SEARCH

SEARCH BY CITATION

Keywords:

  • monoclonal antibody;
  • charge variants;
  • at-line monitoring;
  • HPLC;
  • bioreactor

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods and Materials
  5. Results and Discussion
  6. Conclusions
  7. Acknowledgments
  8. Literature Cited

One major challenge currently facing the biopharmaceutical industry is to understand how MAb microheterogeneity affects therapeutic efficacy, potency, immunogenicity, and clearance. MAb micro-heterogeneity can result from post-translational modifications such as sialylation, galactosylation, C-terminal lysine cleavage, glycine amidation, and tryptophan oxidation, each of which can generate MAb charge variants; such heterogeneity can affect pharmacokinetics (PK) considerably. Implementation of appropriate on-line quality control strategies may help to regulate bioprocesses, thus enabling more homogenous material with desired post-translational modifications and PK behavior. However, one major restriction to implementation of quality control strategies is the availability of techniques for obtaining on-line or at-line measurements of these attributes. In this work, we describe the development of an at-line assay to separate MAb charge variants in near real-time, which could ultimately be used to implement on-line quality control strategies for MAb production. The assay consists of a 2D-HPLC method with sequential in-line Protein A and WCX-10 HPLC column steps. To perform the 2D-HPLC assay at-line, the two columns steps were integrated into a single method using a novel system configuration that allowed parallel flow over column 1 or column 2 or sequential flow from column 1 to column 2. A bioreactor system was also developed such that media samples could be removed automatically from bioreactor vessels during production and delivered to the 2D-HPLC for analysis. With this at-line HPLC assay, we have demonstrated that MAb microheterogeneity occurs throughout the cell cycle whether the host cell line is grown under different or the same nominal culture conditions. © 2013 American Institute of Chemical Engineers Biotechnol. Prog., 30:249–255, 2014


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods and Materials
  5. Results and Discussion
  6. Conclusions
  7. Acknowledgments
  8. Literature Cited

Monoclonal antibodies (MAbs) are proteins with a high specificity for a target molecule known as an antigen. This specificity makes MAbs ideal treatments for a variety of diseases ranging from rheumatoid arthritis, multiple sclerosis, heart disease and cancer. As such, MAbs comprise the largest sector of the biopharmaceutical industry with more than 30 FDA approved therapies currently on the market and global sales of $20.3 billion in 2011.[1]

Though MAbs can be used to treat a variety of diseases effectively, their production costs can make them prohibitively expensive. In recent years, there has been much debate over the cost effectiveness of such treatments.[2-7] For example, Bevacizumab, a monoclonal antibody used to treat metastatic colorectal cancer, has an estimated treatment cost of $39,614 per patient. In 2005, the potential lifetime absolute costs to the Canadian health care system for Bevacizumab treatment were estimated at $299 million, an increase in expenditure of 21% over conventional non-MAb treatments.[2] In 2010, the total cost to treat patients with colorectal cancer in the US was $14.4 billion and is estimated to increase to $16.6 billion by 2020.[8] With such high treatment costs for colorectal cancer, some national health care systems have found that Bevacizumab is not a cost effective treatment, and thus Bevacizumab is not available to patients who might have otherwise been helped by the drug.[3] Similar analyses have found that other MAbs are also not cost-effective treatments.[7, 9, 10]

Biosimilars or generics could reduce the treatment costs associated with monoclonal antibody therapeutics and other biopharmaceuticals and hence make these treatments more accessible. However, issues surrounding the quality of biosimilars and generic biopharmaceuticals remain a topic of debate in the industry. The concerns about the quality of these proteins stem from the presence of protein microheterogeneity. Specifically for MAbs, various chemical and enzymatic post-translational modifications are known to cause MAb microheterogeneity in the form of charge variants.[11] Sialylation, deamidation, and C-terminal lysine cleavage are a few of the chemical and enzymatic post-translational modifications that generate acidic variants of MAbs. Basic variants can result from terminal galactose, C-terminal lysine as well as glycine amidation and tryptophan oxidation. Furthermore, it is known that slight changes to culture conditions can significantly alter intracellular post-translational modifications and result in extreme protein microheterogeneity within batches and from batch to batch of a protein.[12]

One of the major challenges currently facing the biopharmaceutical industry involves understanding how MAb microheterogeneity, and specifically how the presence of charge variants affect therapeutic efficacy, potency, immunogenicity, and clearance. Deliberately modifying the pI of an antibody by approximately one pI unit or more has been shown to alter pharmacokinetics (PK) of an intact MAb.[13, 14] In addition, the relative abundance of basic charge variants has been shown to vary by as much as 20% in commercial lots of some biopharmaceuticals, leading to a 30% difference in antibody-dependent cellular cytotoxicity potency between lots.[15]

On-line quality control strategies implemented appropriately may help to regulate bioprocesses, thus enabling the production of more homogenous material with desired post-translational modifications and PK behavior. However, to date, no online quality control strategies exist for MAb charge variants, or any other biopharmaceutical quality attribute for that matter, mainly due to the limited availability of techniques for obtaining online or at-line measurements of these attributes.

In recent years, more efficient product quality assays, which significantly reduce hands-on sample preparation and overall assay time, have been developed.[16-18] In addition, bioreactor auto-samplers, which can be integrated with some off-line analyzers, are starting to gain popularity since they provide a mechanism to attain at-line monitoring of some attributes, such as metabolite and nutrient concentrations of the media.[19, 20] However, few if any systems are available that can be used to quantify protein quality attributes in real-time.

Here, we describe the development of an at-line assay to isolate MAb charge variants in near real-time, which would ultimately enable the implementation of on-line quality control strategies for MAb production. The assay consists of a two-dimensional high performance liquid chromatography (2D-HPLC) method with sequential in-line Protein A and WCX-10 HPLC column steps. 2D-HPLC was introduced in the mid-1980s, and automated multidimensional liquid chromatography systems have been used to separate complex protein and peptide mixtures.[21-26] However, to the authors' knowledge, none of these multidimensional liquid chromatography systems has the capacity to be integrated with a bioreactor system, and enable real-time monitoring of protein quality attributes. The bioreactor system described in this work has enabled quantification of MAb charge variants in near real-time.

Methods and Materials

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods and Materials
  5. Results and Discussion
  6. Conclusions
  7. Acknowledgments
  8. Literature Cited

Bioreactor system configuration

A parallel bioreactor system (DasGip Biotools, Julich Germany) with 500 ml working volume vessels was integrated with an ARS-M440 (Groton Biosystems, Boxborough MA) and Agilent 1100 Infinity Series HPLC systems through Object Linking and Embedding for Process Control (OPC) software. The DasGip parallel bioreactor scheme is equipped with temperature, pH, DO, and agitation control, as well as a four pump feeding system.

The 1100 Infinity Series HPLC system (Agilent, Santa Clara CA) consists of two quaternary pumps with degassers, auto-sampler (ALS) with thermostat, temperature controlled column compartment for two columns, variable wavelength detector, fluorescence detector, fraction collector, three valve modules and a PC equipped with ChemStation software. The HPLC system configuration has the capacity to use up to eight different buffers or solvents in a single HPLC method. As shown in Figure 1, the valves and tubing of the HPLC system were configured such that flow could originate from either pump 1 or pump 2 and pass in parallel over column 1 or column 2 or in sequence from column 1 to column 2. This novel HPLC system configuration was designed exclusively for the development of the 2D-HPLC method described later in section “At-line 2D-HPLC method to identify MAb charge variants.”

image

Figure 1. Schematic of the valve configuration for the Agilent 1100 2D-HPLC system that allowed for flow that can originate from either pump 1 or pump 2 and pass in parallel over column 1 or column 2 or sequentially from column 1 to column 2. The novel 2D-HPLC configuration enabled at-line detection and separation of MAb charge variants.

Download figure to PowerPoint

The ARS-M440 (Groton Biosystems, Boxborough MA), an automated reactor sampling system, was integrated with the DasGip bioreactor system to enable automated sampling from the bioreactor vessel to the Agilent 1100 Infinity Series 2D-HPLC system as well as automated sample collection for other postrun analysis (e.g., SDS-PAGE or Western analysis). The ARS-M440 has the capacity to remove sample from any port on a bioreactor vessel (maximum of four vessels) and deliver the sample to a maximum of four external analyzers. The design allows for maximum flexibility under software control, and after the initial installation, requires no manual configuration or effort during the bioreactor run. The ARS-M440 system also comes equipped with an OPC client and server that enable real-time data transfer between the analyzers and bioreactor controller software. This unique feature of the ARS allows the implementation of online monitoring and control with analyzers that are not OPC compatible such as the Agilent 1100 Infinity Series HPLC system.

Figure 2 shows a schematic of the material and data flow of the complete bioreactor system with the capacity for at-line HPLC measurements. The ARS-M440 samples the bioreactor vessels, fills a vial in the HPLC ALS with the media sample, then triggers the HPLC to inject the sample and start a pre-specified HPLC method. The ChemStation software was programmed such that the resulting HPLC chromatogram was analyzed automatically post-HPLC method. The peak data generated automatically with the ChemStation software was then transferred in near real-time by the OPC software to the DasGip control software.

image

Figure 2. Material/data flow diagram of bioreactor system with online monitoring and control capabilities. The ARS-M440 samples the bioreactor vessels, fills a vial in the HPLC Auto-sampler with the media sample from the bioreactor vessel, and triggers the HPLC to inject the sample and run a prespecified HPLC method. The ChemStation software is programmed such that the resulting HPLC chromatogram is analyzed automatically post run. The OPC software then transfers the peak data from the ChemStation software to the DasGip control software in near real-time.

Download figure to PowerPoint

HPLC assay materials

The following materials were used to assess MAb charge variants during the development of the 2D-HPLC method: Bio-Monolith Protein A HPLC column, 5.2 × 4.95 mm (Agilent, Santa Clara CA), ProPac WCX-10 weak cation exchange HPLC column, 4 × 250 mm (Thermo Scientific), 49–51% acetic acid solution (Sigma Aldrich), 10 × PBS (Cellgro), glycine HCl (Sigma), Sodium Phosphate monobasic monohydrate (Acros Organics), Sodium Phosphate dibasic heptahydrate (Acros Organics), NaCl (Acros Organics), and Sodium Acetate (Fisher). All buffers were prepared in ultra-pure water and filtered prior to use.

Monoclonal antibody samples

Two IgG1 monoclonal antibodies, MAbA and MAbB, were used during the development and testing of the 2D-HPLC assay. Pure MAbA samples formulated in 5 mM Bis Tris pH 7.0 were used during the development of the HPLC methods. Harvest fluid samples from CHO-K1 cells producing MAbB (gift of Genentech) were used to test and validate the integrated 2D-HPLC assay at-line. Frozen cell stocks were thawed into shake flasks with a 60 ml working volume of CD OPTICHO medium (Life Technologies) supplemented with 4 mM glutamine and shake flasks were placed in a 5% CO2 overlay, 37°C incubator with orbital shaker for agitation. Subsequent passages were supplemented with 25 nM MTX and cultures were passed until enough biomass was generated for shake flask studies or to inoculate the bioreactor vessels with a 500 ml working volume at 7.5 × 105 cells/ml.

Results and Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods and Materials
  5. Results and Discussion
  6. Conclusions
  7. Acknowledgments
  8. Literature Cited

2D-HPLC method to identify MAb charge variants

A 2D-HPLC method was developed to separate MAbs based on variations in charge. As we now discuss, the method required the following two chromatography steps to be performed in series: (1) monoclonal antibody (MAb) purification from harvest fluid using a Protein A affinity and (2) purified MAb separation based on charge variants.

Step 1: MAb Purification from Harvest Fluid

A standard Protein-A affinity column approach was used to purify the MAb from harvest fluid samples. Once the 100 µl harvest fluid sample was loaded onto the protein A column (Biomonolith, Agilent), any MAb protein in the harvest fluid was reversibly bound to the protein-A resin while any nonbinding species was washed from the column with 1 × PBS during the load/wash step. The MAb bound to the Protein-A column was then eluted with 0.5 M Acetic Acid (see Figure 3). The MAb elution peak was collected via a fraction collector such that the purified MAb could then be loaded onto a weak cation exchange column for separation based on charge variants, as we explain next.

image

Figure 3. Step 1 of the 2D-HPLC assay to assess MAb charge variants. Bio Monolithe Protein-A column was used to purify MAbA from harvest fluid samples. Impurities were washed from the sample with 1× PBS and 0.5 M Acetic Acid was used to elute the MAb. The MAb elution peak was collected via a fraction collector such that the purified MAb could subsequently be loaded onto the ProPac WCX-10 HPLC column.

Download figure to PowerPoint

Step 2: Purified MAb Separation Based on Charge Variants

The ProPac WCX-10 weak cation exchange HPLC column was used to separate the purified MAb protein eluent into its charge variants. A gradient-based method was employed with buffers (A) 20 mM Sodium Acetate and (B) 20 mM Sodium Acetate with 1 M NaCl. The separation behavior of charged species was determined at various buffer pH and %B gradients (Figure 4). When buffers at pH 3 were used, MAbA did not bind to the WCX-10 column (See Figures 4a,b). However, the separation of MAbA resulted in two distinct peaks with buffers at pH 6 (see Figure 4c,d). It is important to note the high sensitivity of MAbA variant separation to buffer pH. Even slight variations in buffer pH from pH 4.5 to pH 4 were found to change the WCX-10 chromatogram drastically such that no separation was observed at pH 4 (see Figure 5). Although other gradients were explored to improve the separation of MAbA variants, the +0.5%B per minute gradient, pH 6, resulted is the greatest separation between the two MAb variant peaks (See Figure 4d).

image

Figure 4. Step 2 of the 2D-HPLC assay to assess MAb charge variants. WCX-10 HPLC column separation of MabA using buffers A: 20 mM Sodium Acetate and B: 20 mM Sodium Acetate with 1 M NaCl at various pH and %B gradients: (a) buffer pH 3 with +1.5%B gradient and (b) pH 3 with +0.5%B gradient to (c) pH 6 with +1.5%B gradient and pH 6 with +0.5%B gradient shows that the separation of MAbA variants changed drastically from pH 3 to pH 6. Of the conditions tested, buffers at pH 6 with a gradient of +0.5%B resulted in greatest separation of MabA variants.

Download figure to PowerPoint

image

Figure 5. WCX-10 HPLC column separation of MabA using buffers A: 20 mM Sodium Acetate and B: 20 mM Sodium Acetate with 1 M NaCl at (a) pH 4 and (b) pH 4.5. MAbA variants did not separate at pH 4, but two distinct peaks resulted from separation at pH 4.5. These chromatograms highlight that MAb variant separation via a weak cation exchange HPLC is highly sensitive to buffer pH.

Download figure to PowerPoint

At-line 2D-HPLC method to identify MAb charge variants

To execute the two HPLC methods described in Section “2D-HPLC method to identify MAb charge variants” at-line during MAb production, the protein A column and WCX-10 column HPLC methods were integrated into a single HPLC method such that no manual effort was required at any point during the run. To implement the 2D-HPLC method as a single HPLC method, the Agilent 1100 Infinity Series HPLC system was plumbed to allow flow in parallel over column 1 or column 2 or sequentially from column 1 to column 2 (see Figure 1). With this HPLC system configuration, the elution peak from column 1 could be loaded directly onto column 2 with no manual effort. While this configuration enabled the two individual HPLC column steps to be executed as a single HPLC method, using the buffers that were identified to achieve successful MAb separation as two individual HPLC methods did not enable successful separation of the MAb during the integrated HPLC method.

As discussed in section “2D-HPLC method to identify MAb charge variants,” the MAbA variant separation is highly sensitive to buffer pH, and MAbA did not bind to the WCX-10 column with buffers at pH 3. When the buffers described in section “2D-HPLC method to identify MAb charge variants” were used during the integrated 2D-HPLC method, MAbA did not bind to the WCX-10 column following elution from the protein A column with 0.5 M acetic acid. Prior to executing the HPLC method, the WCX-10 column was equilibrated with 20 mM sodium acetate buffer at pH 6. Despite starting the integrated HPLC method with the WCX-10 column at pH 6, we hypothesize that the low pH of the concentrated 0.5 M acetic acid reduced the apparent pH of the WCX-10 column to less than or equal to pH 3 such that the MAb could not bind to the WCX-10 column. Thus, some rescreening of conditions was needed to enable separation of the charge variants. Ultimately, reducing the acetic acid concentration of the protein A elution buffer to 40 mM and increasing the WCX-10 buffers to pH 7 provided reversible binding of MAbA to the WCX-10 column in the integrated 2D-HPLC method, enabling separation of two distinct peaks, which were observed between retention times 47–53 min (see Figure 6).

image

Figure 6. Resulting chromatogram of 2D separation MAbA via a single HPLC method which combined protein A purification with weak cation exchange separation of charge variants. A brief description of the stages of the HPLC method follows: (a) Harvest fluid is loaded onto the protein A column with 1× PBS and any MAb present in the sample binds; (b) MAb is eluted from Protein A column using 40 mM Acetic Acid and loaded directly onto the WCX-10 column; (c) Purified MAb is then separated and eluted from the WCX-10 column using a +0.125%B gradient with buffers (A) 20 mM Sodium Acetate pH 7 and (B) 20 mM Sodium Acetate pH 7 with 1 M NaCl; (d) The WCX-10 column is regenerated and equilibrated; and finally (e) The Protein A column is regenerated and equilibrated. The appearance of two peaks at retention times between 47 and 53 min suggests that the IgG sample has been separated into its charge variants.

Download figure to PowerPoint

The single integrated 2D-HPLC method consists of a series of five stages described as follows: (a) Harvest fluid was loaded onto the protein A column with 1× PBS (column 1) and any MAb present in the sample binds; (b) MAb was eluted from Protein A column using 40 column volumes of 40 mM Acetic Acid and loaded directly onto the WCX-10 column; (c) Purified MAb was then separated and eluted from the WCX-10 column using a +0.125%B gradient with buffers (A) 20 mM Sodium Acetate pH 7 and (B) 20 mM Sodium Acetate pH 7 with 1 M NaCl; (d) The WCX-10 column was regenerated and equilibrated; and finally (e) The Protein A column was regenerated with glycine-HCl and 0.1 M phosphate buffer containing 1 M NaCl pH 7 then equilibrated with 60 column volumes of 1× PBS (see Figure 6).

To determine assay repeatability, the Chemstation software was calibrated to detect peaks 1 and 2 reproducibly and consistently; then, the integrated 2D-HPLC method was repeated 10 times using the same MabA sample. As shown in Table 1, the peak retention times were consistent, with a 95% confidence interval of 49.0–49.6 min for peak 1 and 50.6–51.2 for peak 2. Also, while some variation was observed in the peak areas among replicates, the ratio between the Peak Area 2: Peak Area 1 was consistent, with a 95% confidence interval of peak ratios between 0.17 and 0.23, thus indicating that the HPLC method is reproducible.

Table 1. Reproducibility of 2D-HPLC Method for MAbA Microheterogeneity Assay Separation (n = 10)
 Retention Time (min)Area (mAu × s)Peak Area Ratio (2:1)
  1. Value in brackets represents the 95% confidence interval.

Peak 149.3 [±0.3]1953 [±147]0.20 [±0.032]
Peak 250.9 [±0.6]397 [±32]

Assessing MAb microheterogeneity at-line

The integrated 2D-HPLC method described in section “At-line 2D-HPLC method to identify MAb charge variants” allows MAbs to be purified directly from an unfiltered harvest fluid sample and separated reproducibly into charged variants. The at-line 2D-HPLC method was used to assess the microheterogeneity of MAbB from shake flask harvest samples. Samples were taken from passages P4-P9 and over the course of a single passage. The 2:1 peak area ratio varied by as much as 6.1% between samples (see Figure 7). The observed variations of 2:1 peak area ratio suggest that slight differences in culture conditions, passage number, and culture duration can influence the peak area ratio. As such, there may be potential to direct the peak area ratio deliberately to a desired value.

image

Figure 7. The 2:1 Peak Area Ratio from at-line 2D-HPLC for determining MAb microheterogeneity. Samples are MAbB Harvest Fluid from various two shake flasks (SF1 and SF2) at different passages (P#) and culture days (D#). Error bars represent the 95% confidence interval (n = 10).

Download figure to PowerPoint

After the integrated 2D-HPLC method was validated for assessing MAb microheterogeneity from shake flask samples, the integrated 2D-HPLC method was implemented at-line with the bioreactor system described in section “Bioreactor system configuration.” CHO-K1 cells producing MAbB were cultivated with 500 mL working volume bioreactor vessels at 37°C, pH 7, 30% DO, and 100 rpm and the integrated 2D-HPLC system was used to assess MAbB microheterogeneity at-line. The separation of MAbB samples from at-line bioreactor media samples resulted in approximately five peaks that can be assigned as either acidic or basic variants relative to the neutral major peak (see Figure 8). The increased number of peaks observed for the separation of MAbB cultivated in the bioreactor versus shake flask cultivation further supports the postulate that culture conditions have a major effect on MAb microheterogeneity.

image

Figure 8. Separation of charge variants of MAbB cultivated with controlled bioreactor using the at-line 2D-HPLC method. MAbB separates into 5 peaks, which is greater than that observed for MabA (see Figure 6) or for MAbB grown in shake flasks. This suggests that MAbB is more heterogeneous than MAb A and the culture conditions can have a major impact to MAb microheterogeneity.

Download figure to PowerPoint

Challenges to implementing at-line 2D HPLC assay for online quality control strategies

The at-line 2D-HPLC assay described in sections “At-line 2D-HPLC method to identify MAb charge variants” and “Assessing MAb micro-heterogeneity at-line” has the potential to be implemented as part of an online quality control strategy during protein production. However, there are some possible challenges that must be considered prior to controller design such as sampling frequency, the time delay between sample extraction and data transmission, and the limit of detection for the assay.

Frequent sampling is preferred for control strategies to improve consistency and efficiency of the controller performance. The time delay of the at-line assay from sample extraction to peak data transmission is 110 min. When taking into account this time delay, MAb charge variants could be sampled within the culture as frequently as every 2 h. However, it is very important to note that sampling too frequently may result in bioreactor media depletion due to the sample volume requirement of the assay. The sample volume extracted from the bioreactor vessel is directly proportional to the length of the sampling line. In the bioreactor system described in section “Bioreactor system configuration,” the ARS-M440 automated bioreactor sampling system extracted a 5 ml sample volume. To prevent media depletion in the 500 ml working volume bioreactor, we chose to sample only once per day. Use of a larger reactor volume or shorter sampling line, however, would allow more frequent sampling of the bioreactor vessel media.

The limit of detection must also be determined for each specific monoclonal antibody prior to successful implementation. Charge variants were distinguished readily using this method for both MAbA and MAbB with media titers as low as 0.02 mg/ml. However, the Agilent ChemStation software required careful calibration prior to each bioreactor run to detect the peaks reliably throughout the culture time. The accuracy of any software's peak detection capabilities improves as media titer increases. Therefore, when using the at-line 2D-HPLC assay described in this work, it may only be possible to implement an online control strategy for MAb charge variants after cells have reached transition phase, and a large enough titer of MAb is produced, such that charge variant peaks can be detected accurately.

Conclusions

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods and Materials
  5. Results and Discussion
  6. Conclusions
  7. Acknowledgments
  8. Literature Cited

One of the major obstacles precluding the development of online quality control strategies for biopharmaceuticals is the lack of availability of real-time measurement capability. Here, we have developed an at-line assay to monitor MAb microheterogeneity during production. The assay required the development of a novel bioreactor system with the capability for interfacing with an at-line 2D-HPLC. The 2D-HPLC assay that was developed to assess MAb microheterogeneity consists of automated sequential Protein A purification of the MAb from harvest fluid samples followed by weak cation exchange separation of the MAb charge variants. To execute the 2D-HPLC method at-line, an 1100 Agilent HPLC system was designed with novel tubing and valve configuration that allowed flow in parallel over column 1 or column 2 or sequentially from column 1 to column 2.

With the at-line assay described in this work, we have demonstrated one of the first near real-time quality measurements for MAbs. Despite the challenges associated with sampling frequency, limit of detection, and time delay, with proper implementation this system can be used to develop and implement novel on-line control strategies that assure quality biopharmaceutical products are produced reproducibly and consistently.

Acknowledgments

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods and Materials
  5. Results and Discussion
  6. Conclusions
  7. Acknowledgments
  8. Literature Cited

The authors gratefully acknowledge the financial support for this work provided by the FDA/ICETECH grant HH2F2232009100101, the NSF- CBET grant 1034213 and the NSF Graduate Research Fellowship- 2008067497. They also acknowledge assistance from Groton Biosystems, DasGip, and Agilent during the establishment of the bioreactor system with at-line HPLC capabilities.

Literature Cited

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods and Materials
  5. Results and Discussion
  6. Conclusions
  7. Acknowledgments
  8. Literature Cited
  • 1
    Aggarwal S. What's fueling the biotech engine-2011 to 2012. Nat. Biotechnol. 2012;30:11911197.
  • 2
    Drucker A, Skedgel C, Virik K, Rayson D, Sellon M, Younis T. The cost burden of trastuzumab and bevacizumab therapy for solid tumours in Canada. Curr Oncol. 2008;15:136142.
  • 3
    Tappenden P, Jones R, Paisley S, Carroll C. The cost-effectiveness of bevacizumab in the first-line treatment of metastatic colorectal cancer in England and Wales. Eur J Cancer 2007;43:24872494.
  • 4
    Cohn DE, Kim KH, Resnick KE, O'Malley DM, Straughn JM. At what cost does a potential survival advantage of bevacizumab make sense for the primary treatment of ovarian cancer? a cost-effectiveness analysis. J Clin Oncol. 2011;29:12471251.
  • 5
    Garrison LP, Veenstra DL. The economic value of innovative treatments over the product life cycle: The case of targeted trastuzumab therapy for breast cancer. Value Health 2009;12:11181123.
  • 6
    Leung HWC, Chan ALF, Leung MSH, Lu CL. Systematic review and quality assessment of cost-effectiveness analysis of pharmaceutical therapies for advanced colorectal cancer. Ann Pharmacother. 2013;47:506518.
  • 7
    Ford J, Cummins E, Sharma P, Elders A, Stewart F, Johnston R, Royle P, Jones R, Mulatero C, Todd R, Mowatt G. Systematic review of the clinical effectiveness and cost-effectiveness, and economic evaluation, of denosumab for the treatment of bone metastases from solid tumours. Health Technol Assess. 2013;17:1+.
  • 8
    Mariotto AB, Yabroff KR, Shao YW, Feuer EJ, Brown ML. Projections of the cost of cancer care in the United States: 2010–2020. J Natl Cancer Inst 2011;103.
  • 9
    Stein JD, Newman-Casey PA, Kim DD, Nwanyanwu KH, Johnson MW, Hutton DW. Cost-effectiveness of various interventions for newly diagnosed diabetic macular edema. Ophthalmology 2013;120:18351842.
  • 10
    Main C, Pitt M, Moxham T, Stein K. The clinical effectiveness and cost-effectiveness of rituximab for the first-line treatment of chronic lymphocytic leukaemia: An evidence review of the submission from Roche. Health Technol Assess. 2010;14:2732.
  • 11
    Khawli LA, Goswami S, Hutchinson R, Kwong ZW, Yang J, Wang X, Yao Z, Sreedhara A, Cano T, Tesar D, Nijem I, Allison DE, Wong PY, Kao Y-H, Quan C, Joshi A, Harris RJ, Motchnik P. Charge variants in IgG1 Isolation, characterization, in vitro binding properties and pharmacokinetics in rats. Mabs 2010;2:613624.
  • 12
    Horvath B, Mun M, Laird MW. Characterization of a monoclonal antibody cell culture production process using a quality by design approach. Mol. Biotechnol. 2010;45:203206.
  • 13
    Sharifi J, Khawli LA, Hornick JL, Epstein AL. Improving monoclonal antibody pharmacokinetics via chemical modification. Q J Nucl Med 1998;42:242249.
  • 14
    Igawa T, Tsunoda H, Tachibana T, Maeda A, Mimoto F, Moriyama C, Nanami M, Sekimori Y, Nabuchi Y, Aso Y, Hattori K. Reduced elimination of IgG antibodies by engineering the variable region. Protein Eng Des Sel 2010;23:385392.
  • 15
    Schiestl M, Stangler T, Torella C, Cepeljnik T, Toll H, Grau R. Acceptable changes in quality attributes of glycosylated biopharmaceuticals. Nat Biotechnol 2011;29.
  • 16
    Chen G, Ha S, Rustandi RR. Characterization of glycoprotein biopharmaceutical products by Caliper LC90 CE-SDS Gel Technology. In Beck A, editor. Glycosylation engineering of biopharmaceuticals: Methods and Protocols. New York: Springer; 2013: Vol. 988, 199209.
  • 17
    Druhmann D, Reinhard S, Schwarz F, Schaaf C, Greisl K, Nötzel T. In Hauser H, editor. Utilizing Roch Cedex bio analyzer for in process monitoring during biotech production, ESACT Meeting on Cell Based Technologies, Vienna, Austria, 2011; BioMed Central Ltd: Vienna, Austria, 2011.
  • 18
    Yin HF, Killeen K. The fundamental aspects and applications of Agilent HPLC-Chip. J Sep Sci. 2007;30:14271434.
  • 19
    Derfus GE, Abramzon D, Tung M, Chang D, Kiss R, Amanullah A. Cell culture monitoring via an auto-sampler and an integrated multi-functional off-line analyzer. Biotechnol Prog. 2010;26:284292.
  • 20
    Scott C. A convergence of new products and technologies changes the game. Biopharm Int Supplement 2010;8:1015.
  • 21
    Takahashi N, Ishioka N, Takahashi Y, Putnam FW. Automated tandem high-performance liquid-chromatographic system for separation of extremely complex peptide mixtures. J. Chromatogr. 1985;326:407418.
  • 22
    Wagner K, Miliotis T, Marko-Varga G, Bischoff R, Unger KK. An automated on-line multidimensional HPLC system for protein and peptide mapping with integrated sample preparation. Anal Chem. 2002;74:809820.
  • 23
    Min JZ, Toyo'oka T, Kurihara T, Fukushima T, Inagaki S. Fully automated two-dimensional high-performance liquid chromatography with electrospray ionization time-of-flight mass spectrometry for the determination of oligosaccharides in glycopeptides after enzymic fluorescence labeling. J Chromatogr A 2007;1160:120127.
  • 24
    Yamashita R, Fujiwara Y, Ikari K, Hamada K, Otomo A, Yasuda K, Noda M, Kaburagi Y. Extracellular proteome of human hepatoma cell, HepG2 analyzed using two-dimensional liquid chromatography coupled with tandem mass spectrometry. Mol Cell Biochem 2007;298:8392.
  • 25
    Kim WD, Tokunaga M, Ozaki H, Ishibashi T, Honda K, Kajiura H, Fujiyama K, Asano R, Kumagai I, Omasa T, Ohtake H. Glycosylation pattern of humanized IgG-like bispecific antibody produced by recombinant CHO cells. Appl Microbiol Biotechnol 2010;85:535542.
  • 26
    Kubelka V, Altmann F, Marz L. The asparagine-linked carbohydrate of honeybee venom hyaluronidase. Glycoconj J 1995;12:7783.