PerspectiveStructural Biology

Unraveling a Flavivirus Enigma

See allHide authors and affiliations

Science  21 Feb 2014:
Vol. 343, Issue 6173, pp. 849-850
DOI: 10.1126/science.1251249

There is growing concern about the spread of flaviviruses, such as dengue virus and West Nile virus, to new geographic areas as they can cause major epidemics and represent global public health threats. Controlling these viruses requires a better molecular understanding of how they infect cells. Nonstructural protein 1 (NS1) is perhaps the most enigmatic flavivirus protein. During infection, NS1 exists in two distinct forms, travels to various compartments, decorates itself with different molecular disguises, and plays numerous roles in its infectious cycle and disease pathogenesis (1). How this protein manages all of this has been a puzzle since its discovery in 1970 (2). Crystallizing NS1 has daunted many researchers because of the heterogeneity of its glycosylation and association with lipids, but as reported on page 881 of this issue, Akey et al. (3) have accomplished this task. The unusual structural details revealed about NS1 may guide the design of compounds that inhibit viral replication and provide clues as to how it contributes to different stages of the virus life cycle and disease.

Flavivirus NS1 is a glycoprotein with a molecular mass of 46 to 55 kD, depending on its glycosylation status. The crystal structures of dengue virus NS1 (3 Å resolution) and West Nile virus NS1 (2.8 Å resolution) exhibit a similar hexameric arrangement of three dimers, confirming the hexameric structure (30 Å resolution) indicated by a cryoelectron-microscopy analysis (4). Each monomer displays an unusual fold consisting of three regions: a “β-roll” domain that dimerizes with that of another monomer; a “wing” domain that resembles a helicase domain; and a “β-ladder” domain that aligns with that of another NS1 molecule to form an extended β-sheet ladder. The ladder forms the plane of the NS1 dimer, with a hydrophobic side (exemplified by a “greasy finger” loop) that can associate with the membrane. The hydrophobic side of each dimer faces the interior of the hexamer. Remarkably, recombinant NS1, which does not possess any transmembrane domain, can convert large liposomes into smaller lipid-protein nanoparticles. This demonstrates that NS1 can directly modulate the lipid membrane without additional cellular proteins. Such lipid-modulation activity and its underlying structure could account for the myriad functions of NS1.

Assist before leaving.

Flavivurus replicates at the ER surface in the infected cell. Viral NS1 protein forms dimers in the ER lumen, yet assist the replication complex on the opposite side of the membrane. Seven nonstructural proteins, together with host proteins (not shown), form the replication complex. Once immature viral particles bud into the secretory pathway, NS1 protein forms hexamers that are secreted as lipoproteins.


After flavivirus entry into a cell by endocytosis, the virus particle is released into the cytoplasm. Viral genomic RNA is translated into proteins and replicated, and virus assembly occurs on the surface of the endoplasmic reticulum (ER). Viral particles bud into the ER and mature as they are transported through the secretory pathway for release from the cell.

NS1 protein is translated from viral RNA and translocated into the ER lumen, where it is glycosylated. NS1 dimers then form and associate with the luminal side of the ER membrane at a virus-induced vesicle packet (see the figure). Although dimeric NS1 is required for viral RNA synthesis, the replication complex resides on the cytoplasmic side of the ER membrane. Two factors could facilitate the recruitment of NS1 to the replication complex: the membrane-association of NS1, and the specific interactions between NS1 and viral transmembrane proteins NS4A and NS4B (5, 6).

What happens after the NS1 dimer has facilitated viral replication? It is eventually released by the infected cell. A model proposes that the assembly of hexameric NS1 (4) is key to this process. Newly synthesized monomeric NS1 is water-soluble. As its concentration and glycosylation increase in the ER lumen, NS1 dimerizes, creating the hydrophobicity needed for its interaction with the membrane (7). Three NS1 dimers juxtapose on the lipid bilayer and pinch off the membrane, resulting in a water-soluble hexamer. Host lipids become trapped within the central channel of the hexamer, forming a lipoprotein particle. The particle is then transported and released from the cell through the secretory pathway.

In dengue virus–infected patients, the concentration of extracellular NS1 can reach 15 µg/ml in sera (1). NS1-based tests have been developed for rapid, point-of-care diagnosis. The concentration of serum NS1 correlates with the amount of the viral RNA present in the patient, and high amounts of circulating dengue virus NS1 early in illness correlate with severe disease outcome (8). Mounting evidence indicates that secreted NS1 modulates disease pathogenesis. Preincubation of hepatocytes with soluble NS1 enhances homologous dengue virus infection (9). Secreted NS1 interacts with host proteins, many of which are involved in the immune complement pathway (10, 11); this may allow flaviruses to evade the immune system. Secreted NS1 also is highly immunogenic. Some antibodies against NS1 are cross-reactive with cellular components; these auto-antibodies may contribute to platelet and endothelial cell damage, leading to vascular leakage, the hallmark of severe dengue hemorrhagic fever and dengue shock syndrome.

The critical roles of NS1 in flavivirus replication and pathogenesis implicate NS1 as an attractive antiviral target. A few tangible approaches can be envisioned. Cells expressing NS1 could be screened for inhibitors of NS1 dimerization and hexamerization, and libraries could be screened for compounds that block the ability of NS1 to convert liposomes into lipoprotein particles. The crystal structure will greatly facilitate structure-based rational design of antiviral compounds. In fact, inhibitors of cellular glucosidases that are required for NS1 glycosylation suppress flavivirus replication in cell culture and in a mouse model (12). Future studies should define how NS1 physically interacts with the replication complex and its specific role in RNA replication. The molecular details remain to be determined as to when, where, and how the conversion of NS1 monomer to dimer and then to hexamer is controlled. One question concerns the NS1 “wing” domain, whose folding is similar to that seen in two proteins [retinoic acid–inducible gene I (RIG-I) and melanoma differentiation–associated gene 5 (MDA5)] that function as viral sensors in the innate immune system. Does this somehow allow flaviviruses to evade the host immune response? Another intriguing question is why, within the family Flaviviridae, only members of the genus Flavivirus encode the NS1 protein; members of the other two genera, Hepacivirus and Pestivirus, do not contain a gene equivalent to NS1. The reason may be that most flaviviruses transfer between insects and mammals. If so, it raises the question of how flavivirus NS1 play distinct roles when replicating in different host cells. Perhaps more interesting is how the essential role of NS1 in flavivirus replication is compensated in hepacivirus and pestivirus. The answers to these questions will unravel more mysteries of this fascinating protein.


View Abstract

Stay Connected to Science

Navigate This Article