The role of TRPM8 in susceptibility to highly pathogenic avian influenza A/H5N1

In the last decade an animal influenza A virus has crossed the species barrier and caused an appreciable number of human cases on five occasions. Highly pathogenic avian influenza A(H5N1) has resulted in very rare episodes of limited person-to-person transmission but has not adapted to humans. Another avian influenza virus (H7N7) caused a large number of, mostly mild, human infections in the Netherlands in 2003 and also demonstrated some ability for limited person-to-person transmission, but was successfully controlled by extensive culling of poultry flocks. The swine H1N1 influenza virus that emerged in 2009 successfully adapted to humans, caused a pandemic, and is now firmly established as a human influenza virus. Another swine-origin virus (H3N2v) has been transmitted occasionally from pigs to humans in the U.S. since 2011 but has not managed to switch hosts and become established in the human population. The ongoing avian influenza A(H7N9) outbreak that was first detected in China in February 2013 has caused more than 130 human cases at the time of writing, and shows some evidence of adaptation to humans.

A central question that arises when an animal influenza virus crosses the species barrier and infects humans is the probability that the virus will adapt to humans and cause a pandemic. Despite much research, the barriers that must be overcome for animal influenza viruses to stably adapt to humans are still not fully understood. Assessing the public health threat of H5N1 and other animal influenza viruses requires a fuller understanding of what currently constrains the ability of animal influenza viruses to infect humans and to transmit between humans.

The characteristics of influenza viruses that are associated with adaptation to humans have been studied in detail. However, the human-host characteristics are much less well studied. Around one third of all human H5N1 cases have occurred in clusters, and 50 of the 54 H5N1 clusters reported as of March 2009 were comprised entirely of blood relatives. We and other authors have concluded that this familial clustering and other aspects of the epidemiology of H5N1 suggest a role for host genetic factors.

To test our hypothesis that host genetic factors may play an important role in restricting susceptibility to H5N1, we conducted a study to look for human genetic factors that might be associated with an increased risk of getting H5N1 infection. We studied 65 people with H5N1 infection (cases) and 2,910 people without H5N1 infection (controls) and found a variant in one particular gene (TRPM8) that was more common in cases compared to controls. We went on to study the effect of this gene on H5N1 infection in cells in the laboratory and in mice. We found that this gene did indeed affect the ability of the H5N1 virus to replicate in cells and to cause disease in mice.

The aim of this new research is to futher determine and characterize the role of the TRPM8 gene in susceptibility to infection with H5N1. We will do this by (1) gathering samples and analysing the genetic code from human H5N1 cases in Indonesia to test if the genetic variant is also present this independent set of samples, (2) fully sequencing the whole TRPM8 gene in cases and controls to identify precisely which genetic variations may be responsible, (3) exploring how different influenza viruses bind to the protein produced by the TRPM8 gene, and (4) further assessing the role of TRMP8 in H5N1 infection in mice.

This research is important since the results will provide new information on the adaptations necessary for animal influenza viruses to succesfully infect humans and may also provide a new target for drugs to prevent or treat influenza infections.

We have previously identified an association between the sialated transmembrane ion channel TRPM8 and susceptibility to H5N1 disease through a genome-wide association study (GWAS) and functional work in cell-lines and a murine model. This grant will support four work-streams that together will provide a full picture of the role of TRPM8 in susceptibility to H5N1.

Work-stream 1 [Horby] will assemble DNA samples from an estimated 50 H5N1 cases in Indonesia as an independent replication sample set. We will use a retrospective case-only study design to assess if the allele distribution in cases differs from that of the known population minor allele frequency. The allele frequency in cases will be assessed by polymerase chain reaction (PCR) at the National Institute of Health Research and Development, Indonesia.

Work-stream 2 [Hibberd] will re-sequence the entire TRPM8 gene of H5N1 cases and controls, which spans a 100Kb region in chromosome 2, in order to identify the full collection of genetic variants within the region of interest. We will sequence all cases (greater than 38) and 4 controls per case, which have sufficient DNA available from the earlier GWAS (50ng/ul x 2 multiplexed pools).

Work-stream 3 [Gamblin and Skehel] will assess the affinity, specificity, and structural properties of binding of the haemagglutinin and neuraminidase of H5N1 and other influenza subtypes to TRPM8. Receptor binding affinity measurements will use the Fortebio Octet and microscale thermophoresis. Structural analysis of complexes formed will be done by EM and X-ray crystallography. The kinetic properties of the neuraminidase will be also be assessed.

Work-stream 4 [Webby] will examine the impact of TRMP8 on H5N1 viral load, virus tissue distribution, lung pathology, and hematopoietic immune responses in a murine model, using TRMP8-null mice and and littermate wild type control animals infected with a sub-lethal dose of A/Chicken/Egypt/1/2006 (H5N1).

The UK National Risk Register 2012 cites pandemic influenza as 'the most significant civil emergency risk', yet assessing the pandemic risk of animal influenza viruses remains difficult. The pandemic potential of highly pathogenic avian influenza (HPAI) H5N1 has been used to justify enormous investments in the control of the epizootic and in pandemic preparedness; it is estimated that the cost of the H5N1 outbreaks since 2003 is in the billions. Yet despite extensive circulation across large parts of Asia for many years, H5N1 has shown no evidence of adaptation to humans. The recent emergence of an avian influenza H7N9 virus that shows some features of adaptation to mammalian hosts represents a similar risk assessment dilemma, but with the added complication of a low pathogenic phenotype in poultry.

If, as hypothesized, TRPM8 is a target for influenza H5N1 virus binding and/or release in humans, this host-pathogen interaction may be an important factor that has constrained the ability of highly pathogenic H5N1 to infect humans and to transmit from person-to-person. Characterization of this interaction may lead to new ways for assessing the pandemic potential of established and new animal influenza viruses. The economic benefits of improved assessments of the probability that any animal influenza virus will stably adapt to humans are potentially enormous. It would allow Governmental and international public health agencies to better target surveillance, assess risk, and develop evidence based public communications and policies. Broader economic benefits would include cost savings resulting from better-calibrated risk assessments, risk mitigation and policy responses to novel influenza viruses. Increasing the reliability of public communications about pandemic risk would improve confidence in public bodies, which was somewhat weakened by the 2009 H1N1 pandemic.

This research may also lead to commercial opportunities through the development of new therapeutics targeting the TRPM8 and similar transmembrane proteins.