The dose was fractionated over 50 raw frames. Currently there is no H7N9-specific prevention or treatment for humans. From a 2013 H7N9 convalescent case occurred in CDKN1C Hong Kong, we isolated four H7 hemagglutinin (HA)-reactive monoclonal antibodies (mAbs) by solitary B cell cloning, with three mAbs directed to the HA globular head website (HA1) and one to the HA stem region PHA-665752 (HA2). Two clonally related HA1-directed mAbs, H7.HK1 and H7.HK2, potently neutralized H7N9 and protected mice from a lethal H7N9/AH1 challenge. Cryo-EM constructions revealed that H7.HK1 and H7.HK2 bind to a 14-centered surface partially overlapping with the antigenic site D of HA1 and disrupt the 220-loop that makes hydrophobic contacts with sialic acid within the adjacent protomer, thus affectively blocking viral access. The more potent mAb H7.HK2 retained full PHA-665752 HA1 binding and neutralization capacity to later H7N9 isolates from 2016C2017, which is consistent with structural data showing the antigenic mutations of 2016C2017 from your 2013 H7N9 only occurred in the periphery of the mAb epitope. The HA2-directed mAb H7.HK4 lacked neutralizing activity but protected mice from your lethal H7N9/AH1 challenge when engineered to mouse IgG2a enabling Fc effector function in mice. Used in combination with H7.HK2 at a suboptimal dose, H7.HK4 augmented mouse protection. Our data exhibited an allosteric mechanism of mAb neutralization and augmented protection against H7N9 when a HA1-directed neutralizing mAb and a HA2-directed non-neutralizing mAb were combined. INTRODUCTION H7N9 is an avian influenza A group 2 virus first transmitted to humans in the spring of 2013 most likely through live poultry market exposure in China (1C3). The virus reemerged in the fall of 2013 and in the winter of later PHA-665752 years, with the largest epidemic reported as the 5th wave in 2016C2017 (4C6). Though there is limited evidence for human-to-human transmission, few mutations in the hemagglutinin (HA) gene of the virus might be sufficient to overcome its inefficiency for human transmission (7C10). Like other influenza virus infections, the most common treatment against H7N9 is the neuraminidase inhibitor oseltamivir, but oseltamivir-resistant strains have emerged (11C13). Intravenous (i.v.) zanamivir, though not clinically approved, has been used on a compassionate basis in some severe cases because of favorable pharmacokinetics and susceptibility against oseltamivir-resistant strains (14, 15); however, the effectiveness of i.v. zanamivir against H7N9 has not been validated in large clinical trials. Despite the use of neuraminidase inhibitors, H7N9 case-fatality rate remains higher than 30%, and currently there is no licensed vaccine against H7N9 for humans. An endonuclease inhibitor baloxavir marboxil, targeting the virus polymerase acid, guarded mice from lethal H7N9 challenge (16), but treatment for human H7N9 contamination with this inhibitor has not been reported. Concerns for a major outbreak and lack of effective treatment warrant further studies to identify and develop human monoclonal antibodies (mAbs) with potent antiviral functions against H7N9. Because HA is the major target for influenza neutralizing antibodies, H7-reactive human mAbs have been isolated and characterized from H7N9 acute infections (17), convalescent cases (18), and H7N9 experimental vaccinees (19C21). The binding sites of these mAbs have been mapped to the HA globular head (HA1) and stem (HA2) domains. A subset of HA1-directed mAbs potently neutralized H7N9 and guarded mice from H7N9 challenges at doses of 0.3, 1, 5 mg/kg or higher (17C20). These HA1-directed mAbs typically neutralized H7N9 by direct interference with or around the receptor (sialic acid) binding site (17, 19, 22). These epitopes correspond to the antigenic sites of A and B as previously mapped on the surface of H3 HA (23C25). Of note, significant antigenic drift has been documented in the HA gene of 2016C2017 H7N9 from the initial 2013 isolates (17, 26, 27). For example, PHA-665752 Huang isolated 17 neutralizing mAbs from four cases infected in 2013 and 2014, yet only three of these mAbs were active against viral isolates from 2016C2017 (17). A broad mAb FluA-20 targeting the HA1 trimer interface did not mediate neutralization backbone (30), we generated the H7N9 2013 pseudotype particles and tested mAb neutralization by a luciferase readout from single round contamination of MDCK cells (Fig. 1C, left). H7.HK1 and.