Next-Generation Bioprocess Optimization an Arrayxpress White Paper August 2014
Total Page:16
File Type:pdf, Size:1020Kb
Next-Generation Bioprocess Optimization An ArrayXpress White Paper August 2014 Pharmaceutical companies are mak- ing significant investments in biolog- ics and dedicating up to 40% of their R&D efforts into their biopharma- ceutical pipelines in lieu of classical, small-molecule drugs (Rader, 2013; Aggarwal, 2014; Evaluate Pharma, Figure 2. Mammalian cell lines used in Figure 1. Biologics in development organized by 2013). Experts forecast continuous biomanufacturing by number of biologics (until product category. Data obtained from PhRMA, 2012). Modified from Kantardjieff and Zhou, 2014. strong market growth with increas- 2013. ing revenue reliance and contribution approach. The goal now is to build post-translational modifications that to gross margin. Within the top 100 in quality starting at the design stage are critical for product efficacy and pharmaceutical products, biologics (Glassey, 2011). This approach relies safety. Currently, 51% of all approved are expected to account for more than on integrative systems and data-driv- biologics are manufactured in mam- 50% of prescription sales by 2018 en methods that contribute to the malian cells, including 83% of all re- (Evaluate Pharma 2013; Figure 1). understanding of biomanufacturing combinant blood factors, 95% of all With the increased sales in biologics, processes, and where critical process monoclonal antibodies, and 74% of dramatic improvements have been parameters are identified, monitored, all other recombinant products (Kan- required throughout the manufactur- and controlled. Ultimately, the goal is tardjieff and Zhou, 2014). Chinese ing process. Over the past several de- to develop processes that are predict- hamster ovary (CHO) cells have be- cades, titers have jumped more than a able, consistent and ensure high prod- come the cell line of choice and the 100-fold, from sub-single digit yields uct quality and titers (FDA, 2004). industry’s workhorse, primarily due (in g/liter) to today’s double-dig- to their proven safety record and ad- it production levels. Early gains in The need to innovate aptation to high-density suspension production capacity were achieved Biomanufacturing performance is de- growth. CHO cells are currently re- simply by using larger bioreactors. termined by the interaction of the BIO sponsible for producing over 60% of Smaller incremental gains resulted and MANUFACTURING compo- all mammalian cell-based biologics from process optimizations in which nents. While significant progress has (Kantardjieff and Zhou 2014; Figure higher cell density, viability, increased been accomplished for the latter, in- 2). Other mammalian cell lines are product expression levels, and higher cluding physical production systems used to a less extent, including baby specific productivities were gradually (bioreactors), media formulations, hamster kidney, mouse myeloma cell, achieved. and process optimization strategies, and human cell lines. much less effort has been dedicated At the same time, initiatives imple- towards the BIO component. The inherent complexity of biological mented by the regulatory agencies systems is the primary contributor such as Quality by Design (QbD) and Because of their molecular complexi- to biomanufacturing process vari- Process Analytical Technology (PAT) ty and unique quality attributes, most ability and inconsistency. Process in- led to improved manufacturing pro- biologics require complex production consistencies are commonplace and cesses and product quality. The pri- systems. Mammalian cell lines are cannot support the expected growth mary objective of these initiatives was ideally suited for this purpose due to in market demand nor the econom- to direct the industry away from the their ability to generate complex hu- ic and regulatory challenges faced by empirical in-process development man-like glycan profiles and other the industry. The biopharmaceuti- cal industry can greatly benefit from line are not likely transferrable to an- gy aims to model and understand the technological innovations that drive other and could not be fully imple- structure and dynamics of the cells’ rapid and adaptive change, provide mented across the entire production functional networks. The increasingly competitive advantage and allow portfolio. Developing an in-depth rapid release of draft genomes for var- it to focus on efficiency, flexibili- understanding of the biology of these ious mammalian systems, including ty, convenience, and quality (Rader, production cell lines is critical for sus- the recent CHO genomes, is expect- 2013; Langer, 2012; Gottschalk, 2012; tained biomanufacturing. ed to further drive the incorporation Gottschalk, 2013; Davidson and Far- of systems biology in bioprocessing id, 2014; Carinhas, 2012). ‘Omics and Systems Biology (Xu et al., 2011; Brinkrolf et al., 2013; ‘Omics-based technologies rely on Lewis et al., 2013). Understanding the system the generation and interpretation of Current methods for cell line devel- high-throughput data from an organ- Integrated Cellular ‘Omics opment and process optimization are ism’s DNA, RNA, proteins and me- Platform - iCOP™ very time consuming, expensive, and tabolites (Figure 3). These technolo- From the need to understand the labor intensive and only lead to in- gies are commonly used to discover production system and technolog- cremental improvements. Commonly novel targets for therapeutics, iden- ical advances in Systems Biology, used approaches in cell line devel- tify biomarkers, pharmacogenomics, ArrayXpress developed iCOP – an opment include gene amplification integrated Cellular ‘Omics Platform strategies and selection of stable, high for Bioprocess Optimization (Fig- expressing clones. Traditionally, pro- ure 4). iCOP was designed using the cess optimization for media and cell principles of Systems Biology to en- culture conditions is achieved either able the systematic and directed en- through Design of Experiment, expe- gineering of production cell lines. rience, or trial-and-error approaches Consistent with the QbD philosophy, (Zhang, 2011). Most importantly, this iCOP is fundamentally a data-driven practice must be repeated for every approach to bioprocessing. The ulti- new production cell line and associat- mate goal is to increase product titer, ed protein product. At best, it results quality, and achieve predictable and in a highly variable and unpredictable stable biomanufacturing, resulting in process, both in terms of productivity, Figure 3. Systems Biology employs a suite of Next-Generation Bioprocessing. ‘Omics technologies to decipher the functional as well as product quality. These heu- dynamics and interactions across the various ristic approaches lack the mechanistic celluar organization levels: DNA, RNA, proteins, iCOP is comprised of two interde- and metabolites. understanding of how and why pro- pendent components: an Integrative cess conditions, or any implemented personalized medicine, and for dis- Systems Biology component and an changes, bring about the desired out- ease diagnosis and classification. Al- Engineering component that drive come. though each has its own application, the Next Generation Bioprocess. The individual ‘Omics-technologies do first component relies on system-wide It is not possible to fully understand not provide a holistic view or capture molecular characterization of the the processes without considering the the complex interactions occurring cell’s functional dynamics that is then cell lines used and their relationship to within the cells. With advances in used to direct cell line development the products they synthesize. An un- processing and computational capaci- and process optimization efforts. derstanding of the BIO component, ty, the concept of Systems Biology has that is, the intracellular processes rel- evolved by combining the individual Integrative Systems Biology evant to biomanufacturing including ‘Omics technologies. As a result, we protein translation, post-translational can now view cellular systems as a Component The task of generating systems-lev- modifications, folding, aggregation, complex network with intricate inter- el data was dramatically simplified trafficking, and secretion is key to actions across their distinct organi- and became cost effective by im- overcoming the inconsistency and zational components that define the provements in the latest generations variability. Without this knowledge, cells (Oltvai and Barabasi 2002; Kita- of ‘Omics technologies. The major any optimizations that lead to pro- no, 2002). More than merely visualiz- obstacle now is the analysis and sub- duction gains observed for one cell ing these interactions, Systems Biolo- sequent biological “data mining” of these very large and highly dimen- targets are manipulated through en- and carbon utilization are critical- sional datasets. Perhaps even more gineering strategies that fall into four ly important for productivity and critical when dealing with Systems main areas: genetic, cellular, metabol- quality. Adverse metabolic reactions Biology is the fact that these various ic, and process engineering. related to any of these processes can ‘Omics datasets are generated from be characterized and identified in the inherently different technologies and In genetic engineering transcrip- Systems Biology component and tar- possess intrinsic characteristics that tomic data may be used to identify geted for modification in order to op- must be properly accounted for us- transcriptional hotspots to support timize central metabolism