E. coli offers short-cut to large-scale antibody production
direct from Escherichia coli bacterium could knock weeks
off conventional mass-production processes that scale up to
mammalian cells, US researchers believe.
Dr George Georgiou and colleagues at the University of Texas at Austin describe their E-clonal antibody system in Nature Biotechnology (advance online publication: 15 April 2007). As the researchers note, isolating antibodies from hybridomas, immortalised B cells or transgenic mice is "time-consuming, not amenable to parallelisation, and does not allow for the tailoring of properties such as antigen-binding affinity, stability or expression level". To date, however, only combinatorial libraries of scFv or FAb antibody fragments have been screened successfully in micro-organisms. Unlike full-length IgG (immunoglobulin G) proteins, these antibody fragments are monovalent and, due to their size and lack of an Fc domain, exhibit very short serum persistence in animals. Earlier studies have shown that immunoglobulin Fc domains as well as functional, full-length IgG antibodies can be expressed in E. coli, the researchers point out. One such experiment achieved high yields of an anti-aTF antibody by expressing the IgG heavy and light chains, each fused to the STII leader peptide, from separate promoters. The bacterially manufactured (and therefore aglycosylated) IgG was shown to bind to the antigen with affinity identical to glycosylated antibodies, as well as exhibiting serum persistence in primates that was indistinguishable from that of fully glycosylated mammalian antibodies. However, the utility of this system for full-length antibodies with specificities for different antigens has not been demonstrated, the researchers added. The University of Texas team set out to develop a library screening system in which IgG did not have to be fused to a display polypeptide, so that isolated clones could be used for preparative antibody production; and high-affinity binders could be isolated quantitatively. They could achieve the latter goal through flow cytometric techniques such as the Anchored Periplasmic Expression (APEx) system developed in the university's laboratories for antibody affinity maturation. However, this required fusion of the library protein to a bacterial inner-membrane anchoring domain. The researchers got around this problem by capturing IgG antibodies expressed in the periplasm of E. coli with an Fc-binding fusion protein (NlpA(1-6)-ZZ), which in turn was tethered to the inner membrane of the bacterium. Full-length heavy and light IgG chains were secreted into the periplasm, where they assembled into aglycosylated antibodies that were captured by the NlpA(1-6)-ZZ protein. Following permealisation of the bacterium's outer membrane, spheroplast clones expressing the E-clonal antibodies, which specifically recognised fluorescently labelled antigen, were selected by flow cytometry. The technique was used to isolate a number of IgGs with nanomolar affinities for the protective antigen of Bacillus anthracis, a bacterium that causes anthrax. The isolation of E-clonal antibodies "combines the many advantages of full-length IgGs with the speed and versatility of protein isolation from combinatorial libraries," the researchers commented. The avidity and stability of full-length immunoglobulins help to isolate binders that "may not be possible to discover using monovalent antibody fragments," they note. Moreover, the technique circumvents the need to isolate antibody fragments and then construct vectors for the expression of full-length immunoglobulins in mammalian cells. From a manufacturing standpoint, the University of Texas team added, expression of recombinant proteins in E. coli is "a robust technology". Importantly, they said, glycosylated IgG produced in bacteria can be used directly to test the effect of antigen neutralisation in animal models, as glycosylation does not exert a major influence on serum persistence.
