The Influenza Virus
Structure of the virus
How the virus gets into the cell
The role of the M2 protein
Commandeering the cell
How the virus replicates
Making the other parts of the virus
Assembly of the virus and how it gains access to a new cell
The disease influenza is caused by a virus (as opposed to a bacterium or a parasite or a fungus). Bacteria, parasites and fungi are clearly living agents, but there is controversy about whether a virus is alive or not, since it doesn’t have the ability to reproduce on its own but must hijack the machinery of a living host cell.
In fact that’s all the virus does, reproduce. From its point of view, you are just a way to make another virus. If making you sick is an efficient way to do that, it will, but if making you too sick interferes with its ability to reproduce, it will change to be nicer to you. Likewise, if your body doesn’t like the way the virus has chosen to use you to reproduce (say by making you cough and sneeze virus for weeks on end), you react in various ways to get rid of it, for example by having your immune system make antibodies to neutralize the virus’s ability to infect you. We have developed defense mechanisms because, just like the virus, without them we wouldn’t get a chance to reproduce and if that happened too often we wouldn’t be here as a species. So mostly we get better, but that doesn’t mean we have ‘won.’ If the virus has found a way to infect another host it has accomplished its purpose, even if you have eliminated or neutralized the ones that took up housekeeping in your respiratory tract. That’s the big picture. What about the details? If the virus has only one end in view, how does it do it?
Influenza viruses are actually a family of viruses (called the orthomyxoviridae) which has three members, types A, B and C. All three types infect people, but type A is the one that causes influenza pandemics, so we will concentrate on it. Influenza A is a medium sized virus, as viruses go, but all viruses are extremely small, much smaller than bacteria. They were first characterized infectious agents so small they could pass through the very fine filters which stopped all other known micro-organisms (hence the older term, ‘filterable viruses’). Compared to almost any other micro-organism viruses are also simple in structure, consisting of little else than genetic material that specifies how the virus is built and any accessory structures it needs to trick a host cell to build a new copy of itself from this blueprint. In effect, it wants to turn the host cell into a xerox machine for influenza virus.
Commandeering a host cell for this purpose is called infection. The consequences for the host cell might be small (new copies of a harmless virus are made) or catastrophic (the host cell dies and viral copies are loosed to infect and kill more cells). Because our bodies try to defend against bad consequences we have defenses which are in turn countered by viral counter-defenses, which the virus builds into its structure. Thus the influenza virus has both the means to replicate itself and some tools and strategies to interfere with the host cell’s ability to prevent it from accomplishing its objective.
↑ top ↑
General structure of the virus
All viruses have a core of genetic material (the blueprint for its structure) which is surrounded by a protective covering of protein (called a capsid). The genetic material is not the same in all viruses. The genetic material in influenza A is a single-stranded ribonucleic acid (ssRNA), in contrast to the genetic material in our cells which is double-stranded deoxyribonucleic acid (DNA), the two strands coiled around each other in the famous double-helix configuration.
In addition to the capsid, many viruses add covering materials (envelopes) that have sugars and fats on them. Influenza A is an enveloped virus. This basic visual model may help those new to the basic structure. When the virus infects a cell its objective is to get the DNA cell’s molecular machinery to make a new ssRNA virus with its genetic material, capsid and envelope, just as if you were to go to someone else’s workshop with a blueprint and get them to put their work aside and use their tools and materials to make many copies of what you wanted at their expense.
To accomplish this the virus must solve several problems. It needs to gain access to the host cell’s equipment and materials, it needs to trick it into working on a virus instead of things useful for the cell and the organism generally, and it needs to get the copies out of the cell and into other cells so they can make copies of themselves in other cells.
↑ top ↑
Getting access to the host cell’s equipment and materials
For the moment we assume the intact viral particle (also called a virion) has arrived in the respiratory tract. It is still on the ‘outside,’ however. It needs to get into a cell where the equipment is located to make a new virus. It is time to take a closer look at the viral envelope which has a protein sticking through it that is important for getting the virus into the host cell.
The envelope is a lipid (fatty) and protein layer containing some sugars that overlies the protein capsid. The hard shell-like capsid is made of the virus’s own protein (designated M1 protein), but the lipid-protein (lipoprotein) layer enveloping it comes from material stolen from the outer cell wall of the previously infected host cell when the virus ‘budded’ through its surface. Technically it is a phospholipid bilayer typical of cell membranes. But the envelope also has viral protein spikes with sugars (carbohydrates) attached sticking through it.
The protein-sugar combination is called a glycoprotein. Three such glycoproteins specified by (‘coded for’) the viral genetic material are called hemagglutinin (HA), neuraminidase (NA), and M2. Since they are on the viral surface, the immune system ‘sees’ them and can make antibodies against them for immunity. There are 16 broad immune classes of HA glycoprotein and 9 different NA immune classes. (Antibodies to M2 also exist, but their role in immunity is not clear at the moment.) The different HA immune classes are designated H1 to H16 and the 9 NA immune classes, N1 to N9. These are the basis for the different influenza virus subtypes like H1N1 or H3N2. The former has HA protein spikes of immune class H1, the latter of immune class H3, and similarly for the NA immune classes.
Within each immune class there may be larger or smaller variation which will affect how well antibodies directed against one H1 glycoprotein will work against another H1 glycoprotein that looks slightly different to the immune system. These differences give rise to the different strains of the virus that arise from year to year, but they are not as large as the differences between the major classes.
Here is an analogy. Dogs come in various broad classes (Saint Bernards, Collies, Dachshunds, etc.) that are very different from each other. They are all dogs, but they don’t look much alike. These are the H1, H2, etc. and N1, N2, etc. classes of glycoproteins. Within each breed there are also smaller variations: some variation in size, hair color, pattern, etc. These are the strains of virus within each H–N subtype combination. If the only dog you know is a black Saint Bernard with a rough coat you may be a bit confused when confronted with a red Saint Bernard with a smooth coat, and if you encounter a chihuahua you may not even recognize the animal before you is a dog. Similarly your immune system may be slow and ineffective in recognizing different strains of the same viral subtype and not even ‘see’ a different subtype at all. Our immune systems have no experience with the H5 subtype of influenza A virus, which is one of the main reasons public health officials are so concerned: one of our main defense systems, the immune system, may be ineffective until it learns to recognize this subtype, and the learning only takes place after infection. In the case of previous infections with the same subtype, even if the strain is different, there may some recognition and the response, while delayed, may still have some effectiveness. But with the H5 subtype, the virus will reproduce unhindered for much longer and your immune system may not have enough time to make antibodies at all.
The importance of the HA glycoprotein goes beyond the immune system response. If the HA and NA proteins only function was to be a target for the immune system the virus would have no use for them and in fact they would be a detriment. The HA protein also functions as the key to getting in to the locked door of the host cell. Here’s how it works.
The HA glycoprotein of the virus has a special region, called the receptor binding site, that can attach to a host cell if that cell has a specific receptor molecule. Host respiratory cells also have lipid bilayers studded with glycoproteins. The receptor molecule for influenza A virus is a host cell glycoprotein with a side chain tipped with a particular kind of sugar called a sialic acid. The specific sialic acid characteristic of human cells is N-acetylneuraminic acid (NeuAc) and when attached to the cell’s glycoprotein via yet another sugar called galactose it is a potential receptor for attachment of the influenza A virus via the HA protein spike. It is ‘potential’ because there is one additional subtlety here. It depends upon how the NeuAc sialic acid is attached to the galactose. There are several possible linkages, of which two, the α-(2,3) and the α-(2,6) linkages, are important for influenza virus recognition. The numbers and the Greek letter tell which atoms are connected to which on the sialic acid and the galactose.
It was thought that bird respiratory and intestinal tract cells had NeuAc- α-(2,3)-Gal linkages while human respiratory tract cells had NeuAc- α-(2,6)-Gal linkages, although it now appears human respiratory tracts have both, although on different cell types. It seems true, however that avian influenza viruses prefer the former and human influenza viruses the latter. Pigs have both types of linkages which is why they are thought to be an important mixing vessel for genetic reassortments and recombinations, since the pig can be co-infected by both human and avian viruses. Thus a change in the HA binding site via a mutation that would allow efficient binding to a human receptor could cause a virus that previously was efficiently transmitted among birds to be transmitted from person to person. This is rather neat, but unfortunately it isn’t the whole story. Other factors also seem to be involved in determining host specificity of influenza viruses. We also know avian viruses can infect humans, although less efficiently. Thus there is more to it and it is an unfinished story.
Things are already a bit complicated. To summarize, the influenza virus has a glycoprotein spike, the HA protein, on its surface that can dock to the right kind of receptor on the host cell’s surface. The HA glycoprotein is thus a key, or at least an important component of a key, that unlocks the door to the inside of the host cell, if it can find and fit the right address. That address is a cell surface glycoprotein with a protein-galactose- α-(2,3/6)-NeuAc (sialic acid) combination waving in the breeze.
The complications don’t end there. Once attached, the virus is engulfed by the host cell in a process where part of the cell surface (the ‘cell membrane’) is induced to form a deepening pit that encloses the virus, eventually pinching off to make a little enclosing bubble that transports the virus into the cell. At some point (exactly when and where isn’t completely clear), to be infective the HA glycoprotein has to be cut (‘cleaved’) into two pieces, HA1 and HA2. This is done by a special kind of enzyme called a protease, of which there are many varieties, some present only in specific tissues, others more common. The site on HA where this must occur (the cleavage site) is quite narrow so the variety of proteases that can trigger infectiveness is limited to those present on one or perhaps a few tissues, primarily respiratory and intestinal tissue in birds and respiratory tissue in humans. Avian viruses of the H5 and H7 subtypes may have mutations that insert extra amino acids (the building blocks of proteins) at the cleavage site, widening it and allowing a larger variety of tissue proteases to perform the cleavage operation. Proteases like to work after specific kinds of amino acids, the ones with basic groups on their side chains (there are three such basic amino acids, lysine, arginine and histidine). The Low Pathogenic H5s have only one basic amino acid at the cleavage site. The ‘polybasic amino acids at the cleavage site’ are characteristic of the Highly Pathogenic Avian Influenza (HPAI) viruses that have devastated poultry in recent years, allowing infection of a wide variety of tissues (nervous system, kidney, heart) by the influenza virus. The extra basic amino acids are arginine and lysine, and the pattern for HPAI H5N1 in poultry has been fairly constant, but there is a variant that is missing either a lysine or an arginine. North Vietnam has such a variant missing an arginine, although it is still “polybasic.”
Again, such mutations are not the only or even necessary determinant of high pathogenicity (more properly virulence). Subtle variations in the configuration of the cleavage site and the presence or absence of sugar sidechains at various places may also play a major part, as recent analysis of remnants of the 1918 virus suggests. At this time, it is not possible to look at the sequence of a virus and predict ahead of time its pathogenicity (ability to cause disease) or virulence (ability to cause severe disease) for humans.
The importance of cleavage of the HA glycoprotein is that it is needed for what happens next. Once inside the bubble (called a vesicle), the internal environment of the vescicle (especially its acidity) causes the viral covering to fuse with the vesicle’s wall which releases the genetic material of the virus into the host cell, finally giving it access to the tool shop to make new copies of itself. This fusing with the vesicle is called ‘uncoating’ and requires cleavage of HA to happen.
↑ top ↑
The role of the M2 protein
Another important component in the uncoating is the viral M2 protein, called an ion channel. It is this protein that is the target of the antiviral drugs amantadine and rimantadine. Unfortunately, the virus can fairly easily change its M2 protein so that these drugs no longer prevent its function, thus producing drug resistance. However resistance in circulating viruses seems uncommon, although it emerges rapidly, often within two or three days of start of treatment. It may be that the mutant M2 is disadvantageous and back mutation also occurs readily. H5 viruses circulating in southeast asia are apparently resistant but not H5s elsewhere. Thus whether amantadine will have any place in treating avian influenza in humans is unclear at this time.
↑ top ↑
Commandeering the host cell’s machinery to make the components of new viruses
The viral construction program, encoded in single stranded RNA, is now inside the cell where it migrates to the nucleus, the location of the host cell’s own DNA blueprint. The uncoated viral RNA, which like DNA specifies proteins via a sequence of three-letter codes (‘the genetic code’), isn’t naked, but is complexed with some important viral proteins in a package referred to as ribonucleoprotein (RNP). One protein component (NP) is involved in stabilizing the shape of the RNA, anchoring it within the capsid before uncoating and transporting it to the right places inside the cell after uncoating. The code for NP is also in the viral RNA, so it is a viral product.
Now the viral RNA must accomplish two things. One is to make a copy of itself (replication). The other is to direct the host cell to manufacture the ten protein pieces of a new virus. We have already seen some of these pieces: the HA, NA, M1, M2 and NP proteins. The RNP complex also has another vital protein, RNA polymerase (abbreviated as PB1). This is an enzyme, also coded for by the virus, that both makes a template (messenger RNA or mRNA) used by the host cell to construct viral proteins; and also participates in making a copy of the viral RNA blueprint. Thus the viral RNA polymerase is key to both tasks, replication and viral protein production.
↑ top ↑
RNA polymerase doesn’t make an exact copy of RNA but instead a ‘code mirror image’ (complementary) copy. Viral RNA is replicated in two steps, one to make a mirror image copy of the RNA and a second step to make a mirror image of the mirror image, i.e., a duplicate of the original viral RNA. Errors can easily occur during this double duplication and the conventional view is that it is these errors that cause the many genetic variations seen as the virus replicates.
In addition to mutation and recombination, there is another process that can produce major genetic variation: reassortment. The viral RNA is actually in the form of eight separate pieces, called segments, analogous to our chromosomes. Six of these segments code a protein each, while two segments code for two proteins each. An intact virus requires all eight RNA segments, but they can come from different viruses. If a host like a pig or a person is infected by both a bird virus and a human or pig virus the segments are in a common pool within the host cell and can become mixed, with some segments coming from one virus and some from another. This can produce another kind of hybrid between a bird and a human influenza virus. The difference between this mechanism and recombination is that the segments move en bloc as a single piece while in recombination bits and pieces of segments might be spliced together as in the human-pig HA example above rather than an HA from a pig or an HA from a human as in reassortment.
Since reassortment involving the HA segment would introduce a new subtype like H5 that doesn’t normally bind to human cells, it would not seem to be a plausible mechanism to start a pandemic by itself if easy transmissibility depended entirely on the receptor binding site. But as we have seen, this is a complicated picture and may depend on new combinations of the other segments. Since eight segments from two different viruses can make 256 different combinations, there is a plausible role for reassortment. Likely there is also some requirement for random mutation or recombination that might affect the binding site of HA. Reassortment is a known mechanism for producing major changes in an influenza virus, which together with other mechanism can produce the kinds of hybrid bird and mammal viruses involved in pandemics. Because we don’t know all the combinations of genetic changes that might produce virulence and their origins, there is more to learn about its relative importance.
↑ top ↑
Making the other parts of the virus
The RNA polymerase not only does the two step replication of viral RNA, but also makes the directions the host cell will use to construct viral proteins. It does this by making a messenger RNA (mRNA) template that moves outside the host cell’s nucleus to the parts of the cell where protein synthesis takes place. The mRNA uses the cell’s machinery to make all the viral proteins, which include, besides the ones we have already discussed, the RNA polymerase itself (PB1); an RNA polymerase accessory protein, PB2, that uses bits of host cell mRNA to get the viral construction process underway; another RNA polymerase component, designated PA, that is necessary for viral replication, but whose function is not yet understood; a protein called the nuclear export factor (NEP, formerly called NS2) needed to get the replicated viral RNA repackaged as RNP and then exported out of the nucleus; and finally a protein, called the non-structural protein, NS, that protects the virus from one of the host’s immediate defense mechanisms by interfering with a non-specific cytokine called interferon.
The other proteins don’t get as much attention as the HA and NA proteins, but it is known that variations in them can be important for virulence. Again, exactly what kinds of changes and how they work to make the virus more or less virulent remains to be worked out.
↑ top ↑
Assembling a new virus and budding from the host cell’s surface to gain access to other cells
At this point the viral RNP and the other viral proteins are now made. They assemble just inside the host cell’s outer wall (the cell membrane) and ‘bud’ through the surface, dragging some of the host cell’s membrane with them to make the viral lipoprotein envelope. This process is far from perfect. Sometimes more than eight segments get incorporated or the wrong eight get packaged together. Those viruses are no longer infective. It has been estimated that there is about 90% ‘spoilage’ in this step, but there are so many viral copies made that this inefficiency can be accommodated.
Now there is only one problem left for the virus to solve. The HA glycoprotein spike on the newly made copy can still grab onto the sialic acid receptor of the just exited cell and in effect try to re-infect it; or the viral particles that have sialic acid stuck to their HA from passing through the cell membrane can stick to each other and become immobilised in clumps. The virus solves this problem with the aid of its other viral glycoprotein spike, neuraminidase (NA), which is an enzyme that can clear sialic acid both from the cell surface and from the HA spike, thus allowing release of the newly produced virions. The neuraminidase inhibiting drugs oseltamivir (Tamiflu) and zanamivir (Relenza) work by preventing NA from releasing the budding virions.
Figure 1. Mechanism of Action of Neuraminidase Inhibitors. Panel A shows the action of neuraminidase
in the continued replication of virions in influenza infection. The replication is blocked by neuraminidase
inhibitors (Panel B), which prevent virions from being released from the surface of infected cells.
Source: Moscona, A. (2005). Neuraminidase Inhibitors for Influenza. N Engl J Med 353: 1363-1373
Unlike the M2 protein inhibited by the antivirals amantadine and rimantadine, the NA enzyme seems to have less flexibility in mutating. Examples of resistant NA so far are also less effective, so the resistant virus itself, while infective, is much less so. Whether this will continue to hold true we don’t know.
The virus is now free to find another host cell to infect and begin this cycle again.