Growth and productivity impacts of periplasmic nuclease expression in an Escherichia coli Fab' fragment production strain

An autonucleolytic suspension HEK293F host cell line for high-titer serum-free AAV5 and AAV9 production with reduced levels of DNA impurity

We sought to engineer mammalian cells to secrete nuclease activity as a step toward removing the need to purchase commercial nucleases as process additions in bioprocessing of AAV5 and AAV9 as gene therapy vectors. Engineering HeLa cells with a serratial nuclease transgene did not bring about nuclease activity in surrounding media whereas engineering serum-free, suspension-adapted HEK293F cells with a staphylococcal nuclease transgene did result in detectable nuclease activity in surrounding media of the resultant stable transfectant cell line, "NuPro-1S." When cultivated in serum-free media, NuPro-1S cells yielded 3.06 × 1010 AAV5 viral genomes (vg)/mL via transient transfection, compared with 3.85 × 109 vg/mL from the parental HEK293F cell line. AAV9 production, followed by purification by ultracentrifugation, yielded 1.8 × 1013 vg/mL from NuPro-1S cells compared with 7.35 × 1012 vg/mL from HEK293F cells. AAV9 from both HEK293F and NuPro-1S showed almost identical ability to transduce cells embedded in a scaffold tissue mimic or cells of mouse neonate brain tissue in vivo. Comparison of agarose gel data indicated that the DNA content of AAV5 and AAV9 process streams from NuPro-1S cells was reduced by approximately 60% compared with HEK293F cells. A similar reduction in HEK293F cells was only achievable with a 50 U/mL Benzonase treatment.

Serum-free lentiviral vector production is compatible with medium-resident nuclease activity arising from adherent HEK293T host cells engineered with a nuclease-encoding transgene

At present lentiviral vector production for cell and gene therapy commonly involves transient plasmid transfection of mammalian cells cultivated in serum-containing media and addition of exogenous nuclease to reduce host cell and plasmid DNA impurities. Switching from serum-containing media to chemically-defined, serum free media, and minimising the number of process additions, are both increasingly regarded as necessary steps for simplifying and potentially automating lentiviral vector bioprocessing in future. Here we adapted human embryonic kidney 293T (HEK293T) cells to grow in serum-free media and also modified these cells with transgenes designed to encode a secreted nuclease activity. Stable transfection of HEK293T cells with transgenes encoding the Staphylococcus aureus nuclease B (NucB) open reading frame with either its native secretion signal peptide, the murine Igκ chain leader sequence or a novel viral transport fusion protein, all resulted in qualitatively detectable nuclease activity in serum-free media. Serum-free transient transfection of human embryonic kidney HEK293T cells stably harbouring the transgene for NucB with its native secretion signal produced active lentivirus in the presence of medium-resident nuclease activity. This lentivirus material was able to transduce the AGF-T immortal T cell line with a green fluorescent protein reporter payload at a level of 2.05 × 10⁵ TU/mL (±3.34 × 10⁴ TU/mL). Sufficient nuclease activity was present in 10 μL of this unconcentrated lentivirus material to degrade 1.5 μg DNA within 2 h at 37 °C, without agitation - conditions compatible with lentivirus production. These observations demonstrate that lentiviral vector production, by transient transfection, is compatible with host cells harbouring a nuclease transgene and evidencing nuclease activity in their surrounding growth media. This work provides a solid basis for future investigations, beyond the scope of this present study, in which commercial and academic groups can apply this approach to therapeutic payloads and potentially omit exogenous nuclease bioprocess additions.

Production of indigo by recombinant bacteria

Indigo is an economically important dye, especially for the textile industry and the dyeing of denim fabrics for jeans and garments. Around 80,000 tonnes of indigo are chemically produced each year with the use of non-renewable petrochemicals and the use and generation of toxic compounds. As many microorganisms and their enzymes are able to synthesise indigo after the expression of specific oxygenases and hydroxylases, microbial fermentation could offer a more sustainable and environmentally friendly manufacturing platform. Although multiple small-scale studies have been performed, several existing research gaps still hinder the effective translation of these biochemical approaches. No article has evaluated the feasibility and relevance of the current understanding and development of indigo biocatalysis for real-life industrial applications. There is no record of either established or practically tested large-scale bioprocess for the biosynthesis of indigo. To address this, upstream and downstream processing considerations were carried out for indigo biosynthesis. 5 classes of potential biocatalysts were identified, and 2 possible bioprocess flowsheets were designed that facilitate generating either a pre-reduced dye solution or a dry powder product. Furthermore, considering the publicly available data on the development of relevant technology and common bioprocess facilities, possible platform and process values were estimated, including titre, DSP yield, potential plant capacities, fermenter size and batch schedule. This allowed us to project the realistic annual output of a potential indigo biosynthesis platform as 540 tonnes. This was interpreted as an industrially relevant quantity, sufficient to provide an annual dye supply to a single industrial-size denim dyeing plant. The conducted sensitivity analysis showed that this anticipated output is most sensitive to changes in the reaction titer, which can bring a 27.8% increase or a 94.4% drop. Thus, although such a biological platform would require careful consideration, fine-tuning and optimization before real-life implementation, the recombinant indigo biosynthesis was found as already attractive for business exploitation for both, luxury segment customers and mass-producers of denim garments.

Graphical Abstract

Supplementary Information

The online version contains supplementary material available at 10.1186/s40643-023-00626-7.

Extraction of recombinant periplasmic proteins under industrially relevant process conditions: Selectivity and yield strongly depend on protein titer and methodology

In this work, we attempted to identify a method for the selective extraction of periplasmic endogenously expressed proteins, which is applicable at an industrial scale. For this purpose, we used an expression model that allows coexpression of two fluorescent proteins, each of which is specifically targeted to either the cytoplasm or periplasm. We assessed a number of scalable lysis methods (high‐pressure homogenization, osmotic shock procedures, extraction with ethylenediaminetetraacetic acid, and extraction with deoxycholate) for the ability to selectively extract periplasmic proteins rather than cytoplasmic proteins. Our main conclusion was that although we identified industrially scalable lysis conditions that significantly increased the starting purity for further purification, none of the tested conditions were selective for periplasmic protein over cytoplasmic protein. Furthermore, we demonstrated that efficient extraction of the expressed recombinant proteins was largely dependent on the overall protein concentration in the cell.

Improving Fab' fragment retention in an autonucleolytic Escherichia coli strain by swapping periplasmic nuclease translocation signal from OmpA to DsbA

Objectives

To reduce unwanted Fab’ leakage from an autonucleolytic Escherichia coli strain, which co-expresses OmpA-signalled Staphylococcal nuclease and Fab’ fragment in the periplasm, by substituting in Serratial nuclease and the DsbA periplasm translocation signal as alternatives.

Results

We attempted to genetically fuse a nuclease from Serratia marcescens to the OmpA signal peptide but plasmid construction failed, possibly due to toxicity of the resultant nuclease. Combining Serratial nuclease to the DsbA signal peptide was successful. The strain co-expressing this nuclease and periplasmic Fab’ grew in complex media and exhibited nuclease activity detectable by DNAse agar plate but its growth in defined medium was retarded. Fab’ coexpression with Staphylococcal nuclease fused to the DsbA signal peptide resulted in cells exhibiting nuclease activity and growth in defined medium. In cultivation to high cell density in a 5 l bioreactor, DsbA-fused Staphylococcal nuclease co-expression coincided with reduced Fab’ leakage relative to the original autonucleolytic Fab’ strain with OmpA-fused staphylococcal nuclease.

Conclusions

We successfully rescued Fab’ leakage back to acceptable levels and established a basis for future investigation of the linkage between periplasmic nuclease expression and leakage of co-expressed periplasmic Fab’ fragment to the surrounding growth media.

Detecting cell lysis using viscosity monitoring in E. coli fermentation to prevent product loss

Monitoring the physical or chemical properties of cell broths to infer cell status is often challenging due to the complex nature of the broth. Key factors indicative of cell status include cell density, cell viability, product leakage, and DNA release to the fermentation broth. The rapid and accurate prediction of cell status for hosts with intracellular protein products can minimise product loss due to leakage at the onset of cell lysis in fermentation. This article reports the rheological examination of an industrially relevant E. coli fermentation producing antibody fragments (Fab'). Viscosity monitoring showed an increase in viscosity during the exponential phase in relation to the cell density increase, a relatively flat profile in the stationary phase, followed by a rapid increase which correlated well with product loss, DNA release and loss of cell viability. This phenomenon was observed over several fermentations that a 25% increase in broth viscosity (using induction-point viscosity as a reference) indicated 10% product loss. Our results suggest that viscosity can accurately detect cell lysis and product leakage in postinduction cell cultures, and can identify cell lysis earlier than several other common fermentation monitoring techniques. This work demonstrates the utility of rapidly monitoring the physical properties of fermentation broths, and that viscosity monitoring has the potential to be a tool for process development to determine the optimal harvest time and minimise product loss. © 2016 The Authors. Biotechnology Progress published by Wiley Periodicals, Inc. on behalf of American Institute of Chemical Engineers, 32:1069-1076, 2016.

Promoter engineering to optimize recombinant periplasmic Fab' fragment production in Escherichia coli

Fab' fragments have become an established class of biotherapeutic over the last two decades. Likewise, developments in synthetic biology are providing ever more powerful techniques for designing bacterial genes, gene networks and entire genomes that can be used to improve industrial performance of cells used for production of biotherapeutics. We have previously observed significant leakage of an exogenous therapeutic Fab' fragment into the growth medium during high cell density cultivation of an Escherichia coli production strain. In this study we sought to apply a promoter engineering strategy to address the issue of Fab' fragment leakage and its consequent bioprocess challenges. We used site directed mutagenesis to convert the Ptac promoter, present in the plasmid, pTTOD-A33 Fab', to a Ptic promoter which has been shown by others to direct expression at a 35% reduced rate compared to Ptac . We characterized the resultant production trains in which either Ptic or Ptac promoters direct Fab' fragment expression. The Ptic promoter strain showed a 25-30% reduction in Fab' expression relative to the original Ptac strain. Reduced Fab' leakage and increased viability over the course of a fed-batch fermentation were also observed for the Ptic promoter strain. We conclude that cell design steps such as the Ptac to Ptic promoter conversion reported here, can yield significant process benefit and understanding with respect to periplasmic Fab' fragment production. It remains an open question as to whether the influence of transgene expression on periplasmic retention is mediated by global metabolic burden effects or periplasm overcapacity. © 2016 American Institute of Chemical Engineers Biotechnol. Prog., 32:840-847, 2016.

'Zipbody' leucine zipper-fused Fab in E. coli in vitro and in vivo expression systems

A small antibody fragment, fragment of antigen binding (Fab), is favorable for various immunological assays. However, production efficiency of active Fab in microorganisms depends considerably on the clones. In this study, leucine zipper-peptide pairs that dimerize in parallel (ACID-p1 (LZA)/BASE-p1 (LZB) or c-Jun/c-Fos) were fused to the C-terminus of heavy chain (Hc, VH-CH1) and light chain (Lc, VL-CL), respectively, to accelerate the association of Hc and Lc to form Fab in Escherichia coli in vivo and in vitro expression systems. The leucine zipper-fused Fab named 'Zipbody' was constructed using anti-E. coli O157 monoclonal antibody obtained from mouse hybridoma and produced in both in vitro and in vivo expression systems in an active form, whereas Fab without the leucine zipper fusion was not. Similarly, Zipbody of rabbit monoclonal antibody produced in in vitro expression showed significant activity. The purified, mouse Zipbody produced in the E. coli strain Shuffle T7 Express had specificity toward the antigen; in bio-layer interferometry analysis, the KD value was measured to be 1.5-2.0 × 10(-8) M. These results indicate that leucine zipper fusion to Fab C-termini markedly enhances active Fab formation in E. coli.

Evaluation of options for harvest of a recombinant E. Coli fermentation producing a domain antibody using ultra scale-down techniques and pilot-scale verification

Ultra scale‐down (USD) methods operating at the millilitre scale were used to characterise full‐scale processing of E. coli fermentation broths autolysed to different extents for release of a domain antibody. The focus was on the primary clarification stages involving continuous centrifugation followed by depth filtration. The performance of this sequence was predicted by USD studies to decrease significantly with increased extents of cell lysis. The use of polyethyleneimine reagent was studied to treat the lysed cell broth by precipitation of soluble contaminants such as DNA and flocculation of cell debris material. The USD studies were used to predict the impact of this treatment on the performance and here it was found that the fermentation could be run to maximum productivity using an acceptable clarification process (e.g., a centrifugation stage operating at 0.11 L/m² equivalent gravity settling area per hour followed by a resultant required depth filter area of 0.07 m²/L supernatant). A range of USD predictions was verified at the pilot scale for centrifugation followed by depth filtration. © 2016 The Authors Biotechnology Progress published by Wiley Periodicals, Inc. on behalf of American Institute of Chemical Engineers Biotechnol. Prog., 32:382–392, 2016

Measuring E. coli and bacteriophage DNA in cell sonicates to evaluate the CAL1 reaction as a synthetic biology standard for qPCR

•

We establish the effect of E. coli cellular material on sensitivity of qPCR for detection and quantitation of a lone genomic sequence.

•

We demonstrate that LRE qPCR matches performance of the conventional Standard Curve qPCR method with respect to absolute quantitation of a genomic E. coli sequence.

•

We characterise the effect of E. coli cellular material on performance of qPCR for detection and quantitation of a bacteriophage DNA sequence.

We measured the impact of the presence of total Escherichia coli (E. coli) cellular material on the performance of the Linear Regression of Efficiency (LRE) method of absolute quantitative PCR (LRE qPCR), which features the putatively universal CAL1 calibration reaction, which we propose as a synthetic biology standard. We firstly used a qPCR reaction in which a sequence present in the lone genomic BirA locus is amplified. Amplification efficiency for this reaction, a key metric for many quantitative qPCR methods, was inhibited by cellular material from bioreactor cultivation to a greater extent than material from shake flask cultivation. We then compared LRE qPCR to the Standard Curve method of absolute qPCR (SC qPCR). LRE qPCR method matched the performance of the SC qPCR when used to measure 417–4.17 × 10⁷ copies of the BirA target sequence present in a shake flask-derived cell sonicates sample, and for 97–9.7 × 10⁵ copies in the equivalent bioreactor-derived sample. A plasmid-encoded T7 bacteriophage sequence was next used to compare the methods. In the presence of cell sonicates from samples of up to OD₆₀₀ = 160, LRE qPCR outperformed SC qPCR in the range of 1.54 × 10⁸–1.54 × 10¹⁰ copies of the T7 target sequence and matched SC qPCR over 1.54 × 10⁴–1.54 × 10⁷ copies. These data suggest the CAL1 standard, combined with the LRE qPCR method, represents an attractive choice as a synthetic biology qPCR standard that performs well even when unpurified industrial samples are used as the source of template material.

Enhancing the selective extracellular location of a recombinant E. coli domain antibody by management of fermentation conditions

The preparation of a recombinant protein using Escherichia coli often involves a challenging primary recovery sequence. This is due to the inability to secrete the protein to the extracellular space without a significant degree of cell lysis. This results in the release of nucleic acids, leading to a high viscosity, difficulty to clarify, broth and also to contamination with cell materials such as lipopolysaccharides and host cell proteins. In this paper, we present different fermentation strategies to facilitate the recovery of a V H domain antibody (13.1 kDa) by directing it selectively to the extracellular space and changing the balance between domain antibody to nucleic acid release. The manipulation of the cell growth rate in order to increase the outer cell membrane permeability gave a small ~1.5-fold improvement in released domain antibody to nucleic acid ratio without overall loss of yield. The introduction during fermentation of release agents such as EDTA gave no improvement in the ratio of released domain antibody to nucleic acid and a loss of overall productivity. The use of polyethyleneimine (PEI) during fermentation was with the aim to (a) permeabilise the outer bacterial membrane to release selectively domain antibody and (b) remove selectively by precipitation nucleic acids released during cell lysis. This strategy resulted in up to ~4-fold increase in the ratio of domain antibody to soluble nucleic acid with no reduction in domain antibody overall titre. In addition, a reduction in host cell protein contamination was achieved and there was no increase in endotoxin levels. Similar results were demonstrated with a range of other antibody products prepared in E. coli.

Expression of recombinant antibodies

Recombinant antibodies are highly specific detection probes in research, diagnostics, and have emerged over the last two decades as the fastest growing class of therapeutic proteins. Antibody generation has been dramatically accelerated by in vitro selection systems, particularly phage display. An increasing variety of recombinant production systems have been developed, ranging from Gram-negative and positive bacteria, yeasts and filamentous fungi, insect cell lines, mammalian cells to transgenic plants and animals. Currently, almost all therapeutic antibodies are still produced in mammalian cell lines in order to reduce the risk of immunogenicity due to altered, non-human glycosylation patterns. However, recent developments of glycosylation-engineered yeast, insect cell lines, and transgenic plants are promising to obtain antibodies with "human-like" post-translational modifications. Furthermore, smaller antibody fragments including bispecific antibodies without any glycosylation are successfully produced in bacteria and have advanced to clinical testing. The first therapeutic antibody products from a non-mammalian source can be expected in coming next years. In this review, we focus on current antibody production systems including their usability for different applications.

High-yield export of a native heterologous protein to the periplasm by the tat translocation pathway in Escherichia coli

Numerous high-value recombinant proteins that are produced in bacteria are exported to the periplasm as this approach offers relatively easy downstream processing and purification. Most recombinant proteins are exported by the Sec pathway, which transports them across the plasma membrane in an unfolded state. The twin-arginine translocation (Tat) system operates in parallel with the Sec pathway but transports substrate proteins in a folded state; it therefore has potential to export proteins that are difficult to produce using the Sec pathway. In this study, we have produced a heterologous protein (green fluorescent protein; GFP) in Escherichia coli and have used batch and fed-batch fermentation systems to test the ability of the newly engineered Tat system to export this protein into the periplasm under industrial-type production conditions. GFP cannot be exported by the Sec pathway in an active form. We first tested the ability of five different Tat signal peptides to export GFP, and showed that the TorA signal peptide directed most efficient export. Under batch fermentation conditions, it was found that TorA-GFP was exported efficiently in wild type cells, but a twofold increase in periplasmic GFP was obtained when the TatABC components were co-expressed. In both cases, periplasmic GFP peaked at about the 12 h point during fermentation but decreased thereafter, suggesting that proteolysis was occurring. Typical yields were 60 mg periplasmic GFP per liter culture. The cells over-expressed the tat operon throughout the fermentation process and the Tat system was shown to be highly active over a 48 h induction period. Fed-batch fermentation generated much greater yields: using glycerol feed rates of 0.4, 0.8, and 1.2 mL h(-1), the cultures reached OD(600) values of 180 and periplasmic GFP levels of 0.4, 0.85, and 1.1 g L(-1) culture, respectively. Most or all of the periplasmic GFP was shown to be active. These export values are in line with those obtained in industrial production processes using Sec-dependent export approaches.