Measurement of uptake curves and adsorption isotherms by automated microscale chromatography pipette tips.

Automated microscale chromatography using pipette tips packed with microliters of resin is seen as an increasingly attractive option for high-throughput screening of purification conditions during early phase biopharmaceutical development. Two types of data that can be produced by these studies are uptake curves and isotherms, providing valuable fundamental separation data to assist with the prediction of larger-scale performance. This chapter will describe an operating protocol for this type of experiment using the example of ovine polyclonal antibodies binding to a multimodal weak cation exchange resin.

Extreme scale-down approaches for rapid chromatography column design and scale-up during bioprocess development.

Chromatography is a ubiquitous protein purification step owing to its unparalleled ability to recover and purify molecules from highly complex industrial feedstocks. Traditionally, column development has been driven by a combination of prior experience and empirical studies in order to make the best choices for design variables. Economic constraints now demand that companies engage with a more systematic exploration of a chromatographic design space. To deliver this capability using purely conventional laboratory columns, however, would require considerable resources to identify practical and economical operating protocols. Hence, recently there has been increased use of extremely small-scale devices that gather data quickly and with minimal feed requirements. Such information can be obtained either during early development for screening and trend-finding purposes or later for more accurate scale-up prediction. This chapter describes some of the key drivers for these small-scale studies and the different types of extreme scale-down chromatography formats that exist and illustrates their use through published case studies. Since extreme scale-down experimentation is linked to fundamental mechanistic engineering approaches as well, the utility of these in delivering process understanding is also highlighted.

Modelling approaches for bio-manufacturing operations.

Fast and cost-effective methods are needed to reduce the time and money needed for drug commercialisation and to determine the risks involved in adopting specific manufacturing strategies. Simulations offer one such approach for exploring design spaces before significant process development is carried out and can be used from the very earliest development stages through to scale-up and optimisation of operating conditions and resource deployment patterns both before and after plant start-up. The advantages this brings in terms of financial savings can be considerable, but to achieve these requires a full appreciation of the complexities of processes and how best to represent them mathematically within the context of in silico software. This chapter provides a summary of some of the work that has been carried out in the areas of mathematical modelling and discrete event simulations for production, recovery and purification operations when designing bio-pharmaceutical processes, looking at both financial and technical modelling.

Strategic Assay Selection for analytics in high-throughput process development: case studies for downstream processing of monoclonal antibodies.

During bioprocess development a potentially large number of analytes require measurement. Selection of the best set of analytical methods to deploy can reduce the analytical requirements for process investigation but currently relies on application of heuristics. This paper introduces a generic methodology, Strategic Assay Selection, for screening a large number of analytical methods to produce a subset of analytics that best suit high-throughput studies. The methodology uses a stochastic ranking approach where analytics are ranked based on their holistic performance in a set of criteria. Strategic Assay Selection can be used to help minimizing the impact of analytics in the generation of bottlenecks often encountered during high-throughput process development studies. This is illustrated by using a typical downstream purification process for a monoclonal antibody product. A list of assays is populated for routinely measured analytes across the different units of operation followed by the calculation of their performances in four criteria. The methodology is then applied to select analytics testing for three analytes and the results are analyzed to demonstrate how it can lead to the selection of analytical methods with the most favorable features.

The hybrid experimental simplex algorithm--an alternative method for 'sweet spot' identification in early bioprocess development: case studies in ion exchange chromatography.

The capacity to locate efficiently a subset of experimental conditions necessary for the identification of an operating envelope is a key objective in many studies. We have shown previously how this can be performed by using the simplex algorithm and this paper now extends the approach by augmenting the established simplex method to form a novel hybrid experimental simplex algorithm (HESA) for identifying 'sweet spots' during scouting development studies. The paper describes the new algorithm and illustrates its use in two bioprocessing case studies conducted in a 96-well filter plate format. The first investigates the effect of pH and salt concentration on the binding of green fluorescent protein, isolated from Escherichia coli homogenate, to a weak anion exchange resin and the second examines the impact of salt concentration, pH and initial feed concentration upon the binding capacities of a FAb', isolated from E. coli lysate, to a strong cation exchange resin. Compared with the established algorithm, HESA was better at delivering valuable information regarding the size, shape and location of operating 'sweet spots' that could then be further investigated and optimized with follow up studies. To test how favorably these features of HESA compared with conventional DoE (design of experiments) methods, HESA results were also compared with approaches including response surface modeling experimental designs. The results show that HESA can return 'sweet spots' that are equivalently or better defined than those obtained from DoE approaches. At the same time the deployment of HESA to identify bioprocess-relevant operating boundaries was accompanied by comparable experimental costs to those of DoE methods. HESA is therefore a viable and valuable alternative route for identifying 'sweet spots' during scouting studies in bioprocess development.

Strategic assay deployment as a method for countering analytical bottlenecks in high throughput process development: case studies in ion exchange chromatography.

High throughput approaches to facilitate the development of chromatographic separations have now been adopted widely in the biopharmaceutical industry, but issues of how to reduce the associated analytical burden remain. For example, acquiring experimental data by high level factorial designs in 96 well plates can place a considerable strain upon assay capabilities, generating a bottleneck that limits significantly the speed of process characterization. This article proposes an approach designed to counter this challenge; Strategic Assay Deployment (SAD). In SAD, a set of available analytical methods is investigated to determine which set of techniques is the most appropriate to use and how best to deploy these to reduce the consumption of analytical resources while still enabling accurate and complete process characterization. The approach is demonstrated by investigating how salt concentration and pH affect the binding of green fluorescent protein from Escherichia coli homogenate to an anion exchange resin presented in a 96-well filter plate format. Compared with the deployment of routinely used analytical methods alone, the application of SAD reduced both the total assay time and total assay material consumption by at least 40% and 5%, respectively. SAD has significant utility in accelerating bioprocess development activities.

Integrated use of ultra scale-down and financial modeling to identify optimal conditions for the precipitation and centrifugal recovery of milk proteins.

This article investigates the integrated application of ultra scale-down (USD) techniques and economic modeling as a means for identifying optimal bioprocess operating conditions. The benefits of the approach are illustrated for the recovery of lactoperoxidase (LPO) from bovine milk. In the process, milk is skimmed to deplete its lipid content, before being subjected to low pH incubation with acetic acid in order to precipitate the primary impurity (casein). Following removal of the solids by disk stack centrifugation, pH adjustment and filtration, cation exchange chromatography is used as a positive mode column step to bind the LPO before it is polished and freeze dried. An economic model of this process was used to identify where greatest product loss occurs and hence where the largest opportunity cost was being incurred. Scale-down analysis was used to characterize the influence of the critical steps, identified as precipitation and centrifugation, upon LPO recovery. A mathematical model was used to relate the centrifuge feed flowrate and discharge interval to the supernatant yield, and it was shown that increasing the centrifugal solids residence time achieved superior solids de-watering and so higher product yield, although this also increased the overall processing time. To resolve this conflict, scale-down data were used again in conjunction with an economic model to determine the most suitable conditions that maximized annual profit and minimized operating costs. The results demonstrate the power of combining USD data with models of economic and process performance in order to establish the best overall operating strategies for biopharmaceutical manufacture.

The simplex algorithm for the rapid identification of operating conditions during early bioprocess development: case studies in FAb' precipitation and multimodal chromatography.

This study describes a data-driven algorithm as a rapid alternative to conventional Design of Experiments (DoE) approaches for identifying feasible operating conditions during early bioprocess development. In general, DoE methods involve fitting regression models to experimental data, but if model fitness is inadequate then further experimentation is required to gain more confidence in the location of an optimum. This can be undesirable during very early process development when feedstock is in limited supply and especially if a significant percentage of the tested conditions are ultimately found to be sub-optimal. An alternative approach involves focusing solely upon the feasible regions by using the knowledge gained from each condition to direct the choice of subsequent test locations that lead towards an optimum. To illustrate the principle, this study describes the application of the Simplex algorithm which uses accumulated knowledge from previous test points to direct the choice of successive conditions towards better regions. The method is illustrated by two case studies; a two variable precipitation example investigating how salt concentration and pH affect FAb' recovery from E. coli homogenate and a three-variable chromatography example identifying the optimal pH and concentrations of two salts in an elution buffer used to recover ovine antibody bound to a multimodal cation exchange matrix. Two-level and face-centered central composite regression models were constructed for each study and statistical analysis showed that they provided a poor fit to the data, necessitating additional experimentation to confirm the robust regions of the search space. By comparison, the Simplex algorithm identified a good operating point using 50% and 70% fewer conditions for the precipitation and chromatography studies, respectively. Hence, data-driven approaches have significant potential for early process development when material supply is at a premium.

An automated packed protein G micro-pipette tip assay for rapid quantification of polyclonal antibodies in ovine serum.

The demands on the biopharmaceutical sector to expedite process development have instigated the deployment of micro-biochemical engineering techniques to acquire manufacturing insight with extremely small sample volumes. In conjunction with automated liquid handlers, this permits the simultaneous evaluation of multiple operating conditions and reduces manual intervention. For these benefits to be sustained, novel ways are now required to accelerate analysis and so prevent this becoming a throughput bottleneck. For example, although Protein G HPLC is used to quantify antibody titres in bioprocess feedstocks, it can be time-consuming owing to the serial nature of its application. Although commercial options are available that can process many samples simultaneously, these require separate, potentially expensive instruments. A more integrated approach is desirable wherein the assay is implemented directly on a robot. This article describes a high-throughput alternative to antibody HPLC analysis which uses an eight-channel liquid handler to control pipette tips packed with 40 µL of Protein G affinity matrix. The linearity, range, limit of detection, specificity and precision of the method were established, with results showing that antibody was detected reliably and specifically between 0.10 and 1.00 mg/mL. Subsequently, the technique was used to quantify the antibody titre in ovine serum, which is used as feed material by BTG PLC for manufacturing FDA-approved polyclonal bio-therapeutics. The mean concentration determined by the tips was comparable to that found by HPLC, but the tip method delivered its results in less than 40% of the time and with the potential for further, substantial time-savings possible by using higher capacity robots.

A microscale approach for predicting the performance of chromatography columns used to recover therapeutic polyclonal antibodies.

A microscale approach is described which screens conditions for recovering polyclonal antibodies from ovine sera by mixed-mode cation-exchange chromatography. The impact of pH and loading buffer salt concentration were assessed using robotically operated 20microL packed pipette tips. Low salt concentrations delivered capacities up to 41mg/mL, while only half this level was obtained at high salt concentrations. Two of the screened conditions were then tested in a 10mL packed bed and overall trends in capacity, yield and purity were found to be retained. Microscale pipette tips thus provided a useful basis for the rapid, approximate definition of a chromatography design space.

Global Sensitivity Analysis for the determination of parameter importance in bio-manufacturing processes.

The present paper describes the application of GSA (Global Sensitivity Analysis) techniques to mathematical models of bioprocesses in order to rank inputs such as feed titres, flow rates and matrix capacities for the relative influence that each exerts upon outputs such as yield or throughput. GSA enables quantification of both the impact of individual variables on process outputs, as well as their interactions. These data highlight those attributes of a bioprocess which offer the greatest potential for achieving manufacturing improvements. Whereas previous GSA studies have been limited to individual unit operations, this paper extends the treatment to an entire downstream process and illustrates its utility by application to the production of a Fab-based rattlesnake antivenom called CroFab [(Crotalidae Polyvalent Immune Fab (Ovine); Protherics U.K. Limited]. Initially, hyperimmunized ovine serum containing rattlesnake antivenom IgG (product), other antibodies and albumin is applied to a synthetic affinity ligand adsorbent column to separate the antibodies from the albumin. The antibodies are papain-digested into Fab and Fc fragments, before concentration by ultrafiltration. Fc, residual IgG and albumin are eliminated by an ion-exchanger and then CroFab-specific affinity chromatography is used to produce purified antivenom. Application of GSA to the model of this process showed that product yield was controlled by IgG feed concentration and the synthetic-material affinity column's capacity and flow rate, whereas product throughput was predominantly influenced by the synthetic material's capacity, the ultrafiltration concentration factor and the CroFab affinity flow rate. Such information provides a rational basis for identifying the most promising strategies for delivering improvements to commercial-scale biomanufacturing processes.

Evaluation of a novel agarose-based synthetic ligand adsorbent for the recovery of antibodies from ovine serum.

This paper evaluates a prototype agarose-based affinity adsorbent utilizing a bound synthetic ligand designed to replace Protein A as an IgG-affinity capture resin and compares its purification characteristics with four commercially available matrices for the recovery of polyclonal antibodies from crude hyperimmune ovine serum. The novel adsorbent was found to show the highest dynamic capacity (29.2 mg/mL) of all matrices under evaluation--30% higher than the other commercial adsorbents evaluated. When using a post-load caprylic acid wash, IgG yields of over 85% and purities of over 90% were achieved consistently over multiple loading cycles. To evaluate bead diffusion, inverted confocal microscopy was used to visualise fluorescent antibody binding on to individual adsorbent beads in real time. The results indicate that the binding characteristics of the prototype adsorbent are similar to those obtained with Protein G Sepharose. This study indicates that the high-capacity prototype matrix is a feasible and potentially cost-effective alternative for the direct capture of antibodies from crude ovine serum and may therefore also be applicable to the purification of other complex industrial feedstocks such as transgenic milk or monoclonal antibodies expressed using recombinant technologies.

A prototype software methodology for the rapid evaluation of biomanufacturing process options.

A three-layered simulation methodology is described that rapidly evaluates biomanufacturing process options. In each layer, inferior options are screened out, while more promising candidates are evaluated further in the subsequent, more refined layer, which uses more rigorous models that require more data from time-consuming experimentation. Screening ensures laboratory studies are focused only on options showing the greatest potential. To simplify the screening, outputs of production level, cost and time are combined into a single value using multi-attribute-decision-making techniques. The methodology was illustrated by evaluating alternatives to an FDA (U.S. Food and Drug Administration)-approved process manufacturing rattlesnake antivenom. Currently, antivenom antibodies are recovered from ovine serum by precipitation/centrifugation and proteolyzed before chromatographic purification. Alternatives included increasing the feed volume, replacing centrifugation with microfiltration and replacing precipitation/centrifugation with a Protein G column. The best alternative used a higher feed volume and a Protein G step. By rapidly evaluating the attractiveness of options, the methodology facilitates efficient and cost-effective process development.

Decision-support software for the industrial-scale chromatographic purification of antibodies.

The high therapeutic and financial value offered by polyclonal antibodies and their fragments has prompted extensive commercialization for the treatment of a wide range of acute clinical indications. Large-scale manufacture typically includes antibody-specific chromatography steps that employ custom-made affinity matrices to separate product-specific IgG from the remainder of the contaminating antibody repertoire. The high cost of such matrices necessitates efficient process design in order to maximize their economic potential. Techniques that identify the most suitable operating conditions for achieving desired levels of manufacturing performance are therefore of significant utility. This paper describes the development of a computer model that incorporates the effects of capacity changes over consecutive chromatographic operational cycles in order to identify combinations of protein load and loading flowrate that satisfy preset constraints of product yield and throughput. The method is illustrated by application to the manufacture of DigiFab, an FDA-approved polyclonal antibody fragment purified from ovine serum, which is used to treat digoxin toxicity (Protherics U.K. Limited). The model was populated with data obtained from scale-down experimental studies of the commercial-scale affinity purification step, which correlated measured changes in matrix capacity with the total protein load and number of resin re-uses. To enable a tradeoff between yield and throughput, output values were integrated together into a single metric by multi-attribute decision-making techniques to identify the most suitable flowrate and feed concentration required for achieving target levels of DigiFab yield and throughput. Results indicated that reducing the flowrate by 70% (from the current level) and using a protein load at the midpoint of the range currently employed at production scale (approximately 200-500 g/L) would provide the most satisfactory tradeoff between yield and throughput.

Purification of antibodies using the synthetic affinity ligand absorbent MAbsorbent A2P.

A protocol for the purification of polyclonal antibodies from ovine serum using the synthetic protein A absorbent MAbsorbent A2P is described. Clarified serum is loaded directly onto the affinity column without prior adjustment and albumin and unwanted serum components are washed from the column using a sodium octanoate buffer before elution of bound antibodies. MAbsorbent A2P was shown to bind approximately 27 mg ml(-1) of polyclonal immunoglobulin under overloading conditions, with eluted IgG purities of >90% and minor levels of albumin (approximately 1%). The anticipated time required to complete the purification protocol is 6-7 h. Although the protocol is similar to methods utilized for antibody purification using chromatography with protein A derived from the cell wall of the microorganism Staphylococcus aureus or protein G from Streptococcus as the affinity ligands, affinity absorbents based on synthetic ligands offer a number of advantages to compounds derived from biological sources, in particular robustness, relatively low cost, ease of sanitization and, in principle, lack of biological contamination.

The integrated simulation and assessment of the impacts of process change in biotherapeutic antibody production.

Growing commercial pressures in the pharmaceutical industry are establishing a need for robust computer simulations of whole bioprocesses to allow rapid prediction of the effects of changes made to manufacturing operations. This paper presents an integrated process simulation that models the cGMP manufacture of the FDA-approved biotherapeutic CroFab, an IgG fragment used to treat rattlesnake envenomation (Protherics U.K. Limited, Blaenwaun, Ffostrasol, Llandysul, Wales, U.K.). Initially, the product is isolated from ovine serum by precipitation and centrifugation, before enzymatic digestion of the IgG to produce FAB and FC fragments. These are purified by ion exchange and affinity chromatography to remove the FC and non-specific FAB fragments from the final venom-specific FAB product. The model was constructed in a discrete event simulation environment and used to determine the potential impact of a series of changes to the process, such as increasing the step efficiencies or volumes of chromatographic matrices, upon product yields and process times. The study indicated that the overall FAB yield was particularly sensitive to changes in the digestive and affinity chromatographic step efficiencies, which have a predicted 30% greater impact on process FAB yield than do the precipitation or centrifugation stages. The study showed that increasing the volume of affinity matrix has a negligible impact upon total process time. Although results such as these would require experimental verification within the physical constraints of the process and the facility, the model predictions are still useful in allowing rapid "what-if" scenario analysis of the likely impacts of process changes within such an integrated production process.