Here are some highlights of completed BioProNET-funded projects – proof of concept funding, business interaction vouchers, workshop funding and scientific exchange awards.
Clicking on each title will open a pdf version of that case study. To see all the case studies in full on this website, click on the purple bar below the titles.
BioProNET early career researchers get more business acumen!
Assessing the production of diagnostic proteins in plant cell cultures
An express route to biopharmaceutical production
A machine learning poly-omics classifier to improve protein production in CHO cells
Scientific exchange visit boosts separation technologies collaboration
Speeding up and slowing down: altering translation speed to enhance protein yield
Hijacking intracellular storage bodies to produce difficult to express proteins
Developing a novel fluorescence-based biopharmaceutical quality control technology
Design & testing of a membrane photobioreactor for advanced biologic production
Collaborative development of glycolipid separation technology to reduce costs
BioProNET funding drives the use of motor proteins for nanopore DNA sequencing
BIV funding grows algae bioprocessing collaboration
BIV funding lights up collaboration on fluorescent protein expression in microalgae
PoC study shows protein synthesis errors can cause activity losses in recombinant protein
Warwick and JEOL Strike Gold in Electron Microscopy Collaboration
Dynamic partnership aims to reduce cell harvest time
Cobra and Lancaster partnership helps unravel new analytical tool for DNA topology
Collaboration creates a recipe for success in cell-free protein synthesis
Edinburgh and Recyclatech Join Forces to Recover Microbial By-Products
Sandpit Meeting Builds Collaboration Workshop
Exchange visit funding seeds early career researcher collaborations
Scissor technology cuts out a collaboration between Bath and Arecor
BioProNET boosted the entrepreneurial skills of three early career researcher members when it funded places for them at the BrisSynBio 4 day MBA (More Business Acumen) course. On this practical and immersive course, participants learned about starting a business – from setting the scene for a business idea, setting up and running a venture, through to the execution and preparation for commercial success. Teams of delegates also took part in a Dragons’ Den-style competition, where they pitched their business ideas developed throughout the course to a panel of real biotech investors, as well as entrepreneurs, academics and industry representatives.
“This training course enabled me to gain more confidence on pursuing a non-academic career focused on the commercialisation of biotechnology and bioprocessing, with my particular interest on how to bring cost effective and disruptive therapies to the market on behalf of patients.” Ricardo Suarez-Heredia, PhD student in biochemical engineering, University College London.
“My participation has had an immediate benefit in terms of enhancing my entrepreneurial thinking and will broaden my skills and knowledge in order to maximise my career outputs and further advance with my career ambitions as a potential biotech leader.” Alejandro Fernandez-Martell, PDRA in cell and bioprocess engineering, University of Sheffield
“I found the whole experience very informative and rewarding. During the course, I met and formed relationships with many well-established researchers in a diverse range of synthetic biology fields, opening the possibility for future collaborations.” Ben Barnes, EngD student, University College London.
Two BioProNET members were part of the team that won the ‘best pre-incorporation prize’ at the pitch event. Their business idea focused on an exosome-based delivery platform. This technology seeks to revive failed drug candidates by improving drug delivery across the blood–brain barrier, and was conceived by combining their academic and industrial expertise in cellular engineering, bioprocess engineering and exosome purification. Over the next 12 months the team plans to conduct preliminary studies to establish intellectual property and conduct further market research. This is a direct consequence of their attendance at the course.
Jose Gutierrez-Marcos from the University of Warwick was awarded Business Interaction Voucher funding from BioProNET to work with Mologic. Their project investigated ways to improve the production in plant cells of human chorionic gonadotropin (hCG)— a protein that could be used in cancer diagnostics.
The challenge – Plant cell cultures have several advantages over other protein expression systems; they are cost-effective, lack toxicity and are amenable to genetic modification. However, the lack of a cost-effective manufacturing process has hampered the use of plant cells to manufacture high-value proteins. In particular, cysteine knot proteins, including hCG, are difficult to produce as recombinant proteins in cultures.
The project – Low levels of protein production, purification and quality are major factors that limit protein production in plant cell cultures. We have assessed the impact that culture media has on the production, purification and quality of recombinant hCG, a protein that has potential use in cancer diagnostics.
Key findings – Expression of hCG in Nicotiana tobacco cell lines was highly dependent on media composition (that is, amounts of salts, vitamins and sugars). By varying the concentrations of these media components, we identified optimal culture conditions for the expression of hCG. We also developed a system to simplify hCG purification that reduces protein aggregation and facilitates rapid protein enrichment. Finally, we determined the quality of the hCG product and found that around 85-90% of purified protein was correctly folded.
The collaboration – The project has provided Jose with a new insight into recombinant protein production using plant cell cultures and in their application in point-of-care diagnostic devices. It has also confirmed that plant cells can fold complex, difficult cysteine-knot proteins more efficiently than most other host cells can. As a result of this project, we have initiated a new collaboration with Mologic’s Centre for Advanced Rapid Diagnostics to improve the expression of other recombinant proteins known to be challenging in other expression systems.
Outcomes and next steps
– A new collaboration has been initiated to improve the expression of seven additional proteins
– Data from the project will be used in a publication
“The purified hCG protein will be used to test Mologic’s proprietary point-of-care and other diagnostic devices, as well as potentially being used as a kit reference standard”
“This is the first demonstration of very high-level export of a biotherapeutic by the Tat pathway in an industrial setting”
Colin Robinson and Kirsty Richards from the University of Kent have used Business Interaction Voucher funding from BioProNET to work with UCB to test a biopharmaceutical production system in an industrial setting.
The challenge – ‘TatExpress’ E. coli cells that overexpress the Tat secretion system have an enhanced ability to export recombinant proteins to the periplasm, making this an attractive system for the industrial production of biopharmaceuticals. However, the TatExpress system had not been tested in the fed-batch fermentation systems that are used by industry.
The project – Kirsty worked at UCB to test the abilities of a TatExpress E. coli strain grown in a fed-batch fermentation system to export human growth hormone (hGH) linked to a Tat signal peptide.
Key findings – The work was a complete success; we showed that the TatExpress strains exported at least as much hGH, on a per-cell basis, as they do under the conditions that we had tested previously tested at Kent (i.e., in shake flasks). Exported hGH was by far the most abundant protein in the periplasm. Moreover, the TatExpress cells exported far more protein that the wild type strains, and export levels were equal to those obtained in industrial applications using the alternative Sec-dependent protein export pathway.
The collaboration – Whilst Kirsty was at UCB, she gained hands-on training on the implementation of a scaled manufacturing process of a biotherapeutic. During the collaboration, UCB and University of Kent colleagues shared knowledge, and the project prompted shared participation in the design of further studies.
Outcomes and next steps
After final experiments, we will submit this work for publication
Data has been used in an EU grant application involving both partners
The collaborators are discussing other avenues for follow-on funding
Business Interaction Voucher funding from BioProNET has enabled Claudio Angione and colleagues from Teesside University to work with the Centre for Process Innovation to develop a computational method that could be used to predict and improve therapeutic protein production.
“As a result of this project, we are currently discussing with CPI further ways to incorporate machine learning in their pipelines.”
The challenge – Chinese hamster ovary (CHO) cells are the most commonly used cell line for the production of therapeutic proteins such as monoclonal antibodies. But like most mammalian cells, they are very inefficient at producing recombinant proteins.
Aims – The objective was to combine novel machine learning techniques with poly-omic analysis, to build a computational method that: (i) accurately identifies whether target cells have optimal conditions for producing the target protein; (ii) if not, predicts genetic modifications that will likely increase protein production.
The collaboration – The Centre for Process Innovation (CPI) provided accesses to experimental bioprocessing data and were involved in key discussions.
Key findings – The researchers integrated experimental data at gene level with data generated in silico via a publicly available genome-scale metabolic model of a CHO cell within an integrated data-driven framework. They trained the poly-omics machine learning method using gene expression data from varying conditions and associated reaction rates in metabolic pathways, reconstructed in silico. They then evaluated this approach through a computational validation based on cross-validation, estimating the average prediction error in general settings. Importantly, they showed that metabolic predictions coupled with gene expression data can significantly improve estimations of lactate production based solely on gene expression.
It is anticipated that the predictive ability of pipeline will to vastly improve with additional omic data (from CPI). Moreover, although the validation focused on lactate production, the proposed methodological framework can be implemented around any target metabolite or protein.
Outcomes and next steps
The computational work was presented at IECM-2 2017 and highlighted on Phys.org: https://sciforum.net/conference/iecm-2/paper/4993
The results will be extended with additional data provided by CPI, and validated in an upcoming project for which the researchers are seeking funding.
Scientific exchange funding from BioProNET has enabled Ben Dolman, a PhD student from the University of Manchester to visit the lab of Maria Cuellar Soares at the Technical University (TU) of Delft in the Netherlands. The visit was co-funded by the High Value Chemicals from Plants NIBB.
During his PhD, Ben has developed a gravity-based separation system that recovers insoluble sophorolipid biosurfactants from a fermentation broth. This technology aligned with the work of the bioseparation technology group from TU Delft, who are working towards the industrially relevant production of insoluble compounds, in particular terpenoids, and ways to separate them from the production media. So Ben’s exchange visit applied the gravity separation device to the production systems used at TU Delft.
Scientific results: The gravity-based separation system could separate insoluble molecules that formed an emulsion with the fermentation broth components, as well as showing that E.coli cells continued to grow (in accordance with a previously described model) after they were returned to the bioreactor after the use of the separator.
“This opens up the application of our system to a whole new range of biomolecules, with the potential to be used as an integrated pre-concentrating step to recover a product-rich emulsion from a bioreactor, and returning the cells and media back to the bioreactor for continued production,” says Ben.
Ben was also able to enhance the understanding of how fermentation time, cell density, cell growth rate and emulsion formation time can have an effect on emulsion stability and the ease of separation, which complemented the ongoing work on this theme being undertaken at TU Delft.
New skills: Ben learned a number of techniques for rigorous analysis of cell behaviour during fermentation. He also attended the AkzoNobel Imagine Chemistry competition finals during his visit, and benefited greatly from hearing first-hand what AkzoNobel’s industrial experts considered the most important steps to be for the separation technology to reach commercialisation.
New collaborations: At the Imagine Chemistry Competition, Ben set up interesting collaborations with two of the other finalist teams.
Exchange of knowledge: He notes that the placement was also incredibly productive from a research commercialisation perspective, as he had the opportunity to meet employees from Delft Advanced Biorenewables to discuss the industrial application of the separation technology.
“We were able to share some of the challenges and opportunities that TU Delft and the University of Manchester are both facing bringing our technology to market,” he says.
One way to increase the yield of recombinant proteins is to increase the speed of translation, through the use of DNA sequences containing codons that are frequently used in the host genome – a process known as codon optimisation. But codon optimisation might reduce activity activity, for example by producing poorly folded protein, especially for proteins that are difficult to express.
This study aimed to investigate if rhythmically alternating slow and fast protein synthesis might improve that activity of a lead to a novel format antibody, by allowing it to fold correctly.
Tobias von der Haar, from the University of Kent has developed computer software that models translation and can be used to design DNA sequences that are translated quickly. This was used to design two codon-optimised fast sequences, as well as a rhythmic sequence, in which slowly decoded regions were introduced between defined codon-optimised antibody domains.
The industrial partner, UCB Celltech, synthesised the DNA sequences, cloned them into CHO cells and then analysed expression levels of the antibody.
Expression of DNA constructs that were fully codon optimised reduced expression of the antibody compared a non-codon optimised DNA construct used by UCB. “This is consistent with the current theory that rapid translation, which normally maximises expression levels, can be counterproductive for difficult to express proteins,” explains Tobias.
Expression levels of the antibody encoded by the rhythmic constructs was comparable to that of the UCB construct, supporting the idea that introducing regions of slower translation can help optimise expression levels in multi-domain proteins.
However, the rhythmic construct did not increase expression levels above those obtained by the UCB construct “This indicates that the main limitation in expression levels for this particular antibody is not at the level of translation, but downstream of this – likely molecule assembly in the ER and/ or secretion,” says Tobias.
The data from this work has provided a starting point for a more systematic exploration of rhythm-based codon optimisation, including the optimal location of slowly translated sequences within the antibody folded domains.
In addition, analysis is currently underway to ascertain whether the introduction of slowly decoded regions has impacted the quality of the expressed antibody – based on levels of aggregated material and antibody. “If the quality of the expressed antibody has been improved, it provides a great opportunity for us to impact our processes,” says Paul Stephens of UCB.
“This was a unique opportunity to gain exposure to the kind of problems encountered in industry when trying to produce novel molecules.”
Proof of Concept funding from BioProNET has enabled Marek Brzozowski and Tim Ganderton from the University of York to investigate a novel protein packaging and secretion system, which might be used to make recombinant proteins that are tricky to produce.
Many potentially useful therapeutic or diagnostic proteins are difficult to express in the mammalian cell lines that are, or could be, used for their production. The proteins may be fragile, misfolded, harmful to the host, or require complicated, time consuming purification strategies.
With the aim of creating a novel mammalian cell-based system for the production of difficult to express proteins, the scientists investigated a naturally occurring system of intracellular storage bodies — known as Weibel Palade bodies. In nature these have a crucial role in the storage and controlled secretion of several haemostatic, and very complex, proteins, most notably von Willebrand factor.
The formation of functional pseudo-Weibel Palade bodies which behave like ‘wild type’ bodies can be induced in certain non-endothelial cell lines by expression of von Willebrand factor. “The ability of intracellular storage bodies to store, protect, and regulate the release of functionally active proteins, has yet to be exploited as part of a novel production strategy for difficult-to-express proteins,” says Tim.
The main part of the work investigated which parts of the von Willebrand factor protein are responsible for promoting accumulation and storage in Weibel-Palade body vesicles “Our aim with is to be able to turn the von Willebrand factor fragment into a Trojan horse for the re-direction and packaging of other proteins into WPB so that it can be hijacked for the production of novel difficult to express proteins,” says Tim.
Using green fluorescent protein as a test protein they identified a ‘minimal critical region’ of von Willebrand factor that was needed to generate functional pseudo-WPB that packaged and secreted the green protein.
As well as demonstrating the feasibility of using Weibel-Palade bodies as part of a protein expression system, the project also generated several novel observations about the fundamental biology of Weibel-Palade bodies. Tim was an invited speaker at the 2017 ESACT-UK meeting, where he presented preliminary results from this project.
Further work studying protein expression levels — to help differentiate stimulated and constitutive protein release — are currently underway. A priority for future studies is to test the system with other proteins. “It will be really interesting to use more relevant proteins — for example, insulin, insulin-like growth factor 1 and erythropoietin to demonstrate effectiveness as a protein expression system,” highlights Tim.
A Business interaction voucher from BioProNET has enabled Christopher Pudney from the University of Bath to partner with Bath ASU to develop a new method for quality control testing of biologics that could be faster and cheaper than current approaches.
There are a limited number of approaches to perform quality control analysis of biological molecules. We will work with partner company Bath ASU to develop a novel technology — based on quantification of the fluorescence edge shift phenomenon — that gives a library of spectroscopic fingerprints for different biopharmaceuticals and accurately quantifies subtle changes to protein structure. We will test a range of biopharmaceuticals, allowing us to develop our technology to improve speed and accuracy and also to establish the limits of sensitivity.
We looked at 12 monoclonal antibodies (including humanised and fully human) using our fluorescence edge shift technology. We benchmarked against circular dichroism and dynamic light scattering and found that fluorescence edge shift technology had a similar sensitivity to both traditional approaches, but that we could combine the measurement power of both approaches. So we could measure unfolding (normally assessed by circular dichroism) and aggregation (normally assessed by dynamic light scattering). Our measurements were not destructive, were faster and were at least as accurate as these approaches. So we found our approach could potentially replace these other approaches making quality control faster and less expensive. Most importantly, we have been able to show our approach can actually predict the stability of antibodies and this opens up a whole new range of possibilities for our approach.
Patent filing (UK Patent Application No. 1604640.1).
Manuscript in review with industrial partner.
Presentation at the 2nd International Antibody Validation Meeting
EPSRC funding as part of seed-corn funding for a GCRF project
EPSRC Impact Acceleration Account award
Short film: https://youtube/kHdZBy5xJ2A
Design consultation and testing of a membrane photobioreactor suitable for advanced biologic production from microalgae
“The fledgling BioProNET-funded collaboration resulted in a successful application to Innovate UK.” Varicon Aqua Solutions
At present, the production of high-value biologics in plankton is uneconomical, primarily due to the demands of mixing and harvesting low biomass concentrations from large volumes. This issue is almost entirely negated by the use of bioreactors that use a biofilm-membrane design, in which the biomass is retained within a high-density, low-volume system. To our knowledge however, there is no commercially available membrane system suitable for the growth of algae as a biofilm.
In this project, Plymouth Marine Laboratory (PML) collaborated with Varicon Aqua Solutions (VAS) to develop a prototype membrane photobioreactor — about 5l volume — for the production of high value biologics from microalgae and assessed biologic production capability using a reference algal system.
We designed and tested a novel solid-state photobioreactor. Growth rates, productivity and algal concentration was measured in several green algal species. Inoculation tests were also carried out to assess the best starting conditions within the bioreactor. We generated a wealth of baseline data for two very different microalgae species and assessed the potential for the system to enter into the competitive market for photobioreactors for the generation of high-value products.
The prototype system will continue to be used at PML for academic research and to provide baseline data to aid VAS with sales of the product. The algal media recipes developed and tested during this work are already generating sales for VAS, and it is envisaged that the PML—VAS relationship will continue to grow. The fledgling BioProNET collaboration resulted in a successful application to Innovate UK to take the technology forward.
BIV funding from BioProNET has enabled James Winterburn and Ben Dolman from The University of Manchester to work with Croda to scale up a new separation technology for the production of high-value glycolipids. Their novel method increases fermentation productivity and yield, and so has potential to reduce costs.
The challenge – Sophorolipids are glycolipid biosurfactants that are used in environmentally friendly cleaning products and cosmetic/personal care products. Yet current fermentation methods are inefficient as the fermentation must be stopped when product concentrations reach a certain value. The use of sophorolipids could be widened — for use in bioremediation and enhanced oil recovery — if production costs can be reduced.
Aims – This project built on the Winterburn Group’s previous work showing that separation of sophorolipids from a fermentation broth during production can give higher overall productivities, (patent application 1610932.4). This collaborative project with Croda tested the functionality of the novel separation process at pilot scale.
The collaboration – A high productivity sophorolipid fermentation protocol, using Candida bombicola as the producer strain and gravity-based separation of the sophorolipid from the fermentation broth, was carried out at the University of Manchester. A gravity-based separator was then designed and built by University of Manchester engineers for pilot scale (30 l) trials at Croda.
Key findings – Sophorolipid recovery was demonstrated for extended periods, and a total of 550 g sophorolipid was recovered per litre of broth. Both the sophorolipid productivity and the yield were higher than those reported in the literature to date. The separator could recover the phase containing the sophorolipid at over 2 g per hour, a separation rate that makes the system applicable to continuous separation from bioreactors of around 600 times the size, or 200 l working volume, an important achievement given many integrated separation systems cannot be translated from lab scale.
Outcomes and next steps
Paper entitled ‘Characterisation and scale up of integrated production and separation of sophorolipids’ in preparation
EPSRC responsive mode grant in preparation ‘Advanced glycolipid biosurfactant processing and separation’
Potential further work with Croda testing the system at larger scales (>100 l) and with other glycolipid biosurfactants.
Possible BBSRC DTP CASE studentship to grow the collaboration
“The BIV-funded work has provided the first evidence that these enzymes can function as motors in a hand-held nanopore–based sequencer”
Business interaction voucher funding from BioProNET has enabled Michael Plevin and James Chong from the University of York to partner with biotech company Oxford Nanopore Technologies. Their project aimed to produce and characterise motor proteins that could be used in a portable hand-held DNA sequencer (see box).
Nanopore-based DNA and RNA sequencing is critically dependent on controlling the movement of the polynucleotide through the pore. The collaborators investigated if a previously untested family of archaeal DNA motor proteins have characteristics that are suitable for use in nanopore-based sequencers.
First, several DNA constructs encoding for different regions of the target protein were made, as well as several mutant variants. Recombinant protein was produced in E. coli, with yields exceeding 50mg per litre. Electron microscopy and other studies showed that the purified proteins adopted the expected structure.
The main outcome was the design and optimisation of a ‘pipeline’ for the production and characterisation of the target proteins. “We incorporated a number of features into this pipeline that would permit the screening of larger numbers of proteins” says Michael. “These included the use of an expression vector that was compatible with ligation-independent cloning, protocols for parallel small-scale expression and solubility tests, structural and oligomeric analyses and newly implemented activity assays.”
Indeed, the new activity assays – based on fluorescence rather than previously used radioactivity – showed that the target motor protein catalysed DNA unwinding and strand separation, indicating that the protein is functional. “The BIV funding has enabled work that provided our first evidence that these enzymes can function as motors in a hand-held nanopore-based sequencer,” says Andrew Heron from Oxford Nanopore Technologies.
In addition to their use in portable DNA sequencers, the pipeline will be useful for identifying and screening uncharacterised motor proteins to explore their biological structure–function relationship.
The BIV has laid the foundations for an ongoing partnership. “This promising progress motivates us to continue the collaboration to explore DNA motor proteins that improve our DNA sequencing technology, and to continue to support the University of York team,” says Andrew. Indeed the partners have been awarded a BBSRC-funded iCASE PhD studentship to investigate the use of hand-held sequencers for characterising DNA motor proteins at single molecule resolution.
They also secured BioProNET for proof of concept funding to develop their target motor proteins — for example in terms of activity, structure and potential to be engineered — for use in portable sequencers.
What is nanopore-based sequencing?
Nanopore-based DNA and RNA sequencing uses devices that incorporate two protein components: a doughnut-shaped nanopore protein and a motor protein. In the sequencing technology, the recombinant protein nanopore is set in a polymer membrane and an ionic current is passed through the nanopore. The motor protein ratchets single-stranded polynucleotides through the nanopore one nucleobase at a time. The current changes as the bases G, A, T or C pass through the pore and the changes in current can be decoded into a DNA sequence using an algorithm.
“The funding allowed us access to expertise and facilities that were otherwise unavailable to AlgaeCytes”
BioProNET funding has enabled two teams of scientists with expertise in algal bioprocessing to work together for the first time. Kevin Flynn, Claudio Fuentes-Grünewald and Deya Gonzalez from Swansea University partnered with AlgaeCytes, an SME that is focused on developing and commercializing healthcare ingredients derived from algae.
Each project partner brought a selected strain of algae into the project to compare the bioactivity of certain extracts. Working together using a business interaction voucher funding enabled Swansea and AlgaeCytes to compare functionality of these bioactive properties from their respective algae. “The funding allowed us access to expertise and facilities that were otherwise unavailable to AlgaeCytes as a SME, but growing company,” says John Dodd, co-founder of AlgaeCytes.
The focus of the project was algal exopolysaccharides a group of high molecular mass polymers that are secreted by microalgae and typically helps to protect themselves against stress. Exopolysaccharides are known to have antiviral, antioxidant and immunomodulatory activity, and so may have useful medical applications.
The collaborators first established protocols for growing two types of algae under stress conditions in order to maximize exopolysaccharide production. They then used novel downstream bioprocessing techniques to concentrate the exopolysaccharides from large volumes of algal growth media.
Although further studies will be needed to further optimize the production of compounds of interest, the project successfully allowed the collaborators to test the biological activity of exopolysaccharides. “We have encouraging results from preliminary tests of bioactives of algal origin that show their inhibitory activity upon several human cancer cell lines,” says Deya. Moreover, the compounds had antioxidant activity on certain cell lines.
“The project has enhanced our research outputs through identification of novel microalgal compounds with potential therapeutic applications,” highlights Kevin. The outcomes of this project provide vital proof-of-concept data for grant applications and it is hoped that the data will eventually form part of a patent application as well as supporting publications. “Importantly the project strengthened the relationship between AlgaeCytes and Swansea University for future projects” concludes John.
“We worked on a commercial protein with a clear route to market; we would not have considered this outside of this collaboration.”
A business interaction voucher from BioProNET has enabled Anil Day from the University of Manchester to partner with biotech company Protein Technologies. Their project focused on the the expression of a novel fluorescent protein — which could have applications in optical imaging studies in laboratory animals — in microalgae.
The production of recombinant proteins in microalgae could offer a lower-cost alternative to mammalian or bacterial systems. The first step to showing that protein production in microalgae is commercially viable is demonstrating that the desired protein can be expressed in microalgae.
The collaboration provided Protein Technologies, which focuses on protein engineering and the manufacture of recombinant proteins, with access to state of the art expertise to express recombinant proteins in microalgae and plants. “This is an area of considerable interest to Protein Technologies but we were unable to do it in-house due to the expertise and resources required,” says Farid Khan of Protein Technologies.
After first generating vectors containing the gene encoding for the infrared protein, the vectors were then transformed into the chloroplasts of the microalgae Chlamydomonas reinhardtii. Western blot analysis on total protein from this transgenic strain showed accumulation, albeit limited, of the infrared protein.
“This first set of transgenic strains of microalgae provides a valuable resource for improving yields by media formulation and changing environmental parameters such as temperature, light intensity and day length,” says Anil.
The project also refined methods for detecting the accumulation of the infrared protein in transgenic algae.
Because much higher levels of recombinant protein expression can be obtained in the chloroplasts of plants, genes encoding the infrared protein were then cloned into a vector to allow expression of the protein in tobacco plants. Anil explains that this additional step is ongoing because it takes longer (typically 4-6 months) to isolate stable transgenic plants.
The project has allowed Anil a new insight into his field, “We worked on a commercial protein that had a clear route to market. We would not have considered outside this collaboration,” he says.
As a result of this project a proposal led by Protein Technologies including the University of Manchester on the industrial biotechnology applications of microalgae has been submitted to the Newton Fund.
“To our knowledge, this is the first direct demonstration of DNA sequence-dependent activity differences”
Proof of concept funding from BioProNET has enabled Tobias von der Haar from the University of Kent and his collaborators to develop a new way of determining the accuracy of protein synthesis. In addition, they were able to use this new technique to show that minor inaccuracies in translation – such as amino acid substitutions – can affect the activity of a recombinant protein.
Cells can be reprogrammed to make many types of recombinant proteins, but this creates additional demand on the cellular protein synthesis machinery that could lead to a decrease in the accuracy of translation and mean that resultant proteins contain more errors compared to endogenous proteins in normal cells. This in turn could lead to changes in the efficacy, bioavailability and immunogenicity of therapeutic and diagnostic proteins.
Working with Cobra Biologics and MRC Technology, Tobias and colleagues sought to establish what effects a loss of translation optimization and decreased protein synthesis accuracy had on the resultant protein. First they developed a new computational tool to generate a database of all possible single-amino acid substitutions in a recombinant protein, as well as LC-MS protocols for analysing mis-incorporated amino acids in a peptide sequence. These tools were then used to analyse recombinant proteins produced in yeast and E. coli – two popular bioprocessing hosts.
The tools could detect minor variations in the amino acid sequence. “The sequence variations would have escaped detection with standard mass spectrometry approaches, but can be reliably visualised using our novel approach,” says Lyne Jossé, who carried out the experimental work.
Many of the observed substitutions were shown to be the result of specific biological mechanisms, such as non-optimal codon usage, that generate specific, predictable translational errors. Interestingly, many other observed errors were universal, occurring in all peptide sequences that were tested from both yeast and E. coli. The source of these latter errors is currently not well understood.
A key aspect of this study was the demonstration that errors in protein synthesis can affect the properties of the resultant protein. Surprisingly, a protein translated from a non-codon-optimised DNA sequence had only about 60% of the specific enzymatic activity of the same protein produced from a codon-optimised DNA sequence in E. coli (but this was not true for yeast). “To our knowledge, this is the first direct demonstration of DNA sequence-dependent activity differences,” highlights Tobias.
“The collaboration has significantly increased our understanding of the potential issues relating to the production of heterologous proteins in E.coli,” says Steve Williams from Cobra Biologics. The study also seeded opportunities for further work – Tobias intents to apply for further funding to investigate the biological mechanisms that cause the observed amino acid substitutions.
“A better understanding of protein export by the TAT system will facilitate better bioprocessing technologies”
Escherichia coli is a popular system for the production of recombinant proteins, but little is known about the distribution and shape of structural elements of E. coli that drive protein expression and export to the periplasm. To investigate this, Corinne Smith from University of Warwick and colleagues used business interaction voucher funding from BioProNET to collaborate with electron microscope specialist JEOL UK.
The collaboration drew on JEOL’s expertise in zero-loss cryo-electron tomography and direct electron detection to investigate the export of human growth hormone by the twin-arginine translocation (TAT) system in E. coli. This system is responsible for the export of fully folded proteins — endogenous and recombinant — from the cytoplasm, across the inner membrane and into the periplasm.
“A better understanding of protein export by the TAT system will facilitate better bioprocessing technologies,” says Corinne.
After first using biochemical studies to show that human growth hormone was exported to the periplasm by the TAT machinery, the collaborators then optimised an immunogold labelling procedure to unambiguously identify human growth hormone in E. coli.
Electron microscopy data of immunogold-labelled growth hormone showed that a proportion of the protein forms inclusion bodies in the cytoplasm, meaning that it cannot be exported and so would affect the yield of protein. The growth hormone that was available for export at the cytoplasmic membrane was randomly distributed throughout membrane, and did not appear to effect the membrane structure.
Sarah Smith, who undertook the experimental work, gained valuable new skills. “This project gave me training in difficult electron microscopy techniques such as imaging of resin-embedded E. coli and electron tomography sections, as well experience of automated image acquisition software, which together which enabled us to gain high resolution data.”
Sarah also showed that a mutant form of growth hormone that cannot be processed for export was randomly distributed in the inner membrane without affecting membrane structure. “In principle this represents a novel way of displaying a protein on the periplasmic face of the E. coli inner membrane, which could have applicability in library screening, protein engineering or whole cell biocatalysts”, she notes.
Moving forward, Corinne and Sarah are collaborating with colleagues at University College London to quantify how much human growth hormone can be made by the system, and hope to combine data with results from this study to publish as a paper on a new method of producing proteins in E. coli.
“This successful project established a working relationship between JEOL and scientists from the University of Warwick, which will be a catalyst for future electron microscopy-based research projects,” concludes Andrew Yarwood from JEOL.
“We would like to collaborate further to develop more sophisticated software for commercial application”
Cell therapy products and recombinant therapeutic proteins that are produced in cellular systems need to be harvested at the end of the production process. Cell harvesting is often achieved using membrane-based systems, which separate intracellular product and cells from unwanted material in the culture medium or their secreted products from cells.
Business interaction voucher funding from BioProNET has enabled Yuhong Zhou from University College London to work with John Philip Gilchrist of BioPro Control Tech on a project that aimed to reduce the time taken to harvest cells. Reducing cell harvest time could result in a better quality of product and reduced costs. Their project initiated work on a computer-based system that could be used to optimally control the flow of cells and culture medium across a membrane-based separation unit.
“We would not have been able to carry out such a project without the collaborating company,” says Yuhong. “The company developed software and hardware to implement the control method, and we did all the wet laboratory experiments at University College London,” she explains.
Their work centred on a cross-flow filtration membrane system (which has two exit streams ) in an ultra-scale down device – so that low volumes (tens of ml) of culture media could be used in the laboratory setting. They aimed to reduce cell harvest time by using the computer-based control system to balance the flux of the culture medium across the membrane against the fouling of the membrane with unwanted material (which could reduce the efficiency of the membrane).
As a simple preliminary test system, the collaborators used a suspension of Baker’s yeast to generate data on the viscosity of the culture medium at several different cell concentrations, which was then used to develop a mathematical model to control flux. An open-source electronics platform was used as the control system hardware and software was written in house to drive the pressure sensor for online monitoring.
“Our results have provided evidence that the control method has the potential to achieve significant process efficiency”, says Yuhong, noting that further studies will be needed to investigate results in industrially relevant feed systems, such as lysates from E. coli or mammalian cell culture broth. Their work also indicates that cost-savings are possible if the control system is integrated into the membrane separation processes.
There are plans to continue the work to further develop the control system and study the application in large scale cross-flow membrane filtration processes. “This work has provided us considerable preliminary data for a new bid for further development of the dynamic control system,” says John Philip. “We would like to collaborate further to develop more sophisticated software for commercial application,” he concludes.
“Raman spectroscopy is sensitive to changes in DNA and RNA structure but is underused in biopharmaceutical analytical R&D”
An increased demand for plasmid DNA in the biopharmaceutical sector — for example, for use in gene therapies — necessitates the use of techniques to analyse the tertiary structure of the DNA, yet current methods are invasive and require a high level of sample preparation. A business interaction voucher from BioProNET has enabled Lorna Ashton from Lancaster University to work with Cobra Biologics to assess a novel method for determining the topology of plasmid DNA.
The project used Ramen spectroscopy; a method for monitoring physiochemical properties of molecules, in which the scattering of light caused by molecular vibrations gives a unique fingerprint of that molecule. It has the the advantages of being non-invasive and providing almost real-time information on molecules.
“Raman spectroscopy is sensitive to changes in DNA and RNA structure but is underused in biopharmaceutical analytical R&D”, explains Lorna.
The business interaction voucher enabled Cobra to explore an alternative to current analytical methods by working with Lorna, who has extensive experience of Ramen spectroscopy, while at the same time allowing Lorna to access otherwise unavailable plasmid DNA samples.
Cobra provided DNA samples in three topological isoforms — supercoiled, nicked (open circle) and linearized forms — that were verified using two current analytical methods (agarose gel electrophoresis and free-solution capillary electrophoresis) at Cobra.
Then, after method optimization, Lorna determined Raman spectra for each of the isoforms of the plasmid DNA. Next, data processing and statistical analysis were performed to assess any clustering of samples with different topologies.
“The acquired Raman spectra revealed different spectral features arising from the supercoiled, open circle and linearised topologies”, says Lorna. “This indicates that Raman spectroscopy can be used to distinguish the different isoforms.”
However, within the duration of the project it was not possible to assess if Raman spectroscopy could provide quantitative data on the relative amounts of each of the topologies in a sample. Although further work is required to move the project forward, Daniel Smith from Cobra notes that the project has provided “encouraging preliminary data, which that will support continuation of the project in a collaborative manner”.
“The most important outcome of the work was that we were able to generate a working cell-free protein synthesis extract from P. pastoris.”
Proof of concept funding from BioProNET has allowed Karen Polizzi and Rochelle Aw from Imperial College London to work with Fufifilm Diosynth Biotechnologies on a project that tested if cellular extracts from the yeast Pichia pastoris could be used to synthesise proteins.
Protein-based drugs are often synthesised in whole cells. However, the use of cell-free protein synthesis systems — that is, the cell’s internal machinery in the absence of the cell wall — has several potential advantages. Compared to whole cell synthesis, this method allows for quicker synthesis, enables the production of proteins that are toxic to living cells and can can be scaled to large volumes more easily.
Currently, cell-free protein synthesis extracts from yeast are not commercially available. “This project has proved the concept that P. pastoris can be used for cell-free protein expression”, says Ian Hodgson from Fujifilm. “To our knowledge is the first time this has been done.”
As a test system, the scientists investigated the synthesis of green fluorescent protein (GFP) and luciferase. The initial phases of the project determined the best way to lyse yeast cells to release the optimum amount of cellular machinery, and developed a recipe to stabilise RNA transcripts and increase the yield of RNA encoding for the reporter proteins.
The main phase of the project showed evidence of combined transcription and translation in the extract from the yeast cells. “The most important outcome of the work was that we were able to generate a working cell-free protein synthesis extract from P. pastoris”, says Karen.
The final titres of GFP and luciferase observed were similar to that observed with a cell extract from another strain of yeast, Saccharomyces cerevisiae, using the same protocol. However, the protein synthesis reaction had a much longer lag phase, and despite initial evidence that the cell-free system was functional, yields of protein were low.
“The project has given us a strong basis to further build upon the results,” highlights Karen. “Optimisation will be key to maximising the productivity of the system.”
In addition, the project has benefited the industrial partner. “The project has also allowed Fujifilm to understand some of the factors that would be important in utilising cell-free extracts for commercial use.”
As a next step, Imperial and Fujifilm hope to continue their collaboration by focusing on the production of a more complex, industrially relevant proteins with the P. pastoris system.
“We have been exposed to challenges that industry faces; we intend to channel such a perspective into our future work to increase its impact.”
A business interaction voucher from BioProNET has enabled scientists from the University of Edinburgh to partner with the SME Recyclatech to investigate a new way of recovering useful products from spent media.
Recyclatech uses industrial biotechnology processes that generate large volumes of spent medium, which contains mycolic acid-producing bacteria that contain high-value glycolipid. The challenge was to develop a simple, cost-effective way to recover the surfactant-containing bacteria from the large volumes biosurfactants of spent medium.
Together the researchers discovered that the bacteria used by Recyclatech have the capacity to stabilise oil-in-water emulsions. The bacteria can become associated with the oil droplets in the emulsion, and so skimming off the oil droplets from the medium allows the bacteria to be captured and recovered.
“This represents an extremely facile and cost-effective procedure to collect bacteria from a batch reaction,” says Joe Tavacoli, an investigator on the project from the University of Edinburgh.The biosurfactant can then be extracted from the bacteria using solvents.
Moreover, the collaborators showed that the capacity of the bacteria to stabilise emulsions and the type of emulsions they could stabilise — oil-in-water or water-in-oil — was probably dependent by the amount of surfactant they hold within their cell walls, which in turn could be controlled by the amount and type of oil that they were fed.
“Working together with the university of Edinburgh has allowed us to demonstrate biosurfactant production and recovery from our novel bacteria, and has indicated further work to generate different surfactants,” says Nick Christofi, Chief Scientific Officer of Recyclatech.
The extracted biosufacants can be used in pharmaceuticals, homecare and other products, while intact bacteria have the potential to clean oil from contaminated soils or water.
The outcomes of this work are promising, with initial data being used to support further grant applications and the possibility of scale-up studies. In addition, the collaboration has forged strong links between the partners. “We have been exposed to challenges that industry faces,” highlights Tavacoli. “We intend to channel such a perspective into our future work to increase its impact,” he says.
In June 2014, BioProNET held its inaugural event, a so-called ‘sandpit’ meeting — an event where scientists from different backgrounds come together to discuss challenges and opportunities — that was attended by about 80 delegates, of which about one-third were from industry.
“We felt it that such a meeting was an important way to bring the bioprocessing community together to eek out challenges and key issues,” says Mark Smales, BioProNET director. “We included lots of time for discussions between attendees,” he highlights.
Key to the success of these discussions was the involvement of two professional facilitators, who were able to maximise interactions and dialogue between attendees, and allow discussions to explore new topics.
The discussions identified several themes that attendees thought could be the focus of follow-on workshops that would build collaborations between industrial and academic scientists. These where: Computational bioprocessing; Continuous processing; Biologic production in microalgae and plants; Analytics and formulation; Synthetic biology tools for bioprocessing; Protein authenticity and translation; Cell-free expression systems; Whole genome tools; Cells as tools; Antibody-drug conjugates.
Indeed, eight workshops were funded BioProNET; the outcomes of four were presented at BioProNET 2nd Annual Scientific meeting in 2015.
Production of pharmaceutical and industrial proteins in microalgae and plants
This workshop was organised by Anil Day (University of Manchester), Jags Pandhal (University of Sheffield) and Yuhong Zhou (University College London) and had attendees from seven universities and six companies, and was jointly funded by Phyconet. The workshop centred on three themes — expression systems; bioreactors, regulation and industry perspective; harvesting and downstream processing — and featured presentations and breakout sessions. Outcomes included a technology assessment, identification of current barriers to progress, the identification of key academic and industry players from both networks, and the establishment of consortia to take projects forward.
Analytics in bioprocessing and formulation
Organised by Paul Dalby (University College London), Gary Montague (Teeside University) and John Liddell (Fujifilm Diosynth Biotechnologies), this workshop had 17 attendees, over half of which were from industry. The group first identified ten key challenges and then grouped these into three themes, which comprised of non-invasive measurements, automated sample preparation and analysis, and data management and predictability. As well as a professionally written report of the meeting (available in the members’ area of this website), other outputs were grant applications to Innovate UK and the EPSRC formulation call.
Cell-free protein synthesis
This workshop, organised by Karen Polizzi (Imperial College London) and Jose Gutierrez-Marcos (University of Warwick) featured a keynote presentation (available here) by Trevor Hallam, chief scientific officer of Sutro BioPharma in the USA. This was followed by discussions on the challenges for large scale manufacturing with cell-free extracts and the use of different cell types.
“We have a new industrial partner that has been very active in our grant application to BioProNET,” says Karen Polizzi “This was largely due to the workshop,” she notes. Discussions — such as UK research capabilities and what is best use of technology — on cell free synthesis are continuing and indeed a follow up workshop is being planned for March 2016.
Recombinant protein authenticity
This workshop was organised by Ian Stansfield (University of Aberdeen), Mick Tuite (University of Kent) and Tobias von der Haar (University of Kent). The plenary lecture, entitled ‘improving heterologous protein production through synthetic biology algorithms’ was given by Manuel Santos, University of Aveiro, Portugal. This was followed by talks and discussions focusing on how the detection and mitigation of mistranslation will provide new routes to optimize recombinant protein expression.
“We established that a collaborative research project between academia and industry in the UK needs to be set up to explore the means of detecting errors in recombinant proteins and designing new error-free expression strategies,” says Mick Tuite.
“This work will allow the set up of a fruitful collaboration between research groups, sharing valuable experiences within the supercritical fluid world.”
Luis Martin, a post-doctoral researcher at the BioComposites Centre, Bangor University, has been awarded scientific exchange funding from BioProNET that has enabled him not only to acquire new skills but also to build collaborations.
Luis works on greener ways to obtain purer fractions of glycolipids from fermentation broths. The purification of glycolipids is the main factor that limits the industrial application of new glycolipids. The driving force of his visit was to investigate the possibility of moving from a batch purification process to a continuous one using specialist equipment that was available at the supercritical fluids research group, directed by Professor Ernesto Reverchon at Università degli Studi di Salerno in Italy.
“As a result of the scientific exchange, I was able to understand and master the technique of supercritical counter current fractionation,” say Luis. “Maybe in the future this technique can be imported to our group at Bangor University to complete the versatility of our laboratories.
Moreover, the exchange strengthened the networking between the two institutions. Luis explains that two Erasmus stays next year have been set up, with two Masters students coming to Bangor University, accounting for a total time of one year. “This work will allow the set up of a fruitful collaboration between research groups, sharing valuable experiences within the supercritical fluid world,” he highlights.
But this not all. Once Luis knew that he had secured funding, he attended the inaugural BioProNET early career researcher meeting, where he continued his collaboration drive. Pravin Badhe, a research assistant at Brunel University met Luis at this event. “The meeting was very helpful for networking; I managed to source access to LC-NMR equipment at Bangor University, which I had been trying to find for nearly 6 months,” he says. Also as a result of the meeting, Kamaljit Moirangthem, a PhD student at the University of Nottingham, was able to set up a collaboration with Luis. “The project is very innovative and has potential to attract future funding,” highlights Moirangthem.
So as a direct result of BioProNET funding and events, the seeds of collaboration for early career researchers are beginning to grow.
Click here or on the title to download the pdf version.
“Further understanding of these effects could lead to the design of insulins that have more rapid effects, which is one of the Holy Grails of the diabetes management.”
Insulin is the mainstay of diabetes therapy, with both long-acting and fast-acting formulations on the market. However, a better understanding of what happens to insulin once it has been injected into the body — into the subcutaneous space underneath the skin — will aid the design of new insulin therapies that could lower the incidence of life-threatening hypoglycaemic episodes.
A business interaction voucher from BioProNET enabled Randall Mrsny from the University of Bath to partner with Jan Jezek from Arecor toinvestigate this. The collaboration brought together expertise in two areas: a new in vitro technique — known as Scissor; Subcutaneous Injection Site Simulator — developed by the University of Bath that models events that occur following insulin injection, and Arecor’s proprietary technologies for stabilising therapeutic proteins.
Because this method of stabilising proteins can alter the pharmacokinetic profile, work carried out under the business interaction voucher used the Scissor system to test the pharmacokinetic profile of Arecor’s formulations of insulin analogues.
Results generated using the Scissor system showed clear differences in the behaviour of different insulin analogues. For example, differences in the precipitation behaviour of long-acting insulin formulations and fast-acting insulin formulations were observed, with the main differences being in the rate and intensity of the precipitation. These results shed light on the effect of formulation components on the fate of insulin in the subcutaneous space, and consequent differences in their bioavailability. “Further understanding of these effects could lead to the design of fast acting formulations of insulins that have more rapid effects, which is one of the Holy Grails of the diabetes management,” says Mrsny.
To disseminate these findings to the wider bioprocessing community, a poster was presented at the BioProNET annual scientific meeting, held in Manchester in October 2015 with almost 180 attendees.
Although the collaborators were unable to optimise the performance of the instrument to follow the release characteristics of long-acting insulin, further studies using an optimised experimental design are being investigated. “The project gave us confidence in the Scissor instrument,” says Jezek. “We are already discussing continuation of the collaboration with Professor Mrsny.”
Click here or on the title to access a pdf version of this story.