Human papillomavirus (HPV), a sexually transmitted virus associated with cervical cancer, remains a significant global health challenge. HPV virus-like particles (VLPs), which mimic the structure of native viral particles without containing genetic material, provide a safe and effective platform for vaccine development. Efficient production of HPV-VLPs requires a robust expression system, and the yeast Komagataella phaffii (K. phaffii) offers advantages including cost-effectiveness, rapid growth, and high recombinant protein productivity. This study aimed to establish an integrated optimization strategy for enhancing HPV16 L1-VLP production in K. phaffii by systematically evaluating host strains, promoter systems, and cultivation conditions. Three commonly used K. phaffii strains (GS115, X-33, and KM71H) expressing HPV16 L1 under either the methanol-inducible AOX1 promoter or the constitutive GAP promoter were compared using shake-flask cultivation and BioLector XT microbioreactor-based process evaluation. The GS115 strain expressing HPV16 L1 under the AOX1 promoter demonstrated the highest VLP yield of 20.50mg/L, representing approximately a 1.9-fold and 3.2-fold increase compared with X-33 (11.07mg/L) and KM71H (6.49mg/L), respectively. Optimized cell disruption conditions further improved intracellular VLP recovery while preserving particle integrity. The produced HPV16 L1-VLPs exhibited structural and biochemical characteristics consistent with intact VLP formation. This study presents a systematic workflow that integrates strain selection, promoter evaluation, and microbioreactor-based process optimization to identify an efficient K. phaffii platform for HPV16 L1-VLP production. The findings provide valuable insights for developing robust and scalable strategies for recombinant VLP manufacturing and other complex biotherapeutic applications.
D-pantothenic acid (DPA) is a commercially vital bioproduct, yet its fermentation titer is often constrained by the lack of precise, real-time process regulation. In this study, a two-stage hybrid modeling framework was developed to achieve real-time prediction and dynamic optimization of Escherichia coli DPAT6 fermentation. First, a forward artificial neural network (ANN) was designed as a soft sensor to map environmental conditions (time, temperature, pH) to critical process state variables. These predicted states were then fed into a support vector machine with a radial basis function kernel (SVM-RBF), which exhibited superior predictive performance (RMSE = 5.03g/L, R2 = 0.982). Leveraging this hybrid model, a three-stage dynamic control strategy for temperature and pH was generated using particle swarm optimization (PSO) and genetic algorithms (GA). Experimental validation confirmed that this model-guided strategy increased the D-pantothenic acid titer to 125.98g/L, a 17.4% improvement over the experimentally optimized static (constant temperature and pH) control strategy (30.0℃, pH 6.80). This work demonstrates the potential of integrating machine learning with dynamic process control to enhance industrial fermentation efficiency.
Efficient preparation and reuse of urease-active bacterial biomass remain important challenges for microbially induced carbonate precipitation (MICP). This study investigated metal ion-induced flocculation of Sporosarcina pasteurii using Ca2+, Fe2+, Fe3+, and Al3+, and evaluated the recovered flocculated biomass as a seed culture for subsequent low-inoculum cultivation and downstream MICP. The results showed that all four metal ions promoted formation and settling of urease-active composite flocs, while abiotic controls indicated that interactions between metal ions and organic components of yeast extract, together with pH-dependent metal hydrolysis, oxidation, and precipitation, contributed substantially to floc formation. Urea addition further altered the pH and settling behavior in a metal-specific manner, with the settled floc volume in the Ca2+ system reduced by up to 88% under the tested conditions. The recovered flocculated biomass retained ureolytic function and supported subsequent cultivation at reduced inoculation ratios. In particular, the Ca2+-treated seed culture maintained higher volumetric and OD600-normalized urease activities than the control during later cultivation, especially at inoculation ratios of 1:500 and 1:1000. Downstream MICP sand column tests further showed that the Ca2+-treated seed culture generally produced the highest CaCO3 content and UCS. These results demonstrate the feasibility of metal ion-induced flocculation for preparing reusable urease-active S. pasteurii seed cultures and linking biomass recovery with subsequent low-inoculum cultivation and MICP treatment.
Astaxanthin, a high-value ketocarotenoid, has gained significant attention for its potent antioxidant properties and wide-ranging applications across the nutraceutical, pharmaceutical, aquaculture, and cosmetic industries. Haematococcus pluvialis remains the most promising natural source for astaxanthin production, but its industrial application is hindered by slow growth, low biomass productivity, and the requirement for stress-induced accumulation. Recent advances in photobioreactor (PBR) technologies have emerged as a crucial strategy to overcome these bottlenecks. This review highlights progress in reactor engineering, including innovations to optimize light distribution, gas–liquid mass transfer, and trophic mode integration. The role of LED illumination in enhancing photosynthesis and astaxanthin biosynthesis, combined with reactor-scale economic analyses, provides insight into the feasibility of large-scale deployment. Furthermore, sustainability aspects such as energy efficiency, carbon footprint reduction, and waste minimization are critically evaluated. This review of advances in bioprocess design offers a comprehensive framework for translating laboratory-scale success into industrial-scale applications. The review aligns with UN SDGs 7 (Affordable and Clean Energy), 9 (Industry, Innovation, and Infrastructure), 12 (Responsible Consumption and Production), and 13 (Climate Action).
Long-chain alkanes in oil shale residue are persistent pollutants that threaten environmental safety. Using n-hexadecane (n-C16) as a model compound, we isolated a hydrocarbon-degrading microbial consortium from the residue and elucidated its metabolic mechanism. Three dominant strains—Pseudomonas stutzeri, Achromobacter xylosoxidans, and Aspergillus ochraceus —were identified and formulated into a composite consortium at an optimized cell abundance ratio of 2.0:3.0:1.0 (CFU basis, P. stutzeri: A. xylosoxidans: A. ochraceus), with a total viable count of~1.0×109 CFU·mL-1. Under conditions determined by response surface methodology, the consortium achieved 98.6% n-C16 degradation within 7 days. GC-MS analysis revealed a complete four-stage pathway: substrate uptake and emulsification, terminal oxidation (n-C16 → hexadecanol → hexadecanal → hexadecanoic acid), β-oxidation, and downstream mineralization. The functional complementarity among consortium members effectively eliminated metabolic bottlenecks, ensuring complete degradation. This work establishes a practical bioremediation framework for oil shale residue, converting it into a detoxified product suitable for safe land application, and offers a sustainable waste management solution for the unconventional energy sector.
Oxytetracycline (OTC) residues in livestock manure pose ecological risks during composting. In this study, montmorillonite-biochar composites prepared at different pyrolysis temperatures were evaluated for OTC removal and composting amendment. WM-500 was selected as a representative composite because it balanced OTC removal performance, biochar yield, and pyrolysis temperature. Independent adsorption tests confirmed its appreciable OTC retention capacity. During sheep manure composting, WM-500 promoted organic matter transformation and humification and accelerated OTC attenuation, particularly during the early stage, with final removal reaching 98.99%. It also reshaped bacterial succession and moderated stage-specific enrichment of tetracycline resistance genes, whereas mobile genetic elements showed marker-dependent responses. Correlation, redundancy analysis, and variation partitioning indicated that ARG dynamics were jointly associated with residual OTC, MGE profiles, bacterial succession, and composting conditions. Overall, WM-500 facilitated OTC attenuation through an adsorption-assisted coupled physicochemical-microbial process while concurrently modifying resistance-related ecological responses, supporting its potential for safer management of antibiotic-contaminated manure.
Glucose isomerase (GI) is a key enzyme in the production of high-fructose corn syrup. Immobilized GI is widely used for continuous production of high-fructose corn syrup. However, simultaneously realizing high enzyme loading, good mechanical strength and long-term operational stability under continuous high-temperature flow remains a considerable challenge. We developed a low-cost diatomite-bentonite core-shell composite carrier for efficient immobilization of glucose isomerase. With acid-oxidation-modified diatomite (DE) as the core, bentonite (BT) was introduced to form a swollen shell layer. After sequential adsorption of polyethyleneimine (PEI) and covalent cross-linking with glutaraldehyde (GA), an efficient immobilized glucose isomerase (GI@DE@BT) was successfully constructed. After systematic optimization of key immobilization parameters, the enzyme activity recovery rate of GI@DE@BT reached 82.1%. The half-life of immobilized GI at 80°C was extended to 38.4h, representing a 1.6-fold increase compared with free GI cells. The immobilized GI retained 80.2% of its initial activity after 40 batches，and the packed-bed reactor maintained a conversion rate above 43% during 100h of continuous operation. The developed novel core-shell immobilization strategy significantly enhances the stability and operational reliability of immobilized GI cells. Nevertheless, pilot-scale reactor data and long-term mechanical stability beyond 40 cycles are not yet available, and further scaled-up investigations are warranted to confirm its industrial potential.
Fermentation is a process with complex mechanism which is expensive to investigate. For this property, effective experimental design strategy is required when modeling and optimizing the process to satisfy different demands. Bayesian optimization (BO) is such an alternative to quickly model the effect and generate promising settings or candidates. Many variants of BO have been developed, but there is not a scheme dealing with the level-mismatch problem. In this situation, the settable levels of the input variables are limited by experimental availability, resulting in an inconsistent number of solutions among the variables. In this paper, a batch BO method is developed to address this problem by introducing a modified penalty coefficient into its acquisition function. The method shows the potential to a generate Pareto front when adopting the output of weighted summation. The method is demonstrated by a simulation case and a real fermentation case. The simulation case is utilized to compare different BO methods, while the real case is used to validate the method by seeking optimal multi-objective configurations of initial fermentation conditions. The results suggest the efficiency and flexibility of the method, showing that it avoids local optimum when dealing with multi-objective problems.
Continuous manufacturing is increasingly implemented for mammalian cell culture processes, but its use with bacterial systems such as Escherichia coli (E. coli) remains limited by the lack of robust technologies for continuous clarification of high-cell-density broths. This study presents a proof-of-concept evaluation of rotating disc dynamic crossflow filtration (DCF) for the continuous clarification of E. coli fermentation broth at approximately 30 g L⁻¹ cell dry weight. Ceramic membranes with nominal pore sizes of 80, 200, and 500 nm were screened with respect to transmembrane pressure (TMP) development, cell retention, and soluble protein transmission using recombinant green fluorescent protein (GFP) as a tracer. All membranes showed recurring cyclic operation, with pore-size-dependent differences in TMP development and visible protein passage. The 200 nm membrane was selected for long-duration operation based on its favorable hydraulic behavior, visible GFP-tracer passage, and its established use as a standard pre-chromatography processing step. Under continuous feed with dynamic inlet biomass concentrations, constant permeate flux was maintained for 48 h by torque- and TMP-triggered discharge control. Process behavior was governed by chamber solids mass, leading to recurring accumulation limits within the investigated runs of approximately 60 g and consistent discharge behavior under the tested torque- and TMP-based control strategies. The results demonstrate proof-of-concept continuous operability of rotating DCF for high-density bacterial clarification and identify chamber-solids accumulation as a key control-relevant process state.
Upgrading existing wastewater treatment plants to achieve higher efficiency requires compact process-intensification strategies, particularly where the construction of additional treatment units is constrained by limited land availability. This study evaluated a compact pilot-scale integrated fixed-film activated sludge (IFAS) membrane photobioreactor treating real municipal wastewater under low carbon-to-nitrogen (C/N) conditions. The system combined suspended activated sludge, attached biofilm, microalgae (Chlorella sp.) and submerged membrane filtration. Two operating periods were investigated by increasing the algae:sludge mass ratio from 1:5 to 1:3. The reactor achieved high carbon removal, with total chemical oxygen demand (COD) removal improving from 91.7 ± 4.3% to 95.9 ± 1.6% despite the higher influent organic load in the second period (1.5 ± 0.8kg COD/m3 d). Ammonium removal also increased from 82.2 ± 26.4% to 96.5 ± 5.5%, indicating enhanced nitrification capacity. In contrast, total nitrogen (TN) removal remained moderate, decreasing from 49.5 ± 23.6% to 39.8 ± 21.4% as carbon availability declined (C/N in period 2: 1.2g soluble COD/TN), confirming carbon limitation as the main bottleneck for complete nitrogen removal. Phosphate removal was poor overall, although a modest improvement was observed at a higher algal fraction. Biomass characterisation showed greater biofilm accumulation and algal proliferation in the second period, with attached biomass doubling and chlorophyll-a increasing from 2.7 ± 0.2 to 4.3 ± 1.5mg/L. Overall, the results demonstrate the feasibility of this IFAS membrane photobioreactor as an advanced proof of concept for low-COD/N wastewater treatment, while highlighting the need for further optimisation to improve denitrification and phosphorus removal.
Anaerobic structured-bed reactors are promising retained biomass systems because they decouple biomass retention from granule formation and provide a large surface area for biofilm development. However, high biomass retention may coexist with poor conversion when biomass becomes spatially compartmentalized. In this study, an up-flow anaerobic sludge blanket (UASB) reactor and an anaerobic structured-bed reactor (AnSTBR) were operated with the same feedstock and nominal operational targets for beet molasses treatment and subsequently challenged by hydraulic retention time reduction to assess how reactor architecture affects process stability and failure mechanisms. Reactor performance was interpreted together with flow cytometry, spatial biomass profiling, and 16S rRNA gene amplicon sequencing to assess whether deterioration coincided with generalized biomass loss or spatial biomass compartmentalization. Under hydraulic intensification conditions, the AnSTBR showed lower COD removal (71 ± 6%) than the UASB (87 ± 2%), despite retaining abundant biomass and showing no clear evidence of generalized physiological collapse. Post-operation inspection revealed pronounced gas entrapment and an estimated liquid containing volume of 1.2 L, indicating higher effective loading. Spatial microbial analyses showed biomass compartmentalization within the polyurethane foam, where recovered cell concentrations were one-to-two orders of magnitude higher than in the liquid phase. Together, these patterns were consistent with reduced hydraulic accessibility of retained biomass, although this was not directly quantified. In contrast, the UASB maintained higher functional stability by concentrating biomass near the influent-contacting sludge bed region. These findings show that biomass retention should be interpreted with spatial biomass distribution, gas disengagement, and effective reactor volume.
L-valine is an essential amino acid and Bacillus subtilis is recognized as a probiotic for humans and other vertebrates, and they are widely used as feedstuff. In order to construct a dual-functional B. subtilis with probiotic function and L-valine production, the feed probiotic B. subtilis ACCC11025 was rationally modified to enhance L‑valine biosynthesis. The strain Bs-Val4.2, which was constructed by substituting the promoters of the key genes in L-valine biosynthetic pathway (i.e., ilvBHCD, ybgE and ywaA) with the strong constitutive promoter Pp43, produced 12.7 ± 0.4 g/L L-valine in shake-flask fermentation. The L-valine production was further increased during inactivation of the branched-chain amino acid exporter regulator AzlB in strain Bs-Val4.2 (from 12.7 ± 0.4 g/L to 16.3 ± 1.2 g/L). In addition, inactivating the enzymes in pyruvate metabolic shunt (i.e., PoxB, IctE and PC) and reducing pdhA expression via replacing its native promoter with the growth phase-dependent promoter PrrnJ were shown to increase L-valine production, whereas overexpression of pyruvate kinase (PK) and malic enzyme (MleA) had little effect. Moreover, the branched-chain α-keto acid dehydrogenase (Bkd) was inactivated to increase L-valine production because of blocking the L-valine degradation. The final strain Bs-Val10.2 produced 43.7 ± 1.2 g/L of L valine in fed-batch fermentation, and showed the excellent performance in antimicrobial activity and antioxidant ability as well as digestive enzyme-producing capacity. In this study, an L-valine producing strain Bs-Val10.2 was construct using feed probiotics as host strain, and thus strain Bs-Val10.2 has the properties of probiotic function and L-valine production. These results of this study also provide a theoretical basis for simplifying the feed production technology.
Recirculating aquaculture systems (RAS) necessitate process-level models that integrate biological conversion with membrane behavior due to the potential accumulation of residual nitrogen, solids, and buffering instability. This study introduces a process-linked moving bed biofilm reactor (MBBR)–nanofiltration (NF) framework, wherein the predicted MBBR effluent serves as the NF feed to elucidate nitrogen transformation, feed conditioning, selective separation, and fouling development in ornamental-fish RAS. Koi wastewater was treated using Kaldnes K5 carriers and an NF90-4040 membrane at MBBR hydraulic retention times (HRTs) of 4, 5, and 6min, with NF recovery maintained at 30%. The model integrates apparent nitrification kinetics, modified by pH, alkalinity, dissolved oxygen, and temperature, with concentration-polarization-corrected rejection and resistance-based fouling. Under the tested conditions, the 5-min HRT demonstrated the optimal trade-off, resulting in permeate ammonia, nitrite, and total suspended solid (TSS) concentrations of 0.17, 0.03, and 0.01mgL-1, respectively. However, alkalinity decreased to 14.31mgL⁻¹ as CaCO3, and permeate ammonia exceeded the koi water-quality criterion. Overall ammonia reduction was limited to 46.3–58.5%, whereas nitrite reduction achieved 86.4–93.5%, and flux declined by 45–56%. Model predictions were consistent with observations for most outputs (r = 0.87–1.00), although near-zero nitrite and TSS permeate values resulted in higher relative errors. Sensitivity and scenario analyses identified TSS-related resistance, nitrification instability, recovery, and buffering as predominant control factors. The framework differentiates biological limitations, particulate carryover, membrane separation, and fouling-related hydraulic resistance, affirming that NF should function as a final polishing barrier following adequate MBBR feed stabilization.
