A combinatorial engineering approach to increase the productivity of CHO cells, and proteomic analysis of cell culture supernatant
Year: 2008
Type: dissertation
Abstract: The optimization of production processes for therapeutic antibodies is a continuing challenge in pharmaceutical biotechnology. It has been demonstrated that vector design and host cell engineering can improve transcriptional and translational efficiency, resulting in the generation of high producer cell lines. However, the introduction of transgenes into production cell lines that regulate protein transport or affect post-translational modifications have been reported in the literature to lead to inconsistent results regarding the improvement of industrial processes. The present study shows that heterologous expression of the spliced form of transcription factor X-box-binding protein 1 (XBP-1(s)) leads to an increase in ER content in adherently growing Chinese hamster ovary (CHO) cells. Furthermore, specific therapeutic antibody productivity was elevated by 60% in CHO cells grown in inoculum suspension cultures. This dose-dependent effect translated into significantly increased overall antibody titers in a fed-batch format where cells were grown in chemically defined serum-free media. Protein-A purified antibody products from mock cells and XBP-1 transfected cells were found to be of comparable quality with regard to glycosylation pattern and physicochemical characteristics. The data demonstrated the potential of XBP-1 engineering to improve mammalian cell culture production processes to yield high amounts of a therapeutic protein product of desired quality. However, only low numbers of cells expressing the transgene at high levels were found during the generation of stable cell lines. Additionally, XBP-1(s) expression in monoclonal cell lines was found to decrease over a prolonged period of seed-stock cultivation. These observations were verified by colony formation assays (CFA) where XBP 1(s) expression led to significantly lower colony counts. In an additional experimental setting it was demonstrated that XBP-1(s) expression enhanced apoptosis in CHO cells. Altogether, the data demonstrated a survival disadvantage upon XBP-1(s) overexpression in engineered CHO cells, hampering a possible commercial cell line development program based on XBP-1 engineered cells. To overcome the XBP-1(s) mediated apoptosis, co-expression of the anti-apoptotic protein x-linked inhibitor of apoptosis (XIAP) and XBP-1(s), combining secretion engineering and anti-apoptosis engineering, was evaluated. The proof-of-concept was demonstrated by colony formation assays (CFA), where concomitant XIAP and XBP-1(s) expression was able to restore colony counts. Moreover, when an mAb-producing CHO cell line was engineered, the bicistronic overexpression of both transgenes showed highest specific productivities and final titers in fed-batch cultures of polyclonal cell lines when compared with control cells. The data demonstrated a path toward optimized mammalian production processes. However, despite the proof-of-concept for the two transgenes approach on a pool level it proved challenging to transfer this idea from a model system to the final stages of industrial cell line development. An additional building block for reproducible high-yielding production processes is sophisticated process monitoring. To be able to identify marker proteins which as a result of their concentration can indicate, for example, transitions between process phases during a production run, an analytical method was developed to investigate changes in the proteome of cell culture supernatant samples. In this approach, the proteome of the supernatant was analyzed by differential two-dimensional gel electrophoresis (2D-DIGE). In order to eliminate interfering effects of the protein product of producer cell lines on the separation method, a mock cell line was generated. This mock cell line was selected to closely simulate production processes with comparable cellular growth characteristics but without product formation. For the concentration of supernatant samples precipitation and ultrafiltration were compared and the filtration method was finally selected. A first analysis of supernatant samples derived from three different process times of a batch culture demonstrated the applicability of the method to separate and finally identify proteins with different abundance in process supernatant samples - a prerequisite for possible process marker proteins. Furthermore, the results of this first study were verified by Western blot analysis for one protein. The established method based on separation of the proteins from the supernatant by 2D-DIGE and subsequent identification by peptide mass fingerprinting using mass spectrometry can now be used as a starting point for further characterization studies. The identification of such process marker proteins is the first step toward next-generation bioprocess monitoring and control and provides an opportunity for further improved high-yielding production processes.
DOI: 10.18419/opus-1784
Author: Becker, Eric
Language: English
Subfield: Molecular Biology
Domain: Life Sciences
Open Access status: green
OpenAlex ID: W1574430039