Bats are excellent hosts for viruses: they are numerous, accounting for 20% of all mammals on Earth, and prolific, existing in colonies of up to 20 million individuals. Bats harbour dangerous pathogens that can spread to domestic animals and humans. Ebola, SARS and Nipah viruses have all crossed from bats to humans, either directly or through intermediate hosts1. The discovery in 20122 that bats harbour influenza A viruses was alarming, because flu viruses are notoriously adept at crossing from animals into humans and causing pandemics that have devastating consequences3. Writing in Nature, Karakus et al.4 show that bat flu viruses infect animals using a host cell receptor that is highly similar across species. The findings are a key step towards quantifying the risk to human and animal health that is posed by flu viruses residing in bats.
Wild birds are the natural reservoir of most influenza A viruses. Avian flu viruses infect birds by binding to sialic acid receptors on the host cells (Fig. 1). The cells that line the human respiratory tract also display sialic acid receptors, but these are slightly different from the receptors in birds. Avian flu viruses can acquire the capacity to pass through the air between humans when they undergo mutations in haemagglutinin, a glycoprotein, which forms the spikes on the virus particle that interact with the sialic acid receptors on host cells. The requirement for optimal receptor binding is a major barrier to infection between species that saves us from frequent flu pandemics originating from birds3.
Until the discovery of bat flu viruses, all known influenza A viruses used sialic acid receptors to infect their hosts. It was a huge surprise when studies revealed that bat flu viruses did not use sialic acid receptors to enter cells, and the hunt was on to identify their receptor.
Given that the elusive receptor had been suggested to be made of protein5, the authors developed two genetic approaches to search for it. One approach was to compare total gene expression in cells that were either resistant or susceptible to infection by an artificial virus bearing the bat flu haemagglutinin at its surface. This led to the identification of messenger RNAs encoding cell-surface proteins that were differentially expressed in resistant and susceptible cells. The second approach was to use the CRISPR gene-editing technique to mutate genes in susceptible cells to prevent these genes from being expressed, and then identify those whose loss of expression prevented the artificial virus from entering. Both approaches led to the same conclusion: the bat flu virus entered host cells by the binding of viral haemagglutinin to a protein complex known as major histocompatibility complex (MHC) class II.
MHC class II proteins are an important component of the immune system. Each complex is composed of one α-chain and one β-chain. The complex displays ‘foreign’ molecules, such as those from invading bacteria and viruses, at the surface of specialized immune cells — a process called antigen presentation. The foreign molecules are then recognized by other cells that develop an immune response against the infectious agent.
Notably, Karakus et al. observed that MHC class II proteins from humans, mice, pigs and chickens all functioned as receptors for bat virus haemagglutinin when expressed in human cells. This finding shows that receptor differences are unlikely to pose a barrier against infection by bat flu viruses between species (Fig. 1). Moreover, it suggests that farm animals might be a possible route by which the newly identified flu virus could pass from bats, with which people have infrequent contact, into the human population. This route is reminiscent of that taken by avian flu viruses when they give rise to human pandemics.
The ability of the bat flu virus to use MHC class II proteins from such a broad range of species is perhaps at first surprising. However, chicken MHC class II α-chains are similar to one type of mammalian α-chain6, and such similarities might provide clues to which molecular domains of the receptor are directly involved in its interactions with the virus haemagglutinin.
The identity of this viral receptor raises several questions. Does receptor choice confer an evolutionary advantage? Hijacking MHC class II as a receptor might allow the virus to evade immune surveillance in infected bats. Indeed, MHC class II proteins are the means by which the Epstein–Barr virus infects certain human immune cells, and binding of the virus to the receptor impairs the immune system’s ability to respond7. Many viruses interfere with the expression of, or destroy, their receptors to stop other virus particles from sticking to cells they have just infected and enable their onward spread. Other flu viruses use another spike protein, called neuraminidase, to remove sialic acid receptors from infected cells. Neuraminidase is present in bat flu viruses, but its function is unclear.
Receptor use often determines which cells and tissues a virus can infect. MHC class II proteins are usually thought of as occurring on immune cells, but Karakus et al. show that bat flu viruses infect mice through MHC class II molecules expressed on epithelial cells that line the upper airways. Whether epithelial cells are the target for infection in the natural host is difficult to establish but relevant to address, because infection of particular tissues in bats might affect the likelihood of animal-to-human transmission.
Interestingly, viruses that spread readily between bat species are also more likely to spread to humans8. Virus excretion in saliva, urine or faeces might make transmission to humans easier than an airborne route. Of note, the expression levels of MHC class II proteins in the respiratory epithelium are usually low, but they increase under certain circumstances, such as during viral infections9. Infection with other viruses could thus affect the susceptibility to flu of people or animals exposed to infected bats.
The likelihood of bat viruses spilling over to other species is also influenced by factors such as the bats’ geographical distribution and the exposure of the recipient hosts to the animals1. We lack surveillance data to tell us how widely distributed bat flu viruses are, and whether they are carried by bat species with which humans or domestic animals have close contact. Given that receptor use does not seem to be host-restricted4, and that the enzyme responsible for replicating the bat flu virus seems to function well in human cells2, the lack of human infections by bat flu so far might be due solely to lack of opportunity.
Nature 567, 35-36 (2019)
Competing Financial Interests
W.S.B. is a member of the UK government advisory group on New and Emerging Respiratory Virus Threats.