Engineering in Life Sciences最新文献

Bioreactor scale-up and scale-down have always been a topical issue for the biopharmaceutical industry and despite considerable effort, the identification of a fail-safe strategy for bioprocess development across scales remains a challenge. With the ubiquitous growth of digital transformation technologies, new scaling methods based on computer models may enable more effective scaling. This study aimed to evaluate the potential application of machine learning (ML) algorithms for bioreactor scale-up, with a specific focus on the prediction of scaling parameters. Factors critical to the development of such models were identified and data for bioreactor scale-up studies involving CHO cell-generated mAb products collated from the literature and public sources for the development of unsupervised and supervised ML models. Comparison of bioreactor performance across scales identified similarities between the different processes and primary differences between small- and large-scale bioreactors. A series of three case studies were developed to assess the relationship between cell growth and scale-sensitive bioreactor features. An embedding layer improved the capability of artificial neural network models to predict cell growth at a large-scale, as this approach captured similarities between the processes. Further models constructed to predict scaling parameters demonstrated how ML models may be applied to assist the scaling process. The development of data sets that include more characterization data with greater variability under different gassing and agitation regimes will also assist the future development of ML tools for bioreactor scaling.

Microplastics (MPs) are major concern due to their potential harm to ecosystems and most research has focused on their presence and fate, with limited attention to their biodegradation in aquatic ecosystems. Nevertheless, MPs act as hotspots for the colonization by a diverse range of microorganisms that can adhere to plastic surfaces, resulting in the subsequent formation of biofilms—a potential threat especially in terms of pathogenicity. This study employed 16S rRNA and 18S rRNA sequencing metagenomic analyses to investigate microbial communities within biofilms on plastic materials exposed to long-term marine and freshwater environments. Three Arcobacter species (Arcobacter nitrofigilis, Arcobacter acticola, and Arcobacter suis) emerged as dominant species in M_MP sample, while Flavobacterium tructae was the predominant species within the F_MP sample. The 18S rRNA sequencing revealed the presence of the fungal phylum Ascomycota and the microalgal species Pseudocharaciopsis ovalis in F_MP. Although, the primary species detected on M_MP and F_MP samples include bacteria previously implicated as pathogen, the predominant species identified in this study were unconnected to MP-associated biofilms or MP degradation. Their presence constitutes a novel discovery, opening promising avenues for the exploration of their potential involvement in the biodegradation of MPs within aquatic environments.

Polyhydroxyalkanoates (PHAs) have gained interest recently due to their biodegradability and versatility. In particular, the chemical compositions of medium-chain-length (mcl)-PHAs are highly diverse, comprising different monomers containing 6–14 carbon atoms. This review summarizes different feedstocks and fermentation strategies to enhance mcl-PHA production and briefly discusses the downstream processing. This review also provides comprehensive details on analytical tools for determining the composition and properties of mcl-PHA. Moreover, this study provides novel information by statistically analyzing the data collected from several reports on mcl-PHA to determine the optimal fermentation parameters (specific growth rate, PHA productivity, and PHA yield from various structurally related and unrelated substrates), mcl-PHA composition, molecular weight (MW), and thermal and mechanical properties, in addition to other relevant statistical values. The analysis revealed that the median PHA productivity observed in the fed-batch feeding strategy was 0.4 g L⁻¹ h⁻¹, which is eight times higher than that obtained from batch feeding (0.05 g L⁻¹ h⁻¹). Furthermore, 3-hydroxyoctanoate and -decanoate were the primary monomers incorporated into mcl-PHA. The investigation also determined the median glass transition temperature (−43°C) and melting temperature (47°C), which indicated that mcl-PHA is a flexible amorphous polymer at room temperature with a median MW of 104 kDa. However, information on the monomer composition or heterogeneity and the associated physical and mechanical data of mcl-PHAs is inadequate. Based on their mechanical values, the mcl-PHAs can be classified as semi-crystalline polymers (median crystallinity 23%) with rubber-like properties and a median elongation at break of 385%. However, due to the limited mechanical data available for mcl-PHAs with known monomer composition, identifying suitable processing tools and applications to develop mcl-PHAs further is challenging.

Corynebacterium glutamicum is used as an industrial platform organism for amino acid production. Previously, the organism was utilized to produce l-histidine with research focusing on metabolic engineering approaches to increase titer and yield. Only a few studies have been published that provide information on bioprocess development, with media optimization and fed-batch cultivation procedure being particularly promising areas. In this work, we show how experimental setups such as miniature cultivation technology, dynamic and time-optimized LC-MS/MS metabolic footprinting tools, and automated workflows for the detection of local and global metabolic patterns can significantly accelerate bioprocess development. Potential media bottlenecks in form of phosphate and magnesium availability were identified by sensitivity analysis in parallelized microscale cultivation assisted by lab automation. A rapid dilute-and-shoot flow-injection-analysis tandem mass spectrometry approach was used to cope with the resulting cultivation throughput and allowed to quantify amino acids with 1 min per sample. We were able to increase the l-histidine titer of a C. glutamicum random mutagenesis mutant by a factor of 5.8 through process optimization while also identifying both known and previously unknown targets for additional strain improvements. The presented methodology can be seen as a supplement to traditional approaches in the field of bioprocess development.

Innovations in viral vaccine manufacturing are crucial for pandemic preparedness and to meet ever-rising global demands. For influenza, however, production still mainly relies on technologies established decades ago. Although modern production shifts from egg-based towards cell culture technologies, the full potential has not yet been fully exploited. Here, we evaluate whether implementation of state-of-the-art technologies for cell culture-based recombinant protein production are capable to challenge outdated approaches in viral vaccine process development. For this, a fully automated single-cell cloning strategy was established to generate monoclonal suspension Madin-Darby canine kidney (MDCK) cells. Among selected cell clones, we could observe distinct metabolic and growth characteristics, with C59 reaching a maximum viable cell concentration of 17.3 × 10⁶ cells/mL and low doubling times in batch mode. Screening for virus production using a panel of human vaccine-relevant influenza A and B viruses in an ambr15 system revealed high titers with yields competing or even outperforming available MDCK cell lines. With C113, we achieved cell-specific virus yields of up to 25,000 virions/cell, making this cell clone highly attractive for vaccine production. Finally, we confirmed process performance at a 50-fold higher working volume. In summary, we present a scalable and powerful approach for accelerated development of high-yield influenza virus production in chemically defined medium starting from a single cell.

Digitalization with integrated devices, digital and physical assistants, automation, and simulation is setting a new direction for laboratory work. Even with complex research workflows, high staff turnover, and a limited budget some laboratories have already shown that digitalization is indeed possible. However, academic bioprocess laboratories often struggle to follow the trend of digitalization. Due to their diverse research circumstances, high variety of team composition, goals, and limitations the concepts are substantially different. Here, we will provide an overview on different aspects of digitalization and describe how academic laboratories successfully digitalized their working environment. The key aspect is the collaboration and communication between IT-experts and scientific staff. The developed digital infrastructure is only useful if it supports the laboratory worker and does not complicate their work. Thereby, laboratory researchers have to collaborate closely with IT-experts in order for a well-developed and maintainable digitalization concept that fits their individual needs and level of complexity. This review may serve as a starting point or a collection of ideas for the transformation toward a digitalized laboratory.

Process Analytical Technologies (PATs) are taking a key role in the run for automatization in the biopharmaceutical industry. Spectroscopic methods such as Raman spectroscopy or mid-infrared (MIR) spectroscopy are getting more recognition in the recent years for inline monitoring of bioprocesses due to their ability to measure various molecules simultaneously. However, their dependency on laborious model calibration making them a challenge to implement. In this study, a novel one-point calibration that requires a single reference point prior to the inline monitoring of glucose and lactate in bioprocesses with MIR spectroscopy is assessed with 22 mammalian cell perfusion (PER) processes in two different scales and four different products. Concentrations are predicted over all PERs runs with a root mean square error (RMSE) of 0.29 g/L for glucose and 0.24 g/L for lactate, respectively. For comparison conventional partial least square regression (PLSR) models were used and trained with spectroscopic data from six bioreactor runs in two different scales and three products. The general accuracy of those models (RMSE of 0.41 g/L for glucose and 0.16 g/L for lactate) are in the range of the accuracy of the one-point calibration. This shows the potential of the one-point calibration as an approach making spectroscopy more accessible for bioprocess development.