November 15, 2006 (Vol. 26, No. 20)
Focused Methods to Improve Performance
Production of recombinant therapeutic proteins in mammalian cells represents a significant segment of the pharmaceutical market, and therefore striving for increased productivity of these lines represents a major investment of resources.
Historically the methods for increasing productivity have involved the manipulation of nutritional biochemistry combined with statistical approaches for the analysis of many variables and, to some extent, the collective experiences of the scientists working to improve the process. These methods have served to increase productivity greatly over the past 50 plus years, but are limited due to their inherent inefficiency and lack of predictive capabilities.
In order to design more focused and efficient experiments to test for enhancement of productivity, SAFC (www.sigmaaldrich.com) harnessed the power of both genomic and proteomic approaches to identify relevant pathways within these cell cultures. These pathways can then be either manipulated to drive the cells to produce more or be used as biomarkers to indicate productive formulations in high-throughput systems.
The elucidation of biologically important markers for recombinant protein production is a major emphasis of this type of research. These markers could potentially be used in a variety of ways to improve culture conditions, including active approaches to agonize/antagonize important pathways within a medium formulation or diagnostic approaches indicative of improved conditions, leading to decreased development time. Alternatively, standard overexpression or knock-out paradigms could be used to generate clones with the desired phenotype (figure 1). Due to the vast amount of data generated by omics approaches, many applications will undoubtedly become apparent as we gain more knowledge of the systems and pathways involved in the regulation of basic in vitro cell biology.
The genetic characteristics that define a high-producing animal cell line are surprisingly poorly understood. Current genomic tools such as microarrays allow us to take a holistic approach to studying gene function. With the advent of enhanced sequencing technologies capable of quickly generating entire genomic sequences, we are now able to build custom microarrays specific to any mammalian host cell line.
This wealth of genetic information can quickly be used in a reverse-engineering approach to improve and maintain recombinant protein productivity and for biomarker discoveries that, when combined with predictive modeling algorithms, can be used to rapidly select an optimal performance medium for any desired phenotype.
For over ten years, microarray technologies have proven their utility in life science research. This technology has been successfully applied in cell line engineering to study and identify key genes involved in various pathways between cultures. The information generated from these studies is of considerable value as the intracellular environment can be modified through genetic engineering to alter the pathway in culture in order to extend cell viability and/or maintain protein production.
By empowering microarrays for industrial cell line engineering we are not limited to studying solely gene expression. Chromosomal abnormalities are frequently associated with recombinant CHO cell lines, but it is not clear if they are more or less favorable for the stable expression of the recombinant proteins. Microarray comparative genomic hybridization (array CGH) is a technique that can be used to measure the genome-wide DNA variations associated with stable protein production.
Similarly, little is understood regarding the genetic mechanisms involved with the regulation of transcriptionally active transgenes. A high-producing clone is dependent on the integration site of the expression vector within the host’s genome. Epigenetic silencing of gene expression is thought to occur when the site of integration is associated with heterochromatin. Microarray-based DNA methylation profiling can be used to identify epigenetic signatures associated with CpG islands.
These techniques can be used to study and map favorable integration sites, which exhibit high levels of transcriptional activity without being subjected to silencing modifications. These sites can then be targeted using a site-specific integration system to enhance stable gene expression.
In the past decade, academic and industry researchers have successfully applied proteomic techniques, such as 2-D gel electrophoresis (2-DE), liquid chromatography (LC) and mass spectrometry (MS), to investigate protein-expression profile changes in different cell culture processes and conditions.
These proteomic techniques are useful tools to study mechanisms of action of conventional industrial cell culture methods to boost protein expression, such as hypothermal and hyperosmolality culture conditions. The well-established workflow of protein identification and relative quantification includes 2-DE analysis, spot selection and excision, tryptic digestion, followed by MALDI-TOF and tandem (MS/MS) mass spectrometry.
CHO, NS0 myeloma, and hybridomas have been studied for protein-expression profiles with hundreds of proteins identified to be up- or down-regulated in various functional groups, from structural proteins and endoplasmic reticulum chaperones to metabolic enzymes. These hits provide potential biomarkers and targets for rational cell line engineering approaches.
Compared to genomic methods, proteomic methods are lower throughput and more labor intensive. Some emerging new techniques in protein separation, identification, and quantification provide new tools for increasing throughput. 2D-Difference Gel Electrophoresis (2D-DIGE) labels up to three different protein extract samples with different fluorescent dyes, mixed and separated concurrently by 2-DE. The images from different dyes are merged, and the differences analyzed.
Quantitative MS methods that utilize stable isotope tags, such as ICAT (isotope-coded affinity tags), and isobaric amine specific tags, such as iTRAQ™, allow for relative protein quantity without the labor-intensive process of 2-DE. Up to four samples can be analyzed simultaneously by iTRAQ.
Robotic spot-picking has greatly increased the throughput of protein identification. In addition, regulatory proteins are often less abundant than structural proteins. New approaches such as membrane protein fractionation by surface biotinylation and enriched microsomal fractionation may capture changes in low-abundance protein expression and shed light on pathway studies.
As in other disciplines, analysis and interpretation of proteomic data in cell culture applications should always be coupled with genomic data and adequate biological replication. Follow-up and confirmatory studies using Western blotting and functional assays are essential to rule out false positives in the proteomic hits, leading to more successful identification of biomarkers for cell culture development.
Metabolomics has become an area of increased focus for functional genomics. Metabolomics is defined as the quantitative complement of low-molecular-weight metabolites (metabolome) present in a cell under a given set of physiological conditions.
Relative to the transcriptome and proteome, the individual components of the metabolome have generally more complex functions in the cell. Regarded as complementary to transcriptomics and proteomics, the metabolic network will facilitate genome phenotyping, finally bridging the genotype-to-phenotype gap. Developing databases of metabolite concentrations in cells that are grown in well-defined conditions and integrating metabolomic data meaningfully with data from the other levels of functional-genomic analysis will make a significant contribution to cell culture.
While still in its infancy, metabolomics is already beginning to make a significant impact in the fields of functional genomics, nutrition, plant science, and pharmaceutics. For example, in the functional genomics field, metabolic-profiling techniques have been used to explore the function of genes in a variety of different species. Based on the classified metabolic profiles, functions of some orphan genes have been successfully characterized. By comparison of metabolic profiles between normal samples and diseased samples, metabolic biomarkers have also been identified effectively.
Although the application of metabolomics to cell culture is in an early stage, it has a predictable and remarkable future in this field. With the rapid progress of functional genomics, the knowledge of transcriptomics and proteomics has led to an increased number of tools that can be applied to cell culture.
As a downstream omics application, metabolomics will provide more detailed, direct, and precise information to better predict the behavior of cells in culture. The meaningful integration of metabolic information and the other levels of functional-genomic knowledge will facilitate a better understanding of the biochemical pathways involved in cellular regulation machinery.
There are many possible approaches to gain a better understanding of the black box of in vitro cell biology. Each of these approaches has clear benefits and drawbacks, but in order to generate applicable solutions to increase protein production, each of these approaches will play a role. The large amount of information gleaned will have to be coalesced into plausible solutions that can be tested and applied effectively. This is no small task, given the obvious complexity of the information.
Application of the Data
One commonly overlooked component associated with the combination of these technologies is the downstream bioinformatics platforms that are required to handle such extensive data analysis. Undoubtedly, the manner in which this task in undertaken will play a central role in its outcome. Despite the difficulty of the task, the increased level of knowledge will lead to more efficient pharmaceutical production strategies in the future.
Daniel W. Allison, Ph.D., is principal R&D scientist, Christoph L. Bausch, Ph.D., is senior R&D scientist, Nan Lin, Ph.D., is principal R&D scientist, Min Zhang, Ph.D., is senior R&D scientist, and Kevin J. Kayser, Ph.D., is R&D manager at SAFC Biosciences. Web: www.sigmaaldrich.com. Phone: (314) 289-8496. E-mail: dallison@ sial.com.