(C) PLOS One
This story was originally published by PLOS One and is unaltered.
. . . . . . . . . .



Cellular receptors for mammalian viruses [1]

['Ana Valero-Rello', 'Institute For Integrative Systems Biology', 'Consejo Superior De Investigaciones Científicas-Universitat De València', 'Paterna', 'València', 'Carlos Baeza-Delgado', 'Iván Andreu-Moreno', 'Rafael Sanjuán']

Date: 2024-02

The interaction of viral surface components with cellular receptors and other entry factors determines key features of viral infection such as host range, tropism and virulence. Despite intensive research, our understanding of these interactions remains limited. Here, we report a systematic analysis of published work on mammalian virus receptors and attachment factors. We build a dataset twice the size of those available to date and specify the role of each factor in virus entry. We identify cellular proteins that are preferentially used as virus receptors, which tend to be plasma membrane proteins with a high propensity to interact with other proteins. Using machine learning, we assign cell surface proteins a score that predicts their ability to function as virus receptors. Our results also reveal common patterns of receptor usage among viruses and suggest that enveloped viruses tend to use a broader repertoire of alternative receptors than non-enveloped viruses, a feature that might confer them with higher interspecies transmissibility.

Specific interactions between viruses and cellular entry factors are a critical initial step in viral infection. The identification of virus receptors and other key entry factors is important for understanding viral tropism, pathogenesis and cross-species transmissibility, as well as for the design of antiviral drugs. Here, we provide a comprehensive meta-analysis of our current knowledge of virus receptors and attachment factors. We use the newly assembled dataset to reveal general patterns of receptor use across viruses and to derive predictions about the propensity of each cell surface protein to serve as a virus receptor. This work may assist future research on receptor discoveries, and suggests new implications of virus-receptor interactions, including viral emergence.

Funding: This work was financially supported by Advanced Grant 101019724—EVADER from the European Research Council, grant PID2020-118602RB-I00—ZooVir from the Spanish Ministerio de Ciencia e Innovación, and grant 202130-31 from Fundació La Marató de TV3 to R.S. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. R.S. and A.V-R. received a salary from the European Research Council. C.B-D. received a salary from the Fundació La Marató de TV3.

Currently, virus receptors are collected in a few databases, such as ViralZone ( viralzone.expasy.org ), KEGG (genome.jp/kegg), and VTHunter ( db.cngb.org/VThunter ). Moreover, previous articles have relied on this information in combination with different search strategies to investigate viral entry factors [ 18 – 21 ]. Overall, these databases and previous works report about 100 cellular receptors for a similar number of viruses. However, they may not provide a comprehensive view of the literature, and their content is probably biased towards human and economically relevant viruses. Here, we aim to provide a more thorough analysis of the actual diversity of known receptors and attachment factors used by mammalian viruses. To achieve this goal, we implemented systematic and semi-automated search strategies that allowed us to double the amount of information extracted from the literature compared to previous work, and to pinpoint the role of each cellular factor in viral entry. Using machine learning, we identify cell surface proteins that are more likely to function as virus receptors, as well as common features of these proteins. We also explore how the repertoire of cellular receptors varies according to viral species, families and other viral features, and show that this repertoire correlates with viral cross-species transmissibility.

However, virus receptor studies can be technically challenging, particularly due to the complex nature of viral entry, which often involves redundant receptors, co-receptors, and accessory receptors, as well as different attachment factors. The use of multiple functional receptors by viruses has been extensively documented, two examples being SARS-related [ 12 , 13 ] and Zika [ 14 ] viruses. Strategies for the identification of cellular factors determining viral entry include systematic perturbation methods such as RNAi, CRISPR-Cas, and overexpression of candidate genes, as well as biochemical and biophysical methods used to demonstrate and quantify virus-receptor binding, including protein microarrays, affinity-purification mass spectrometry, biolayer interferometry, and plasmon resonance [ 15 , 16 ]. Virus receptor inference can involve full viruses or use pseudotypes, an approach that focuses precisely on viral entry and allows handling non-culturable viruses [ 17 ].

Mammals can be infected by thousands of viruses belonging to tens of different families. Cellular receptors and other entry factors critically determine infectivity and play a major role in viral cross-species transmission [ 1 – 4 ]. There are numerous examples showing the evolution of key mutations in viral receptor-binding proteins that promote transmissibility. For instance, certain changes in the influenza virus hemagglutinin determine sialic acid preferences and the ability of avian strains to infect humans [ 5 , 6 ]. Similarly, changes in the affinity of the viral spike protein for human ACE2 have been instrumental in the emergence and evolution of SARS-CoV-2 [ 7 , 8 ]. Conversely, several receptor-coding genes have evolved under virus-driven selection, such NPC1 in bats [ 9 ] and TFRC in rodents [ 10 ], among others. Therefore, the identification of cellular factors involved in viral entry is a cornerstone in our understanding of viral tissue tropism and pathogenesis. The discovery and characterization of virus receptors also facilitate the development of entry inhibitors and allow the targeting of therapeutic viruses to specific cells [ 11 ].

Results

Overview of the patterns of receptor use across viruses Among the total 201 individual proteins or protein complexes identified, some were used by many viruses, the most frequent being different integrin subunits, followed by CD209 (DC-SIGN) and CLEC4M (C-type lectins), HAVCR1 (TIM-1), TFRC and AXL, each associated to >10 different viruses belonging to more than five families (Fig 2). As previously noted [3,22], this underscores that viral entry frequently exploits cellular functions related to cell-cell adhesion (e.g. integrins), carbohydrate-mediated signalling (lectins), and autophagy (HAVCR1, AXL), and shows that non-proteinaceous components of the virion surface can play a central role in this process. For instance, carbohydrates in viral surface glycoproteins can bind lectins [23–25], and HAVCR1 can interact with the lipid membrane of many enveloped viruses to promote viral endocytosis in a process known as apoptotic mimicry [26,27]. PPT PowerPoint slide

PNG larger image

TIFF original image Download: Fig 2. Heat map of the 50 cellular proteins most frequently used as receptors by mammalian viruses. HUGO gene names are shown, except for the HLA, VGGC, and integrin (ITGs) complexes. Shades of grey indicate the number of viruses known to use each protein as a receptor. Bars on the left show the total number of viruses (light grey) and viral families (dark grey) using each protein (top three values indicated). Brackets on the right indicate protein clusters obtained in a hierarchical cluster analysis, using the cosine similarity metric to group proteins according to levels of virus sharing. Three such clusters (C1-C3) are highlighted. In columns, viruses are aggregated by viral families. Baltimore group, taxonomical order, and whether the virus is enveloped are indicated. For each family, grey bars at the bottom show the number of known virus-receptor pairs (light grey; top values indicated), and the number of distinct receptors used (dark grey). Ro: Rowavirales; Se: Sepolyvirales; Zu: Zurhausenvirales; Ch:Chitovirales; He: Herpesvirales; Pc: Piccovirales; Bl: Blubervirales; Or: Ortervirales; Re: Reovirales; Pi: Picornavirales; Am: Amarillovirales; Ma:Martellivirales; Ni: Nidovirales; Ar: Articulavirales;Bu: Bunyavirales; Mo: Mononegavirales. https://doi.org/10.1371/journal.ppat.1012021.g002 A hierarchical cluster analysis in which entry factors were grouped according to virus sharing suggested some general patterns (Fig 2). A large group (C1) was formed by host proteins with heterogeneous functions used predominantly by plus-strand RNA viruses, particularly flaviviruses, coronaviruses, and togaviruses. Another cluster (C2) included AXL, TYRO3, HAVCR1, and NPC1, which are frequent receptors for flaviviruses, togaviruses, arenaviruses, and filoviruses. These viruses are often internalized nonspecifically in cells by apoptotic mimicry and use downstream specific receptors that mediate membrane fusion in the endosome, such as NPC1. Other receptor clusters were associated with a given viral family, such as solute carriers and immune signalling proteins for retroviruses (C3).

Variations in the type and number of receptors used by different viruses Our tentative functional classification of receptors suggested that non-enveloped viruses tend to rely on a single receptor more often than enveloped viruses, which on the contrary are more likely to use alternative receptors, each sufficient for entry (Table 2 and S2 Fig). We also found that non-enveloped viruses are three times more often reported to use carbohydrate moieties associated with undefined proteins than enveloped viruses (34.6% versus 11.7% of all interactions, respectively). Dependence on such moieties is seemingly strongest for caliciviruses and polyomaviruses (60.0% and 80.0% of the total interactions, respectively) and weakest for retroviruses (1.3%). We then examined in more detail how many different cellular proteins are used as receptors by each virus. We found ample variation across viral families, with over 50 described for flaviviruses, retroviruses, and herpesviruses, whereas other families such as Circoviridae, Nairoviridae, Bornaviridae, and Polyomaviridae showed five or fewer (Fig 2). However, these observations can be strongly biased by the research effort dedicated to each virus. To address this, we used the number of publications in PubMed as a proxy of research effort, and estimated the effect of this variable on the known number of proteins used as receptors by each virus, using a generalized linear model (GLM; Fig 4A). This allowed us to identify viruses that use more receptors than expected from research effort alone, such as hepatitis C virus, dengue virus, HIV-1 and HIV-2, and Ebola virus (Fig 4B). In some cases, such as HIV-2, the excess of known receptors could be explained by knowledge acquired from a related virus, whereas other viruses may be subject to research biases not accounted for here, or might be truly promiscuous in terms of receptor usage. Some highly studied viruses did not exhibit particularly high numbers of host proteins used as receptors, such as influenza viruses, and some enteric viruses. Specifically, according to our search, no cellular proteins have been found to serve as receptors for influenza B or Norwalk viruses despite these being well-studied viruses research and the fact that several receptors have been described for the related influenza A virus and murine norovirus, respectively. PPT PowerPoint slide

PNG larger image

TIFF original image Download: Fig 4. Variation in the number of known receptors across viruses. A. Relationship between research effort, measured the number of PubMed records and the number of different proteins used as receptors for each virus. The dashed line shows the expected number of receptors obtained from a GLM (null model), and dotted lines correspond to 95% confidence intervals. B. Known receptors versus the expected number under the null model shown in panel A. The dashed line indicates no deviation from the model, and the dotted lines the 95% confidence interval. Viruses shown in yellow exceed the upper limit of the confidence interval and have more than 10 known receptors. Viruses shown in blue fall below the lower limit of the confidence interval, and their expected number of receptors according to the model is higher than five. CVB: Coxsackievirus B (enterovirus B); DENV: dengue virus; EBOV: Zaire ebolavirus; HCV: hepatitis C virus (hepacivirus C); HSV-1: herpes simplex virus 1 (human alphaherpesvirus 1); PV: poliovirus (enterovirus C); SINV: Sindbis virus; WNV: West Nile virus. https://doi.org/10.1371/journal.ppat.1012021.g004 To test for more general patterns, we added to our GLM two more factors: the viral family, and whether or not the virus is enveloped. This showed that, overall, enveloped viruses use a wider variety of host proteins as receptors than non-enveloped viruses, and also more receptors that are sufficient for entry according to the literature (main or alternative receptors; P < 0.001). We inferred from this model that, after controlling for research effort, enveloped viruses use on average 2.4 times more cellular proteins as receptors than non-enveloped viruses. This excess was more marked for phenuiviruses, togaviruses, and filoviruses, whereas polyomaviruses and caliciviruses showed particularly low numbers of such receptors (S3 Fig).

[END]
---
[1] Url: https://journals.plos.org/plospathogens/article?id=10.1371/journal.ppat.1012021

Published and (C) by PLOS One
Content appears here under this condition or license: Creative Commons - Attribution BY 4.0.

via Magical.Fish Gopher News Feeds:
gopher://magical.fish/1/feeds/news/plosone/

You are viewing proxied material from magical.fish. The copyright of proxied material belongs to its original authors. Any comments or complaints in relation to proxied material should be directed to the original authors of the content concerned. Please see the disclaimer for more details.