Hemagglutinin
Created by Amanda House
Causing an estimated 50 million deaths worldwide, the 1918/1919 H1N1 influenza A virus is responsible for causing the worst infectious pandemic in history (1,2). Its many descendants, which include "drifted" H1N1 viruses and reassorted H2N2 and H3N2 viruses, have since caused almost every recorded influenza A outbreak (1). Understanding the origins of the 1918/1919 virus, particularly the basis for its exceptional virulence, may aid in the prediction of future influenza pandemics (3).
In 1918, the cause of influenza was unknown and, consequently, no intact virus survived (4). Fragments of the viral genome, discovered in Alaskan victims buried in the permafrost, have since provided what is known about the 1918/1919 virus. For decades, virologists have been trying to piece together why this strain exhibited such pathogenic properties (5). Many theories have been proposed, the most popular of which suggests that any molecular evidence for these deadly characteristics is most likely encoded within its hemagglutinin (pdb id: 1ruz), the viral protein responsible for infection.
Hemagglutinin (HA), the major surface glycoprotein of influenza viruses, is responsible for virus attachment to host receptors, internalization of the virus, and subsequent pH-dependent membrane-fusion events that occur within the endosomal pathway of the infected cell (4). Synthesized as a single polypeptide precursor (HA0) (pdb id: 1rd8) in host cell endoplasmic reticulua, this type I transmembrane glycoprotein quickly assembles into a homotrimer and is transported to the Golgi network (6). Once inside the Golgi, HA0 undergoes extensive posttranslational modifications, including palmitoylation and proteolytic cleavage, before being sent to the plasma membrane, its final destination.
Proteolytic cleavage by host cell trypsin-type endoproteases creates two, disulfide-linked subunits: HA1 (328 residues, Image 1) and HA2 (160 residues, Image 2) (7). Due to HA’s trimeric structure, a total of six chains contribute to three structurally distinct regions: a large, globule head comprised of anti-parallel beta-sheets that form a beta-sandwich and jelly-roll fold; a central, coiled-coil, alpha-helical stalk; and a globular foot comprised of anti-parallel beta-sheets (inter). These three structural domains make up four functional subdomains (8). Without such cleavage, HA would fail to cause infection (7).
Distal to the membrane, HA1 (chains H, J, and L) plays two structural roles: its N-terminus provides the central strand in the 5-stranded, globular foot domain and the rest of it makes its way to the 8-stranded, globular headpiece, which contains the HA1/HA2 cleavage site (7). Proximal to the membrane, the C-terminus of HA2 (chains I, K, and M) anchors the protein and provides the remaining strands of the globular foot. The rest of HA2 forms two alpha-helical stalks, which complete the triple-stranded central stalk, which functions to stabilize the trimer.
Upon viral budding, HA functions as lectin. With its three sialic acid-binding sites (one per monomer), HA exploits cellular surface glycans, treating their terminal sialic (N-acetyl-neuraminic acid residues as receptors (represented by analogs: NAG and NDG) (8). Following attachment, the cell imports the virus via endocytosis, creating an endosome, which it then begins to acidify for digestion (9). Unfortunately for the unsuspecting cell, HA takes advantage of another cellular process by using this drop in pH to drive membrane-fusion. As the pH approaches 6.0, HA begins to denature, exposing a portion of its chain known as the fusion peptide. This short, hydrophobic N-terminus of HA2, created during proteolytic cleavage of HA0, facilitates the fusion of the viral and endosomal membranes, releasing the viral genome into the cell (10).
Due to its hydrophobic nature, at normal pH, the fusion peptide remains tucked inside the interior of the protein stalk. When exposed during denaturation, the fusion peptide stabs itself into the endosomal membrane to "take cover." As the interior of the endosome continues to acidify, another conformation of HA soon becomes more stable and, subsequently, HA refolds, retracting the fusion peptide, which, like a grappling hook, pulls the endosomal membrane to the virion’s side. Positioned right next to one another, the viral and endosomal membranes fuse. This fusion signals the disassembly of the viral core, thus releasing the viral content into the host cell.
Correlating with species specification, HAs preferentially bind to two types of sialic acid-linkages found on cellular surfaces (8). Avian influenza viruses, which include all sixteen known HA serotypes, attach to alpha 2,3-linked sialosaccharides, the predominate form in avian enteric tracts. Human influenza viruses (H1, H2, and H3 subtypes), on the other hand, prefer alpha 2,6-linked sialosaccharides, the major form throughout the human respiratory tract. Interestingly, swine influenza viruses bind to alpha 2,6-linked sialosaccharides, as well as, sometimes alpha 2,3 linkages, both of which are found in porcine tracheae. Such linkage specificity makes HA the primary determinant of host range.
While phylogenetic analyses of the 1918/1919 strain places it at the base of the human H1 branch, some distance away from avian strains, structurally, the virus is distinctly avian (5). Unlike H2 and H3 viruses, which require two amino acid switches to change species preference (Gln226Leu and Gly228Ser; avian to human), H1viruses are able to bind to human cells without such modifications (8). Supporting the popular, avian-origin theory for the 1918/1919 virus, research suggests that structural changes in the receptor binding domain of an avian H1 HA, particularly involving a shift in its binding domain loops, may be responsible (4).
Susceptible to many influenza subtypes, pigs are often considered the likely intermediate for such avian-origin to occur (8). When compared to the HAs of the first isolated influenza viruses, a swine-strain from 1930 (pdb id: 1rd8) and a human-strain from 1934 (pdb id: 1ru7), the 1918/1919 HA was found to be more similar to the former (superimposition of H chains), further suggesting the possibility of an intermediate host.
As the most abundant antigen on the viral surface, HA plays an important role in biological recognition (10). Four prosthetic group-binding sites (three on HA1, one on HA2) and one receptor-binding site make up the primary neutralizing epitopes targeted by antibodies [prosthetic groups: 1, 2 (with amino acids), 3 (with amino acids), 4 (with amino acids)] (8). By binding to these sites, host antibodies disable the virus, preventing infection (10). By studying the antigenic properties of this protein, virologists hope to improve vaccines and other treatments, not only for influenza, but other viral diseases as well.