Inclusion bodies (IBs) are protein microparticles produced by bacterial overexpression of recombinant proteins. Because of the protein functionality and secretion properties, IBs have been used in biotechnology and biomedicine. We have developed artificial IBs that preserve the features of natural versions but also have a controlled composition, ensuring biocompatibility. Our focus lies in pointing out the performance and functional improvement of the protein released from both IBs and artificial IBs.

Structural genomics has encouraged the development of systematic approaches to achieve the expression and purification of proteins that have not been previously studied. Such approaches are increasingly pertinent, as unbiased “omic” studies in many disease areas discover new possible targets for interventions, many of which have not been investigated at the protein and chemical level. Target2035 is an ambitious plan, aiming to generate chemical tools to all potential drug targets in the human genome. I will discuss how this could be done - and where innovative technologies as well as an international coordinated effort will be needed.

We have designed and engineered a set of purification-enabling mutations into specific regions of multi-specific antibody chains that enables a highly effective, rapid and high-throughput, all affinity-based purification scheme for many different formats. This innovation can help accelerate the early identification of lead candidate molecules in research by allowing simple and fast isolation of highly pure material from mixtures of product-related impurities.

Recombinant protein production is crucial in contemporary drug discovery, contributing to target identification, screening, selectivity, and structural biology studies. Swift and top-quality protein production is crucial for successful drug development, where efficient recombinant protein production becomes necessary. Our team is continuously improving technology, utilising a medium-scale recombinant protein purification platform to enhance sophistication, throughput, optimisation, and finally, reducing time and costs at the early stages of drug discovery.

• Benefits of testing multiple constructs in parallel. How can we produce the full length protein?
• How many and which expression systems should a lab set up to produce a variety of proteins (intracellular, secreted, membrane)?
• HTP expression screening in multiple hosts: What scale, tags, conditions, equipment, readout?
• Challenges of working in HTP: What conditions to test first to increase success?

• Lab digitalisation - how far are we and where are the hurdles?
• How far can we go with automation of experimental approaches in R&D?
• How far can lab automation and model based approaches decrease the number of experiments and thus save time and costs?
• Where do you see the general opportunities and current limits of AI in bioprocess development?
• What do you think of the value of small scale high throughput experimental approaches for larger scale processes?
  

Gene therapy design and manufacturing platforms incorporate many technologies originally created for recombinant protein production. This talk will discuss how this technology flow is now moving in the opposite direction, driven by the innovation required to bring these complex product formats to market. Using recent examples from my academic and industry lab, I will present how we are harnessing our ATMP technological innovations to improve biomanufacturing of DTE protein products.

mRNA vaccines are in the spotlight, creating an opportunity to reinforce the expertise in mRNA manufacturing technologies. Built upon the tunable character of ionic liquids and able to achieve enhanced extractions and keep the stability of nucleic acids, these compounds are being investigated by a CICECO team (Augusto Pedro, Francisca Silva, Mara Freire, Maria Sousa, and Luis Silva) to integrate the production and clarification steps in the mRNA manufacturing process.

Utilising high-throughput systems and an ability to gain a holistic view of both product and process challenges is critical in the process development of modalities such as recombinant proteins. Approaching process development with molecular design, upstream production, downstream purification, and analytical characterisation in a single focused effort can allow for improvements in process understanding and expedite timelines in a product's journey from bench to clinic.

Quality by Design is cost-effective only when knowledge is transferred from one drug candidate to the next, from one scale to the other. In this contribution we show how transfer learning methods that originate from the area of machine-learning can be used to transfer knowledge between products and scale, allowing reduction of experimental effort and acceleration of process development.