Frontiers in Bioengineering and Biotechnology

Human brain organoids have emerged as powerful models for studying the physiology, pathology, and evolution of the human brain. When combined with single-cell approaches, they offer the potential to directly probe the dynamics and mechanisms underlying cell fate decisions. A major technical bottleneck, however, remains the ability to reliably visualize and manipulate individual cells within their dense and heterogeneous tissue environment. Microinjection has proven effective for this purpose, allowing direct delivery of membrane-impermeable probes into single cells within intact tissue. Despite its versatility, widespread adoption of microinjection has been limited by its technically demanding and low-throughput nature. Automated microinjection systems developed for murine tissue have demonstrated that robotics can overcome these limitations, enabling systematic single-cell lineage tracing at scale. Here, we present a vision-guided robotic system capable of imaging organoid slices, identifying tissue boundaries, and targeting specific single cells for microinjection. This approach is generalizable across murine and human tissues and enables high-throughput single-cell manipulation, opening new avenues for studying human brain development at scale.

Lignin is an underutilised global resource with significant potential for the production of chemicals that are currently derived from fossil resources. However, biotechnological lignin valorisation faces various challenges including its recalcitrance and the toxicity of aromatic intermediates, products, and by-products like formaldehyde. While biochemical production from lignin-derived monomers has been demonstrated by disrupting native lignin degradation pathways, this approach required co-feeding additional carbon sources such as glucose for growth. This dependence on additional carbon sources can create competition with the food industry and undermine the economic sustainability of the bioprocess. Here, we report growth of a protocatechuate production strain of Pseudomonas putida EM42 on the by-products from p-coumarate and ferulate valorisation, achieving carbon efficiencies of up to 78 %. Additional flux balance analysis identified C1 assimilation pathways, including two novel pathways, beneficial for the growth on the formaldehyde by-product from ferulate degradation leading to improved carbon utilisation. This study demonstrates how by-product utilisation from lignin conversion can eliminate the need for co-feeding additional carbon sources, thereby potentially improving the efficiency of lignin valorisation.

Liver transplantation remains the only curative treatment for end-stage liver failure, yet its impact is constrained by organ shortages and graft non-utilisation. Machine perfusion (MP) enables ex-vivo assessment of donated livers; however, existing viability criteria rely on intermittent sampling, reducing temporal resolution and accuracy. Indocyanine green (ICG), a clinically validated dye cleared exclusively by hepatocytes, provides a continuous index of hepatic function beyond initial injury. Accordingly, we present a non-invasive, clamp-on optical sensor that enables continuous, real-time quantification of ICG clearance during MP. The sensor consists of a clamp-on module with an 808nm laser and phototransistor connected to a microcontroller-based unit and computer for real-time plotting. The raw phototransistor signal was linearised to a unitless absorbance signal proportional to perfusate ICG; bi-exponential fitting yielded plasma disappearance rate (PDRbi, %/min) and the 15-minute residual fraction (R15). Across 10 whole and 3 split human livers (45 boluses; 13 paired with spectrophotometry), the sensor closely matched spectrophotometric measurements (mean R2 = 0.994; range 0.983-0.999). The sensor resolved expected physiological trends: ICG clearance increased with temperature (PDRbi: 8.2%/min (subnormothermic MP) to 22.6%/min (normothermic MP) (n=4); 9.3%/min (32{degrees}C) to 11.9%/min (36{degrees}C) (n=1)). The sensors continuous signal traces also revealed early mixing dynamics and medication-related effects that are missed by intermittent sampling. This optical sensor enables accurate, real-time monitoring of ICG clearance during ex-vivo perfusion. The ex-vivo setting is uniquely positioned to validate ICG clearance models and enhance clinical interpretation.

The development of autonomous agents in bioporcess development is crucial for advancing biopharma innovation, as it can significantly reduce the time and resources required to transition from product to process. While robotics and machine learning have greatly accelerated drug discovery and initial screening, the later stages of development have primarily benefited from experimental automation, lacking advanced computational tools for experimental planning and execution. For example, in the development of new monoclonal antibodies, the search for optimal upstream conditions (such as feeding strategy, pH, temperature, and media composition) is often conducted using sophisticated high-throughput (HT) mini-bioreactor systems, while the integration of machine learning tools for experimental design and operation in these systems have not matured accordingly. In this work, we introduce an integrated user-friendly software framework that combines a Bayesian experimental design algorithm, a cognitive digital twin of the cultivation system, and an advanced 24-parallel mini-bioreactor perfusion experimental setup. This results in an autonomous experimental machine capable of: (1) embedding existing process knowledge, (2) learning during experimentation, utilizing information from similar processes, (4) predicting future events, and (5) autonomously operating the parallel cultivation setup to achieve challenging objectives. As proof of concept, we present experimental results from 27-day-long cultivations operated by the autonomous software agent, which successfully achieved challenging goals such as increasing the viable cell volume (VCV) and maximizing the viability throughout the experiment.

Bioprocess development is commonly characterized by long development times, especially in the early screening phase. After promising candidates have been pre-selected in screening campaigns, an optimal operating strategy has to be found and verified under conditions similar to production. Cultivating cells with pulse-based feeding and thus exposing them to oscillating feast and famine phases has shown to be a powerful approach to study microorganisms closer to industrial bioreactor conditions. In view of the large number of strains and the process conditions to be tested, high-throughput cultivation systems provide an essential tool to sample the large design space in short time. We have recently presented a comprehensive platform, consisting of two liquid handling stations coupled with a model-based experimental design and operation framework to increase the efficiency in High Throughput bioprocess development. Using calibrated macro-kinetic growth models, the platform has been successfully used for the development of scale-down fed-batch cultivations in parallel mini-bioreactor systems. However, it has also been shown that parametric uncertainties in the models can significantly affect the prediction accuracy and thus the reliability of optimized cultivation strategies. To tackle this issue, we implemented a multi-stage Model Predictive Control (MPC) strategy to fulfill the experimental objectives under tight constraints despite the uncertainty in the parameters and the measurements. Dealing with uncertainties in the parameters is of major importance, since constraint violation would easily occur otherwise, which in turn could have adverse effects on the quality of the heterologous protein produced. Multi-stage approaches build up scenario tree, based on the uncertainty that can be encountered and computing optimal inputs that satisfy the constrains despite of such uncertainties. Using the feedback information gained through the evolution along the tree, the control approach is significantly more robust than standard MPC approaches without being overly conservative. We show in this study that the application of multi-stage MPC can increase the number of successful experiments, by applying this methodology to a mini-bioreactor cultivation operated in parallel.

Recent technological advances in the field of Additive Manufacturing (AM) and the increasing need in Regenerative Medicine (RM) for devices that better and better mimic native tissues architecture are showing limitations in the current scaffolds fabrication techniques. A switch from the typical layer-by-layer approach is needed to achieve precise control on fibers orientation and pores dimension and morphology. In this work a new AM apparatus, the RAVEN (Robot-Assisted Volumetric ExtrusioN) system, is presented. RAVEN is based on a 7-DOF robotic arm and an FDM extruder and allows for volumetric extrusion of polymeric filaments. The development process, namely the robotic motion optimization, the optimization towards small-scale trajectories, the custom-made hardware/software interfaces, and the different printing capabilities are hereby presented. The successful results are promising towards future advanced applications such as in vivo bioprinting, in which the ability of the robot to change its configuration while printing will be crucial.

We present three quantitative methods: 1) estimation of exoskeleton mechanical power and energy ratio from published data, 2) a systematic review of the exoskeleton literature on reported energy ratios, and 3) timing correction analysis of the replication experiment, to address concerns raised by Collins et al. (2026) about Luo et al. (2024). Together, these analyses support the reported metabolic reductions and the validity of exoskeleton control via learning in simulation. The critique rests on an unsupported premise: that exoskeleton energy ratios above 4 are physiologically implausible. This premise of Collins et al. (2026) is not supported by the cited evidence, and the error originates in their own cited source. Sawicki and Ferris (2009), the paper they invoke as authority for the limit of 4, state explicitly that "reported values of the muscular efficiency range from 0.10 to 0.34, with many sources assuming an average of [~]0.25." The value of 4 corresponds to this average, it is not a physiological ceiling. Treating an average as a physiological upper limit is a fundamental error. The published exoskeleton literature further contradicts the claim, including work by the authors of the critique themselves (Collins et al., 2015: 4.3; Young et al., 2017: 5.0) and independent work (Malcolm et al., 2013: 4.8; Seo et al., 2017: 6.7). In contrast, our walking energy ratio is 2.4, calculated directly from Fig. 4 of our paper. Our device delivers higher peak torque (14.1 Nm vs. 10.9 Nm, Lim et al., 2019) and achieves a slightly larger metabolic reduction (24.3% vs. 21%). Independent groups have since demonstrated meaningful metabolic reductions using learning-in-simulation frameworks, including Barati et al. (2026, 15.2% mean and 22.5% maximum) and Zhou et al. (2025, [~]20% during running). The claim of Collins et al. (2026) that this problem "remains unsolved" is directly contradicted by these independent results. The experiment in the critique is not a valid replication of our method. Our controller is a neural network with [~]10,000 parameters learned through deep reinforcement learning in musculoskeletal simulation; the critique instead applies a pre-programmed fixed torque curve with no learnable parameters. Beyond this, the replication contains three methodological errors: 1) a heel-strike timing assumption producing offsets up to 30% of the gait cycle; 2) an averaged torque profile that discards subject-specific control; and 3) a device [~]50% heavier than ours (4.8 kg vs. 3.2 kg) without measuring the metabolic penalty of the added weight. The critique also misreports Samsung data, with reported values approximately double those in the original publication, errors that directly underpin their physiological limit argument.

Human pluripotent stem cells (hPSCs) hold significant promise for regenerative medicine, yet optimizing their expansion in three-dimensional bioreactor systems remains challenging due to complex interactions between mechanical forces, metabolic constraints, and aggregate formation dynamics. This study developed and validated a mechanistic mathematical model to predict hPSC proliferation dynamics in vertical-wheel bioreactor (VWBR) systems, incorporating the effects of shear stress and energy dissipation rate (EDR) on cell growth and aggregate dynamics. Seven model variants employing different kinetic formulations for shear stress and energy dissipation rate effects were systematically evaluated through model selection, identifiability analyses, and experimental validation. Experimental data from six bioreactor conditions varying in initial cell density (2 x 104-15 x 104 cells/mL), agitation rate (30-60 RPM), and working volume (100-500 mL) were used for model calibration and selection. Bayesian Information Criterion analysis identified a model combining Michaelis-Menten kinetics for shear stress inhibition with a EDR-mediated aggregate detachment formulation as the best-performing variant, achieving a Mean Relative Prediction Error of 13.97%, comparable to the experimental variability of 16.29%. Independent validation experiments using leave-out data gathered under different media exchange schedules confirmed model accuracy with prediction errors below 14%, consistent with observed experimental variability around 12%. The validated model was used to optimize the media exchange protocol, leading to a 37.5% reduction in media consumption with only a 13.5% reduction in final cell yield, demonstrating its utility for prospective, quantitative bioprocess design in VWBR systems.

Gene editing technologies such as zinc finger nuclease (ZFN), transcription activator-like effector nuclease (TALEN), and clustered regularly interspaced short palindromic repeats (CRISPR) have revolutionized genetic engineering and now are being used in clinical gene therapy. We systematically reviewed gene editing clinical trials from ClinicalTrials.gov using a searching strategy that included all different gene editing technologies, followed by two rounds of independent assessment based on the inclusion and exclusion criteria, data extraction, and review of the background publications. 76 trials met our inclusion criteria including 30 studies on genetically engineered T-cell therapies for cancer, 19 studies on virus infections, and 26 studies on monogenic diseases. We have also analyzed the proportions to which different gene editing and gene delivery methods are used. We observed a growing trend of registered CRISPR-based trials indicating a raising interest in developing new therapeutic methods based on this technology. Overall, our study showed that there are promising phase-I and -II trials testing the safety and feasibility of gene editing in different clinical settings.

Animal models have been used to investigate urinary tract catheters, but most do not accurately recapitulate human bladder physiology nor support different types of catheters. Recent advances in three-dimensional models mimicking tissue microenvironments have improved our understanding of human tissue development and allowed us to model many diseases. Here, our objective was to characterise bladder urothelial mucoadhesion and microtrauma associated with intermittent catheterisation. We employed our three- dimensional Urine-tolerant Human Urothelial model (3D-UHU) to investigate two different intermittent catheter types, a polyvinylpyrrolidone (PVP)-coated catheter (PVP-CC), and a coating-free integrated amphiphilic surfactant (IAS) catheter. We showed that pressing catheters onto the models for two minutes caused compression of the 3D-UHU model in contact regions, with disruption of urothelial umbrella cells detected by immunofluorescence staining of surface markers uroplakin III, cytokeratin-20 and chondroitin sulphate, alongside occasional effacement of apical surfaces exposing intermediate layers. PVP-CC had a significantly higher number of urothelial cells adhered to its surface after catheter removal compared with IAS. Moreover, application and removal of PVP-CC caused a decrease in transepithelial electrical resistance, suggesting barrier disruption, whereas IAS did not cause a statistically different effect. However, no change in paracellular permeability rates assessed by FITC-dextran were observed in models after application of catheter pieces. Finally, compression of models induced trauma-triggered inflammatory cytokine responses. Specifically, we observed increased secretion of the pro- inflammatory cytokine IL-1{beta} as well as the cell adhesion molecule CEACAM1 after exposure to PVP-CC compared with IAS. Our findings suggest that the IAS catheter damaged the epithelium less than the PVP-CC catheter. These data demonstrate that the 3D-UHU model holds promise as an alternative to animal models for investigating the effects of urinary catheters on the urothelium.

Kinetic metabolic models of central metabolism have been proposed to understand how Saccharomyces cerevisiae navigates through nutrient perturbations. Yet, these models lacked important variables that constrain metabolism under relevant physiological conditions and thus have limited operational use such as in optimization of industrial fermentations. In this work, we developed a physiologically informed kinetic model of yeast glycolysis connected to central carbon metabolism by including the effect of anabolic reactions precursors, mitochondria and the trehalose cycle. A parameter estimation pipeline was developed, consisting of a divide and conquer approach, supplemented with regularization and global optimization. We show how this first mechanistic description of a growing yeast cell captures experimental dynamics at different growth rates and under a strong glucose perturbation, is robust to parametric uncertainty and explains the contribution of the different pathways in the network. Our work suggests that by combining multiple types of data and computational methods, complex but physiologically representative and robust models can be achieved.

There is an urgent need to ameliorate the transport of the reconstituted vaccines to the vaccination sites to improve the COVID-19 vaccination campaigns. The maintenance of the integrity of the mRNA of the different COVID-19 reconstituted vaccines after continuous movement at room temperature during at least three hours ensures the safety of a ground transportation.

Process systems engineering methods and tools have been difficult to apply in bioprocess engineering, mainly due to the high complexity of biological systems and the low reproducibility of the experiments. High throughput robotic cultivation platforms in combination with computational tools for experimental design, resource scheduling, and operation, are rapidly gaining popularity. One important contribution being the generation of data in high throughput needed to overcome this lack of data with high information content and the worrying reproducibility crisis in life sciences. In this work, directed acyclic graphs are used to represent, manage and track all experimental workflows in a robotic platform. They support data provenance and enable traceability and reproducibility of workflows in robotic facilities. The experimental workflows are automated using Apache Airflow enabling to manage all necessary steps for fed-batch cultivations, including sampling, sample transport by a mobile robot, feed additions, data collection, storage in a SQL database and model fitting. The added value of this system is demonstrated in scale-down experiments, where E. coli BL21 (DE3), producing elastin like proteins, exhibits robustness towards glucose oscillations that mimic industrial cultivation conditions.

Formate is a promising energy carrier that could be used to transport renewable electricity. Some acetogenic bacteria, such as Eubacterium limosum, have the native ability to utilise formate as a sole substrate for growth, which has sparked interest in the biotechnology industry. However, formatotrophic metabolism in acetogens is poorly understood, and a systems-level characterization in continuous cultures is yet to be reported. Here we present the first steady-state dataset for E. limosum formatotrophic growth. At a defined dilution rate of 0.4 d-1, there was a high specific uptake rate of formate (280{+/-}56 mmol/gDCW/d), however, most carbon went to CO2 (150{+/-}11 mmol/gDCW/d). Compared to methylotrophic growth, protein differential expression data and intracellular metabolomics revealed several key features of formate metabolism. Upregulation of pta appears to be a futile attempt of cells to produce acetate as the major product. Instead, a cellular energy limitation resulted in the accumulation of intracellular pyruvate and upregulation of Pfl to convert formate to pyruvate. Therefore, metabolism is controlled, at least partially, at the protein expression level, an unusual feature for an acetogen. We anticipate that formate could be an important one-carbon substrate for acetogens to produce chemicals rich in pyruvate, a metabolite generally in low abundance during syngas growth.

When striving for maximal throughput at minimal volumes while cultivating close to industrial conditions, simple and robust feeding strategies offer important advantages. Enzyme-mediated glucose cleavage from dextrin is an easy way of imitating continuous fed-batch in the small scale, with no complex equipment required. While the release rate - and thus the feed rate - can be controlled by adapting the enzyme concentration, it strongly depends on the concentration of the involved substances and the environmental conditions. Thus, it is a challenge to use the technology for controlling the specific growth rate, as it is commonly done with feed pumps. For solving this problem, we present here a mathematical model that extends simple Michaelis-Menten kinetics by considering different substrate fractions and can be applied to control the glucose release rate even in high throughput experiments. The fitted model was used during automated microbial cultivations to control the growth rate in quasi-continuous fed-batch processes and to realize different exponential growth rates by intermittent additions of enzyme and dextrin by a liquid handling robot system. We thus present an approach for defined biocatalytically controlled glucose supply of small-scale systems, where - if at all - continuous feeding was only possible with low accuracy or high technical efforts until now.

The introduction of the bufotoxin-secreting cane toad (Rhinella marina) to Queensland in 1935 has had a devastating impact on wildlife in the Australian tropics. Having evolved for millions of years in the absence of cane toads or other bufotoxin-secreting organisms, many of the Australias native predators that include cane toads in their diet suffered large population declines following cane toad invasion to their habitat. One marsupial species, the northern quoll (Dasyurus hallucatus), is now classified as endangered (IUCN Red List) largely due to bufotoxin ingestion. This study aimed to introduce bufotoxin resistance into a marsupial cell line by editing part of the ATP1A1 gene encoding the extracellular H1-H2 domain - the binding target of bufotoxin. To this end, CRISPR prime editing was used to replace the part of the wildtype ATP1A1 gene encoding the H1-H2 domain in fibroblasts of a related marsupial model, the fat-tailed dunnart (Sminthopsis crassicaudata), with modifications known to be associated with bufotoxin resistance. The genetically modified cell population showed a >45-fold increase in resistance to bufalin (an active component of bufotoxin) compared to wild type. This study provides a proof of concept towards engineering genetic resistance in the northern quoll to halt or even reverse its current population decline.

Gene therapy has been the subject of extensive research following the advent of gene-editing technologies. Genetic disorders with difficult-to-target tissues, such as cystic fibrosis (CF), still face many challenges in developing efficacious gene therapy. The potential universal approach of gene replacement involves inserting a functional CFTR gene after generating DNA double strand breaks using gene editors such as CRISPR/Cas9. However, this strategy has not achieved clinical significance, as CRISPR/Cas9-mediated integration of CFTR is limited primarily by the infrequent activity of the homology-directed repair (HDR) pathway. To circumvent this limitation and improve CFTR transgene integration and expression, we explored a method of adding a second integration site, which we termed the dual-locus-targeting method. Using a helper-dependent adenoviral vector (HDAd)-delivered CRISPR/Cas9 system in porcine epithelial cells, we found that sequential delivery of two vectors, one targeting the CFTR locus and the other the genomic safe harbour site GGTA1, enhanced the integration efficiency of lacZ and CFTR donor genes to 16.5% and 3.4%, respectively. These results demonstrated a potential strategy to improve the efficacy of CFTR replacement for the development of a universal and permanent gene therapy treatment for CF lung disease. GRAPHICAL ABSTRACT O_FIG O_LINKSMALLFIG WIDTH=200 HEIGHT=76 SRC="FIGDIR/small/731381v1_ufig1.gif" ALT="Figure 1"> View larger version (17K): org.highwire.dtl.DTLVardef@1774590org.highwire.dtl.DTLVardef@1782915org.highwire.dtl.DTLVardef@1d13b12org.highwire.dtl.DTLVardef@17d3f93_HPS_FORMAT_FIGEXP M_FIG C_FIG

Biomanufacturing can play a pivotal role in the transition away from fossil fuel dependence for the production of chemicals and fuels. There is growing interest in alternative bioproduction feedstocks to conventional sugars that do not compete for land use with food production. Ethylene glycol, a C2 compound that can be recovered from plastic waste or derived from CO2 with increasing efficiency, is gaining attention as a carbon source for microbial processes. Here we review the natural and synthetic metabolic pathways currently available for ethylene glycol assimilation. The pathways are compared in terms of their maximum theoretical yields for biomass and value-added products, thermodynamic favourability, minimum enzyme costs, and orthogonality to central carbon metabolism. We find that synthetic pathways outperform their natural counterparts in terms of higher thermodynamic driving forces, reduced enzyme costs, and higher theoretical yields for the majority of bioproducts analyzed as well as for biomass. However, natural assimilation pathways are equally or even more orthogonal to growth-associated reactions than synthetic pathways. Given these tradeoffs, the optimal EG assimilation pathway may depend on product and process choice.

The silent pandemic of antibiotic resistance is thriving, prompting the urgent need for the development of new antibacterial drugs. However, within the preclinical pipeline, in vitro screening conditions can differ significantly from the final in vivo settings. To bridge the gap between in vitro and in vivo assays, we developed a pig skin-based bioluminescent ex vivo burn wound infection model, enabling real-time assessment of antibacterials in a longitudinal, non-destructive manner. We provide a proof-of-concept for A. baumannii NCTC13423, a multidrug-resistant clinical isolate, which was equipped with the luxCDABE operon as a reporter using a Tn7-based tagging system. This bioluminescence model provided a linear correlation between the number of bacteria and a broad dynamic range (104 to 109 CFU). This longitudinal model was subsequently validated using a fast-acting enzybiotic as an antibacterial. Since this model combines a realistic, clinically relevant yet strictly controlled environment with real-time measurement of bacterial burden, we put forward this ex vivo model as a valuable tool to assess the preclinical potential of novel antibacterial compounds. Summary statementHere, we demonstrate the potential of a bioluminescence-based ex vivo model for the longitudinal assessment of antibacterials. Moreover, we also provide a proof-of-concept with an engineered lysin.

Methanol is a promising renewable energy carrier that can be used as a favourable substrate for biotechnology, due to its high energy efficiency conversion and ease of integration within existing infrastructure. Some acetogenic bacteria have the native ability to utilise methanol, along with other C1 substrates such as CO2 and formate, to produce valuable chemicals. Continuous cultures favour economically viable bioprocesses, however, the performance of acetogens has not been investigated at the molecular level when grown on methanol. Here we present steady-state chemostat quantification of the metabolism of Eubacterium limosum, finding maximum methanol uptake rates up to 640{+/-}22 mmol/gDCW/d, with significant fluxes to butyrate. To better understand metabolism of acetogens under methanol growth conditions, we sampled chemostats for proteomics and metabolomics. Changes in protein expression and intracellular metabolomics highlighted key aspects of methanol metabolism, and highlighted bottleneck conditions preventing formation of the more valuable product, butanol. Interestingly, a small amount of formate in methylotrophic metabolism triggered a cellular state known in other acetogens to correlate with solventogenesis. Unfortunately, this was prevented by post-translation effects including an oxidised NAD pool. There remains uncertainty around ferredoxin balance at the methylene-tetrahydrofolate reductase (MTHFR) and at the Rnf level.