Available online at www.sciencedirect.com
ScienceDirect Editorial overview: Pharmaceutical biotechnology: Engineering cells for high quality biopharmaceuticals production Beth H Junker and Jamey D Young Current Opinion in Biotechnology 2014, 30:viii–x For a complete overview see the Issue Available online 29th October 2014 http://dx.doi.org/10.1016/j.copbio.2014.10.002 0958-1669/# 2014 Elsevier Ltd. All rights reserved.
Beth H Junker
Bioprocess Development, Merck Research Laboratories, Kenilworth, NJ, USA e-mail: [email protected] Beth H Junker is currently a director in Bioprocess Development at Merck Research Laboratories in Kenilworth, NJ, USA. She obtained her Ph.D. from the Massachusetts Institute of Technology in biochemical engineering. She has been working at Merck for the past 25 years in varied technical roles to develop and commercialize diverse products such as VAQTATM, CancidasTM, and most recently, KeytrudaTM. Her interests include multi-scale production systems, quality-by-design, operational excellence, and sustainable bioprocessing.
Jamey D Young
Chemical and Biomolecular Engineering and Molecular Physiology and Biophysics, Vanderbilt University, Nashville, TN, USA e-mail: [email protected] Jamey D Young is an Assistant Professor of Chemical and Biomolecular Engineering and Current Opinion in Biotechnology 2014, 30:viii–x
Introduction Biopharmaceuticals, encompassing vaccines and protein therapeutics, comprise the fastest growing class of drug compounds on the market. This growth has spurred innovation in manufacturing to improve product titers and reduce the cost of goods sold. There is now a proliferation of success stories in the biotechnology industry where recombinant protein titers have been improved up to and beyond the 5 g/L level. Although further improvements in product titers are still achievable, a broader goal for host cell and protein engineering that encompasses product quality and consistency, in addition to high yield, has now firmly taken root. Furthermore, with a wave of billion-dollar biologics coming off patent in the next several years, the impetus to produce biosimilars with quality metrics ‘equivalent’ to established biopharmaceutical compounds is rapidly mounting. It is likely that future cell culture engineering efforts will place greater emphasis on controlling quality over the shear amount of product that can be manufactured by a recombinant host. In this issue dedicated to Pharmaceutical Biotechnology, we highlight several new technologies that improve our ability to predict, control, and monitor the quality of biotherapeutics derived from cell culture.
Systems and synthetic biology tools Before product quality can be predicted or controlled, the genetic factors that determine quality have to be understood. A major recent breakthrough in this regard was the publication of genome sequences for Chinese hamster and several Chinese hamster ovary (CHO) host cell lines. CHO cells account for over 70% of recombinant protein therapeutics currently on the market [1], and CHO sequence information provides a Rosetta Stone for translating knowledge and methodologies developed in other organisms to this important production host. Baik and Lee summarize ongoing developments in CHO genome sequencing as well as efforts to consolidate and distribute genome information through the CHOgenome.org website (http://www.chogenome.org/). They point to the importance of establishing a communitywide CHO reference genome with functional annotation to enable systematic studies of genotype-phenotype relationships. Vishwanathan et al. further highlight the ways that genome sequence information is also bolstering the application of transcriptomic profiling to understand the basis of hyper-productivity in CHO cells and to elucidate genes that influence the glycosylation pattern of product proteins. Similarly, Heffner et al. point to the important role of publicly available sequence databases, as well as advances in mass spectrometry approaches and bioinformatics tools, to www.sciencedirect.com
Editorial overview Junker and Young ix
Assistant Professor of Molecular Physiology and Biophysics at Vanderbilt University. The central theme of his research is metabolic engineering and systems biology. He received his Ph.D. from Purdue University in 2005 and continued his training as a postdoctoral fellow at MIT until joining Vanderbilt in 2008. He was awarded the NSF CAREER Award in 2010 and the DOE Early Career Award in 2012. He has authored over 30 articles and book chapters describing the application of mathematical modeling and 13 C flux analysis to a variety of research topics, including E. coli metabolic engineering, liver physiology, cancer cell metabolism, photosynthesis, and cell culture engineering.
facilitate proteomics studies in CHO and E. coli hosts. In addition to bioprocess applications, they also review how proteomics has been applied in the drug development pipeline to identify new molecular targets and discover novel monoclonal antibodies. High-titer protein production requires upregulation of bioenergetic pathways to provide the energy and reductant required to maximize growth and protein synthesis. Furthermore, protein glycosylation requires adequate supply of nucleotide-sugar precursors derived from central metabolism. Therefore, it is no surprise that alterations in central metabolism are strongly associated with changes in cell growth, recombinant protein productivity, and glycan profile. Dickson describes how metabolomics analysis can provide a global perspective on the metabolic phenotypes associated with high-titer protein expression and can uncover metabolic variations among different clones. Furthermore, metabolomics can be used to identify overflow metabolites, such as sorbitol or glycerol, that may accumulate in the media under conditions when pathway flux is unbalanced. When combined with 13C flux analysis, this information can be critical for guiding media optimization and host cell engineering efforts. Young describes how recent advances in 13C flux analysis techniques have led to a surge in applications to both mammalian and microbial expression hosts. These studies have identified alterations in citric acid cycle and pentose phosphate pathway fluxes that are closely tied to enhancements in recombinant protein expression. However, the extent to which these metabolic alterations are necessary to drive high-titer protein expression is still unclear, and, furthermore, little is known about the effects that re-engineering central carbon metabolism can have on glycan profiles and other product quality attributes. Controlling glycosylation represents a major challenge in the biopharmaceutical industry since glycosylation can affect in vivo efficacy, serum halflife, or antigenicity of recombinant proteins. Once the factors that influence protein glycosylation are understood at a systems level, it is possible to construct predictive approaches to design host cells that achieve a desired glycan profile. Spahn and Lewis provide an overview of modeling approaches that have been used to guide such glycoengineering efforts. Implementation of these and other types of cellular engineering strategies is facilitated by the ability to precisely manipulate host genomes. However, unlike bacteria and yeast hosts, engineering of mammalian cells has typically relied on random vector integration to amplify expression of target genes or knockdown gene expression via shRNA. This is a crude approach that requires extensive cell selection and screening and cannot achieve complete knockout or replacement of target genes. Cheng and Alper describe recent developments in the application of TALENs, CRISPR/Cas9, zinc-finger nucleases, and other genome editing approaches to achieve targeted genome integrations in mammalian cells. These techniques, in combination with rapidly advancing omics capabilities and systems biology tools, are heralding a new era of rational genome engineering in mammalian cell biotechnology.
Glycosylation engineering in biopharmaceutical hosts Achieving acceptable product glycosylation involves, first and foremost, the selection of a suitable production host. However, Butler and Spearman point out that even widely used biopharmaceutical production hosts such as CHO and baby hamster kidney (BHK) cells are capable of expressing non-human glycoforms that may provoke an immune response in some patients. For this reason, several companies have turned to the use of human cell lines for recombinant protein expression, based on their prior success in vaccine production, to eliminate these undesirable glycoforms. Alternatively, www.sciencedirect.com
Current Opinion in Biotechnology 2014, 30:viii–x
x Pharmaceutical biotechnology
genetic engineering or chemical inhibition of glycosylation enzymes can be used to promote human-like glycoforms or to produce antibodies with enhanced cytotoxicity. Strasser et al. describe how these glycoengineering strategies are now enabling the use of plant cells as hosts for producing therapeutic glycoproteins, the first of which are now reaching the market. There have also been ongoing efforts to produce glycoengineered strains of S. cerevisiae and P. pastoris to suppress high-mannose glycoforms and promote the attachment of terminal sugars characteristic of human glycoproteins (such as N-acetylglucosamine, galactose, and sialic acid). Meehl and Stadheim describe how these glycoengineered yeast strains have been developed and used, not only for protein manufacturing, but also for screening candidate molecules by yeast surface display. Finally, the most ambitious feats of glycoengineering have involved the transfer of a functional N-glycosylation pathway into E. coli. Jaffe´ et al. describe the significant developments and remaining challenges in the pursuit to establish E. coli as an expression host for glycoprotein production.
Approaches for controlling protein quality and stability Tight control of protein quality and stability is closely linked to the success of any recombinant protein production system. New and viable approaches focus on improved understanding of several fundamentals. Young and Robinson offer insights about which pathways and mechanisms correlate to quality, anticipating future integration of protein engineering strategies with host cell manipulation. Roberts provides an overview of how and why different aggregated states of proteins occur, potential impacts on product quality, fundamental approaches to control formation, and practices currently used in industry. Dorai and Ganguly bridge the current understanding of glycoprotein heterogeneity to future host cell lines that minimize both intracellular and extracellular degradation of their protein products. Brorson and Jia present examples of N-terminal and C-terminal heterogeneities of commercially produced antibodies, methods of detection, and their impact on product quality. Despite efforts to produce a uniform therapeutic product, Harris and Kilby remind us that sometimes multiple
Current Opinion in Biotechnology 2014, 30:viii–x
species are produced due to amino acid misincorporation, proposing that quality specifications for recombinant proteins are best based on misincorporation data for native human proteins. Yang and Ambrogelly demonstrate that engineering of the IgG4 hinge region to resemble that of an IgG1 dramatically reduces half molecule exchange, impacting specificity since a heavy chain-light chain unit is exchanged between two IgG4 molecules. Raju and Lang describe the linkage between the structure and function of terminal sialic acid residues of glycoconjugates, highlighting diverse effects on biological activity, serum half-life, and structural stability. In contrast, Ju and Jung offer an alternative solution to reducing glycan heterogeneity by simply removing the glycans, then applying cellular and biopropress engineering to raise productivity and antibody engineering to reinstate therapeutic quality. In parallel with advancements in the prediction and control of product quality, advancements in the monitoring of product quality are underway. Hogwood et al. turn our attention to process and molecular impacts on the production and subsequent removal of host cell proteins, emphasizing the critical role of current and emerging measurement techniques to achieve an accurate understanding across the process. Pais et al. remind us that relevant information about post-translational modifications needs to be available close to real-time. Thus, new or re-designed analytical, automation, and multivariate tools are required to achieve monitoring and control during the manufacturing process.
Conclusion This special section of Current Opinion in Biotechnology focuses attention on a critical need in pharmaceutical biotechnology: the ability to understand, predict, and control product quality. We have invited 21 groups from around the world to review recent developments, future prospects, and ongoing challenges to engineer cells for high-quality biopharmaceuticals production. We would like to thank all the contributors and peer reviewers for their views on this rapidly evolving field of research.
References 1.
Kim JY, Kim YG, Lee GM: CHO cells in biotechnology for production of recombinant proteins: current state and further potential. Appl Microbiol Biotechnol 2012, 93:917-930.
www.sciencedirect.com
ScienceDirect Editorial overview: Pharmaceutical biotechnology: Engineering cells for high quality biopharmaceuticals production Beth H Junker and Jamey D Young Current Opinion in Biotechnology 2014, 30:viii–x For a complete overview see the Issue Available online 29th October 2014 http://dx.doi.org/10.1016/j.copbio.2014.10.002 0958-1669/# 2014 Elsevier Ltd. All rights reserved.
Beth H Junker
Bioprocess Development, Merck Research Laboratories, Kenilworth, NJ, USA e-mail: [email protected] Beth H Junker is currently a director in Bioprocess Development at Merck Research Laboratories in Kenilworth, NJ, USA. She obtained her Ph.D. from the Massachusetts Institute of Technology in biochemical engineering. She has been working at Merck for the past 25 years in varied technical roles to develop and commercialize diverse products such as VAQTATM, CancidasTM, and most recently, KeytrudaTM. Her interests include multi-scale production systems, quality-by-design, operational excellence, and sustainable bioprocessing.
Jamey D Young
Chemical and Biomolecular Engineering and Molecular Physiology and Biophysics, Vanderbilt University, Nashville, TN, USA e-mail: [email protected] Jamey D Young is an Assistant Professor of Chemical and Biomolecular Engineering and Current Opinion in Biotechnology 2014, 30:viii–x
Introduction Biopharmaceuticals, encompassing vaccines and protein therapeutics, comprise the fastest growing class of drug compounds on the market. This growth has spurred innovation in manufacturing to improve product titers and reduce the cost of goods sold. There is now a proliferation of success stories in the biotechnology industry where recombinant protein titers have been improved up to and beyond the 5 g/L level. Although further improvements in product titers are still achievable, a broader goal for host cell and protein engineering that encompasses product quality and consistency, in addition to high yield, has now firmly taken root. Furthermore, with a wave of billion-dollar biologics coming off patent in the next several years, the impetus to produce biosimilars with quality metrics ‘equivalent’ to established biopharmaceutical compounds is rapidly mounting. It is likely that future cell culture engineering efforts will place greater emphasis on controlling quality over the shear amount of product that can be manufactured by a recombinant host. In this issue dedicated to Pharmaceutical Biotechnology, we highlight several new technologies that improve our ability to predict, control, and monitor the quality of biotherapeutics derived from cell culture.
Systems and synthetic biology tools Before product quality can be predicted or controlled, the genetic factors that determine quality have to be understood. A major recent breakthrough in this regard was the publication of genome sequences for Chinese hamster and several Chinese hamster ovary (CHO) host cell lines. CHO cells account for over 70% of recombinant protein therapeutics currently on the market [1], and CHO sequence information provides a Rosetta Stone for translating knowledge and methodologies developed in other organisms to this important production host. Baik and Lee summarize ongoing developments in CHO genome sequencing as well as efforts to consolidate and distribute genome information through the CHOgenome.org website (http://www.chogenome.org/). They point to the importance of establishing a communitywide CHO reference genome with functional annotation to enable systematic studies of genotype-phenotype relationships. Vishwanathan et al. further highlight the ways that genome sequence information is also bolstering the application of transcriptomic profiling to understand the basis of hyper-productivity in CHO cells and to elucidate genes that influence the glycosylation pattern of product proteins. Similarly, Heffner et al. point to the important role of publicly available sequence databases, as well as advances in mass spectrometry approaches and bioinformatics tools, to www.sciencedirect.com
Editorial overview Junker and Young ix
Assistant Professor of Molecular Physiology and Biophysics at Vanderbilt University. The central theme of his research is metabolic engineering and systems biology. He received his Ph.D. from Purdue University in 2005 and continued his training as a postdoctoral fellow at MIT until joining Vanderbilt in 2008. He was awarded the NSF CAREER Award in 2010 and the DOE Early Career Award in 2012. He has authored over 30 articles and book chapters describing the application of mathematical modeling and 13 C flux analysis to a variety of research topics, including E. coli metabolic engineering, liver physiology, cancer cell metabolism, photosynthesis, and cell culture engineering.
facilitate proteomics studies in CHO and E. coli hosts. In addition to bioprocess applications, they also review how proteomics has been applied in the drug development pipeline to identify new molecular targets and discover novel monoclonal antibodies. High-titer protein production requires upregulation of bioenergetic pathways to provide the energy and reductant required to maximize growth and protein synthesis. Furthermore, protein glycosylation requires adequate supply of nucleotide-sugar precursors derived from central metabolism. Therefore, it is no surprise that alterations in central metabolism are strongly associated with changes in cell growth, recombinant protein productivity, and glycan profile. Dickson describes how metabolomics analysis can provide a global perspective on the metabolic phenotypes associated with high-titer protein expression and can uncover metabolic variations among different clones. Furthermore, metabolomics can be used to identify overflow metabolites, such as sorbitol or glycerol, that may accumulate in the media under conditions when pathway flux is unbalanced. When combined with 13C flux analysis, this information can be critical for guiding media optimization and host cell engineering efforts. Young describes how recent advances in 13C flux analysis techniques have led to a surge in applications to both mammalian and microbial expression hosts. These studies have identified alterations in citric acid cycle and pentose phosphate pathway fluxes that are closely tied to enhancements in recombinant protein expression. However, the extent to which these metabolic alterations are necessary to drive high-titer protein expression is still unclear, and, furthermore, little is known about the effects that re-engineering central carbon metabolism can have on glycan profiles and other product quality attributes. Controlling glycosylation represents a major challenge in the biopharmaceutical industry since glycosylation can affect in vivo efficacy, serum halflife, or antigenicity of recombinant proteins. Once the factors that influence protein glycosylation are understood at a systems level, it is possible to construct predictive approaches to design host cells that achieve a desired glycan profile. Spahn and Lewis provide an overview of modeling approaches that have been used to guide such glycoengineering efforts. Implementation of these and other types of cellular engineering strategies is facilitated by the ability to precisely manipulate host genomes. However, unlike bacteria and yeast hosts, engineering of mammalian cells has typically relied on random vector integration to amplify expression of target genes or knockdown gene expression via shRNA. This is a crude approach that requires extensive cell selection and screening and cannot achieve complete knockout or replacement of target genes. Cheng and Alper describe recent developments in the application of TALENs, CRISPR/Cas9, zinc-finger nucleases, and other genome editing approaches to achieve targeted genome integrations in mammalian cells. These techniques, in combination with rapidly advancing omics capabilities and systems biology tools, are heralding a new era of rational genome engineering in mammalian cell biotechnology.
Glycosylation engineering in biopharmaceutical hosts Achieving acceptable product glycosylation involves, first and foremost, the selection of a suitable production host. However, Butler and Spearman point out that even widely used biopharmaceutical production hosts such as CHO and baby hamster kidney (BHK) cells are capable of expressing non-human glycoforms that may provoke an immune response in some patients. For this reason, several companies have turned to the use of human cell lines for recombinant protein expression, based on their prior success in vaccine production, to eliminate these undesirable glycoforms. Alternatively, www.sciencedirect.com
Current Opinion in Biotechnology 2014, 30:viii–x
x Pharmaceutical biotechnology
genetic engineering or chemical inhibition of glycosylation enzymes can be used to promote human-like glycoforms or to produce antibodies with enhanced cytotoxicity. Strasser et al. describe how these glycoengineering strategies are now enabling the use of plant cells as hosts for producing therapeutic glycoproteins, the first of which are now reaching the market. There have also been ongoing efforts to produce glycoengineered strains of S. cerevisiae and P. pastoris to suppress high-mannose glycoforms and promote the attachment of terminal sugars characteristic of human glycoproteins (such as N-acetylglucosamine, galactose, and sialic acid). Meehl and Stadheim describe how these glycoengineered yeast strains have been developed and used, not only for protein manufacturing, but also for screening candidate molecules by yeast surface display. Finally, the most ambitious feats of glycoengineering have involved the transfer of a functional N-glycosylation pathway into E. coli. Jaffe´ et al. describe the significant developments and remaining challenges in the pursuit to establish E. coli as an expression host for glycoprotein production.
Approaches for controlling protein quality and stability Tight control of protein quality and stability is closely linked to the success of any recombinant protein production system. New and viable approaches focus on improved understanding of several fundamentals. Young and Robinson offer insights about which pathways and mechanisms correlate to quality, anticipating future integration of protein engineering strategies with host cell manipulation. Roberts provides an overview of how and why different aggregated states of proteins occur, potential impacts on product quality, fundamental approaches to control formation, and practices currently used in industry. Dorai and Ganguly bridge the current understanding of glycoprotein heterogeneity to future host cell lines that minimize both intracellular and extracellular degradation of their protein products. Brorson and Jia present examples of N-terminal and C-terminal heterogeneities of commercially produced antibodies, methods of detection, and their impact on product quality. Despite efforts to produce a uniform therapeutic product, Harris and Kilby remind us that sometimes multiple
Current Opinion in Biotechnology 2014, 30:viii–x
species are produced due to amino acid misincorporation, proposing that quality specifications for recombinant proteins are best based on misincorporation data for native human proteins. Yang and Ambrogelly demonstrate that engineering of the IgG4 hinge region to resemble that of an IgG1 dramatically reduces half molecule exchange, impacting specificity since a heavy chain-light chain unit is exchanged between two IgG4 molecules. Raju and Lang describe the linkage between the structure and function of terminal sialic acid residues of glycoconjugates, highlighting diverse effects on biological activity, serum half-life, and structural stability. In contrast, Ju and Jung offer an alternative solution to reducing glycan heterogeneity by simply removing the glycans, then applying cellular and biopropress engineering to raise productivity and antibody engineering to reinstate therapeutic quality. In parallel with advancements in the prediction and control of product quality, advancements in the monitoring of product quality are underway. Hogwood et al. turn our attention to process and molecular impacts on the production and subsequent removal of host cell proteins, emphasizing the critical role of current and emerging measurement techniques to achieve an accurate understanding across the process. Pais et al. remind us that relevant information about post-translational modifications needs to be available close to real-time. Thus, new or re-designed analytical, automation, and multivariate tools are required to achieve monitoring and control during the manufacturing process.
Conclusion This special section of Current Opinion in Biotechnology focuses attention on a critical need in pharmaceutical biotechnology: the ability to understand, predict, and control product quality. We have invited 21 groups from around the world to review recent developments, future prospects, and ongoing challenges to engineer cells for high-quality biopharmaceuticals production. We would like to thank all the contributors and peer reviewers for their views on this rapidly evolving field of research.
References 1.
Kim JY, Kim YG, Lee GM: CHO cells in biotechnology for production of recombinant proteins: current state and further potential. Appl Microbiol Biotechnol 2012, 93:917-930.
www.sciencedirect.com
