Rational Design of Antibody Cocktails to Treat Disease Caused By Bordetellae

In spite of near-universal vaccination, pertussis (aka whooping cough) remains a significant public health problem with no specific therapy for established disease. Because disease is not driven by bacterial infection at the time pertussis is diagnosed, antibiotic therapy is often ineffective. High-titer polyclonal antibody preparations (IVIG) have been previously evaluated in mice and humans, with demonstrated ability to protect mice up to seven days after infection but less clear outcomes in infants. We hypothesize that cocktails of anti-toxin antibodies, selected to bind key protective epitopes with high affinity will be more effective at ameliorating disease in a passive immunotherapy setting and these results can guide design of next generation vaccines.

To develop a potential passive immunization therapy, we turned to two well-characterized pertussis-toxin binding monoclonal antibodies developed by the Sato group in the 1980s, 1B7 and 11E6. We have characterized their mechanisms of action, which explains their observed synergy: 11E6 inhibits the toxin-receptor interaction, while 1B7 can bind receptor-bound toxin and re-direct intracellular trafficking of the toxin. These murine antibodies have been humanized via exclusively in silico methods and expressed with human immunoglobulin constant regions in CHO cells Full biochemical characterization shows high binding affinity, increased stability and high in vitro toxin neutralizing activity. The humanized antibodies were evaluated in a mouse model in which the animals were infected with a virulent pertussis strain isolated from a critically ill newborn.  The antibodies, individually and in combination, enabled the mice to continue to gain weight, diminished bacterial colonization in the lungs, and completely mitigated the rise in white blood cell count (WBC) that is characteristic of disease. Importantly, they were superior to the IVIG previously used in human clinical trials. We have further evaluated the antibodies’ ability to protect baboons when administered on day 3 after infection when the white blood cell count is elevated. Here, antibody-treated but not control animals saw a rapid decrease in WBC, coughing and a modest decrease in bacterial colonization.

We conclude that these antibodies are protectiv ein both animal models of disease and that they may be protective in humans as well. We have initiated GMP manufacturing processes in anticipation of IND filing in 2015.