Whooping cough caused by Bordetella pertussis remains a childhood scourge despite widespread vaccination since the 1950's. While the bacteria produces 20 different virulence factors, the best correlate of protection is pertussis toxin (PT), to which immunity is conferred via whole protein immunization or DNA encoding subunit S1. PT is comprised of 5 dissimilar subunits (S1 – S5), with the S1 subunit catalyzing ADP-ribosylation of the Gi regulatory proteins, inhibiting various metabolic and physiological responses in the host cell, while S2-S5 aid in cell adhesion. Hence appropriately engineered anti-pertussis toxin antibodies augur well as therapeutics to complement conventional antibiotic and acellular vaccine therapy. Advances in antibody engineering have facilitated generation of humanized antibodies from murine hybridomas to reduce and eventually obviate human anti-mouse antibody (HAMA) responses elicited by murine components during therapy. Our goal is to construct PT neutralizing humanized antibodies. Murine and humanized versions of 1B7 and 11E6 (binding the S1 and B oligomer respectively) have been cloned, expressed recombinantly (scFv /scAb) and characterized. Humanization by CDR grafting resulted in an affinity loss for both antibodies as measured by surface plasmon resonance, necessitating directed evolution (error prone PCR, phage display technology) to recover affinity. Optimization of Escherichia coli culture conditions improved soluble expression levels of scAbs by 6 (3.24g/l - Hu1B7) and 2.4-fold (6.25g/l - Hu11E6) enabling structural studies. We have successfully grown single crystals of Hu11E6 and efforts to obtain Hu1B7 crystal are underway. X-ray crystallographic data will aid in identification of crucial residues to recover affinity loss, in addition to revealing the structural basis of the PT-neutralizing antibody interactions.