Screening with bioinformatics for new dengue cross-serotype antibodies

From the Vaccine and Infectious Disease Division and the Chan Zuckerberg Biohub.

Dengue virus (DENV) is a mosquito-borne Flavivirus that causes 400 million infections yearly, facilitated by the Aedes mosquito, whose geographic distribution covers half the world. Much DENV research focuses on designing a vaccine to protect against infection by the four distinct DENV serotypes, which can lead to a spectrum of disease severity, ranging from asymptomatic to death. However, the immune response to DENV is not straightforward. Most natural infections create a beneficial immune memory response, protecting the host from subsequent infections. In contrast, sequential DENV infections with different serotypes can lead to severe disease. This phenomenon has been partly attributed to to pre-existing antibodies from the initial infection that are cross-reactive but are not able to block or "neutralize" infection by the  the second virus. Instead, during secondary infection, these antibodies have the potential to facilitate host cell uptake of the virus in a process called antibody-dependent enhancement (ADE). Therefore, a safe and effective vaccine must elicit broadly neutralizing antibodies (bNAb) that simultaneously block all four serotypes.

Although vaccines designed with a combination of immunogens derived from all four serotypes are in clinical development, their efficacy relies on the assumption that tetravalency will elicit an equal neutralizing response to each serotype. However, previous results suggested that these vaccines have failed to protect equally against the DENV serotypes, and have in fact worsened disease outcome following subsequent DENV infection in some vaccine recipients. In a recent study published in eLife, Dr. Leslie Goo, a new investigator in the Vaccine and Infectious Disease Division, and her colleagues at the Chan Zuckerberg Biohub and Stanford University discovered new DENV bNAbs targeting an epitope on the envelope protein conserved on all DENV serotypes. She hopes that, as an alternative to the tetravalent vaccine strategy, this epitope can eventually be used to engineer a single immunogen to elicit bNAbs to protect against all DENV serotypes. “Epitope-focused vaccine design has taken off for other antigenically diverse viruses such as HIV and influenza. Many bNAbs with defined epitopes on HIV and influenza have been described, providing tremendous insight into how highly functional antibodies recognize these viruses. In contrast, prior to our study, there was only one known class of bNAbs against flaviviruses, so we wanted to know whether there were others that could help us build a better map of the 'vulnerable' sites on the DENV envelope protein to guide vaccine design,” Dr. Goo explained.

Schematic representation of a monoclonal antibody that can neutralize all four dengue virus serotypes (DENV1-4) by targeting a conserved epitope on the envelope protein depicted by a yellow star on each virus particle.”
Schematic representation of a monoclonal antibody that can neutralize all four dengue virus serotypes (DENV1-4) by targeting a conserved epitope on the envelope protein depicted by a yellow star on each virus particle.” Image provided by Leslie Goo.

Instead of using traditional laborious methods to discover neutralizing antibodies by screening hundreds of B cells, Dr. Goo and colleagues took advantage of their previous study1, in which they had analyzed the single B cell transcriptomes of DENV-infected patients. Out of more than 350 single B cells analyzed, they focused on 38 that were bioinformatically predicted to encode DENV-specific neutralizing antibodies for functional characterization in their new study. The authors cloned the paired heavy and light chain sequences from each B cell into expression vectors and transfected mammalian cells for recombinant antibody production. The antibodies were then tested for neutralization against a panel of flaviviruses. Promisingly, the authors identified seven antibodies that neutralized all four DENV serotypes, suggesting that their screening process had successfully identified bNAbs.

To further profile these bNAbs, the authors performed dose-dependent neutralization assays to assess their potency. Two somatically related antibodies, named J8 and J9, not only neutralized all four DENV serotypes with up to 60-fold greater potency than the previously identified class of bNAbs, but also did so regardless of the maturation state of the DENV particle, which can be heterogeneous and can affect antibody recognition. Neutralization potency is important because ADE can occur when antibody concentrations wane. The authors used an ADE assay to demonstrate that compared to other antibodies, the ability of J8 and J9 to facilitate ADE was restricted to a narrow range of very low antibody concentrations, beyond which neutralization is observed. Thus, when induced by a vaccine, these bNAbs could afford protection against all DENV serotypes with a potentially minimal risk of ADE. Furthermore, J8 and J9 could block infection even after DENV has attached to host cells. As DENV uses different cell surface proteins to attach to different cell types, the ability of J9 and J8 to neutralize at a post-attachment step suggests that these bNAbs could be effective at blocking infection of multiple cell types.

Finally, the authors interrogated the epitope targeted by J8 and J9  by testing their ability to neutralize a library of DENV variants encoding envelope protein mutants. They found that mutations at sites distinct from those recognized by the previously characterized class of DENV bNAbs disrupted J8 and J9 neutralizing activity, demonstrating that these bNAbs are specific for different, previously uncharacterized envelope protein epitopes. This result suggests that J8 and J9, in conjunction with previously identified bNAbs, could provide complimentary neutralization coverage against diverse circulating DENV variants.

Dr. Goo summarized their work, explaining that “in addition to revealing a new epitope on the virus envelope protein that can ultimately be exploited for vaccine design, these antibodies could serve as a foundation for the development of monoclonal antibody therapy to protect against severe dengue disease”. She also posited that their bioinformatics-based approach has potentially to be “broadly applicable to the rapid discovery of highly functional antibodies against other viruses.”

This work was supported by the National Institutes of Health, the Chan Zuckerberg Biohub, the National Science Foundation, the Dr. Ralph and Marian Falk Medical Research Trust, Stanford University, and the Fred Hutch.

Durham ND, Agrawal A, Waltari E, Croote D, Zanini F, Fouch M, Davidson E, Smith O, Carabajal E, Pak JE, Doranz BJ, Robinson M, Sanz AM, Albornoz LL, Einav S, Quake SR, McCutcheon Km, Goo L. 2019. Broadly neutralizing human antibodies against dengue virus identified by single B cell transcriptomics. Elife. 2019 Dec 10;8. pii: e52384. doi: 10.7554/eLife.52384.

1 Zanini et al., 2018