Gary J. Lye

Room-temperature ionic liquids as replacements for organic solvents in multiphase bioprocess operations.

Organic solvents are widely used in a range of multiphase bioprocess operations including the liquid-liquid extraction of antibiotics and two-phase biotransformation reactions. There are, however, considerable problems associated with the safe handling of these solvents which relate to their toxic and flammable nature. In this work we have shown for the first time that room-temperature ionic liquids, such as 1-butyl-3-methylimi- dazolium hexafluorophosphate, [bmim][PF(6)], can be successfully used in place of conventional solvents for the liquid-liquid extraction of erythromycin-A and for the Rhodococcus R312 catalyzed biotransformation of 1, 3-dicyanobenzene (1,3-DCB) in a liquid-liquid, two-phase system. Extraction of erythromycin with either butyl acetate or [bmim][PF(6)] showed that values of the equilibrium partition coefficient, K, up to 20-25 could be obtained for both extractants. The variation of K with the extraction pH was also similar in the pH range 5-9 though differed significantly at higher pH values. Biotransformation of 1,3-DCB in both water-toluene and water-[bmim][PF(6)] systems showed similar profiles for the conversion of 1,3-DCB initially to 3-cyanobenzamide and then 3-cyanobenzoic acid. The initial rate of 3-cyanobenzamide production in the water-[bmim][PF(6)] system was somewhat lower, however, due to the reduced rate of 1,3-DCB mass transfer from the more viscous [bmim] [PF(6)] phase. It was also shown that the specific activity of the biocatalyst in the water-[bmim] [PF(6)] system was almost an order of magnitude greater than in the water-toluene system which suggests that the rate of 3-cyanobenzamide production was limited by substrate mass transfer rather than the activity of the biocatalyst.

Application of in situ product-removal techniques to biocatalytic processes

Biocatalytic processes for the manufacture of small, highly functionalized molecules frequently have limited productivity. A common reason for this is the presence of the reaction products that can cause inhibitory or toxic effects (making poor use of the enzyme) or promote unfavourable equilibria (giving low conversions). In each case, the product needs to be removed as soon as it is formed in order to overcome these constraints and hence increase the productivity of the biocatalytic process. Here, we review the need for in situ product removal and the process research required for its implementation.

Accelerated design of bioconversion processes using automated microscale processing techniques

Microscale processing techniques are rapidly emerging as a means to increase the speed of bioprocess design and reduce material requirements. Automation of these techniques can reduce labour intensity and enable a wider range of process variables to be examined. This article examines recent research on various individual microscale unit operations including microbial fermentation, bioconversion and product recovery techniques. It also explores the potential of automated whole process sequences operated in microwell formats. The power of the whole process approach is illustrated by reference to a particular bioconversion, namely the Baeyer-Villiger oxidation of bicyclo[3.2.0]hept-2-en-6-one for the production of optically pure lactones.

Reactor Operation and Scale-Up of Whole Cell Baeyer-Villiger Catalyzed Lactone Synthesis

The recombinant whole cell biocatalyst Escherichia coli TOP10 [pQR239], expressing cyclohexanone monooxygenase from Acinetobacter calcoaceticus NCIMB 9871, was used in 1.5‐ and 55‐L fed‐batch processes to oxidize bicyclo[3.2.0]hept‐2‐en‐6‐one to its corresponding regioisomeric lactones, (–)‐(1 S,5 R)‐2‐oxabicyclo[3.3.0]oct‐6‐en‐3‐one and (–)‐(1 R,5 S)‐3‐oxabicyclo[3.3.0]oct‐6‐en‐2‐one. By employing a bicyclo[3.2.0]hept‐2‐en‐6‐one feed rate below that of the theoretical volumetric biocatalyst activity (275 μmol·min−1·L−1), the reactant concentration in the bioreactor was successfully maintained below the inhibitory concentration of 0.2–0.4 g·L−1. In this way approximately 3.5 g·L−1 of the combined regioisomeric lactones was produced with a yield of product on reactant of 85–90%. The key limitation to the process was shown to be product inhibition. This process was scaled up to 55 L, producing over 200 g of combined lactone product. Using a simple downstream process (centrifugation, adsorption to activated charcoal, 5‐fold concentration with ethyl acetate elution, and silica gel chromatography), we have shown that the two regioisomeric lactone products could be isolated and purified at this scale.

Microwell Engineering Characterization for Mammalian Cell Culture Process Development

Experimentation in shaken microplate formats offers a potential platform technology for the rapid evaluation and optimization of cell culture conditions. Provided that cell growth and antibody production kinetics are comparable to those found in currently used shake flask systems then the microwell approach offers the possibility to obtain early process design data more cost effectively and with reduced material requirements. This work describes a detailed engineering characterization of liquid mixing and gas–liquid mass transfer in microwell systems and their impact on suspension cell cultures. For growth of murine hybridoma cells producing IgG1, 24‐well plates have been characterized in terms of energy dissipation (P/V) (via Computational Fluid Dynamics, CFD), fluid flow, mixing and oxygen transfer rate as a function of shaking frequency and liquid fill volume. Predicted kLa values varied between 1.3 and 29 h−1; liquid‐phase mixing time, quantified using iodine decolorization experiments, varied from 1.7 s to 3.5 h; while the predicted P/V ranged from 5 to 35 W m−3. CFD simulations of the shear rate predicted hydrodynamic forces will not be detrimental to cells. For hybridoma cultures however, high shaking speeds (>250 rpm) were shown to have a negative impact on cell growth, while a combination of low shaking speed and high well fill volume (120 rpm, 2,000 µL) resulted in oxygen limited conditions. Based on these findings a first engineering comparison of cell culture kinetics in microwell and shake flask formats was made at matched average energy dissipation rates. Cell growth kinetics and antibody titer were found to be similar in 24‐well microtiter plates and 250 mL shake flasks. Overall this work has demonstrated that cell culture performed in shaken microwell plates can provide data that is both reproducible and comparable to currently used shake flask systems while offering at least a 30‐fold decrease in scale of operation and material requirements. Linked with automation this provides a route towards the high throughput evaluation of robust cell lines under realistic suspension culture conditions. Biotechnol. Bioeng. 2010; 105: 260–275.

Framework for the rapid optimization of soluble protein expression in Escherichia coli combining microscale experiments and statistical experimental design.

A major bottleneck in drug discovery is the production of soluble human recombinant protein in sufficient quantities for analysis. This problem is compounded by the complex relationship between protein yield and the large number of variables which affect it. Here, we describe a generic framework for the rapid identification and optimization of factors affecting soluble protein yield in microwell plate fermentations as a prelude to the predictive and reliable scaleup of optimized culture conditions. Recombinant expression of firefly luciferase in Escherichia coli was used as a model system. Two rounds of statistical design of experiments (DoE) were employed to first screen (D‐optimal design) and then optimize (central composite face design) the yield of soluble protein. Biological variables from the initial screening experiments included medium type and growth and induction conditions. To provide insight into the impact of the engineering environment on cell growth and expression, plate geometry, shaking speed, and liquid fill volume were included as factors since these strongly influence oxygen transfer into the wells. Compared to standard reference conditions, both the screening and optimization designs gave up to 3‐fold increases in the soluble protein yield, i.e., a 9‐fold increase overall. In general the highest protein yields were obtained when cells were induced at a relatively low biomass concentration and then allowed to grow slowly up to a high final biomass concentration, >8 g·L−1. Consideration and analysis of the model results showed 6 of the original 10 variables to be important at the screening stage and 3 after optimization. The latter included the microwell plate shaking speeds pre‐ and postinduction, indicating the importance of oxygen transfer into the microwells and identifying this as a critical parameter for subsequent scale translation studies. The optimization process, also known as response surface methodology (RSM), predicted there to be a distinct optimum set of conditions for protein expression which could be verified experimentally. This work provides a generic approach to protein expression optimization in which both biological and engineering variables are investigated from the initial screening stage. The application of DoE reduces the total number of experiments needed to be performed, while experimentation at the microwell scale increases experimental throughput and reduces cost.

Lowering oxygen tension enhances the differentiation of mouse embryonic stem cells into neuronal cells

Embryonic stem cells (ESC) are capable of proliferating indefinitely in vitro whilst retaining their ability to differentiate into cells of every adult lineage. Efficient, high yield processes, which direct differentiation of ESC to specific lineages, will underpin the development of cost‐effective drug screening and cell therapy products. The aim of this study was to investigate whether laboratory oxygen tension currently used for the neuronal differentiation of ESC was suboptimal resulting in inefficient process yields. An adherent monolayer protocol for the neuronal differentiation of mouse ESC (mESC) was performed in oxygen controlled chambers using a chemically defined media over an 8 day period of culture. When exposed to oxygen tensions more appropriate to in vivo neuronal development (2% O2), there was a 34‐fold increase in the yield of viable cells from the differentiation process. Low oxygen tension inhibited cell death during an early phase (48 to 96 h) and toward the end (120 to 192 h) of the process. The percentage of cells expressing neuronal markers was determined by flow cytometry, revealing a small rise in the βIII tubulin and a threefold increase in the MAP2 populations at 2% O2. The total increase in the yield of viable cells expressing neuronal markers was shown to be 55‐fold for βIII tubulin and 114‐fold for MAP2. In conclusion, this study revealed that low oxygen tension can be used to enhance the yield of neuronal cells derived from ESCs and has implications for the development of efficient, cost‐effective production processes.

Design and characterization of a microfluidic packed bed system for protein breakthrough and dynamic binding capacity determination

A 1.5 μL ion exchange chromatography column to accommodate resins used for biopharmaceutical processing has been designed to produce breakthrough curves and to quantify dynamic and maximum protein binding capacities. Channels within a glass chip were fabricated using photolithography and isotropic etching. The design includes a 1 cm long microfluidic column in which compressible, polydispersed porous agarose beads (70 μm mean diameter) were packed using a keystone method where particles aggregate in a narrow channel. The depth of the column is such that two bead layers exist. The fabrication technique used forms Cartesian geometries as opposed to circular cross sections found in standard columns. The voidage was therefore higher than standard values when measured by 3D confocal microscopy. In conjunction with microscopic techniques, the column allows visualization of events within the bed such as adsorption profiles that would otherwise be difficult to observe. In this work, the binding of fluorescently labeled protein during isocratic loading was used to generate breakthrough from the microcolumn. Useful breakthrough curves were achieved using mobile phase velocities from 60 to 270 cm h−1. Calculated dynamic binding capacities were compared well with previously published data on conventional scale columns. The microfluidic chromatography column described here thus allows study of process scale chromatography behavior at scales 20,000 times smaller than in current practice. The work described in this article is representative of the proof of principle of a potentially powerful tool for the generation of microfluidic process bed data for the biopharmaceutical industry.

Studies on the interaction of fermentation and microfiltration operations : Erythromycin recovery from Saccharopolyspora erythraea fermentation broths

Changes in fermentation media not only affect the performance of the fermentation itself (with regard to the kinetics of biomass and product formation and the yields obtained) but also the initial product-recovery operations downstream of the fermentor. In this work, microfiltration experiments to remove Saccharopolyspora erythraea biomass from fermentation broth and to recover erythromycin were carried out using two fundamentally different media; a soluble complex medium (SCM) and an oil-based process medium (OBM). Small-scale batch fermentations of 14-L working volume were carried out in triplicate using both media. Broth samples were taken from each fermentation at regular intervals from the end of the exponential-growth phase onwards. These were then processed using a Minitan II (acrylic), tangential crossflow-filtration module, fitted with a single 60 cm(2) Durapore hydrophilic 0.2 microm membrane, operated in concentration mode. The OBM fermentations produced higher titers of erythromycin but required longer fermentation times due to increased lag phases and slower maximum-growth rates. The OBM also increased the loading on the membrane; at maximum product titers residual oil concentrations of 3 g. L(-1), antifoam concentrations of 2 g. L(-1) and flour concentrations estimated at approximately 10 g/L(-1) were typical. It was found that both the permeate flux and erythromycin transmission were affected by the choice of medium. The OBM had significantly lower values for both parameters (12.8 Lm(-2) h(-1) and 89.6% respectively) than the SCM (35.9 Lm(-2) h(-1) and 96.7% respectively) when the fermentations were harvested at maximum erythromycin titers. Transmission of erythromycin stayed approximately constant as a function of fermentation time for both media, however, for the OBM the permeate flux decreased with time which correlated with an increase in broth viscosity. The relatively poor microfiltration performance of the OBM medium was, however, offset by the higher titers of erythromycin that were achieved during the fermentation. The filtration characteristics of the SCM broth did not show any correlation with either broth viscosity or fermentation time. Image-analysis data suggested that there was a correlation between hyphal morphology (main hyphal length) and permeate flux (no such correlation was found for the OBM broth). Moreover, it has been shown for the OBM broth that the residual flour had a profound effect on the microfiltration characteristics. The influence of the residual flour was greater than that imposed by the morphology and concentration of the biomass. The understanding of the factors governing the interaction of the fermentation and microfiltration operations obtained in this work provides a first step towards optimization of the overall process sequence.

Lipase catalysed resolution of the Lotrafiban intermediate 2,3,4,5-tetrahydro-4-methyl-3-oxo-1H-1,4-benzodiazepine-2-acetic acid methyl ester in ionic liquids: comparison to the industrial t-butanol process

The Candida antarctica lipase B (Novozyme 435) catalysed resolution of 2,3,4,5-tetrahydro-4-methyl-3-oxo-1H-1,4-benzodiazepine-2-acetic acid methyl ester (SB-235349), a key Lotrafiban intermediate, has been investigated in six ionic liquids including [BMIM][PF6] and [BMIM][N(SO2CF3)2]. The initial rate and final yield of the reaction have subsequently been determined in [BMIM][PF6] as a function of initial substrate concentration (5–40 g L−1), temperature (25–100 °C) and initial water content (3–15% H2O w/w). In each case the results have been compared to those obtained for the optimised industrial process operated in t-butanol (88% v/v). Simply replacing the organic solvent with an ionic liquid under otherwise identical reaction conditions reduced the rate of conversion. However, exploiting the increased solubility of the substrate in ionic liquids and the ability to operate at higher temperatures increased the overall rate of reaction by a factor of four while maintaining the same overall yield of 47%. In each case the ee of the product was 99%. Further experiments demonstrated the ability to re-use the enzyme over 10 reaction cycles and suggested that solute mass transfer in ionic liquids might be an issue for reactions carried out at larger scale. Overall the results suggest that ionic liquids can be very favourable reaction media for industrial bioconversion processes, which also overcome many of the safety and environmental concerns of conventional organic solvents.