All relevant data are within the manuscript and its Supporting Information files. Introduction {#sec001} ============ The emergence of Zika virus (ZIKV) as a vector-borne infectious disease, with a suspected causal link to congenital malformations and neurological disorders, has sparked a global public health emergency. The unprecedented magnitude of the outbreak and the significant health risks to pregnant women have raised an urgent need for a safe and effective vaccine \[[@pntd.0007687.ref001]\]. ZIKV belongs to the genus Flavivirus, a group of arboviruses transmitted mainly by mosquitoes. Since the first isolation of ZIKV in 1947 from a rhesus monkey in the Zika forest of Uganda, it has been reported as endemic in West Africa \[[@pntd.0007687.ref002]--[@pntd.0007687.ref004]\]. ZIKV transmission is usually restricted to regions of tropical Africa and Southeast Asia, and its dispersion to other geographic areas occurred recently \[[@pntd.0007687.ref005]\]. Its arrival in Brazil in 2015 resulted in a large outbreak, the first time ZIKV reached urban areas \[[@pntd.0007687.ref006]\]. The number of infections has since increased to \>1 million cases, and more than 24 countries have been affected, among them French Polynesia, American Samoa, and the Federated States of Micronesia \[[@pntd.0007687.ref007]\]. ZIKV is transmitted to humans mainly via *Aedes* mosquitoes, including the most common and widespread species *Aedes aegypti* (Linnaeus 1762) and *Aedes albopictus* (Skuse 1894). Humans are generally considered the main ZIKV reservoir \[[@pntd.0007687.ref008], [@pntd.0007687.ref009]\]. The rapid spread of ZIKV is related to the capacity of *Ae*. *aegypti* as a vector species and anthropophilic behavior in transmitting ZIKV to people and their ability to disseminate in urban areas and to feed on people as well as other vertebrate hosts \[[@pntd.0007687.ref010]--[@pntd.0007687.ref012]\]. Because of the increasing number of ZIKV cases in humans and associated risks to public health, a potential vaccine against ZIKV is urgently needed. In order to develop a vaccine, it is important to define the correlates of protection for humans and animals. Both cellular and humoral immunity have been observed in humans exposed to ZIKV. However, the relative importance of the cellular and humoral responses and their impact on protection remain to be determined \[[@pntd.0007687.ref013], [@pntd.0007687.ref014]\]. Immune sera from infected persons may be used to evaluate the correlates of protection against ZIKV infection \[[@pntd.0007687.ref015]\]. For example, human neutralizing antibodies against ZIKV correlated with protection of rhesus macaques against a lethal ZIKV challenge \[[@pntd.0007687.ref016]\]. Neutralizing antibodies to ZIKV have also been detected in sera from exposed individuals from different ZIKV-affected countries \[[@pntd.0007687.ref015], [@pntd.0007687.ref017]--[@pntd.0007687.ref020]\], with neutralizing antibody titers in pregnant women being an important correlate of protection \[[@pntd.0007687.ref017]\]. In addition, the presence of neutralizing antibodies in sera of humans who were previously exposed to the virus has been shown to correlate with protective efficacy \[[@pntd.0007687.ref015]\]. Moreover, the level of ZIKV neutralizing antibodies in humans was highly correlated with the reduction of virus titer in blood \[[@pntd.0007687.ref020]\]. Given the importance of the humoral response in protection, several approaches have been explored to develop a vaccine against ZIKV and to boost the immune response, including recombinant proteins, inactivated whole viruses, recombinant virus-like particles (VLPs), live-attenuated or chimeric flaviviruses \[[@pntd.0007687.ref021]\]. In several studies, vaccinated animals and humans were protected against challenge with ZIKV, and in a mouse model a protective role was also shown for the cellular immune response \[[@pntd.0007687.ref014], [@pntd.0007687.ref022]\]. Immunization with a chimeric virus has also provided protection against ZIKV infection in mouse models and can also protect against ZIKV infection in primates \[[@pntd.0007687.ref023]\]. Live-attenuated virus vaccines have been shown to be effective against ZIKV in animal models, and the attenuated vaccine viruses can protect against viremia, viral shedding, and congenital malformations in mice \[[@pntd.0007687.ref022], [@pntd.0007687.ref024]--[@pntd.0007687.ref026]\]. However, for the development of vaccines against arboviruses, vaccines against ZIKV will likely need to induce neutralizing antibodies to protect against both vaccine-attenuated and wild-type virus \[[@pntd.0007687.ref022], [@pntd.0007687.ref024]\]. One vaccine candidate based on ZIKV-specific monoclonal antibodies (mAbs) has been tested in mice \[[@pntd.0007687.ref027]\]. This study showed that one of these mAbs, called mAb114, was able to protect against ZIKV infection in mice, indicating that monoclonal antibodies (mAbs) could be useful for developing ZIKV vaccine candidates. Previously, we have isolated several neutralizing mAbs against ZIKV from a human-immunized mouse. These mAbs bind to the envelope (E) protein in the pre-fusion conformation and neutralize ZIKV in vitro \[[@pntd.0007687.ref028], [@pntd.0007687.ref029]\]. The mAbs that neutralize ZIKV block the viral fusion protein from undergoing a conformational change \[[@pntd.0007687.ref028], [@pntd.0007687.ref029]\]. It is assumed that mAbs may protect against infection by inhibiting virus entry through the fusion activity of the pre-fusion E protein in the viral envelope. In this study, we present a structural analysis of a protective mAb that potently neutralizes ZIKV and determine the mAb-binding epitope. Materials and methods {#sec002} ===================== Cloning, expression, and purification of E proteins {#sec003} --------------------------------------------------- The amino acid sequence of the ZIKV envelope protein (Env, residues 1--501 of ZIKV SPH2015 strain) from the GeneBank database (accession No. KX087101) was used for analysis of the antibody-binding epitope. The three different domains of the E protein (domain I, residues 1--132; domain II, residues 132--329; and domain III, residues 329--501) and the full-length E protein (residues 1--501) were PCR-amplified and cloned into a modified pFastBac1 vector (Invitrogen, San Diego, CA, USA) with an N-terminal myc epitope (MEQKLISEEDL) and a C-terminal His6-tag. The primer sequences for amplification of domain I, domain II, domain III, and the full-length E protein are shown in [S1 Table](#pntd.0007687.s004){ref-type="supplementary-material"}. The genes were subcloned into the pFastBac1 vector between the BamHI and NotI sites (for domain I and domain III) or between the BamHI and NcoI sites (for domain II). The sequence was confirmed by DNA sequencing (Genomic Unit at InVivo Biotech). The construct was transformed into competent DH10Bac cells for baculovirus amplification \[[@pntd.0007687.ref030]\]. Protein expression was performed as described previously \[[@pntd.0007687.ref031]\]. Briefly, the recombinant baculoviruses were generated by transfection of Sf9 cells with the *E*. *coli* DH10Bac harboring the transfer plasmid and the *Autographa californica* multicapsid nucleopolyhedrovirus genome by Lipofectamine (Invitrogen) for expression of the ZIKV Env protein in Sf9 cells (Invitrogen). Sf9 cells were cultured in SF-900 II SFM (Invitrogen) with 10% fetal bovine serum (Invitrogen) at 26°C. Protein expression was induced by adding 50 ng/mL of tetracycline. After 48 h, the cell culture supernatant containing the expressed proteins was collected by centrifugation at 5,000 rpm for 10 min at 4°C. To enrich for secreted protein, the cell supernatant was filtered using a 0.2 μm filter (Millipore). The target protein was then purified using a HisTrap HP column (GE Healthcare, Uppsala, Sweden) equilibrated with His buffer A (20 mM Tris-HCl, pH 8.0, 150 mM NaCl). Expression and purification of mAbs {#sec004} ----------------------------------- The immunization of mice was performed as described previously \[[@pntd.0007687.ref028]\]. In brief, mice were inoculated intramuscularly three times at a 7-day interval with purified ZIKV E protein. After the last immunization, the spleen was removed from the immunized mice and single splenocyte cells were fused with myeloma cells using PEG (Sigma-Aldrich, St. Louis, MO, USA). Hybridomas were selected by HAT and IPTG/HAT selection. Culture supernatants were screened for IgG antibodies against ZIKV E protein. Culture supernatants of hybridoma cells were then purified using a HiTrap Protein G column (GE Healthcare). The concentration of antibodies was determined by measuring the absorbance at 280 nm (1 mg