Epigenetics is a hot new research field, but it seems that the influenza virus already has it figured out. By mimicking epigenetic regulation in human cells, one flu strain suppresses the expression of antiviral genes. See Article p.428
Influenza virus infections are sometimes severe, or even deadly, but most people recover within a few days, helped by their body's immune system. Yet this apparently simple interaction between the virus and its host is deceptive. Influenza viruses are, in fact, masters at circumventing their hosts' defences — they co-opt cellular protein-synthesis pathways to produce viral proteins, for example, and antagonize aspects of the immune response. On page 428 of this issue, Marazzi et al.1 describe another viral evasion tactic. They show that an influenza protein called NS1 mimics a host-cell histone protein that is involved in regulating gene expression. By hijacking this regulatory machinery, the virus inhibits the cell's production of antiviral proteins*.
Cells tightly regulate the expression of specific genes according to an organism's developmental stage and physiological state, and in response to environmental stimuli such as infection. Gene expression can be altered in various ways, including by transcription-factor proteins, which bind to DNA and regulate the transcription of certain genes, or through epigenetic modification — heritable chemical changes to DNA and associated proteins that do not affect the DNA sequence but alter the rate of transcription. Epigenetics has rapidly become a much-studied topic since these regulatory mechanisms were first reported around 20 years ago.
A common epigenetic alteration is the addition of methyl or acetyl groups to histone proteins, which are the 'building blocks' of chromatin, the scaffold that packages DNA into chromosomes. The four main histone proteins (H2A, H2B, H3 and H4) consist of a globular domain and an unstructured amino-terminal tail region that is dominated by the positively charged amino acids lysine and arginine. These amino acids can be modified by enzymes that add or remove methyl or acetyl groups2. Such modifications alter the histone–DNA interaction, which in turn leads to dynamic changes in chromatin that allow specific areas of the chromatin to unfold, or other proteins to be recruited to the site. Thus, by regulating the accessibility of a particular region of DNA, epigenetic modifications can regulate gene expression.
Some epigenetic changes are generally associated with a site of active gene transcription3. An example of such a change is methylation at the lysine residue (K) of the amino acid sequence designated ARTK in histone H3 (known as the H3K4 modification). Marazzi et al.1 found that the H3N2 influenza virus contains an amino-acid sequence (ARSK) very similar to the histone's ARTK sequence (Fig. 1). This mimic sequence is found in the virus's NS1 protein, which is not essential for viral structure but is known to have other roles in evading the host's immune system4. The authors also show that this sequence similarity is functional — the NS1 tail can serve as a substrate for the histone-modifying enzyme Set1, a lysine methyltransferase.
Marazzi and colleagues also demonstrate that H3N2's histone-mimic tail binds directly to the transcription-elongation complex PAF1C (Fig. 1). PAF1C accompanies the transcription enzyme RNA polymerase II during the formation of messenger RNA, and is thought to recruit the Set1 enzyme that leads to H3K4 methylation5. The authors propose that NS1 binding to PAF1C interferes with gene transcription in the host cell. Indeed, the researchers found that when cells from human lung tissue were infected with the H3N2 virus, transcription of rapidly inducible genes was halted. But when the cells were infected with H3N2 viruses that had been mutated so that they could no longer bind PAF1C, transcription of these genes was unaffected. Furthermore, the authors provide compelling evidence that it is specifically the transcription of antiviral genes that is interfered with by the interaction between NS1's histone-like tail and PAF1C. Thus, histone mimicry seems to support viral infection, and may give a selective advantage to viruses that have this ability.
Is this mechanism unique to H3N2 viruses, and could it be a recent evolutionary adaptation? Moreover, does possession of this mechanism have a role in the variation in infectivity and disease severity that is seen among influenza strains? The H3N2 virus is an emerging influenza subtype that has caused only a limited number of human infections during the 2011–12 flu season. However, they seem to cause illness of around the same severity as that induced by standard seasonal influenza viruses, which do not contain the NS1 histone-mimic tail1. It will be necessary to assess other emerging influenza strains for histone mimicry to determine whether this mechanism is evolutionarily new, and to study more influenza cases to predict its potential impact on public health. Even if histone mimicry is not widespread and does not dramatically alter disease severity, Marazzi and colleagues' discovery may provide a novel target for antiviral drugs aimed at modulating the host's antiviral response.
The finding raises another intriguing question: what would happen if a person is concurrently infected with H3N2 virus and a more virulent influenza virus, such as the highly pathogenic avian H5N1? Or what might occur if an H5N1 virus acquires the NS1 sequences required for the protein to function as a histone mimic? The combined effect of a suppressed antiviral response and a highly virulent virus might lead to devastating consequences for the host. Alternatively, if an overly active host response is itself the main culprit in causing severe influenza, as has been suggested6, a suppressed antiviral response could actually reduce disease pathology.
Research aimed at addressing these questions may be hampered by an increasing wariness about generating recombinant H5N1 viruses7. But ongoing influenza research remains essential. Although epigenetic gene regulation is new to scientists, other mechanisms of gene regulation no doubt remain to be discovered, and we can be certain that viruses have already evolved to interfere with these regulatory processes, and to use them to their advantage.
*This article and the paper1 under discussion were published online on 14 March 2012.