Genome-Wide Knockout Screen Identifies Human Sialomucin CD164 as an Essential Entry Factor for Lymphocytic Choriomeningitis Virus

Abstract

Lymphocytic choriomeningitis virus (LCMV) is a well-studied mammarenavirus that can be fatal in congenital infections. However, our understanding of LCMV and its interactions with human host factors remains incomplete. Here, host determinants affecting LCMV infection were investigated through a genome-wide CRISPR knockout screen in A549 cells, a human lung adenocarcinoma line. We identified and validated a variety of novel host factors that play a functional role in LCMV infection. Among these, knockout of the sialomucin CD164, a heavily glycosylated transmembrane protein, was found to ablate infection with multiple LCMV strains but not other hemorrhagic mammarenaviruses in several cell types. Further characterization revealed a dependency of LCMV entry on the cysteine-rich domain of CD164, including an N-linked glycosylation site at residue 104 in that region. Given the documented role of LCMV with respect to transplacental human infections, CD164 expression was investigated in human placental tissue and placental cell lines. CD164 was found to be highly expressed in the cytotrophoblast cells, an initial contact site for pathogens within the placenta, and LCMV infection in placental cells was effectively blocked using a monoclonal antibody specific to the cysteine-rich domain of CD164. Together, this study identifies novel factors associated with LCMV infection of human tissues and highlights the importance of CD164, a sialomucin that previously had not been associated with viral infection. IMPORTANCE Lymphocytic choriomeningitis virus (LCMV) is a human-pathogenic mammarenavirus that can be fatal in congenital infections. Although frequently used in the study of persistent infections in the field of immunology, aspects of this virus's life cycle remain incomplete. For example, while viral entry has been shown to depend on a cell adhesion molecule, DAG1, genetic knockout of this gene allows for residual viral infection, implying that additional receptors can mediate cell entry. The significance of our study is the identification of host factors important for successful infection, including the sialomucin CD164, which had not been previously associated with viral infection. We demonstrated that CD164 is essential for LCMV entry into human cells and can serve as a possible therapeutic target for treatment of congenital infection.

Keywords: CD164; CRISPR screen; lymphocytic choriomeningitis virus; viral entry.

FIG 1

Genome-wide CRISPR loss-of-function screen in human cells identify host factors important for LCMV infection. (A) Schematic of CRISPR-based KO screen done in A549 lung epithelial cells for the identification of LCMV host factors. (B) Gene enrichment for CRISPR screen of rLCMV-mCherry infection. Enrichment scores were determined by MaGECK analysis, and genes were colored by biological function. Dotted line indicates −log₁₀(enrichment score) = 4. All genes and their enrichment scores can be found in Table S1. (C) Percentage of infected cells as determined by flow cytometry following infection of A549 homozygous knockouts (CD164, SPR14, IL2RA, and KHNYN) or heterozygous knockouts (ARFRP1, YKT6, ACKR4, RAB10, EMC1, and SYS1) with rLCMV-mCherry. Wild-type cells were used as normalization controls. Cells were infected at an MOI of 1 and harvested at 24 hpi. Error bars indicate standard errors from three independent experiments. (D) Quantification of viral infection in WT, ΔCD164, ΔCD164 complemented with human CD164 (ΔCD164 + hCD164), and ΔCD164 complemented with mouse Cd164 (ΔCD164 + mCd164) in A549, 293T, and 3T6 cell type backgrounds. Cells were infected with rLCMV-mCherry at an MOI of 1 and harvested at 24 hpi. Error bars indicate standard errors from three independent experiments.

FIG 2

Infection of KO cell lines with a panel of mammarenavirus GP pseudotyped virus. (A to E) Percent infection of ΔCD164, ΔDAG1, and ΔCD164 ΔDAG1 double-KO cells relative to WT in either A549 or 293T cell type backgrounds following inoculation with low-DAG1-affinity LCMV strain Armstrong 53b-GP (A) or WE2.2-GP (B) and high-DAG1-affinity strain Armstrong clone 13-GP (C), W54-GP (D), or WE-GP (E) pseudotyped virus as determined by flow cytometry for GFP positivity. Cells were infected at an MOI of 1 and measured 24 hpi. Error bars indicate standard errors from three independent experiments. (F to H) Percent infection of ΔCD164, ΔDAG1, and ΔCD164 ΔDAG1 double-KO cells relative to WT in either A549 or 293T cell type backgrounds following inoculation with LASV-GP (D), GTOV-GP (E), or MACV-GP (F) pseudotyped virus as determined by flow cytometry for GFP positivity. Cells were infected at an MOI of 1 and measured 24 hpi. Error bars indicate standard errors from three independent experiments.

FIG 3

CD164 functional region determination through antibody binding, domain deletion, and alanine mutagenesis. (A, left) Schematic of wild-type, ΔCD164, ΔCD164 + hCD164(ΔE1), ΔCD164 + hCD164(ΔE1-2), ΔCD164 + hCD164(ΔE1-3), ΔCD164 + hCD164(ΔE1-4), ΔCD164 + hCD164(ΔE1-5), and ΔCD164 + hCD164(ΔE2-6). (Right) Complemented A549 and 293T cells were challenged with rLCMV-mCherry (MOI, 1) and infection was measured by flow cytometry at 24 hpi. Percent infection was normalized to wild type. Error bars represent standard errors from three independent experiments. (B, left) Schematic of wild type, ΔCD164 KO + hCD164(N72A), ΔCD164 + hCD164(N77A), ΔCD164 + hCD164(N94A), and ΔCD164 + hCD164(N104A). (Right) Complemented A549 and 293T cells were challenged with rLCMV-mCherry (MOI, 1) and infection was measured by flow cytometry at 24 hpi. Percent infection was normalized to the wild type. Error bars represent standard errors from three independent experiments. (C) Amino acid similarities of the cysteine-rich region in human CD164 and mouse Cd164 determined using the ClustalW program on SnapGene. Yellow circles indicate cysteine residues, red N symbolizes N-linked glycosylation sites, and identical amino acids are highlighted in green. (D) Blockade of LCMV infection with serial dilutions of anti-human CD164 monoclonal mouse antibody clone N6B6 or mouse IgG2a-κ isotope control in wild-type A549 cells. Cells were infected at an MOI of 1 and infection measured at 24 hpi. Error bars indicate standard errors from three independent experiments.

FIG 4

Characterization of CD164 as a therapeutic target in human placenta. (A) Double immunofluorescence staining for CD164 and CK7 or isotype staining followed by counterstaining with DAPI in villous trophoblastic tissue. Original images were taken by confocal microscopy at ×100 magnification. Scale bar represents 20 μm. (B) Immunofluorescence imaging of JEG-3 placenta cells preincubated with various concentrations of anti-CD164 MAb N6B6 and infected with r3LCMV-mCherry at an MOI of 0.5. Cells were fixed and imaged at ×10 magnification 24 hpi. Scale bar represents 20 μm. (C) Quantification of percent infection of JEG-3 placenta cells preincubated with various concentrations of anti-CD164 MAb N6B6 and infected with r3LCMV-mCherry at MOI of 0.5. Analysis was done on 4 fields of view in 2 independent infections and normalized to infection control.