laying a solid foundation for the future advancement of sustainable technologies and synthetic biology.

Methods

Chemicals and materials

All chemical reagents were used without further purification. Alginic acid sodium salt (viscosity 500–1000 mPa·s), carrageenan (E407), calcium chloride anhydrous (CaCl₂), sodium citrate dihydrate ( ≥ 99%), sodium chloride (NaCl), acetone, glutaraldehyde (25% in water), tungsten trioxide (WO₃), (3-(3,4-dichlorophenyl)−1,1-dimethylurea) (DCMU), lithium phenyl (2,4,6-trimethylbenzoyl) phosphinate (LAP), Luria-Bertani (LB) medium, and poly (ethylene glycol) diacrylate (PEGDA, Mw = 1000) were purchased from Shanghai Aladdin Biochemical Technology CO., Ltd. Tris-acetate-phosphate (TAP) medium was purchased from Institute of Hydrobiology, Chinese Academy of Sciences. Agar was purchased from Bei-jing Coolaber Technology Co., Ltd. Propidium iodide (PI) was purchased from Shanghai Aladdin Biochemical Technology CO., Ltd. SYTOX Green Nucleic Acid Stain was purchased from Thermo Fisher Scientific (Waltham, MA, USA). Glutathione S-transferase (GST) activity assay kit (BC0355) was purchased from Solarbio Science & Technology Co., Ltd (Beijing, China). High-purity water with a resistivity of 18.2 MΩ·cm was obtained from a Millipore water purification system.

Characterization

The rheological behavior of hydrogels were characterized on an Anton Paar MCR302 rotational rheometer. UV-vis absorption measurements were recorded with a PerkinElmer LAMBDA1050 + UV/Vis/NIR spectrophotometer. Microscopic and fluorescence images were obtained using a Nikon ECLIPSE Si biomicroscope and an Olympus IX73 inverted fluorescence microscope, respectively. Z-stack fluorescence images were collected with a Leica STELLARIS 5 CLSM. The viability of microalgal cells was assessed using Agilent Technologies’ NovoCyte Advanteon flow cytometry and steady-state transient fluorescence spectrometer (Fluorolog-3). Scanning electron microscopy was performed on a Phenom XL system (Thermo Fisher Scientific) at an accelerating voltage of 10 kV. The photosynthetic activity was assessed with a pulse amplitude modulation chlorophyll fluorescence imaging (Imaging-PAM). Hydrogen was measured using a calibrated gas chromatograph (GC 7900) with a column packed with molecular sieve 13X using nitrogen as carrier gas.

Cells

Chlamydomonas reinhardtii (CC5325, denoted as CC) were kindly provided by the Shenzhen Key Laboratory of Marine Microbiome Engineering at Shenzhen University. C. reinhardtii was cultured at 23 °C with agitation at 150 rpm and a continuous illumination of 50 µmol photons m^-2 s^-1 in a shaker incubator. Green fluorescent protein (GFP)-labeled Bacillus subtilis (Bacillus subtilis 168-GFP) was obtained from Beijing Biobw Co., Ltd (China).

Preparation of MSNs

Due to the different bioink composition and nozzle parameters during coaxial printing, the carrageenan concentration was investigated for the core bioink (3% SA) and the shell bioink (3% SA + 10% PEGDA). For visualization and structural characterization, microalgae or bacteria were separately incorporated into the bioinks, and their fluorescence signals were used to examine the printed filament structures under a fluorescence microscope. Filament diameters were further quantified by image analysis to assess the relationship between carrageenan concentration and nozzle diameter.

Based on the obtained formulations, 10 mL C. reinhardtii cells (1 × 10⁷ cells mL^–1) at the exponential growth stage were centrifuged (4000 g, 5 min) and concentrated into 75 µL (4000 g, 6 min), and then suspended in 10 mL of 30 mg mL^–1 sodium alginate (SA) solution, where TAP medium was used to dissolve SA. Then, 450 mg carrageenan was added to the mixture to obtain core bioink for the printing process.

Next, different amounts of B. subtilis (5 mL, 10 mL, 15 mL, 20 mL, 25 mL) in the exponential growth period (1 × 10⁸ cells mL^–1) were centrifuged and concentrated to 75 µL (8000 g, 6 min) and suspended in 5 mL of 60 mg mL^–1 SA solution, where LB medium was used to dissolve SA. The cell number ratio of C. reinhardtii to B. subtilis in MSNs was 1:0, 1:5, 1:10, 1:15, 1:20, and 1:25, respectively. The MSN with C. reinhardtii: B. subtilis = 1:0 refers to a hydrogel system containing only C. reinhardtii (i.e., shell bioink without bacteria). Then 5 mL of 20% PEGDA and 5 mg of photoinitiator LAP were added. Then, 900 mg of carrageenan was added to the mixture to obtain shell bioink for the printing process.

The core and shell bioinks were transferred into 5 mL and 10 mL syringes matched with the 3D printer (Bio-architect® SR, Regenovo, Hangzhou, China), respectively. The core and shell bioinks were distributed pneumatically through nozzles with pressures of 0.3 MPa and 0.4 MPa, respectively. Customized 3D structures with different sizes and shapes were designed using printer software, and the diameter of the material was adjusted by using different nozzles. Finally, MSNs were cross-linked in 100 mM CaCl₂ solution for 10 min and cured under UV light for 30 s (UV light with a wavelength of 365 nm and a power density of 3 mW/cm²).

Preparation of Homo system

To prepare the Homo system, 10 mL C. reinhardtii cells (1 × 10⁷ cells mL^–1) at the exponential growth stage were centrifuged and concentrated into 75 µL (4000 g, 5 min), and then suspended in 10 mL of 30 mg mL^–1 sodium alginate (SA) solution, where TAP medium was used to dissolve SA. Next, 15 mL B. subtilis at exponential growth period (1 × 10⁸ cells mL^–1) were centrifuged and concentrated to 75 µL (8000 × g, 6 min) and suspended in 5 mL of 60 mg mL^–1 SA solution, where LB medium was used to dissolve SA. Then 5 mL of 20% PEGDA and 5 mg of photoinitiator LAP were added. The two solutions were combined and thoroughly mixed. Finally, 1800 mg carrageenan was added to the mixture to obtain the bioink for the printing process. The bioinks were distributed pneumatically through nozzles with a pressure of 0.4 MPa. The Homo system was cross-linked in 100 mM CaCl₂ solution for 10 min and cured under UV light for 30 s.

Determination of the chlorophyll content

The method established in reference was used to quantitatively determine the chlorophyll content of microalgae in MSNs^53,54. Specifically, the freeze-dried MSNs were ground into powder and soaked in acetone in a thermostatic mixer (600 rpm, 37 °C, Eppendorf ThermoMixer C) until the cells were colorless. The mixture was then centrifuged (4000 g, 5 min) and the chlorophyll concentration in the supernatant were determined by the absorbance values at 665 nm (A₆₆₅) and 649 nm (A₆₄₉). The total concentration of chlorophyll (C, mg L^–1) was calculated according to the equations below:

$$\mathrm{Chl\; a}=13.70\times {\mathrm{A}}_{665}-5.76\times {\mathrm{A}}_{649}$$ 1

$$\mathrm{Chl\; b}=25.80\times {\mathrm{A}}_{649}-7.60\times {\mathrm{A}}_{665}$$ 2

$$\mathrm{Chl}\left(\mathrm{total}\right)=6.10\times {\mathrm{A}}_{665}+20.04\times {\mathrm{A}}_{649}$$ 3

Photosynthetic performance measurements

The photosynthetic performance was measured using a pulse amplitude modulation chlorophyll fluorescence imaging (Imaging-PAM). Specifically, a certain amount of C. reinhardtii and B. subtilis was inoculated in liquid medium or hydrogel in the same volume, and the samples were then placed in darkness for 30 min. The steady-state chlorophyll fluorescence (F₀) and maximum chlorophyll fluorescence (F_m) were obtained by using a modulated light and a brief pulse of saturating light, respectively. The parameter F_v/F_m indicated the maximum quantum efficiency of PSII. This factor was determined by calculating the ratio of (F_m – F₀) to F_m.

Determination of cell viability by flow cytometry

To avoid interference among fluorescence signals from chlorophyll, GFP, and dyes (e.g., propidium iodide and SYTOX Green), monoculture hydrogels containing microalgae or bacteria were prepared to evaluate the viability of each component. The assay of dead cells for C. reinhardtii and B. subtilis was conducted by incubating with 10 μM SYTOX Green and 5 μM propidium iodide in the dark at room temperature for 15 minutes, respectively. The FITC channel is used to detect microalgal dead cells labeled with SYTOX Green dye, while the PE channel is used to detect bacterial dead cells labeled with PI. To avoid the interference of the green fluorescence of B. subtilis-GFP with other dyes, we used the native B. subtilis in the analysis of dead cells.

Qualitative analysis of hydrogen using tungsten oxide

Tungsten oxide (WO₃) can react with hydrogen to produce tungsten bronze (2 xH_ad + WO₃ ↔ WO_3−x·xH₂O) and therefore changes color from pale yellow to blue. It was used for perceivable hydrogen detection. Initially, N₂ was used to remove oxygen from the headspace of the vial containing MSN and tungsten oxide (WO₃) powder. One group was incubated in dark, while the other group was exposed to continuous illumination (50 µmol photon m⁻² s⁻¹) for 48 hours.

Preparation of SM system and measurement of H₂ production

For hydrogen production in SM system, C. reinhardtii and B. subtilis at exponential growth stage were suspended in a conical vial containing 10 mL TAP medium at a cell number ratio of 1:15 (C. reinhardtii: B. subtilis). The initial cell numbers of C. reinhardtii and B. subtilis per millilitre of TAP were consistent with those of MSN, and the headspace volume of the hydrogen production experiments was the same. The headspace was purged with nitrogen for three minutes to remove oxygen to accelerate hydrogen production, and culture conditions (23 °C, 50 µmol photon m^–2 s^–1) were the same as those of liquid-free culture. The amount of hydrogen produced is expressed in mL L^–1, and the liters in the unit refer to the volume of the TAP medium. The collection and detection methods of hydrogen are consistent with those of the MSNs system.

Collection of C. reinhardtii and B. subtilis from hydrogel systems

For cells from the hydrogel for imaging and transcriptome sequencing experiments, MSNs were dissolved with 0.9% NaCl solution containing 100 mM sodium citrate. The undissociated hydrogel (shell part) was removed by gravity sedimentation, and the remaining suspension was centrifuged (4000 g, 5 min) to collect the C. reinhardtii. Subsequently, the supernatant was centrifuged again (8000 g, 5 min) to collect the B. subtilis. Note: the hydrogel samples for flow cytometry assay was prepared without adding photoinitiator (i.e., only calcium chloride cross-linking), and the dissociation of the samples was performed in accordance with the steps described above.

Rheological performance testing

The rheological properties of bioink were tested using a MCR 302 rheometer (Anton Parr, Germany), with a gap height of 1 mm at 25 °C. The variation of shear stress with shear strain (the strain range was from 0 to 100%) was tested by oscillatory amplitude sweeps. The dynamic strain sweep experiment has a fixed frequency of 10Hz, and the deformation range is 0.01% to 100%.

Mechanical performance testing

The tensile, compressive, and cyclic compression tests of hydrogels were performed using a universal tensile testing machine (Instron 5967, Instron Electronics Instrument Co., Ltd., USA). For the tensile test, a dumbbell-shaped specimen with a filling interval of 0, a thickness of 4 mm, a length of 30 mm, and a width of 10 mm was printed. For the compression test, a cylindrical specimen with a filling interval of 0, a diameter of 14 mm, and a height of 20 mm was printed. The tensile and compression speeds were both set at 10 mm min^-1. Three specimens were prepared for each mechanical property experiment.

Determination of glutathione S-transferase activity

To assess the potential effects of photopolymerization on microalgae and bacteria, three types of MSN hydrogels were prepared (i.e., CC:BS = 1:0; CC:BS = 0:1; CC:BS = 1:1). The experimental group was cross-linked in 100 mM CaCl₂ solution for 10 min and cured under UV light for 30 s (UV light with a wavelength of 365 nm and a power density of 3 mW/cm²). The control group was cross-linked in the same concentration of CaCl₂ solution for 10 minutes. Then, MSNs were dissolved with 0.9% NaCl solution containing 100 mM sodium citrate and centrifuged (4000 g, 5 min) to collect the cells. Subsequently, glutathione S-transferase activity assay kit (BC0355, Solarbio) was used to analyze the glutathione S-transferase activity according to the manufacturer’s instructions.

Measurement of dissolved oxygen

The dissolved oxygen (DO) was measured according to the previous method with modifications²⁹. Specifically, the MSNs (CC:BS = 1:15), C. reinhardtii-containing hydrogels, and B. subtilis-containing hydrogels were respectively incubated in the water, and the oxygen level was monitored using a dissolved oxygen sensor (PyroScience OXF1100).

Culture of MSNs and measurement of H₂ and O₂

In liquid-free culture for hydrogen production, MSNs was placed in a sealed bottle containing an agar film at the bottom, where the agar film supplied water and maintained the humidity for the MSN culture. The headspace was purged with nitrogen for three minutes to remove oxygen to accelerate hydrogen production. All the experiments were purged after introducing the MSNs into the setup or at the beginning of each cycle, except for the experiment investigating the time-dependent changes of hydrogen and oxygen concentrations without nitrogen purging. The culture environment was maintained at 38% relative humidity and 23 °C, and 50 µmol photon m^–2 s^–1 was continuously illuminated. For hydrogen production in SM system, C. reinhardtii and B. subtilis at logarithmic growth phase were suspended in a conical bottle containing TAP medium at a cell number ratio of 1:15. The initial cell numbers of C. reinhardtii and B. subtilis were consistent with MSNs. The headspace was purged with nitrogen for three minutes to remove oxygen to accelerate hydrogen production, and culture conditions (23 °C, 50 µmol photon m^–2 s^–1) are the same as those of liquid-free culture. At different time intervals, 1 mL of the headspace was extracted with an airtight syringe and injected into a calibrated gas chromatograph (GC 7900) with a column packed with molecular sieve 13X using nitrogen as carrier gas. The concentrations of H₂ and O₂ in the headspace were both determined by reported methods^27,42. Specifically, the packed column was held at 50 °C and the thermal conductivity detector (TCD) was set to 120 °C for H₂ analysis. The volumes of hydrogen and oxygen gas were calculated using the external standard method. The amount of hydrogen produced is expressed in mL L^–1, and the liters in the unit refer to the volume of the MSN. For hydrogen production at a cell number ratio of 1:15, the chlorophyll content was measured as 20 ± 3 mg Chl L⁻¹. Therefore, the hydrogen yield of 1763 ± 98 mL H₂ L^–1 can be converted to 3985 ± 327 μmol H₂(mg Chl)^–1.

The effect of DCMU on MSNs for hydrogen production

The MSNs with DCMU were prepared in the same protocol as the forementioned MSNs, except that DCMU was dissolved in TAP medium at different concentrations (8 μM and 16 μM) when the core ink was formulated.

Continuous H₂ production

H₂ production was analyzed in repeated batches. When no obvious increase of hydrogen was observed, the reaction bottle was purged with nitrogen again to replace the gas to re-establish oxygen-free conditions. At different time intervals, 1mL of headspace was extracted with an airtight syringe and injected into a gas chromatograph to analyze the hydrogen content.

Transcriptome sample preparation

In order to ensure the rationality and validity of the experiment, three groups of samples with different culture methods (i.e., conventional SM system, Homo system, and MSN system) were designed. The control group was the C. reinhardtii in the SM and Homo system, while the experimental group was the C. reinhardtii in the MSN system. Three duplicate samples were set in each group to enhance the reliability of the experimental results. Microalgae (i.e., C. reinhardtii) under different culture methods were collected and washed with PBS to remove impurities. The samples were then quickly transferred to a freezer tube and centrifuged (4000 g, 5 min) to remove the supernatant. To maintain the activity and sequencing performance of the samples, the sample was quickly frozen in liquid nitrogen for 15 minutes, after which the samples were transferred to a refrigerator at −80 °C for long-term storage.

RNA sequencing and transcriptome analysis

Total RNA from microalgal samples was extracted using TRIzol reagent (Invitrogen, USA) according to the manufacturer’s protocol with minor modifications to accommodate the characteristics of microalgal cells. Briefly, fresh microalgal samples were thoroughly ground in liquid nitrogen, and 1 mL of TRIzol reagent was added per 50–100 mg of tissue to ensure complete cell lysis. The mixture was incubated at room temperature for 5 min, followed by the addition of 0.2 mL chloroform. After vigorous shaking, the samples were centrifuged at 13,000 g for 10 min at 4 °C. The aqueous phase was carefully transferred to a new RNase-free tube, and total RNA was precipitated with an equal volume of isopropanol, washed with 75% ethanol, air-dried, and dissolved in RNase-free water. The concentration, purity, and integrity of RNA were assessed using a NanoDrop spectrophotometer and agarose gel electrophoresis, and the qualified RNA samples were subsequently used for transcriptome sequencing. The final cDNA library was obtained by PCR enrichment. Enrichment analysis was performed using the clusterProfiler package, which applied a hypergeometric test to calculate P values. The sequences were annotated using the Kyoto Encyclopedia of Genes and Genomes (KEGG) database to identify functional genes and perform metabolic pathway analysis.

Normalization of gene expression data and heatmap generation

The original FPKM values were Z-score normalized using the formula:

$$z=\frac{x-\mu }{\sigma }$$ 4

where x is the original expression value, and μ and σ are the mean and standard deviation of that gene across all samples, respectively.

Statistics and reproducibility

All studies were performed with three independent replications. All samples used in the study were randomly allocated into different experimental groups. The average and standard deviation were calculated in Prism 9.5 software (GraphPad Software, USA). Data are presented as mean ± standard deviation from at least three biological replicates.

Reporting summary

Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.

Author contributions

J.L., J.J.Z., X.S.D., and X.Y.L. conceived and designed the experiments. X.Y.L. performed the major experimental work and wrote the manuscript. Q.L., M.W.J., W.W.T., N.D., and X.Q.L. offered guidance and assistance for 3D printing and microalgae cultivation. H.S., X.H., H.L., and X.J.L. helped with the analysis of the results. All authors discussed the results and contributed to manuscript writing and editing.

Peer review

Peer review information

Nature Communications thanks Lisa Utschig, who co-reviewed with Audrey Short; and the other, anonymous, reviewer(s) for their contribution to the peer review of this work. A peer review file is available.

Data availability

RNA-seq data are available in the NCBI database under accession PRJNA1431651. Source data are provided with this paper.

Competing interests

The authors declare no competing interests.

Supplementary information

The online version contains supplementary material available at 10.1038/s41467-026-70988-x.