3.2.1 Summary of the response

Influenza is a viral disease which generally starts with an infection of the upper respiratory tract and develops with systemic symptoms including fever and lassitude over the course of 1–2 weeks. Although it is debilitating, most people make a full recovery. However it may be fatal for older people and those with weaker immune systems, particularly if the virus disposes them to concurrent bacterial infections such as pneumonia.

The disease is caused by a group of myxoviruses, influenza A, B and C, although the most serious infections are caused by influenza A, and the following discussion refers to influenza A. The virus is seen in this transmission electron micrograph with proteins projecting from the viral envelope. Very many different variants of influenza have been identified since the virus was first discovered in 1935, and the variants are responsible for the different infections which occur in successive years. This is the reason that we can suffer from flu several times during our lives – in effect we become infected on each occasion with a different strain of virus, which our immune system has not encountered before and to which we are not immune.

Look at the overall structure of influenza A. The myxoviruses are all enveloped, that is to say that they have an outer membrane or envelope, which has been derived from the plasma membrane of an infected host cell. Two major viral proteins are present in the envelope, namely the haemagglutinin and the neuraminidase. Neuraminidase is an enzyme which cleaves sialic acid residues on glycoproteins and this protein performs an important function, in allowing the virus to bud off from the infected cell and spread through the body. The haemagglutinin is also essential for viral infection. It binds to carbohydrate groups present on glycophorins, molecules which occur on the surface of cells of the body. Binding of haemagglutinin to the surface of host cells is the first step in infection. Since the glycophorins are widely distributed on different cell types, influenza is able to cause widespread infection. As we shall see later, the antibody response against the haemagglutinin of the virus is critical in protecting us against on-going infection, although antibodies to the neuraminidase can also contribute to immunity. Within the viral envelope, the viral capsid is formed from the ‘M-protein’, which contains the virus’ genetic material, RNA, nucleoproteins, and a number of enzymes needed for replication.

The viral genome consists of 8 separate strands of RNA. Because the genome is fragmented in this way, it means that the different strains of virus can re-assort their genes relatively easily, and this is an important source of new viral strains.

Look now at the epidemic pattern of influenza over a number of years. The graph shows the number of isolates of different strains of influenza, in laboratories in the USA between 1995 and 1997. You can see that in temperate latitudes the infection rates follow a seasonal pattern, with more people developing the disease in the winter months. Examination of the structure of the haemagglutinin in successive years shows that minor mutations occur in the primary structure between the virus strains prevalent in each year. Although these do not affect the ability of the haemagglutinin to bind to host cells, the change is sufficient to prevent antibodies specific for a previous strain from binding to the new strain’s haemagglutinin. In effect, last year’s antibodies are unable to protect us from this year’s strain of flu.

The structure of the viral haemagglutinin can be seen in this model, which shows the backbone of the chain of amino acids. We will look at the way in which a neutralising antibody binds to an epitope on the haemagglutinin. The epitope is formed by amino acid residues in the loops at the exposed part of the molecule. Three loops which contribute residues to the epitope are highlighted here by space-filling the residues. Two domains of the heavy chain of the neutralising antibody are shown in yellow. The constant and variable domains are clearly visible, and the three hypervariable loops which contribute to the binding site are picked out in colour. You can see how the epitope and the heavy chain are complementary in shape. To complete the picture, we will add the light chain to the model. In this case, it is clear that the light chain contributes very little to the antibody combining site. When the space-filling model is completed, you can see that the residues forming the epitope are buried in the centre of the binding site and it is precisely these residues which are most likely to mutate between one viral strain and another. This progressive but limited change in the structure of the haemagglutinin is called genetic drift. Occasionally, perhaps once every 10–20 years, a major new strain of influenza appears which is radically different from those of previous years. Such a change is called genetic shift. The appearance of such new strains is associated with a worldwide serious epidemic of flu, called a pandemic. For example the pandemic strain of ‘Spanish flu’ which developed in 1918, had a different haemagglutinin and neuraminidase from the previously dominant strain, and this outbreak is thought to have caused the deaths of 20 million people world-wide. The picture shows a ward at the time. This epidemic particularly affected young fit people. One doctor wrote ‘it is only a matter of hours until death comes. It is horrible.’

The origin of new pandemic strains has been much debated, but it appears most likely that a human strain of flu exchanges genetic material with an animal strain of flu for example from ducks, or pigs. Such a reassortment of genes could occur if two different flu viruses simultaneously infect the same cell, and produce new viruses containing some gene segments from each type – remember that the flu virus has a segmented genome which allows this to occur.

The major pandemic strains are distinguished according to which haemagglutinin they have and which neuraminidase. So, for example in 1957 the dominant strain H1, N1 changed both its haemagglutinin and neuraminidase, and the new dominant strain H2N2 persisted until 1968. At present, (that is, in 2001) there are two dominant strains of influenza A in circulation H1N1 and H3N2.

One of the problems of producing vaccine for flu is that we do not know what next year’s major strain will be, and there is only a limited amount of time available before an epidemic spreads. The map shows the way in which the epidemic of Asian flu spread in 1957 from its origin in China in February, through South-East Asia by April and from there to all parts of the world by the end of the year. The figures on the map indicate the months in which the virus was isolated in different areas.

Nowadays, laboratories throughout the world track the appearance of new variants, and aim to identify the current circulating strains and any potentially new pandemic strains. Having decided the composition of the vaccine, there are just a few months to prepare it for the next flu season. Have a look at the current vaccine, it contains examples of the two major strains of influenza A H1N1 and H3N2 which are circulating and vaccine for the main current strain of influenza B. As there is insufficient time and resource to produce vaccine for everyone, the vaccine is usually recommended for older people and high risk groups, such as health professionals.

In addition to the antibody response, cytotoxic T cells are important in clearing virally-infected cells. The cytotoxic T cells recognise peptide fragments of several of the viral proteins, including the internal proteins such as the M-protein nucleoproteins and polymerases which are genetically stable. The bar chart shows the ability of lymphocytes from a single donor to kill cells which have been transfected with a single flu antigen. This is a measure of the prevalence of cytotoxic T cells for each of the proteins. Most of the cytotoxicity is directed against internal proteins shown on the green bars. Clones of cytotoxic cells against internal proteins usually recognise several of the major strains of flu. In contrast, those which recognise the external proteins are often, but not always strain-specific.

Even when a cytotoxic T cell recognises the same antigen as an antibody, it usually recognises a different portion. For example cytotoxic T cells specific for the haemagglutinin often recognise internal fragments of the antigen rather than the external epitopes recognised by antibodies. Moreover, since cytotoxic T cells recognise antigen presented by MHC molecules, and since MHC molecules are different in each individual, the T cells in each individual recognise different parts of the antigen. These bar charts show the response of two individuals against peptides from the haemagglutinin. T cells from the first person respond to four different regions of the molecule, with highly antigenic peptides centred on residues 100, 180, 300 and 400. T cells from the second individual recognise different regions of the haemagglutinin.

There is some evidence that individuals with specific MHC haplotypes may be more efficient than others at recognising and destroying influenza-infected cells.

In summary, influenza A gives us an example of extreme genetic variability, where successive dominant strains of flu emerge. The new strains are not susceptible to control by antibodies in the host population, and so individuals may suffer from repeated infections. Indeed the general immunity in the host population provides the selective pressure for the emergence of the new strains.