CD209L/L-SIGN and CD209/DC-SIGN Act as Receptors for SARS-CoV-2

Abstract

As the COVID-19 pandemic continues to spread, investigating the processes underlying the interactions between SARS-CoV-2 and its hosts is of high importance. Here, we report the identification of CD209L/L-SIGN and the related protein CD209/DC-SIGN as receptors capable of mediating SARS-CoV-2 entry into human cells. Immunofluorescence staining of human tissues revealed prominent expression of CD209L in the lung and kidney epithelia and endothelia. Multiple biochemical assays using a purified recombinant SARS-CoV-2 spike receptor-binding domain (S-RBD) or S1 encompassing both N termal domain and RBD and ectopically expressed CD209L and CD209 revealed that CD209L and CD209 interact with S-RBD. CD209L contains two N-glycosylation sequons, at sites N92 and N361, but we determined that only site N92 is occupied. Removal of the N-glycosylation at this site enhances the binding of S-RBD with CD209L. CD209L also interacts with ACE2, suggesting a role for heterodimerization of CD209L and ACE2 in SARS-CoV-2 entry and infection in cell types where both are present. Furthermore, we demonstrate that human endothelial cells are permissive to SARS-CoV-2 infection, and interference with CD209L activity by a knockdown strategy or with soluble CD209L inhibits virus entry. Our observations demonstrate that CD209L and CD209 serve as alternative receptors for SARS-CoV-2 in disease-relevant cell types, including the vascular system. This property is particularly important in tissues where ACE2 has low expression or is absent and may have implications for antiviral drug development.

Figure 1

CD209L is expressed in lung as well as renal epithelial and endothelial cells: PFA fixed human lung, renal, and temporal arteriole tissues were subjected to immunofluorescence staining. Lung tissue stained with anti-MUC1, anti-CD31, and anti-CD209L antibodies. (A) Type II alveoli epithelial cells of alveoli were positive for CD209L (red) and MUC1 (green). (B) Pulmonary arteriole endothelial cells were positive for CD31 (green) and CD209L (red). (C) Endothelial cells of the temporal arteriole were stained with CD209L (green) and CD31 (red). (D) Renal endothelial cells were positive for CD209L (red) and CD31 (green). White arrowhead points to the alveolar cell, AS = alveolar space, and white dotted line corresponds to alveolar septa. Image magnification, white bars = 50 μm.

Figure 2

Endothelial cells are permissive to SARS-CoV-2 virus, and soluble CD209L neutralizes viral entry. (A) HUVEC-TERT cells seeded in chamber slides were mock-infected or infected with SARS-CoV-2 at the indicated MOIs. Cells were fixed at 1 day post infection and stained with an antibody directed against the viral nucleoprotein, N (green). Cell nuclei were stained with DAPI (blue). Image magnification 50 μm. (B) HUVEC-TERT cells (2 × 10⁴/well, 96-well plate, and quadruple/group) were infected with pseudovirus with different concentrations. After 24 h, cells were processed and subjected to a luciferase activity assay, and representative data are shown. (C) Schematic of Myc-tagged soluble CD209 (sCD209L-Myc) and Western blot analysis of sCD209L-Myc. (D) HUVEC-TERT cells (2 × 10⁴/well, 96-well plate, and triplicate/group) were infected with mock, pseudotyped virus (10 ng/mL) with control conditioned medium (CM) or CM containing sCD209L (approximate concentration 1.5 μg/mL). After 24 h, cells were analyzed for luciferase activity. P < 0.05.

Figure 3

CD209L mediates SARS-CoV-2 entry and infection in endothelial cells. (A) HUVEC-TERT cells expressing control shRNA or CD209L-shRNA (2 × 10⁴/well, 96-well plate, and triplicate/group) were infected with different amounts of SARS-CoV-2 pseudotyped lentivirus. After 24 h, cells were processed and subjected to luciferase activity, and representative data are shown. (B) HUVEC-TERT cells expressing control shRNA or CD209L-shRNA seeded in chamber slides (triplicate/group) were infected with SARS-CoV-2 at the indicated MOIs. Cells were fixed at 1 day post infection and stained with an antibody directed against the viral nucleoprotein, N (green). Cell nuclei were stained with DAPI (blue). Image magnification 50 μm. Quantification of N protein positive cells is shown. P < 0.05. (C) Western blot analysis showing expression of CD209L, CD209, and ACE2. (D) HEK-293 cells expressing CD209L, CD209, or ACE2 (2 × 10⁴/well, 96-well plate, and quadruple wells/group) were infected with pseudotyped virus (10 ng/mL) or mock virus. After 24 h, cells were analyzed for luciferase activity. P < 0.05.

Figure 4

CD209L and CD209 bind to SARS-CoV-2-S-RBD. (A) Schematic of Fc-CoV-2-S-RBD; expression of Fc-CoV-2-S-RBD and CD209L in HEK-293 is shown. (B) Immunoprecipitation assay demonstrates the binding of CD209L with Fc-CoV-2-S-RBD. (C) Schematic and Coomassie blue stain of CoV-2-S-RBD-HIS are shown. (D) Far-Western blot analysis shows the binding of HIS-STRP-tagged CoV-2-S-RBD with CD209L. (E) Western blot analysis of CD209L, CD209, and Fc-ACE2 ectopically expressed in HEK-293 cells. (F) CoV-2-S-RBD-HIS applied onto a PFVD membrane with varying concentrations as indicated in the figure legend via a dot blot apparatus. The membranes, after blocked with 5% BSA, were incubated with cell lysates derived from HEK-293 cells expressing Fc-ACE-2-Myc, CD209L-Myc, or CD209-Myc, and the binding of ACE2, CD209L, and CD209 to CoV-2-S-RBD was detected with anti-Myc antibody. Quantification of the dot blots is shown. AU, arbitrary unit. ImageJ software was used to quantify the dot blots.

Figure 5

CD209L is N-glycosylated, and N-glycosylation modulates its interaction with SARS-CoV2 spike protein. (A) HCD MS/MS spectrum of m/z 1505.1470, assigned as the [M + 2H]²⁺ of the glycopeptide ⁸⁷DAIYQNLTQLK⁹⁷ + HexNAc₂Hex₈, corresponding to CD209L site N92 bearing a Man₈ glycan. Stepped-collision energy (15/25/35% NCE) was used. The fragment ions do not retain the glycan moiety. p = peptide, N = N-acetylhexosamine (HexNAc), H = hexose (Hex). (B) Dot blot of CD209L/HEK-293 cell lysate under control conditions (no treatment) or treated with PNGase F or Endo H. Dot blot of HEK-293 cell lysate was used as a negative control. Following treatment of lysate, the spike protein receptor-binding domain (S-RBD-HIS) was incubated with the immobilized lysate, followed by detection with anti-HIS antibody. (C) Quantification of blots is shown. AU, arbitrary unit. ImageJ software was used to quantify the dot blots.