Engineering Robust Production Microbes for Large-Scale Cultivation
2019; Elsevier BV; Volume: 27; Issue: 6 Linguagem: Inglês
10.1016/j.tim.2019.01.006
ISSN1878-4380
AutoresMaren Wehrs, Deepti Tanjore, Thomas Eng, Jeff Lievense, Todd Pray, Aindrila Mukhopadhyay,
Tópico(s)Viral Infectious Diseases and Gene Expression in Insects
ResumoStrain engineering in the laboratory often does not consider process requirements in larger-scale bioreactors. Systems and synthetic biology can be applied to design microbial strains that allow reliable and robust production on a large scale. Commercial microbial platforms should be selected and developed based on their relevance to final process goals. Systems biology and synthetic biology are increasingly used to examine and modulate complex biological systems. As such, many issues arising during scaling-up microbial production processes can be addressed using these approaches. We review differences between laboratory-scale cultures and larger-scale processes to provide a perspective on those strain characteristics that are especially important during scaling. Systems biology has been used to examine a range of microbial systems for their response in bioreactors to fluctuations in nutrients, dissolved gases, and other stresses. Synthetic biology has been used both to assess and modulate strain response, and to engineer strains to improve production. We discuss these approaches and tools in the context of their use in engineering robust microbes for applications in large-scale production. Systems biology and synthetic biology are increasingly used to examine and modulate complex biological systems. As such, many issues arising during scaling-up microbial production processes can be addressed using these approaches. We review differences between laboratory-scale cultures and larger-scale processes to provide a perspective on those strain characteristics that are especially important during scaling. Systems biology has been used to examine a range of microbial systems for their response in bioreactors to fluctuations in nutrients, dissolved gases, and other stresses. Synthetic biology has been used both to assess and modulate strain response, and to engineer strains to improve production. We discuss these approaches and tools in the context of their use in engineering robust microbes for applications in large-scale production. It is well recognized that microbes can produce a vast range of compounds, from fuels and commodity chemicals to pharmaceuticals and fine chemicals [1Zhang M.M. et al.Engineering microbial hosts for production of bacterial natural products.Nat. Prod. Rep. 2016; 33: 963-987Crossref PubMed Google Scholar, 2Carbonell P. et al.An automated design-build-test-learn pipeline for enhanced microbial production of fine chemicals.Commun. Biol. 2018; 1: 66Crossref PubMed Scopus (114) Google Scholar, 3Beller H.R. et al.Natural products as biofuels and bio-based chemicals: fatty acids and isoprenoids.Nat. Prod. Rep. 2015; 32: 1508-1526Crossref PubMed Google Scholar]. Correspondingly, efforts to broaden the scope and demonstrate proof-of-concept pathways for new molecules are gaining in number and scope [2Carbonell P. et al.An automated design-build-test-learn pipeline for enhanced microbial production of fine chemicals.Commun. Biol. 2018; 1: 66Crossref PubMed Scopus (114) Google Scholar, 4Casini A. et al.A pressure test to make 10 molecules in 90 days: external evaluation of methods to engineer biology.J. Am. Chem. Soc. 2018; 140: 4302-4316Crossref PubMed Scopus (88) Google Scholar, 5Campodonico M.A. et al.Generation of an atlas for commodity chemical production in Escherichia coli and a novel pathway prediction algorithm, GEM-Path.Metab. Eng. 2014; 25: 140-158Crossref PubMed Scopus (124) Google Scholar]. However, despite the development of microbial platforms to convert nearly any carbon source into any desired product, a rather modest number of these cases have seen successful transition to industrial-scale processes and marketed products. Economic competitiveness with established chemical or biosynthetic routes is an important factor. Low titers and yields in the laboratory setting also need to be overcome to proceed with the scale-up. However, at the level of core bioconversion technology, a dominant cause for this dearth of implementation is the challenge associated with predicting the strain performance in industrial-scale bioreactors. The environment of a commercial-scale bioreactor is drastically different from that of laboratory-scale cultures such as shake flasks. It has long been recognized that most strains do not perform the same way in the two scenarios [6Humphrey A. Shake flask to fermentor: what have we learned?.Biotechnol Progress. 1998; 14: 3-7Crossref Scopus (118) Google Scholar]. Thus, strain development for industrial scale-up necessarily goes beyond pathway engineering [7Willrodt C. et al.Guiding efficient microbial synthesis of non-natural chemicals by physicochemical properties of reactants.Curr. Opin. Biotechnol. 2015; 35: 52-62Crossref PubMed Scopus (23) Google Scholar, 8Chubukov V. et al.Synthetic and systems biology for microbial production of commodity chemicals.NPJ Systems Biol. Applicat. 2016; 2: 16009Crossref PubMed Scopus (152) Google Scholar]. A recent review places the cost of scaling-up a large-volume chemical in the range of $100 million to $1 billion USD [9Crater J.S. Lievense J.C. Scale-up of industrial microbial processes.FEMS Microbiol. Lett. 2018; 365 (fny138)Crossref PubMed Scopus (98) Google Scholar]. The risk and cost of failure is high, and de-risking of the microbial chassis itself should be an important aspect of strain engineering. Process engineering practices for industrial scale-up of a microbial production system are fairly well established, with multiple reviews written on the topic [10Junker B.H. Scale-up methodologies for Escherichia coli and yeast fermentation processes.J. Biosci. Bioeng. 2004; 97: 347-364Crossref PubMed Scopus (173) Google Scholar, 11Young T.B. Fermentation scaleup: industrial experience with a total environmental approach.Ann. N. Y. Acad. Sci. 1979; 326: 165-180Crossref Scopus (9) Google Scholar, 12Schmidt F.R. Optimization and scale up of industrial fermentation processes.Appl. Microbiol. Biotechnol. 2005; 68: 425-435Crossref PubMed Scopus (221) Google Scholar]. In contrast, the response of microbial phenotype to fluctuating bioreactor conditions, genotypic drift during long-term cultivation, and other physiological factors has been less explored, and the potential impacts of such factors on scale-up performance have only recently been addressed [8Chubukov V. et al.Synthetic and systems biology for microbial production of commodity chemicals.NPJ Systems Biol. Applicat. 2016; 2: 16009Crossref PubMed Scopus (152) Google Scholar, 13Rugbjerg P. et al.Diverse genetic error modes constrain large-scale bio-based production.Nat. Commun. 2018; 9: 787Crossref PubMed Scopus (88) Google Scholar]. The best efforts in this area are in scale-down models, where large numbers of strains are generated and tested under conditions representative of large-scale growth and production, allowing better selection of strains that will scale up more predictably [14Neubauer P. Junne S. Scale-down simulators for metabolic analysis of large-scale bioprocesses.Curr. Opin. Biotechnol. 2010; 21: 114-121Crossref PubMed Scopus (138) Google Scholar, 15Baez A. et al.Simulation of dissolved CO(2) gradients in a scale-down system: a metabolic and transcriptional study of recombinant Escherichia coli.Biotechnol. J. 2011; 6: 959-967Crossref PubMed Scopus (20) Google Scholar, 16Caspeta L. et al.The effect of heating rate on Escherichia coli metabolism, physiological stress, transcriptional response, and production of temperature-induced recombinant protein: a scale-down study.Biotechnol. Bioengineer. 2009; 102: 468-482Crossref PubMed Scopus (60) Google Scholar, 17de Jonge L.P. et al.Scale-down of penicillin production in Penicillium chrysogenum.Biotechnol. J. 2011; 6: 944-958Crossref PubMed Scopus (71) Google Scholar, 18Lara A.R. et al.Living with heterogeneities in bioreactors: understanding the effects of environmental gradients on cells.Mol. Biotechnol. 2006; 34: 355-381Crossref PubMed Scopus (281) Google Scholar]. However, it remains difficult to develop strains that will maintain the desired production characteristics when scaled up. Advances in systems and synthetic biology [8Chubukov V. et al.Synthetic and systems biology for microbial production of commodity chemicals.NPJ Systems Biol. Applicat. 2016; 2: 16009Crossref PubMed Scopus (152) Google Scholar, 19Lee S.Y. Kim H.U. Systems strategies for developing industrial microbial strains.Nat. Biotechnol. 2015; 33: 1061-1072Crossref PubMed Scopus (332) Google Scholar] may be able to provide solutions in this space. Beyond improvements such as the use of strains without plasmids or inducer requirement, there are some ambitious questions to be posed. Can we create genetically and phenotypically stable strains to prevent a phenotypic drift during subculturing of the seed cultures? Can we reliably increase the duration of the production phase to increase the profitability of the process? Can we track subpopulations and the causal parameters for phenotypic drift or heterogeneity within the culture? Can we dynamically regulate genes and pathways to respond to inhibitory byproducts and fluctuating stresses to reduce negative selections and enhance robust strain performance? There are examples of studies that address these questions individually and which, when presented collectively, reveal the full potential of the approach. Most ambitiously, we envision microbial engineering approaches that can predict how a strain will behave in a commercial-scale bioreactor environment and that will enable us to pre-emptively design strains that meet those needs (Figure 1). In this review, we focus on systems and synthetic biology methods that may allow a better understanding of physiological changes during scale-up and under industrial conditions, and synthetic biology methods that may provide ways to mitigate or control these responses. Though these methodologies are as yet underutilized, there are some excellent examples that illustrate the potential of these approaches in predictive design of engineered strains. A wide range of chemical, physical, and biological factors can negatively impact microbial growth and product formation during bioprocess scale-up from microtiter plates, to shake flasks, to bench-scale bioreactors, and ultimately to commercial bioreactors. Compounding this issue is the fact that mimicking the environments encountered in these different stages of scale-up during high-throughput phenotypic screening in the initial strain engineering phase is not straightforward. Moreover, the associated challenges that need to be addressed during scale-up vary for different host microbes and cultivation products. Furthermore, many measurements performed in microtiter plates and shake flasks are different from the analyses conducted during bioreactor studies (Figure 2). This exacerbates the risk that laboratory-scale strain optimization does not address or resolve challenges that will subsequently be seen as bioprocesses are scaled toward commercialization. One conspicuous difference between laboratory-scale and large-scale cultivation is the operating pressure. As bioreactor volumes increase, the height of the water column in a bioreactor creates an increasing hydrostatic pressure gradient. In principle, pressure can influence biological properties, including enzyme activity and cell membrane permeability that are important to cell viability and metabolic flux [18Lara A.R. et al.Living with heterogeneities in bioreactors: understanding the effects of environmental gradients on cells.Mol. Biotechnol. 2006; 34: 355-381Crossref PubMed Scopus (281) Google Scholar]. The increased hydrostatic pressure at the bottom of a commercial-scale bioreactor raises the concentration of dissolved gases. For instance, dissolved CO2 (dCO2) is in equilibrium with bicarbonate and carbonate ions which contribute to medium osmolarity and affect broth pH. Thus, the physiological effect of dCO2 could be a result of the accumulation of the dissolved gas itself, of osmotic pressure changes, of pH changes, or even a combination of all three factors [20Zanghi J.A. et al.Bicarbonate concentration and osmolality are key determinants in the inhibition of CHO cell polysialylation under elevated pCO(2) or pH.Biotechnol. Bioeng. 1999; 65: 182-191Crossref PubMed Scopus (60) Google Scholar]. Other colloidal or thermodynamic properties, such as broth bulk viscosity, emulsion stability, and product or biomass settling, could also be influenced by pressure, among other factors. Adverse effects on the cultivation may also persist during product harvest and impact the efficiency and quality of the downstream process (e.g., solid–liquid separation, chromatography, extraction, crystallization, distillation, chemical upgrading, and polymerization), resulting in higher production costs and reduced product quality [21Kumar D. Murthy G.S. Impact of pretreatment and downstream processing technologies on economics and energy in cellulosic ethanol production.Biotechnol. Biofuels. 2011; 4: 27Crossref PubMed Scopus (250) Google Scholar]. With increasing bioreactor volume, the mixing time increases from a few seconds (laboratory-scale) to several minutes (hundreds of cubic meters) [6Humphrey A. Shake flask to fermentor: what have we learned?.Biotechnol Progress. 1998; 14: 3-7Crossref Scopus (118) Google Scholar, 12Schmidt F.R. Optimization and scale up of industrial fermentation processes.Appl. Microbiol. Biotechnol. 2005; 68: 425-435Crossref PubMed Scopus (221) Google Scholar]. Furthermore, other bioreactor internal components, such as spargers, baffles, and cooling coils, can create dead zones with poor mixing, heat transfer, and gas–liquid mass transfer. In large-scale production processes imperfect mixing results in microenvironments and inhomogeneities, resulting in gradients of bioprocess parameters such as pH, temperature, dissolved oxygen (DO), dCO2, and the concentration of nutrients. Combined, these variations could lead to transient or permanent insults such as oxidative damage, nutrient limitations, and other stress responses that reduce microbial viability, stability, and productivity [22Deparis Q. et al.Engineering tolerance to industrially relevant stress factors in yeast cell factories.FEMS Yeast Res. 2017; 17 (fox036)Crossref PubMed Scopus (108) Google Scholar]. Additional chemical stresses in large-scale industrial cultivations arise from raw material and microbial contaminants. To reduce production costs, many processes utilize crude raw materials which are utilized with little or no refinement and thus introduce impurities that accumulate in the broth to inhibitory/toxic levels. Some examples of these real-world feedstocks include corn slurry, corn steep liquor, sugar cane juice, molasses, sugar beet juice, agricultural residues, food-processing waste, municipal solid waste, and waste gases. Microbial contamination of the raw materials, harsh raw material pretreatment conditions, and microbial contamination of the cultivation can further increase the levels of deleterious impurities and reduce product titer rate yield (TRY) metrics [22Deparis Q. et al.Engineering tolerance to industrially relevant stress factors in yeast cell factories.FEMS Yeast Res. 2017; 17 (fox036)Crossref PubMed Scopus (108) Google Scholar, 23Serate J. et al.Controlling microbial contamination during hydrolysis of AFEX-pretreated corn stover and switchgrass: effects on hydrolysate composition, microbial response and fermentation.Biotechnol. Biofuels. 2015; 8: 180Crossref PubMed Scopus (33) Google Scholar, 24Fischer C.R. et al.Selection and optimization of microbial hosts for biofuels production.Metab. Engineer. 2008; 10: 295-304Crossref PubMed Scopus (305) Google Scholar, 25Mohamed E.T. et al.Generation of a platform strain for ionic liquid tolerance using adaptive laboratory evolution.Microb. Cell Factor. 2017; 16: 204Crossref PubMed Scopus (48) Google Scholar, 26Mukhopadhyay A. Tolerance engineering in bacteria for the production of advanced biofuels and chemicals.Trends Microbiol. 2015; 23: 498-508Abstract Full Text Full Text PDF PubMed Scopus (160) Google Scholar]. These factors can and should be addressed in host strain selection and strain-screening assays, but the priority may be underappreciated and thus the sensitivities understudied at the early stages of a strain/cultivation development program where sterile technique and very pure raw materials are commonly used during small-scale screening and optimization. If strain engineering and strain screening could account for these risks and mitigate them from the outset, process scaling would become more predictable, especially in the case of developing novel microbes as production platforms. Exposure to heterogeneous conditions in large fermenters can trigger genetic and physiological responses in production microbes [18Lara A.R. et al.Living with heterogeneities in bioreactors: understanding the effects of environmental gradients on cells.Mol. Biotechnol. 2006; 34: 355-381Crossref PubMed Scopus (281) Google Scholar, 27Enfors S.O. et al.Physiological responses to mixing in large scale bioreactors.J. Biotechnol. 2001; 85: 175-185Crossref PubMed Scopus (348) Google Scholar, 28Takors R. Scale-up of microbial processes: impacts, tools and open questions.J. Biotechnol. 2012; 160: 3-9Crossref PubMed Scopus (126) Google Scholar]. High-throughput analytics, such as automated DNA sequencing and −omics technologies (fluxomics, transcriptomics, proteomics, and metabolomics), as well as high-throughput screening platforms, enable these to be characterized by simulating industrial conditions in scaled-down laboratory experiments. Mathematical models of large-scale fermenters provide important guidance in designing the scaled-down conditions and in forecasting the impact of small-scale learnings in the large-scale environment. This approach increases the probability that key performance metrics, such as TRY, will translate fully and quickly from the laboratory to the manufacturing plant, thereby minimizing start-up costs [8Chubukov V. et al.Synthetic and systems biology for microbial production of commodity chemicals.NPJ Systems Biol. Applicat. 2016; 2: 16009Crossref PubMed Scopus (152) Google Scholar, 29Delvigne F. et al.Bioprocess scale-up/down as integrative enabling technology: from fluid mechanics to systems biology and beyond.Microb. Biotechnol. 2017; 10: 1267-1274Crossref PubMed Scopus (40) Google Scholar]. Heterogeneity in gas mixing, in substrates and nutrient, as well as in the inherent heterogeneity in a microbial population have all been examined in this context. Gas mixing, such as for CO2 and O2, is a key parameter resulting in measurable differences in microbial cell response in a bioreactor. In recombinant Escherichia coli cultures grown in batch cultivation, increasing dCO2 in the broth resulted in a slower growth rate and higher by-product accumulation [30Castan A. et al.Oxygen enriched air supply in Escherichia coli processes: production of biomass and recombinant human growth hormone.Enzyme Microb. Technol. 2002; 30: 847-854Crossref Scopus (42) Google Scholar]. Another study found that fluctuating dCO2 increased the time needed for E. coli to produce green fluorescent protein (GFP), and to observe a transcriptional response of glutamate decarboxylase and α‐ketoglutarate dehydrogenase [15Baez A. et al.Simulation of dissolved CO(2) gradients in a scale-down system: a metabolic and transcriptional study of recombinant Escherichia coli.Biotechnol. J. 2011; 6: 959-967Crossref PubMed Scopus (20) Google Scholar]. A comparative analysis of transcript levels revealed a fast transcriptional response of Corynebacterium glutamicum to alternating CO2 and bicarbonate levels, which appeared to be strongly correlated with both strength and duration of the pCO2 stimulus [31Buchholz J. et al.CO₂/HCO₃⁻ perturbations of simulated large scale gradients in a scale-down device cause fast transcriptional responses in Corynebacterium glutamicum.Appl. Microbiol. Biotechnol. 2014; 98: 8563-8572Crossref PubMed Scopus (38) Google Scholar]. The authors observed increased expression of genes encoding enzymes necessary for oxaloacetate production and hypothesized this to be the result of rapid enhanced bicarbonate availability for PEP/pyruvate carboxylation. The first findings of the impact of oxygen supply were reported by Oosterhuis et al. [32Oosterhuis N.M. et al.Scale-down and optimization studies of the gluconic acid fermentation by Gluconobacter oxydans.Biotechnol. Bioengineer. 1985; 27: 711-720Crossref PubMed Scopus (72) Google Scholar]. In these early studies, Gluconobacter oxydans was cycled between zones with low and high oxygen availability to study the effect on gluconic acid production. More recent studies used similar approaches to study the effect of oxygen availability on intracellular processes in E. coli and C. glutamicum [18Lara A.R. et al.Living with heterogeneities in bioreactors: understanding the effects of environmental gradients on cells.Mol. Biotechnol. 2006; 34: 355-381Crossref PubMed Scopus (281) Google Scholar, 33Lara A.R. et al.Fast dynamic response of the fermentative metabolism of Escherichia coli to aerobic and anaerobic glucose pulses.Biotechnol. Bioengineer. 2009; 104: 1153-1161Crossref PubMed Scopus (57) Google Scholar, 34Limberg M.H. et al.Metabolic profile of 1,5-diaminopentane producing Corynebacterium glutamicum under scale-down conditions: blueprint for robustness to bioreactor inhomogeneities.Biotechnol. Bioengineer. 2017; 114: 560-575Crossref PubMed Scopus (36) Google Scholar, 35Käß F. et al.Assessment of robustness against dissolved oxygen/substrate oscillations for C. glutamicum DM1933 in two-compartment bioreactor.Bioprocess Biosyst. Engineer. 2014; 37: 1151-1162Crossref PubMed Scopus (37) Google Scholar, 36von Wulffen J. et al.Rapid sampling of Escherichia coli after changing oxygen conditions reveals transcriptional dynamics.Genes (Basel). 2017; 8: 90Crossref Scopus (13) Google Scholar]. In these studies, transcriptional changes between aerobic and anaerobic environments explained observed alterations in stoichiometry and kinetics, as well as production of ethanol and organic acids, indicating that E. coli responds rapidly to dissolved oxygen gradients [33Lara A.R. et al.Fast dynamic response of the fermentative metabolism of Escherichia coli to aerobic and anaerobic glucose pulses.Biotechnol. Bioengineer. 2009; 104: 1153-1161Crossref PubMed Scopus (57) Google Scholar]. Fu et al. [37Fu Z. et al.Exometabolome analysis reveals hypoxia at the up-scaling of a Saccharomyces cerevisiae high-cell density fed-batch biopharmaceutical process.Microb. Cell Fact. 2014; 13: 32Crossref PubMed Scopus (33) Google Scholar] carried out exometabolome measurements to study the physiological effects of scale-up of a high-density fed-batch process of Saccharomyces cerevisiae which revealed reduced oxygen availability at commercial scale, resulting in overflow metabolism. As one approach to mitigate the effects of oxygen availability in large bioreactors, Liu et al. introduced VHb (Vitreoscilla hemoglobin, a membrane protein facilitating O2 transport) into a fatty-acid-producing strain of E. coli to promote oxygen supply and energy metabolism. The resulting strain yielded a 70% higher fatty acid titer as compared to its parental strain [38Liu D. et al.Enhancing fatty acid production in Escherichia coli by Vitreoscilla hemoglobin overexpression.Biotechnol. Bioeng. 2017; 114: 463-467Crossref PubMed Scopus (23) Google Scholar]. While host response to such parameters in large-scale production remains a challenge (see Outstanding Questions), the selection of a host that is natively resilient to such fluctuations [35Käß F. et al.Assessment of robustness against dissolved oxygen/substrate oscillations for C. glutamicum DM1933 in two-compartment bioreactor.Bioprocess Biosyst. Engineer. 2014; 37: 1151-1162Crossref PubMed Scopus (37) Google Scholar] and designing processes to overcome metabolic bottlenecks [39Wehrs M. et al.Production efficiency of the bacterial non-ribosomal peptide indigoidine relies on the respiratory metabolic state in S. cerevisiae.Microb. Cell Fact. 2018; 17: 193Crossref PubMed Scopus (23) Google Scholar] or laboratory evolution [40Sandberg T.E. et al.Laboratory evolution to alternating substrate environments yields distinct phenotypic and genetic adaptive strategies.Appl. Environ. Microbiol. 2017; 83e00410-17Crossref PubMed Scopus (52) Google Scholar] are a few approaches that may be used to address this issue. Besides leading to suboptimal gas transfer effects, imperfect mixing in large-scale bioreactors can impose substrate supply gradients on bioreactor cultures. Recent findings have shown that E. coli can adopt both short- and long-term strategies to withstand stress conditions regarding changing nutrient availability [41Löffler M. et al.Engineering E. coli for large-scale production – strategies considering ATP expenses and transcriptional responses.Metab. Engineer. 2016; 38: 73-85Crossref PubMed Scopus (52) Google Scholar, 42Löffler M. et al.Switching between nitrogen and glucose limitation: unraveling transcriptional dynamics in Escherichia coli.J. Biotechnol. 2017; 258: 2-12Crossref PubMed Scopus (12) Google Scholar, 43Nieß A. et al.Repetitive short-term stimuli imposed in poor mixing zones induce long-term adaptation of E. coli cultures in large-scale bioreactors: experimental evidence and mathematical model.Front. Microbiol. 2017; 8: 1195Crossref PubMed Scopus (19) Google Scholar, 44Simen J.D. et al.Transcriptional response of Escherichia coli to ammonia and glucose fluctuations.Microb. Biotechnol. 2017; 10: 858-872Crossref PubMed Scopus (27) Google Scholar]. In a scale-down study of penicillin production in Penicillium chrysogenum, the influence of substrate gradients on process performance and cell physiology was investigated by imposing an intermittent feeding regime on a laboratory-scale culture [17de Jonge L.P. et al.Scale-down of penicillin production in Penicillium chrysogenum.Biotechnol. J. 2011; 6: 944-958Crossref PubMed Scopus (71) Google Scholar]. The authors observed that, while the biomass yield remained the same, the production of penicillin was reduced significantly in the intermittently fed cultures and based this observation on a reduction in flux of the pathway which was possibly caused by fluctuations in intracellular ATP and AMP levels. In S. cerevisiae, genome-wide analysis of transcriptional cross-regulation by different environmental parameters under different nutrient limitations (C,N,P,S) in both aerobic and anaerobic laboratory-scale chemostat cultures identified 155 oxygen-responsive genes and several other genes responsive to different macronutritional limitations [45Boer V.M. et al.The genome-wide transcriptional responses of Saccharomyces cerevisiae grown on glucose in aerobic chemostat cultures limited for carbon, nitrogen, phosphorus, or sulfur.J. Biol. Chem. 2003; 278: 3265-3274Crossref PubMed Scopus (264) Google Scholar, 46Tai S.L. et al.Two-dimensional transcriptome analysis in chemostat cultures: combinatorial effects of oxygen availability and macronutrient limitation in Saccharomyces cerevisiae.J. Biol. Chem. 2005; 280: 437-447Abstract Full Text Full Text PDF PubMed Scopus (120) Google Scholar]. Data obtained from studies mirroring cellular responses to typical large-scale stimuli could be used to derive and validate adequate models for in silico predictions of commercial-scale performance. These predictions could be used to optimize the bioreactor hardware design configuration as well as the process operating conditions. This would mitigate potential reductions in performance and TRY metrics and could be used to derive design criteria to engineer production strains for improved robustness during commercial-scale manufacturing. Recently, a metabolically structured kinetic model developed for penicillin production by P. chrysogenum showed sufficient accuracy to enable the simulation of dynamic metabolic processes at relevant timescales for bioreactor mixing (which can range from seconds to minutes) in commercial-scale fed-batch cultivations [47Tang W. et al.A 9-pool metabolic structured kinetic model describing days to seconds dynamics of growth and product formation by Penicillium chrysogenum.Biotechnol. Bioeng. 2017; 114: 1733-1743Crossref PubMed Scopus (32) Google Scholar]. Nitrogen-responsive regulation can be another source of culture heterogeneity or variation, as the C/N balance is known to impact metabolic output. One study in E. coli identified that relA, the key regulator responsible for the synthesis of signal molecule guanosine tetraphosphate (ppGpp) during the stringent response, is also activated during nitrogen starvation and thus the two major bacterial stress responses are coupled to manage conditions of nitrogen limitation [48Brown D.R. et al.Nitrogen stress response and stringent response are coupled in Escherichia coli.Nat. Commun. 2014; 5: 4115Crossref PubMed Scopus (105) Google Scholar]. Michalowski et al. combined mechanistic knowledge from this and several other studies to engineer a strain of E. coli which maintained a constant ppGpp pool independently of nutritional supply, thus allowing for increased intracellular pyruvate accumulation and greater metabolic flux towards a desired final product [49Michalowski A. et al.Escherichia coli HGT: engineered for high glucose throughput even under slowly growing or resting conditions.Metab. Engineer. 2017; 40: 93-103Crossref PubMed Scopus (30) Google Scholar]. Stochasticity of gene expression is commonly considered the main source of phenotypic heterogeneity in a microbial population, due to cell-to-cell variation in promoter expression, unequal transporter distribution, or bet-hedging after a diauxic shift [2
Referência(s)