Yeast cell surface display system for determination of humoral response to active immunization with a monoclonal antibody against EpCAM.

Even though an immunogenic formulation of the murine monoclonal anti-EpCAM (epithelian cell adhesion molecule) antibody Mab 17-1A, has been shown to evoke a strong humoral immune response in both, monkey studies and early clinical trials, conventional anti-EpCAM ELISA could not identify anti-EpCAM immune response in relation to treatment with Mab17-1A. In contrast, usage of cellulose membranes prepared by SPOT technology presenting overlapping EpCAM peptides allowed the unequivocal determination of EpCAM related antibodies present in monkeys sera after immunization with IGN101. Based on such contradictory results, it was of high interest to compare obtained data to a different method for better assessment of their possible interpretation. Therefore, in the present studies, some EpCAM peptides, determined as reactive by binding of IgG isolated from sera of treated monkeys on membranes prepared by SPOT technology, were represented on yeast surface using the pYD1 yeast display vector system. Binding of biotinylated IgG from sera was detected with streptavidin-FITC and quantity of binding was determined by FACS measurement. Though using this completely different method, experiments with pre-immune and immune sera of four monkeys exemplarily are comparable to the results obtained by analysis with synthetic peptide arrays.

Generation of bioactive peptides by biological libraries.

Biological libraries are powerful tools for discovery of new ligands as well as for identification of cellular interaction partners. Since the first development of the first biological libraries in form of phage displays, numerous biological libraries have been developed. For the development of new ligands, the usage of synthetic oligonucleotides is the method of choice. Generation of random oligonucleotides has been refined and various strategies for random oligonucleotide design were developed. We trace the progress and design of new strategies for the generation of random oligonucleotides, and include a look at arising diversity biases. On the other hand, genomic libraries are widely employed for investigation of cellular protein-protein interactions and targeted search of proteomic binding partners. Expression of random peptides and proteins in a linear form or integrated in a scaffold can be facilitated both in vitro and in vivo. A typical in vitro system, ribosome display, provides the largest available library size. In vivo methods comprise smaller libraries, the size of which depends on their transformation efficiency. Libraries in different hosts such as phage, bacteria, yeast, insect cells, mammalian cells exhibit higher biosynthetic capabilities. The latest library systems are compared and their strengths and limitations are reviewed.

Peptides derived from a secretory yeast library restore factor VIII activity in the presence of an inhibitory antibody.

The development of autoantibodies against factor VIII represents one of the major complications in the treatment of hemophilia A patients. We have employed a novel library system to obtain peptides that specifically neutralize the interaction between factor VIII and these inhibitors. The random peptides are presented as carboxy-terminal extensions of the eukaryotic initiation factor 5a, an intracellular protein with a molecular mass of 18 kDa. These random peptides formed an unique binding site, as demonstrated by molecular simulations using the computer programs InsightII and GROMACS. The library was screened to identify peptides binding to the murine monoclonal anti-factor VIII antibody ESH8 and to inhibitors derived from patients with factor VIII antibodies. Ten peptides binding to ESH8 were identified. Their specificity was confirmed by displacement assays. Two peptides with the sequences STKTLGRPLHGPAGPVEGGALAGVAEDADLVTAVSGR and YHCKREDLTDRDATCALRQPPQAVRGLGPRVTAVSGR showed the ability to restore the factor VIII activity from 33% up to approximately 90% in functional tests performed in vitro. Three candidates for binding to factor VIII antibodies derived from four different patient's sera were achieved. Three fusion proteins with the peptide sequences PQLGSRRSTTPSLTFQNASWFPAGGPCARSNRG, SGSRQVCKLARSLQPF and WERGRRVGAQVRHARHLVARVLDGAGHQARLTAVNGP bound to inhibitors derived from different patients. Furthermore, two of the obtained fusion proteins with the peptide sequences RHWTALGPAPTHTCADLNYPLLS and WERGRRVGAQVRHARHLVARVLDGAGHQARLTAVNGP did also bind to the monoclonal antibody ESH8. This study demonstrates the potential of this system to identify peptides that inhibit the activity of potent inhibitory antibodies and also shows potential as a method for screening of bioactive peptides.

Generic method for quantification of FLAG-tagged fusion proteins by a real time biosensor.

Availability of rapid quantitative protein-expression analysis is often the bottleneck in high throughput screening applications. A real time biosensor was employed to establish a quantitative assay for FLAG fusion proteins using FLAG-tagged bacterial alkaline phosphatase as standard. A range of FLAG-tagged bacterial alkaline phosphatase concentrations were injected over the anti-FLAG M2 antibody surface of the biosensor and used as standards to determine the concentration of different FLAG-tagged proteins with a molecular mass of 18.1 kDa respectively 49.3 kDa from yeast culture supernatants. The M2 immobilized chip was found to retain binding capacity following regeneration for at least 120 cycles. This real time biosensor method allows the quantitation of proteins from culture supernatants using a calibration curve obtained with a different protein. Further benefits include the short assay time of approximately 5 min, the small amount of sample required (35 microl per injection) and the ability to monitor the binding event in real time.