(48a) Development of a Bacterial System for the Expression of Full-Length Monoclonal Antibodies

Monoclonal antibodies (mAbs) have proven to be an excellent basis for developing diverse and powerful clinical therapeutics. Owing to their complex multimeric structure that requires oxidation of multiple disulfide bonds, mAbs are typically produced in mammalian cell systems. However, full-length mAb expression has been reported in Escherichia coli and required targeting to the periplasmic space, where disulfide bond formation/oxidation is favored. This strategy, while successful, relies on membrane co-localization and assembly of heavy and light chains in the periplasm, which due to its small volume is unlikely to be optimal for high-level expression of large, multi-subunit proteins like mAbs. Moreover, periplasmic expression of full-length mAbs also requires the energy intensive inner membrane translocation of both heavy and light chains, which likely presents another barrier to optimal mAb production.

Expression in the cytoplasm of E. coli can potentially provide a means to circumvent these limitations. However, the glutaredoxin and thioredoxin pathways provide a major obstacle to cytoplasmic expression and assembly of full-length mAbs. In wild type E. coli, these pathways maintain thiol groups on cysteines in their reduced, non-disulfide bonded state. To address this complication, E. coli strain SHuffle™T7 Express was engineered for the cytoplasmic expression of proteins requiring multiple disulfide bonds for proper folding. The SHuffle™T7 Express strain carries mutations in the oxidoreductase genes of the thioredoxin and glutaredoxin pathways (trxB, gor), a growth suppressor mutation (ahpC*), as well as a chromosomal copy of the periplasmic disulfide bond isomerase DsbC lacking its native signal sequence. Collectively, these mutations render the cytoplasm a favorable environment for the correct folding of proteins requiring disulfide bonds, such as mAbs. Here, we report a new expression platform called MAXmAb (Membrane-Autonomous eXpression of monoclonal Antibodies) that utilizes E. coli SHuffle™T7 Express cells for cytoplasmic production of mAbs. Using MAXmAb, full-length immunoglobulin G (IgG) antibodies in three formats - fully mouse, mouse/human chimeric, and humanized - specific for E. coli maltose binding protein (MBP) were expressed and assembled in the cytoplasm. The cytoplasmically-expressed IgGs, which we call cytoclonal antibodies (cytAbs), displayed strong binding activity for MBP that was comparable to that for IgGs expressed from hydridoma cells. Additionally, we reprogrammed the humanized anti-MBP IgG using a variable domain sequence substitution strategy. This resulted in a panel of new cytAbs that exhibited altered antigen specificity, in particular, for Bacillus anthracis protective antigen (PA63), the cardiac glycoside digoxin (Dig), the leucine zipper domain of the yeast transcriptional activator Gcn4 (Gcn4LZ), a peptide derived from the influenza hemagglutinin protein (HAG), and human mitogen activated protein kinase 9 (JNK2). Each of the newly generated cytAbs exhibited high in vitro binding affinity for their respective antigens and comparable expression to the original anti-MBP cytAb.  We also show the potential advantage of the cytoplasmic localization and assembly of light and heavy IgG chains system over the existing periplasmic strategy. Finally, building upon a previous study, we demonstrate the recovery of immunological activity, specifically binding activity for FcγRI/CD64, by site-directed mutagenesis of the CH3 domain of cytAb heavy chains.  Overall, our study opens the possibility of developing E.coli as a robust and rapid platform for the development, discovery, and engineering of mAbs for therapeutic and diagnostic applications.