Host-virus Interactions: Phylogenetic sleuthing reveals pair of paralogous genes

  1. Jamie E Henzy
  2. Welkin E Johnson  Is a corresponding author
  1. Boston College, United States

The genes involved in innate immunity face special challenges that other genes do not. For example, in order to target invading viruses, the immune system needs to distinguish between viral ("non-self") components and "self" components that are meant to be in the host. Complicating this task is the wide assortment of potential pathogens that the host may encounter, combined with the ability of most pathogens to readily evolve to escape the host's immune system. This "moving target" requires evolutionary innovation on the part of the host, which is a tall order for vertebrates because they evolve sluggishly.

Now, in eLife, Matthew Daugherty of the Fred Hutchinson Cancer Research Center (FHCRC) and co-workers report how they have used a combination of phylogenetic sleuthing and a novel assay in yeast to explore the evolution of a family of antiviral proteins called IFIT proteins (Daugherty et al., 2016). These proteins are components of the innate immune system in mammals (Fensterl and Sen, 2015).

When a virus is detected, a signaling protein called interferon is released and it induces an antiviral state by upregulating a whole slew of interferon-stimulated genes. IFIT genes are among the most highly upregulated of these genes, and IFIT proteins target a wide range of viruses (Diamond and Farzan, 2013). It is known that IFIT genes are encoded by a complex multigene locus that varies among different mammalian lineages: for example, a common mouse has six intact copies of IFIT-like genes, whereas a rat has four (Liu et al., 2013). However, despite their importance to innate immunity, there is much we do not know about IFIT proteins.

One puzzle involves the ability of an IFIT protein previously known as IFIT1 to distinguish between cellular (mammalian) mRNAs and viral mRNAs. Mammalian mRNA advertises its selfness by adding a methyl group to the first ribose sugar of the mRNA molecule to produce a "cap1" structure; the unmethylated mRNA molecule is said to have a "cap0" structure (Figure 1; Banerjee, 1980). However, many viruses have evolved ways to fly under the radar either by "snatching" cap1 structures from host mRNAs, or by encoding a methyltransferase enzyme that allows them to make their own cap1 structures (Hyde and Diamond, 2015).

A yeast assay to discriminate between the antiviral proteins IFIT1 and IFIT1B.

IFIT1 and IFIT1B both inhibit the growth of budding yeast (Saccharomyces cerevisiae; top row, middle and right). Co-expressing human Cap1-methyltransferase (which converts cap0-mRNA to cap1-mRNA by adding a methyl group to the first ribose sugar; left) rescued yeast cells expressing IFIT1B (bottom row, right), but not yeast cells expressing IFIT1B (bottom row, middle). Daugherty et al. used this assay to study a wide range of IFIT1/IFIT1B homologs. IFIT: interferon-induced tetratricopeptide repeats.

This disguise appears to work against mouse IFIT1 protein, which exclusively targets cap0 mRNAs, leaving cap1 mRNAs unscathed (Diamond, 2014). Human IFIT1 protein, by contrast, is not tricked by the disguise and targets both cap0 and cap1 viral mRNAs. This suggests that human IFIT1 protein recognizes viral mRNA by a different and as-yet-unknown means. A simple explanation for this difference could be that functional specialization, perhaps driven by challenges from different pathogens, drove the IFIT1 gene to diverge in the mouse and human lineages. This would make the IFIT1 genes orthologs.

Daugherty and co-workers – Aaron Schaller and Harmit Malik, both from FHCRC, and Adam Geballe from FHCRC and the University of Washington School of Medicine – examined this assumption by comparing the IFIT locus among a wide panel of mammalian genomes. This analysis pointed to a gene duplication event that occurred at the IFIT1 locus early in mammalian evolution, more than 100 million years ago. This means that mouse IFIT1 and human IFIT1 are paralogs (caused by a gene duplication event) rather than orthologs (caused by a speciation event), so the authors renamed the mouse gene IFIT1B.

Armed with this new insight into the relationship between human IFIT1 and mouse IFIT1B, Daugherty et al. demonstrated that these two paralogs indeed have distinct functions. To do this, they made use of the fact that budding yeast produces only cap0 structures (Banerjee, 1980). When they expressed mouse IFIT1B in yeast, it targeted cap0 mRNA, in turn inhibiting growth of the yeast, as expected. Growth was rescued when human cap1-methyltransferase was co-expressed with IFIT1B (Figure 1). Human IFIT1, on the other hand, behaved differently. Like mouse IFIT1B, it also inhibited cap0 mRNA and inhibited the growth of the yeast: however, growth was not rescued by co-expressing cap1-methyltransferase. This suggests that IFIT1 depends on something other than cap status to find its target mRNAs. Interestingly, orthologs of IFIT1B from two primates – gibbons and African green monkeys – exhibited the same cap0-targeting behavior as mouse IFIT1B, despite rodents and primates having diverged more than 100 million years ago. Human IFIT1B, by contrast, has lost this particular function.

This study illustrates how a family of vertebrate genes involved in pathogen resistance is continually shaped by the dynamic nature of the multigene locus in which it resides. Such loci frequently undergo duplication, recombination, and gene conversion events (Eirín-López et al., 2012Nei et al., 1997) that can significantly alter the immune repertoire of various species. Gene duplication events, in particular, can result in relaxed selective pressure on one of the copies, allowing the acquisition of novel functions (Lynch and Conery, 2000; Ohno, 1970). As shown here, such an event is the likely basis for the different functions of IFIT1 and IFIT1B. Indeed, based on this study, these two genes appear to have been the most active of the mammalian IFIT genes, undergoing frequent gene birth and loss, as well as recurrent homogenization.

A number of questions remain unanswered: for instance, how does IFIT1 recognize viral mRNA? And what is the function of human IFIT1B (if, indeed, it has a function)? However, the insights that Daugherty et al. glean from applying phylogenetic analyses to functional studies elegantly underscore the value of multigene loci as facilitators of adaptive responses in the ongoing interplay between vertebrates and their pathogens.

References

    1. Banerjee AK
    (1980)
    5'-terminal cap structure in eucaryotic messenger ribonucleic acids
    Microbiological Reviews 44:175–205.

Article and author information

Author details

  1. Jamie E Henzy

    Biology Department, Boston College, Chestnut Hill, United States
    Competing interests
    The authors declare that no competing interests exist.
    ORCID icon "This ORCID iD identifies the author of this article:" 0000-0002-1499-4857
  2. Welkin E Johnson

    Biology Department, Boston College, Chestnut Hill, United States
    For correspondence
    welkin.johnson@bc.edu
    Competing interests
    The authors declare that no competing interests exist.
    ORCID icon "This ORCID iD identifies the author of this article:" 0000-0001-5991-5414

Publication history

  1. Version of Record published: May 31, 2016 (version 1)

Copyright

© 2016, Henzy et al.

This article is distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use and redistribution provided that the original author and source are credited.

Metrics

  • 1,111
    views
  • 135
    downloads
  • 0
    citations

Views, downloads and citations are aggregated across all versions of this paper published by eLife.

Download links

A two-part list of links to download the article, or parts of the article, in various formats.

Downloads (link to download the article as PDF)

Open citations (links to open the citations from this article in various online reference manager services)

Cite this article (links to download the citations from this article in formats compatible with various reference manager tools)

  1. Jamie E Henzy
  2. Welkin E Johnson
(2016)
Host-virus Interactions: Phylogenetic sleuthing reveals pair of paralogous genes
eLife 5:e17224.
https://doi.org/10.7554/eLife.17224

Further reading

    1. Microbiology and Infectious Disease
    Alejandro Prieto, Luïsa Miró ... Antonio Juarez
    Research Article

    Antimicrobial resistance (AMR) poses a significant threat to human health. Although vaccines have been developed to combat AMR, it has proven challenging to associate specific vaccine antigens with AMR. Bacterial plasmids play a crucial role in the transmission of AMR. Our recent research has identified a group of bacterial plasmids (specifically, IncHI plasmids) that encode large molecular mass proteins containing bacterial immunoglobulin-like domains. These proteins are found on the external surface of the bacterial cells, such as in the flagella or conjugative pili. In this study, we show that these proteins are antigenic and can protect mice from infection caused by an AMR Salmonella strain harboring one of these plasmids. Furthermore, we successfully generated nanobodies targeting these proteins, that were shown to interfere with the conjugative transfer of IncHI plasmids. Considering that these proteins are also encoded in other groups of plasmids, such as IncA/C and IncP2, targeting them could be a valuable strategy in combating AMR infections caused by bacteria harboring different groups of AMR plasmids. Since the selected antigens are directly linked to AMR itself, the protective effect extends beyond specific microorganisms to include all those carrying the corresponding resistance plasmids.

    1. Microbiology and Infectious Disease
    Hideo Fukuhara, Kohei Yumoto ... Katsumi Maenaka
    Research Article

    Canine distemper virus (CDV) belongs to morbillivirus, including measles virus (MeV) and rinderpest virus, which causes serious immunological and neurological disorders in carnivores, including dogs and rhesus monkeys, as recently reported, but their vaccines are highly effective. The attachment glycoprotein hemagglutinin (CDV-H) at the CDV surface utilizes signaling lymphocyte activation molecule (SLAM) and Nectin-4 (also called poliovirus-receptor-like-4; PVRL4) as entry receptors. Although fusion models have been proposed, the molecular mechanism of morbillivirus fusion entry is poorly understood. Here, we determined the crystal structure of the globular head domain of CDV-H vaccine strain at 3.2 Å resolution, revealing that CDV-H exhibits a highly tilted homodimeric form with a six-bladed β-propeller fold. While the predicted Nectin-4-binding site is well conserved with that of MeV-H, that of SLAM is similar but partially different, which is expected to contribute to host specificity. Five N-linked sugars covered a broad area of the CDV-H surface to expose receptor-binding sites only, supporting the effective production of neutralizing antibodies. These features are common to MeV-H, although the glycosylation sites are completely different. Furthermore, real-time observation using high-speed atomic force microscopy revealed highly mobile features of the CDV-H dimeric head via the connector region. These results suggest that sugar-shielded tilted homodimeric structure and dynamic conformational changes are common characteristics of morbilliviruses and ensure effective fusion entry and vaccination.