Two biotechnology companies, located some 15 miles apart in the SanFrancisco Bay Area, this week independently claimed the Holy Grailthat immunologists have been seeking since 1980.CellGenesys, Inc. (CGI), of Foster City, and GenPharm International(GPI), of Mountain View, each reported, in print, creation of atransgenic mouse that produces fully human monoclonal antibodies.GPI's paper in today's Nature (April 28) bore the title: "Antigen-specific human antibodies from mice comprising four distinct geneticmodifications." And in the May issue of Nature Genetics (an editoriallyseparate spin-off of Nature), CGI announced: "Antigen-specific humanmonoclonal antibodies from mice engineered with human Ig heavy andlight chain YACs."Ever since 1975, when Nobelists George Koehler and Csar Milsteinbrought monoclonal antibodies into existence, writers and researchershave been calling them "silver bullets" because they home inunerringly on their specific target antigen. But like lead bullets,monoclonals sometimes backfire. To this day, the most practical way toproduce them is by immunizing a mouse against the desired targetantigen, then generating the antibody in hybridoma cells.So, to this day, firing the silver bullet at a tumor or autoimmuneantigen, for example, is ruled out because the antibody's payloadincludes residual mouse proteins that might sensitize a patient'simmune system to reject a second round of the otherwise idealimmunotherapy.Now, almost unbeknowst to each other, GPI and CGI havesimultaneously equipped transgenic mice with the kit of tools that thehuman immune system uses to fabricate antibodies on demand, toneutralize a threatening antigen. At the same time, they have knockedout the transgenic animal's own murine antibody tool kit."What we did, over-all," GPI's vice-president of research anddevelopment, Robert Kay, told BioWorld, "was to perform a differentgene manipulation in each of four mouse strains, then crossbreed theirprogeny to obtain a human-monoclonal-making animal that GPI hastrademarked as "HuMAb-Mouse."Employing an almost identical strategy, CGI engineered a similartransgenic rodent it calls "Xenomouse." That firm's director ofmolecular immunology is Aya Jakobovitz, principal author of theHuman Genetics paper. She told BioWorld, "About 15 years ago, whenbiotechnology started, we were all excited that we were able toengineer bacteria to produce human proteins. Now we are looking at atechnology that allows us to genetically engineer mice to producethousands of human proteins."Both firms begin by using gene targeting _ homologousrecombination _ to deprive a first mouse strain of its genes for makingantibody heavy chains. A second strain loses its genes for the kappalight chain. A third receives the human gene locus for turning outheavy chains. The fourth gets the locus for human kappa chains.Cross-breeding all four together results in a HuMAb-Mouse inMountain View; and a Xenomouse in Foster City.To insert their bulky packages of DNA sequences, 100 or morekilobase pairs long, into the mouse embryos, CGI use YACs _artificial yeast chromosomes. GPI employs many mini-loci constructedof selected fragments of the human genes. The transgene comprises,besides crucial control elements, the V, D, J and C segments thatencode the antibody's variable and constant regions. Confronted with astrange antigen, these genes reassemble in a virtually infinitepermutation to come up with a matching monoclonal active site.The main difference between the GenPharm and Cell Genesys reportsis that the former's HuMAb-Mouse, Kay said, has reacted with a trulyhuman antigen, namely CD4. This led first to a primary, relativelyweak IgM-class antibody, which then matured into a secondary-response, class IgG "disease-fighting antibody," Kay said. Besides itsenhanced affinity and specificity, this IgG molecule is completely freeof any leftover mouse proteins.CGI's paper reported a comparable experiment, but pitting itsXenomouse's monoclonals against a non-human tetanus toxinfragment. This yielded a "fully human" primary-response IgMmonoclonal. Jakobovitz's team has since produced human antibodiesagainst a human protein, IgE antibody.Both firms are counting the months until their presumably rejection-proof, all-human antibodies can begin human clinical trials. "We'relooking at late 1995," said Kay, "to find out how these therapeutics doin immunosuppression, eventually colon and ovarian tumors."Eventually, the mouse-made all-human antibodies should be able toneutralize the immune-system cells that attack the body's own tissues."Some of the areas we are looking at," said Jakobovitz, are the chronicdiseases _ inflammation, autoimmunity, organ transplantation, cancer_ where the big advantage of using all-human antibodies is that theycan be administered repeatedly." It's too early to predict when CGI willbe ready for clinical trials, she added "but probably in the next twoyears."To reach that goal, she explained, the Xenomouse technology is beingtransferred for further preclinical studies to Xenotech L.P., a 50-50worldwide joint venture between CGI and JT Immunotech USA, amedical subsidiary of Japan Tobacco. (See BioWorld Today, Feb. 10,1993, p.1)

-- David N. Leff Science Editor

(c) 1997 American Health Consultants. All rights reserved.