(C) PLOS One [1]. This unaltered content originally appeared in journals.plosone.org.
Licensed under Creative Commons Attribution (CC BY) license.
url:https://journals.plos.org/plosone/s/licenses-and-copyright

------------



A genome-wide CRISPR/Cas9 gene knockout screen identifies immunoglobulin superfamily DCC subclass member 4 as a key host factor that promotes influenza virus endocytosis

['Yangming Song', 'College Of Veterinary Medicine', 'Gansu Agricultural University', 'Lanzhou', 'People S Republic Of China', 'State Key Laboratory Of Veterinary Biotechnology', 'Harbin Veterinary Research Institute', 'Chinese Academy Of Agricultural Sciences', 'Harbin', 'Haixiang Huang']

Date: 2022-01 

Influenza virus infection is dependent on host cellular factors, and identification of these factors and their underlying mechanisms can provide important information for the development of strategies to inhibit viral infection. Here, we used a highly pathogenic H5N1 influenza virus to perform a genome-wide CRISPR/Cas9 gene knockout screen in human lung epithelial cells (A549 cells), and found that knockout of transmembrane protein immunoglobulin superfamily DCC subclass member 4 (IGDCC4) significantly reduced the replication of the virus in A549 cells. Further studies showed that IGDCC4 interacted with the viral hemagglutinin protein and facilitated virus internalization into host cells. Animal infection studies showed that replication of H5N1 virus in the nasal turbinates, lungs, and kidneys of IGDCC4-knockout mice was significantly lower than that in the corresponding organs of wild-type mice. Half of the IGDCC4-knockout mice survived a lethal H5N1 virus challenge, whereas all of the wild-type mice died within 11 days of infection. Our study identifies a novel host factor that promotes influenza virus infection by facilitating internalization and provides insights that will support the development of antiviral therapies.

Influenza virus infection is initiated by the attachment of the viral HA protein to sialic acid receptors on the host cell surface; most of the virus particles enter cells through clathrin-mediated endocytosis (CME). However, it is still largely unknown which protein(s) play(s) a role in transmitting the signal of viral binding across the plasma membrane to initiate CME. In this study, we found that the single-pass type I transmembrane protein immunoglobulin superfamily DCC subclass member 4 (IGDCC4) plays a key role in influenza virus internalization into host cells. IGDCC4 knockout dramatically increased the ability of mice to resist influenza virus infection. Our study provides insights that will support the development of therapies against influenza.

Funding: The study was supported by the grant 31521005 from the Innovative Research Group Project of National Natural Science Foundation of China to HC. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

Copyright: © 2021 Song et al. This is an open access article distributed under the terms of the Creative Commons Attribution License , which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.

Viruses rely on the host cellular machinery to complete their life cycles, and some host factors play pivotal roles in viral replication. Such host factors are ideal drug targets or therapeutic targets because they will not mutate under drug- or vaccine-mediated selective pressure. Screening and identifying key host factors required for viral replication will provide important insights into potential antiviral drug targets. Genome-wide screening methods have been widely used, including small interfering RNA (siRNA), proteomics, and insertional mutagenesis screening, to identify host factors involved in influenza virus infection and infections by other viruses such as HIV [ 14 – 20 ]. Since 2013, CRISPR/Cas9 technology has been gradually used to perform genome-wide screens for host factors of interest in mammalian cells [ 21 , 22 ]. Unlike RNAi inhibition, CRISPR/Cas9 decreases gene expression at the DNA level, and can be used to study certain phenotypes that require knockout levels, or it can be used to study non-transcriptional regions and non-coding RNAs. CRISPR/Cas9 genome-wide screens have proven to be more efficient and reliable to identify host factors required for viral replication [ 23 – 28 ].

Vaccination and antiviral drugs are effective ways to prevent influenza virus infection. Although vaccination provides adequate protection against seasonal influenza infection, the appearance of antigenic variants caused by antigen shift and antigen drift is faster than the development of new vaccines. At the beginning of an emerging pandemic, antiviral drugs will be on the frontline to prevent virus infection while a corresponding specific vaccine is developed [ 5 ]. There have been only three licensed classes of antiviral drugs for influenza virus infection: M2 channel inhibitors (e.g., amantadine), neuraminidase inhibitors (e.g., oseltamivir and zanamivir), and viral RNA polymerase inhibitors (e.g., favipiravir). But the emergence of drug-resistant viruses has become a serious problem. Adamantanes are no longer clinically used because most of viruses are resistant [ 6 , 7 ]. Moreover, some seasonal viruses with high levels of resistance to oseltamivir and oseltamivir-resistant A(H7N9) viruses in humans have been identified [ 12 ]. Favipiravir is a broad-spectrum antiviral that was recently developed to prevent influenza virus infection, but it also still faces the issue of the emergence of resistant variants with its extensive use. Two mutations in the PB2 and PA genes that provide robust resistance to favipiravir have been identified in the laboratory setting [ 13 ]. Therefore, there is an urgent need to identify novel targets for influenza antiviral drug development.

Influenza virus continuously evolves in nature and severely threatens both human and animal health with considerable impact on the global economy. Influenza virus causes human seasonal epidemics with 290,000 to 650,000 deaths annually worldwide, and caused serious pandemics with high morbidity and mortality in humans in 1918, 1957, 1968, and 2009 [ 1 , 2 ]. Highly pathogenic avian influenza viruses, including H5 and H7 viruses, are always present in waterfowl and wild birds, and have sporadically crossed the species barrier to threaten public health [ 3 – 5 ]. H5N1 viruses have caused disease outbreaks in poultry and wild birds in more than 60 countries across three continents since 2003 [ 6 – 8 ], and over 850 human infections have been reported in 16 countries with a mortality rate of nearly 52% [ 9 ]. H7N9 viruses caused over 1,560 human infections in China and a mortality rate of 39% [ 10 , 11 ].

Results

Establishment of Genome-wide CRISPR/Cas9 screening in A549 cells To identify host factors required for influenza virus infection, we performed a genome-wide CRISPR/Cas9 screen in A549 cells as described previously [21]. An A549-Cas9 AAVS1 cell line stably expressing Cas9 protein was purchased from GeneCopoeia (Cat: SCL-76321-G2), and subcloned to obtain clones with high Cas9 protein expression based on copGFP intensity by using FACS sorting. We transduced the A549-Cas9 cells with lentivirus pools for human GeCKO v2 single guide RNA (sgRNA) libraries A and B, respectively. The two libraries contained, in total, 123,411 unique sgRNAs targeting 19,050 genes (www.addgene.org). To ensure the integrity of the cell libraries, each sgRNA covered at least 1,000 cells. Each cell library was cultured for 14 days with puromycin pressure to remove cells without transduced sgRNA. We then performed high-throughput sequencing for plasmid libraries with one amplification in E. coli and the transduced cell libraries to confirm library integrity (Fig 1A). The plasmid libraries A and B contained 99.10% and 99.93% sgRNAs, respectively, compared with the reference libraries from Addgene. Some sgRNAs may target host genes essential for cell survival, so the transduced cell libraries A and B were just maintained with 89.98% and 82.55% sgRNAs, respectively (Fig 1B). Thus, we generated pooled GeCKO libraries with sufficient coverage of sgRNAs in A549 cells (A549-GeCKO cell libraries) to be used to screen host factors essential for influenza virus infection. PPT PowerPoint slide

PNG larger image

TIFF original image Download: Fig 1. Genome-wide CRISPR/Cas9 screening of A549 cells for host factors required for highly pathogenic H5N1 influenza virus infection. (A) Schematic of the generation of the A549-GeCKO libraries by using lentivirus sgRNA libraries and A549 cells expressing Cas9 (A549-Cas9). A549-Cas9 cells were transduced with lentivirus containing sgRNA libraries A and B, respectively, and were maintained in puromycin for 14 days to generate the A549-GeCKO library, which was then deep sequenced. (B) Coverage of amplified library plasmids and GeCKO library cells compared with sgRNA lists from Addgene detected by high-throughput sequencing. (C) Schematic of screening host factors associated with H5N1 virus infection. (D) The heat map plots showing sgRNA abundance of the selected host factors in different screens. (E) H5N1 virus replication titers in different gene knockout cells. The cells were infected with H5N1 virus at an MOI of 0.01; the supernatants were collected at 48 h post-infection for viral titration in chicken embryos. The data shown are the means ± SDs of three biological repeats. The two-tailed unpaired t-test was used for the statistical analysis. **, p < 0.01, ***, p < 0.001. https://doi.org/10.1371/journal.ppat.1010141.g001

Screening and identifying host factors required for H5N1 influenza virus infection Next, we infected the pooled A549-GeCKO cell libraries with A/Anhui/1/2005(H5N1) virus (H5N1 virus) at a multiplicity of infection (MOI) of 3; this virus was isolated from a patient with a fatal outcome in China in 2005 [29]. Surviving cells were reseeded and expanded for two more rounds of H5N1 virus infection. Then genomic DNA was extracted from the surviving cells, and integrated sgRNAs were amplified by PCR and sequenced (Fig 1C). We repeated the screening three times, and identified robust enrichment of 979 sgRNAs (≥15 reads) targeting 648 human genes (S1 Table), of which 10 host genes with the highest enrichment were identified in three independent experiments (Fig 1D). Two of these ten genes, SLC35A1 and ST3GAL4, have previously been identified by others [30, 31]. Host dependency factors acquired by using genome-scale CRISPR knockout screens must be further validated to exclude false positives due to bad sequencing readouts, copy number variants, or off-target sgRNAs. H5N1 virus replicated similarly in A549 cells transduced with lentivirus bearing the non-targeted gRNA (A549-NT) and wild-type A549 cells, indicating that lentivirus transduction does not affect influenza virus replication in A549 cells (S1 Fig). To further validate whether our candidate genes are important for influenza virus replication, we selected five of the top 10 candidates, including the previously reported SLC35A1 [30], and constructed their polyclonal knockout cell lines. The polyclonal cells for the five genes were infected with H5N1 virus at an MOI of 0.01. Supernatants were collected for virus titration at 48 h post-infection (p.i.). Viral titers in the SLC35A1-, IGDCC4-, and ZFAT-knockout cells were significantly lower than those in the control A549 cells, but knockout of PACSIN2 and ZNF471 did not significantly affect H5N1 virus replication (Fig 1E).

IGDCC4 is important for the early step of influenza virus replication IGDCC4 may encode two proteins via alternative splicing (www.uniprot.org): isoform 1 is a single-pass type I transmembrane protein, and isoform 2 lacks a signal peptide and may be an intracellular protein. Compared with isoform 1, isoform 2 lacks 270 amino acids at the N-terminus, and its 61 amino acids at the N-terminus differ from amino acids 271–331 of isoform 1 (Fig 2A). The function of IGDCC4 in viral replication has never been studied. We therefore further investigated the effect of IGDCC4 on influenza virus infection. An IGDCC4 knockout A549 cell clone (IGDCC4-KO) was generated by using CRISPR/Cas9 and knockout of both IGDCC4 isoform 1 and isoform 2 was confirmed by Western blotting with a mouse anti-IGDCC4 antibody (Fig 2B). IGDCC4 knockout had no effect on cell viability as measured by using a luminescent cell viability assay (Fig 2C). PPT PowerPoint slide

PNG larger image

TIFF original image Download: Fig 2. IGDCC4 is important for the early stage of H5N1 virus replication. (A) Schematic illustration of the proteins encoded by IGDCC4. The information on the proteins encoded by the IGDCC4 gene was acquired from the UniProtKB website (accession number: Q8TDY8) (www.uniprot.org). The amino acids 62 to 980 in isoform 2 were identical to amino acids 332 to 1250 in isoform 1. (B) IGDCC4 expression level in A549 cells and IGDCC4-KO cells. Both the membrane proteins and the total proteins of the A549 cells and IGDCC4-KO cells were assessed by Western blotting to determine the IGDCC4 protein expression level. (C) The viability of IGDCC4-KO cells was measured by using the CellTiter-Glo assay and compared with that of the control A549 cells. (D) Replication of H5N1 virus in IGDCC4-KO and control A549 cells. IGDCC4-KO and A549 cells were infected with H5N1 virus at an MOI of 0.01; supernatants were collected at the indicated timepoints for virus titration in chicken embryos. (E) The viability of A549 cells transfected with the indicated siRNA was measured by using the CellTiter-Glo assay. (F) Knockdown of IGDCC4 by siRNA in A549 cells. The mRNA level of IGDCC4 in A549 cells transfected with siRNA targeting IGDCC4 was measured by qRT-PCR at 36 h post-transfection and was compared with that in A549 cells transfected with scrambled siRNA. (G) IGDCC4 expression level in A549 cells transfected with different siRNA. (H) Replication of H5N1 virus in IGDCC4 knockdown cells and A549 cells. A549 cells transfected with siRNA targeting IGDCC4 or scrambled siRNA were infected with H5N1 virus at an MOI of 0.01; supernatants were collected at the indicated timepoints for virus titration in MDCK cells by using plaque assays. (I). The effect of IGDCC4 on viral transcription and viral genome replication of H5N1 virus. IGDCC4-KO and A549 cells were infected with H5N1 virus at an MOI of 3. The mRNA and vRNA levels of the NP gene were detected by using qRT-PCR and were normalized with the value of A549 cells at the indicated timepoints. The data shown in panels B–I are the means ± SDs of three independent experiments or replicates. The two-tailed unpaired t-test was used for the statistical analysis. **, p < 0.01, ***, p < 0.001. https://doi.org/10.1371/journal.ppat.1010141.g002 To investigate the effect of IGDCC4 on influenza virus replication, IGDCC4-KO and A549 cells were infected with H5N1 virus at an MOI of 0.01. The supernatants were collected for viral titration at different timepoints after infection. The titers of H5N1 virus in IGDCC4-KO cells were 14.7-fold, 21.5-fold, 26.1-fold, and 31.6-fold lower than those in the A549 cells at 12, 24, 36, and 48 h p.i., respectively (Fig 2D). To further confirm these results, we synthesized and tested an IGDCC4 siRNA (Fig 2E–2G), and found that the replication of H5N1 virus in the IGDCC4 siRNA-treated cells was 5.4-fold, 10.3-fold, 11.8-fold, and 8.8-fold lower than that of the virus in the A549 cells at 12, 24, 36, and 48 h p.i., respectively (Fig 2H). These results demonstrate that the cellular protein IGDCC4 promotes influenza virus replication. To determine which stage of the influenza virus life cycle is affected by IGDCC4, IGDCC4-KO and control A549 cells were infected with H5N1 virus at an MOI of 3, and the levels of vRNA and mRNA of the viral NP gene in the cells were measured at 2, 4, 6, and 8 h p.i. by using quantitative reverse transcription PCR (qRT-PCR). At these four timepoints, the levels of both types of viral RNA in IGDCC4-KO cells were significantly lower than those in A549 cells (Fig 2I). At 2 h p.i., the early stage of the viral life cycle, the vRNA in the cells was mainly derived from infecting virus rather than progeny virus, and the lower level of vRNA in IGDCC4-KO cells at this timepoint indicates that the host factor IGDCC4 was most likely involved in the early stage of the viral life cycle.

Knockout of IGDCC4 reduces the nuclear targeting of NP We investigated the cellular distribution of IGDCC4 and viral NP protein at the early stage of virus infection by using confocal fluorescent microscopy. A549 cells and IGDCC4-KO cells infected with the H5N1 virus at an MOI of 5 were incubated at 4°C for 1 h and then at 37°C for 2 h. The cells were fixed at different incubation timepoints and stained with specific antibodies to observe the localization of the IGDCC4 and viral NP proteins. IGDCC4 was detected in the uninfected and infected A549 cells at 1 h p.i., it was detected at a reduced level at 2 h p.i. (1 h post-incubation at 37°C), and was nearly undetectable at 3 h p.i. (2 h post-incubation at 37°C) (Fig 3A). In contrast, IGDCC4 was not detected in IGDCC4-KO cells at any timepoint (Fig 3A). Viral NP was clearly visible in the A549 cells at 1 h p.i., and accumulated in the nucleus of 28% and 77% of the A549 cells at 2 h (1 h post-incubation at 37°C) and 3 h p.i. (2 h post-incubation at 37°C), respectively (Fig 3B). In contrast, viral NP was only detected in the nucleus of 4% and 13% of IGDCC4-KO cells at 2 h (1 h post-incubation at 37°C) and 3 h p.i. (2 h post-incubation at 37°C), respectively (Fig 3B). These results further demonstrate that IGDCC4 is involved in the early stage of the influenza virus life cycle. PPT PowerPoint slide

PNG larger image

TIFF original image Download: Fig 3. Colocalization of IGDCC4 and viral NP protein in A549 cells and IGDCC4-KO cells infected with H5N1 virus at different timepoints. (A) Cells were infected with H5N1 virus at an MOI of 5 and incubated at the indicated temperature and for the indicated time as described in the text. They were then fixed and stained with a rabbit anti-NP antibody and a mouse anti-IGDCC4 antibody, followed by incubation with Alexa Fluor 633 goat anti-rabbit IgG(H+L) (red) and 488 donkey anti-mouse IgG(H+L) (green). The nuclei were stained with DAPI (blue). (B) Quantitative analysis of NP localization in virus-infected cells. The ratio of cells showing colocalization of virus NP and nucleus was calculated from 100 virus-infected cells. https://doi.org/10.1371/journal.ppat.1010141.g003

IGDCC4 promotes influenza virus internalization Binding of HA protein to the sialic acid receptors on the cell surface is the first step for influenza virus to invade the host cell [32]. Because wheat germ agglutinin (WGA) specifically recognizes sialic acid moieties [33], we used Alexa Fluor 647-conjugated WGA and flow cytometry to examine the expression of sialic acid receptor on the surface of IGDCC4-KO and A549 cells treated with or without neuraminidase (NA). We found that the expression level of sialic acid receptors on the surface of NA-treated cells was clearly lower than that on the untreated cells, but the level of sialic acid receptors on the surface of IGDCC4-KO and A549 cells was comparable (Fig 4A). PPT PowerPoint slide

PNG larger image

TIFF original image Download: Fig 4. IGDCC4 is required for influenza virus internalization. (A) Sialic acid expression in different cells. IGDCC4-KO and A549 control cells treated with or without neuraminidase (NA) were stained with Alexa Fluor 647 conjugated with wheat germ agglutinin, and after three washes with PBS, the cells were suspended to detect WGA binding by use of flow cytometry. (B) NP protein of viruses attached to the surface of IGDCC4-KO and A549 cells. IGDCC4-KO and A549 cells were incubated with H5N1 virus at an MOI of 5 at 4°C for 1 h and then washed with cold PBS (pH = 7.2) or cold acidic PBS (pH = 1.5), which can elute virus particles on the cell surface that have not been internalized. The cells were collected and the NP protein of the viruses attached to the cells was evaluated by Western blotting. (C) NP protein of viruses internalized into IGDCC4-KO and A549 cells. A549 cells treated with or without dynasore and IGDCC4-KO cells were incubated with H5N1 virus at an MOI of 5 at 4°C for 1 h and at 37°C for 1 h. After being washed with cold PBS (pH = 7.2) or acidic PBS (pH = 1.5), the cells were collected and the NP protein level of the total attached and internalized viruses (normal PBS washed) or the internalized viruses (acidic PBS washed) was determined by using Western blotting. (D) Overexpression of IGDCC4 restores the internalization of influenza virus into IGDCC4-KO cells. A549 cells and IGDCC4-KO cells transfected with the indicated plasmids were infected with H5N1 virus at an MOI of 5. After being incubated at 4°C for 1 h and 37°C for 1 h, the cells were washed three times with ice-cold acidic PBS. The washed cells were then collected to detect the NP protein level by Western blotting, and (E) the RNA level by qRT-PCR. The band intensities of the Western blots from three assays were quantified by using ImageJ software, and the relative gray values of NP to GAPDH are presented. The data shown represent or are from three independent experiments or replicates (means ± SDs). The two-tailed unpaired t-test was used for the statistical analysis. ***, p < 0.001. https://doi.org/10.1371/journal.ppat.1010141.g004 To test whether knockout of IGDCC4 affects the attachment of influenza virus to the cell membrane, IGDCC4-KO and A549 cells were incubated with H5N1 virus at an MOI of 5 at 4°C for 1 h and then washed with cold PBS (pH = 7.2) or cold acidic PBS (pH = 1.5), which can elute virus particles on the cell surface that have not been internalized [14]. The cells were collected and the NP protein of viruses attached to the cells was evaluated by Western blotting. As shown in Fig 4B, in the PBS (pH = 7.2) washed cells, the level of NP protein attached to the surface of IGDCC4-KO cells was comparable to that of A549 cells, but in the cold acidic PBS washed cells, the level of NP protein of both cells was nearly not detectable, indicating that IGDCC4 is not needed for H5N1 virus to attach the cell surface. To examine the effect of IGDCC4 on virus internalization, A549 cells treated with or without dynasore, a specific endocytosis inhibitor, and IGDCC4-KO cells were incubated with H5N1 virus at an MOI of 5 at 4°C for 1 h and at 37°C for 1 h. After being washed with cold PBS (pH = 7.2) or acidic PBS (pH = 1.5), the cells were collected and the NP protein level of the total attached and internalized viruses (normal PBS washed) or the internalized viruses (acidic PBS washed) was determined by using Western blotting. In the PBS (pH = 7.2) washed cells, the NP level was comparable; in the acidic PBS washed cells, the NP level in IGDCC4-KO and dynasore-treated A549 cells was comparable, but clearly lower than that in control A549 cells (Fig 4C), indicating that most of the influenza virus particles were not internalized into the IGDCC4-KO or dynasore-treated A549 cells. We further investigated whether overexpression of IGDCC4 could restore the endocytosis of influenza virus in IGDCC4-KO cells. We found that the NP protein level (Fig 4D) and RNA level (Fig 4E) of the viruses internalized in the pCAGGS-IGDCC4 transfected IGDCC4-KO cells were comparable to those in A549 cells transfected with pCAGGS, but significantly higher than those in IGDCC4-KO cells transfected with pCAGGS. These results demonstrate that IGDCC4 indeed promotes the internalization of influenza virus.

IGDCC4 does not affect the internalization of vesicular stomatitis virus (VSV) and transferrin A549 cells and IGDCC4-KO cells infected with VSV-green fluorescent protein (VSV-GFP) at an MOI of 10 were inoculated for 1 h at 4°C or for 1 h at 4°C and then 1h at 37°C for evaluating the attached viruses or internalized viruses. We found that the RNA level of VSV-GFP virus attached to IGDCC4-KO and A549 cells was comparable (Fig 7A), and that the RNA level of VSV-GFP virus internalized to IGDCC4-KO cells and A549 cells was also comparable (Fig 7B). To further investigate whether IGDCC4 is important for endocytosis of cargo in general, we compared the internalization of transferrin in A549 and IGDCC4-KO cells, and we found that there was no difference in the internalization of transferrin between the A549 cells and the IGDCC4-KO cells (Fig 7C and 7D). These results suggest that IGDCC4-mediated endocytosis may be cargo-specific. PPT PowerPoint slide

PNG larger image

TIFF original image Download: Fig 7. IGDCC4 does not affect the internalization of vesicular stomatitis virus-green fluorescent protein (VSV-GFP) and transferrin. The A549 cells and IGDCC4-KO cells infected with VSV-GFP at an MOI of 10 were incubated and washed differently before being collected for detecting viral attachment (A) or (B) internalization by using qRT-PCR, the values of IGDCC4-KO cells were normalized with that of A549 cells. Data shown are from three replicates (means ± SDs). (C) and (D) Internalization of transferrin in IGDCC4-KO and A549 cells were evaluated by incubating cells with 50 μg/ml fluorescently labeled transferrin at 37°C for 30 min before receiving an acid wash to quench extracellular transferrin and were subsequently processed for confocal microscopy. Automatic image analysis quantified a total of 100 cells in three independent experiments and normalized to the A549 cells. The data shown in A, B, and D are from three independent experiments or biological replicates (means ± SDs). https://doi.org/10.1371/journal.ppat.1010141.g007

[END]

[1] Url: https://journals.plos.org/plospathogens/article?id=10.1371/journal.ppat.1010141

(C) Plos One. "Accelerating the publication of peer-reviewed science."
Licensed under Creative Commons Attribution (CC BY 4.0)
URL: https://creativecommons.org/licenses/by/4.0/


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