Membrane proteins are important therapeutic targets as they transduce signals into cells, transport molecules, bind to surfaces, and catalyze reactions. However, researchers face challenges when developing antibody-based drugs targeting membrane proteins, such as monoclonal antibody and chimeric antigen receptor (CAR) T cell therapies, because the proteins are embedded within the membrane, leaving only a few surface-exposed epitopes accessible to therapeutics.
To discover potentially therapeutic antibodies, researchers employ antibody display systems where antibody fragment libraries are inserted into cells or phages that then “display” the proteins on their surfaces. Early display systems used microorganisms to express mammalian proteins. Newer mammalian systems improve upon bacteria, yeast, and phage display by producing antibody fragments via endogenous eukaryotic secretion machinery.1 This ensures that mammalian antibodies fold properly and are compatible with downstream mammalian cell production systems.
Typical mammalian display systems can only screen antibody libraries with up to 107 variants due to low transfection efficiency in mammalian cells. This limitation prevents researchers from screening naïve antibody libraries that are orders of magnitude larger. To circumvent this problem, scientists must perform a preliminary screen using another technology such as phage display, which can handle larger libraries. However, this solution introduces additional experimental steps and may isolate antibodies incompatible with mammalian systems.
Another major drawback of mammalian antibody discovery systems is that the cells expressing the antibody library incubate with soluble target antigen, either free in solution or bound to beads. Because the targets are purified and linearized proteins, researchers often isolate antibodies that bind to membrane protein epitopes that are typically hidden in the membrane and conformations that are not physiologically relevant. Ideally, these experiments would only uncover antibodies that bind to extracellular domains in their folded state.
The OXGENE Self-Labelling Integral Membrane (SLIMTM) mammalian display antibody discovery platform solves the problems associated with mammalian display systems and membrane proteins. With this platform, Chinese hamster ovary (CHO) cells are engineered to express the membrane protein antigen. The target antigen embeds in the cell membranes in its native configuration with its post-translational modifications intact, removing the need for purification. Synthetic antibody libraries are then introduced into the engineered CHO cells via highly efficient lentiviral transduction, and the antibody fragments bind to the surface epitopes of their target in their physiological state. The fragments are tagged with human influenza hemagglutinin (HA) for easy isolation by magnetic and FRET-based flow sorting.
Researchers recently used this system to identify novel antibody variable fragments that bound to EpCAM, a cell surface glycoprotein that mediates cell-cell adhesion and can promote tumor growth.2 Using the SLIM system, researchers can screen libraries with more than 109 variants to discover novel binders, perform affinity maturation, and identify candidates for CAR T and other antibody-based therapeutics.
- M. Ho et al., “Isolation of anti-CD22 Fv with high affinity by Fv display on human cells,” PNAS, 103:9637-42, 2006.
- N. Robertson et al., “Development of a novel mammalian display system for selection of antibodies against membrane proteins,” J Biol Chem, 295:18436-48, 2020.