PERSPECTIVE Industrial Systems Biology Jose �� Manuel Otero, Jens Nielsen Systems Biology, Department of Chemical and Biological Engineering, Chalmers University of Technology, SE-412 96 Goteborg, �� Sweden telephone: ��46-31-772-8633 fax: ��46-31-16-0062 e-mail: nielsenj@chalmers.se Received 15 June 2009 revision received 2 October 2009 accepted 14 October 2009 Published online 4 November 2009 in Wiley InterScience (www.interscience.wiley.com). DOI 10.1002/bit.22592 ABSTRACT: The chemical industry is currently undergoing a dramatic change driven by demand for developing more sustainable processes for the production of fuels, chemicals, and materials. In biotechnological processes different micro- organisms can be exploited, and the large diversity of metabolic reactions represents a rich repository for the design of chemical conversion processes that lead to efficient production of desirable products. However, often micro- organisms that produce a desirable product, either naturally or because they have been engineered through insertion of heterologous pathways, have low yields and productivities, and in order to establish an economically viable process it is necessary to improve the performance of the microorgan- ism. Here metabolic engineering is the enabling technology. Through metabolic engineering the metabolic landscape of the microorganism is engineered such that there is an efficient conversion of the raw material, typically glucose, to the product of interest. This process may involve both insertion of new enzymes activities, deletion of existing enzyme activities, but often also deregulation of existing regulatory structures operating in the cell. In order to rapidly identify the optimal metabolic engineering strategy the industry is to an increasing extent looking into the use of tools from systems biology. This involves both x-ome tech- nologies such as transcriptome, proteome, metabolome, and fluxome analysis, and advanced mathematical modeling tools such as genome-scale metabolic modeling. Here we look into the history of these different techniques and review how they find application in industrial biotechnology, which will lead to what we here define as industrial systems biology. Biotechnol. Bioeng. 2010 105: 439���460. �� 2009 Wiley Periodicals, Inc. KEYWORDS: systems biology industrial biotechnology metabolic engineering Introduction The term ������industrial biotechnology������ first widely appeared in the literature in the early 1980s when genetic engineering, propelled by recombinant DNA technology, was searching for applications beyond health care and medical bio- technology (Ferrandiz-Garcia, 1982 Pass, 1981). Industrial biotechnology today represents an established field with significant government, corporate, and academic invest- ment. Formally, industrial biotechnology is the bioconver- sion, either through microbial fermentation or cell-free biocatalysis, of organic feedstocks extracted from biomass or their derivatives to chemicals, materials, and/or energy. Biomass is the result of photosynthetic carbon fixation by plants to form organic polymers that may be digested, enzymatically or chemically, to carbohydrate, protein, and lipid monomers. Industrial biotechnology, often referred to as white biotechnology in Europe (Maury et al., 2005), aims to provide cost-competitive, environmentally friendly, sustainable alternatives to existing or newly proposed petrochemical processes. Processes that exploit industrial biotechnology have recently garnered increasing global attention with traditional petrochemical processing under scrutiny due to increasing raw material costs, environmental constraints, and decreasing sustainability. Industrial biotechnology has experienced unprecedented growth with bio-based production processes representing 5% of the total chemical production sales volume. By 2010, several studies have estimated that the total fraction will increase to 20%, representing US$310 billion of a projected total sales volume of US$1,600 billion. Industrial biotechno- logy will continue to capture significant sales volume percentages in the arenas of basic chemicals and commodities (2���15%), specialty and added-value chemicals (2���20%), and polymers (1���15%). However, the greatest percentage gain is likely to occur in the fine chemical market (16���60%), where industrial biotechnology platforms enable complex chemistry that are presently produced via complex synthetic or combinatorial routes (Hirche, 2006). Furthermore, industrial biotechnology is enabling new products including novel therapeutic agents such as polyketides, and specialty chemi- cals not previously identified such as the diverse poly- unsaturated fatty acids and biopolymers produced by microalgae (Gavrilescu and Chisti, 2005). In its relatively short history, industrial biotechnology commercialization of fermentation processes for antibiotics Correspondence to: Prof. J. Nielsen Additional Supporting Information may be found in the online version of this article. �� 2009 Wiley Periodicals, Inc. Biotechnology and Bioengineering, Vol. 105, No. 3, February 15, 2010 439

(penicillin production by Penicillium chrysogenum annual market size exceeding US$1.5 billion), vitamins (L-ascorbic acid production by the Reichstein process and biocatalysis by Gluconobacter oxydans annual market size exceeding US$600 million), organic acids (citric acid production by Aspergillus sp. annual market size exceeding US$1.5 billion), and amino acids (L-glutamate and L-lysine production by Corynebacterium glutamicum annual production exceeding 600,000 tons) are well established and successful (Gavrilescu and Chisti, 2005). In each of these examples, host organisms well suited for production of the target compound were naturally isolated. Furthermore, under controlled environ- ments, random mutagenesis followed by screening, selec- tion, and traditional bioprocess development were used to enhance production yields, titers, productivities, and robustness. Despite the fact that this method provides little to no mechanistic understanding of which specific genetic perturbations lead to improved strains so that they could be further exploited, it has proven to be commercially successful as illustrated by the more than 1,000-fold improvement in penicillin titer by P. chrysogenum (Nielsen, 1995). The significant increases in research and development, and commercialization at industrial scales of biotechnolo- gical processes may be attributed to several key factors, which can be grouped into four broader factors that are important to consider in connection with development of a new bio-based process: (1) process economics, (2) biotechnology process development, (3) environmental impact, and (4) sustainability and self-sufficiency. Each of these broad factors involve several identifiable and quantitative drivers fueling the application of industrial biotechnology to processes previously exclusive to the petrochemical industry or for the production of new chemicals. Figure 1 outlines an overview of how each of these four factors needs to be considered and evaluated before development of any industrial biotechnology process. This review aims to provide a historical perspective of industrial biotechnology process development, and in particular, focus on the rapid deployment of metabolic engineering and systems biology technologies that first emerged from academic research groups driven by the human health and medical biotechnology sectors. Specifically, mature, recently launched and in-development examples of products that have benefited from systems biology will be highlighted for motivation. Based on such examples, and de novo processes presently in proof-of- concept, we here define a new term, industrial systems biology, acknowledging that tools established in the rapidly growing field of systems biology, often applied to metabolic engineering, are prevalent in two forms. Enterprises are reshaping existing or forming new process development groups with industrial systems biology capabilities and expertise, or they are out-sourcing process development to small, recently formed entities that specialize in industrial systems biology. Examples of such enterprises focused on providing industrial systems biology expertise to more traditional process development groups include METabolic EXplorer (www.metabolic-explorer.com, France, founded in 1999), Genomatica (www.genomatica.com, USA, founded in 2000), Fluxome Sciences (www.fluxome.com, Denmark, founded in 2002), Amyris Biotechnologies (www.amyris.com, USA, founded in 2003), and Microbia Precision Engineering (www.microbia.com, USA, a sub- sidiary of Ironwood Pharmaceuticals, formerly Microbia). Although relatively small (US$ 50 million 2009 total revenue), these companies have significant collaborations Figure 1. Industrial biotechnology drivers. The above figure summarizes the four key factors that are often evaluated when considering substitution of a petro- chemical process with a biotechnology process, or its implementation for production of a novel chemical. Process economics, as compared to petrochemical equivalents or other benchmarking processes, are critical in establishing commercial viability, with particular focus paid to long-term operating costs. Next, biotechnology development costs, resources, and development efforts are considered, with initial analysis focused on establishing pilot-plant scale proof-of-concept. The final two factors to be critically evaluated include sustainability and self-sufficiency, and environmental impact. Sustainability and self-sufficiency not only relate to process specific con- siderations, such as feedstock availability, or the opportunities for further expansion through biorefinery integration, but also includes focus on public perception and the socio-political landscape. A careful consideration of these four general sectors, will ultimately determine whether proceeding with biotechnology process development is warranted or not. While not immediately obvious to most research and development scientists or engineers, it is critical to not divorce the impact these considerations may have on process development, particularly in designing strategies for construction of a microbial cell factory. It is such analysis that often defines the constraints, boundaries, targets, and viable metabolic engineering strategies, including which systems biology approaches should be exploited to experimentally demonstrate proof-of-concept. 440 Biotechnology and Bioengineering, Vol. 105, No. 3, February 15, 2010

with many of the major chemical manufacturing, nutra- ceutical, pharmaceutical, and petrochemical companies. Industrial Systems Biology Systems biology is the quantitative analysis, often through the use of predictive mathematical models, of biological systems. Often it involves collection, analysis, and integra- tion of whole genome scale data sets with the objective to gain a quantitative phenotypic description of the biological system. With genome sequences becoming readily available for production organisms, process development has been a benefactor of the scientific achievements in systems biology, particularly in the areas of transcriptomics, proteomics, metabolomics, and fluxomics. Such developments today encompass a systems biology toolbox that may be further exploited for production of metabolic intermediates that often serve as desirable precursors in the petrochemical sector. Figure 2 provides a road-map for how industrial systems biology may be applied to microbial cell factories of interest. Figure 2. Industrial systems biology. Industrial systems biology is a dynamic interaction between various disciplines and approaches. At the core is a platform technology based on a production host, for which a genome sequence is available, and subsequent annotations based on existing literature review, database query, comparative genomics, and experimental data, where available, are completed. The annotations may vary in types of functional genomics data assigned to specific fields however, a standard skeleton syntax structure of defining a gene, the gene product (e.g., metabolic enzyme), the metabolites serving as reactants and products (including any co-factors and intermediates), and the resulting stoichiometry is often applied. This framework, referred to as a genome-scale metabolic network reconstruction, may then be used for stoichiometric or kinetic modeling. Often, because kinetics parameters such as the forward and reverse reaction rates at physiologically relevant conditions have not been experimentally determined for a significant fraction of the network, flux balance analysis (FBA) is used for predictive modeling as it only depends on the stoichiometry and network constraints (e.g., precise stoichiometric definition of biomass, ATP maintenance terms, glucose uptake rate). Once a high-probability of success metabolic engineering strategy has been identified, often requiring gene over-expression, deletion, or non-native pathway reconstruction, genetic engineering is performed on the production host, yielding a modified strain. The modified strain is initially characterized, and may undergo directed evolution or other non-targeted approaches to yield an improved phenotype. The resulting modified strain is then characterized under well-controlled fermentation conditions, where physiological parameters, such as maximum specific growth rate, substrate consumption rates, product yields and titers, by-product formation, and morphology are determined. Furthermore, functional genomics characterization, often requiring transcriptome, proteome, metabolome, and fluxome measurements is completed. Bioinformatics, coupled with data integration, are then required for analysis of the resulting modified strain, and to identify opportunities for a second round of metabolic engineering. Furthermore, the analysis should lead to a revised model with improved predictive power that may yield promising strategies for further phenotype improvement. While this approach has often been referred to as the metabolic engineering cycle, we here compliment the traditional cycle to include integrative approaches and data sets from systems biology. Together, when applied to industrial biotechnology products, this is referred to as industrial systems biology. Otero and Nielsen: Industrial Systems Biology 441 Biotechnology and Bioengineering