Antibody-dependent infection of human macrophages by severe acute respiratory syndrome coronavirus

Abstract

Background: Public health risks associated to infection by human coronaviruses remain considerable and vaccination is a key option for preventing the resurgence of severe acute respiratory syndrome coronavirus (SARS-CoV). We have previously reported that antibodies elicited by a SARS-CoV vaccine candidate based on recombinant, full-length SARS-CoV Spike-protein trimers, trigger infection of immune cell lines. These observations prompted us to investigate the molecular mechanisms and responses to antibody-mediated infection in human macrophages.

Methods: We have used primary human immune cells to evaluate their susceptibility to infection by SARS-CoV in the presence of anti-Spike antibodies. Fluorescence microscopy and real-time quantitative reverse transcriptase polymerase chain reaction (RT-PCR) were utilized to assess occurrence and consequences of infection. To gain insight into the underlying molecular mechanism, we performed mutational analysis with a series of truncated and chimeric constructs of fragment crystallizable γ receptors (FcγR), which bind antibody-coated pathogens.

Results: We show here that anti-Spike immune serum increased infection of human monocyte-derived macrophages by replication-competent SARS-CoV as well as Spike-pseudotyped lentiviral particles (SARS-CoVpp). Macrophages infected with SARS-CoV, however, did not support productive replication of the virus. Purified anti-viral IgGs, but not other soluble factor(s) from heat-inactivated mouse immune serum, were sufficient to enhance infection. Antibody-mediated infection was dependent on signaling-competent members of the human FcγRII family, which were shown to confer susceptibility to otherwise naïve ST486 cells, as binding of immune complexes to cell surface FcγRII was necessary but not sufficient to trigger antibody-dependent enhancement (ADE) of infection. Furthermore, only FcγRII with intact cytoplasmic signaling domains were competent to sustain ADE of SARS-CoVpp infection, thus providing additional information on the role of downstream signaling by FcγRII.

Conclusions: These results demonstrate that human macrophages can be infected by SARS-CoV as a result of IgG-mediated ADE and indicate that this infection route requires signaling pathways activated downstream of binding to FcγRII receptors.

Figure 1

Anti-Spike but not control serum triggered infection of human macrophages by SARS-CoVpp. (A) Monocyte-derived macrophages were incubated for 1 hour with SARS-CoVpp carrying a luciferase reporter gene in the presence of a 1:1000 dilution of either anti-Spike or control serum and infected cells were detected by fluorescence staining of firefly luciferase (green), as described under Materials and methods. Cell nuclei were labeled with DAPI (blue). SARS-CoVpp infection of macrophages occurred only in the presence of anti-Spike serum. Images are representatives of five independent experiments using macrophages from five donors. (B) Infection was quantified by counting the percentage of immunofluorescence-positive cells in randomly chosen fields using Metamorph software. Results are shown as means ± SEM. Scale bar: 20 μm.

Figure 2

Anti-Spike serum enhanced SARS-CoV infection in human monocyte-derived macrophages. (A) Monocyte-derived macrophages were infected with SARS-CoV (strain HK39849) at an MOI of 1 in the presence of a 1:1000 dilution of either anti-Spike (black bars) or control (white bars) serum for 1 hour and then cultured with fresh medium for 6 hours as described under Materials and methods. After fixation, cells were permeabilized for intracellular staining of SARS-CoV nucleocapsid protein, which was revealed by TRITC-conjugated goat anti-mouse antibody (orange) while cell nuclei were stained with DAPI (blue). Images shown are representatives of the results obtained with macrophages from three donors at 6 hours post infection. Scale bar: 20 μm. (B) Infectivity of SARS-CoV was determined by calculating the percentage of nucleocapsid-positive cells in five randomly selected fields using Metamorph software. Presence of anti-Spike serum increased infectivity of SARS-CoV by 3-4-fold in macrophages of two out of three donors. Note that the ordinate’s scale for Donor 3 is different for ease of visualization. Results are shown as means ± SEM.

Figure 3

Nucleocapsid and ORF1b gene expression in SARS-CoV infected human macrophages. Monocyte-derived macrophages from three independent donors were infected with SARS-CoV as described in the legend to Figure 2 in the presence of anti-Spike (black bars) or control serum (white bars). Positive and negative RNA strands were determined for SARS-CoV nucleocapsid (A) and ORF1b (B) at the specified time points (hours) post infection (h.p.i.) using real-time qPCR. Data are shown as the number of viral RNA copies normalized to that of 10⁸ copies of 18S rRNA in the corresponding sample. Overall, RNA levels were higher in infected cells in presence of anti-Spike compared to control serum. Results are shown as means ± SEM of four measurements from two separate runs for each donor and gene.

Figure 4

Purified IgGs from heat-inactivated mouse anti-Spike serum mediate ADE of infection of Raji cells by SARS-CoVpp. Serial dilutions of purified IgGs (solid bars) and protein G-sepharose column flow-through (hatched bars) from either heat-inactivated mouse anti-Spike (dark grey) or control (light grey) serum were incubated with SARS-CoVpp for 1 hour and opsonized particles were then used to infect Raji cells as described under Materials and methods. SARS-CoVpp infection was assessed three days post infection by expressing luciferase activity as Relative Luminescence Units (RLU), as described under Materials and methods. IgG concentrations were used to calculate the dilution factor and adjust the amount of the corresponding flow-through (FT, hatched bars) from the purification columns. Results are shown as means ± SEM of 6 measurements from two independent experiments. *P < 0.05 (by the unpaired Student’s t test).

Figure 5

Susceptibility to ADE is dependent on an intact cytosolic domain. (A) Schematic representation of wild-type and mutated constructs that were utilized to produce stably transduced ST486 cell lines (see Materials and methods); from top to bottom: human FcγRIIA-His131 (hFcγRIIA-H), FcγRIIA-Arg131 (hFcγRIIA-R), FcγRIIB1 (hFcγRIIB1) and FcγRIIB2 (hFcγRIIB2). B) Surface expression in stably transduced ST486 cell lines was evaluated by flow cytometry using a mouse monoclonal anti-FcγRII antibodies (dark grey) or isotype control (light grey), and expressed as mean fluorescence intensity in Arbitrary Units (AU) as described under Materials and methods. Constructs are identified by Arabic numbers and grouped from top to bottom as indicated in A. C) Immune complex-binding ability of wild-type, truncated and chimeric FcγR receptors. SARS-CoVpp were incubated with purified mouse anti-Spike IgG (dark grey) or control IgG (light grey) to form immune complexes, which were then added to the ST486 transfectants (see Materials and methods). After washing and fixation, bound immune complexes were detected by flow cytometry with an FITC-conjugated goat anti-mouse F(ab’)2, and results shown as in B. Constructs are identified by Arabic numbers and grouped from top to bottom as indicated in A. D) Susceptibility of ST486 transfectants to ADE of infection by SARS-CoVpp. SARS-CoVpp were incubated with purified mouse anti-Spike IgG (dark grey) or control IgG (light grey) and then used to infect cells (see Materials and methods). Cells were washed and three days post infection incubation was stopped by adding luciferase substrate to measure enzymatic activity, which was expressed as Relative Luminescence Units (RLU). Constructs are identified by Arabic numbers and grouped from top to bottom as indicated in A. Results are shown as the means ± SEM of three (B and C) or 4–5 independent experiments (D). *P < 0.05; **P < 0.01 by the unpaired Student’s t test.