The human antibody response to dengue virus infection

Abstract

Dengue viruses (DENV) are the causative agents of dengue fever (DF) and dengue hemorrhagic fever (DHF). Here we review the current state of knowledge about the human antibody response to dengue and identify important knowledge gaps. A large body of work has demonstrated that antibodies can neutralize or enhance DENV infection. Investigators have mainly used mouse monoclonal antibodies (MAbs) to study interactions between DENV and antibodies. These studies indicate that antibody neutralization of DENVs is a "multi-hit" phenomenon that requires the binding of multiple antibodies to neutralize a virion. The most potently neutralizing mouse MAbs bind to surface exposed epitopes on domain III of the dengue envelope (E) protein. One challenge facing the dengue field now is to extend these studies with mouse MAbs to better understand the human antibody response. The human antibody response is complex as it involves a polyclonal response to primary and secondary infections with 4 different DENV serotypes. Here we review studies conducted with immune sera and MAbs isolated from people exposed to dengue infections. Most dengue-specific antibodies in human immune sera are weakly neutralizing and bind to multiple DENV serotypes. The human antibodies that potently and type specifically neutralize DENV represent a small fraction of the total DENV-specific antibody response. Moreover, these neutralizing antibodies appear to bind to novel epitopes including complex, quaternary epitopes that are only preserved on the intact virion. These studies establish that human and mouse antibodies recognize distinct epitopes on the dengue virion. The leading theory proposed to explain the increased risk of severe disease in secondary cases is antibody dependent enhancement (ADE), which postulates that weakly neutralizing antibodies from the first infection bind to the second serotype and enhance infection of FcγR bearing myeloid cells such as monocytes and macrophages. Here we review results from human, animal and cell culture studies relevant to the ADE hypothesis. By understanding how human antibodies neutralize or enhance DENV, it will be possible to better evaluate existing vaccines and develop the next generation of novel vaccines.

Keywords: antibody; antibody dependent enhancement; dengue virus; neutralization.

Figure 1

Structure of dengue virus envelope protein (A) and the dengue virus particle (B). (A) E protein on the mature virion is a homodimer and each subunit has three domains designated I (red), II (yellow) and III (blue). (B) Arrangement of E proteins on the surface of the virion. Both images A and B are from [12].

Figure 2

Maturation of dengue virions. Dengue virions bud into the endoplasmic reticulum as immature, non-infectious particles. The surface of immature particles has a jagged appearance because prM and E proteins are initially assembled as trimeric spikes that protrude away from the envelope. In the trans-Golgi compartment a cellular protease cleaves prM protein to generate the mature M protein, which also results in the rearrangement of E protein trimers to form dimers that lie flat on the surface of the envelope creating the smooth surface observed in mature, infectious virions released into the extracellular space. The images reproduced here are from [19].

Figure 3

Location of mouse MAb epitopes on DENV E protein. The figure is based on the structure of the ectodomain of DENV3 E protein solved by Modis and colleagues [7]. (A) The image depicts the major regions on domains I (red) and II (yellow) recognized by mouse MAbs. (B) An enlarged view of domain III (blue) displaying the lateral ridge and A strand epitopes recognized by strongly neutralizing mouse MAbs. Figure originally published in [22].

Figure 4

Quaternary structure epitope recognized by West Nile human MAb CR4354. The figure depicts a raft of three E protein homo dimers, which is the basic building block of the flavivirus envelope. The domains of E protein are color coded as described in figure 1. The approximate footprint of human MAb CR4354 is circled. Note that the foot print encompasses EDI/II and EDIII from adjacent dimers. One hundred and twenty CR4354 epitopes are predicted to be available on a virion for antibody binding. Figure adapted from [57,59].