SARS-CoV-2 Delta (B.1.617.2) variant replicates and induces syncytia formation in human induced pluripotent stem cell-derived macrophages

Cell cultures

Human induced pluripotent stem cells (hiPSCs; ATCC ACS-1019) were seeded in 6-well plates pre-coated with Matrigel (Cat# 354234; Corning, Corning, NY, USA). Cells were cultured in mTeSR-1 medium (Cat# 85850; StemCell Technologies, Vancouver, CA), which was changed daily. At about 70–80% confluency, cells were detached with Accutase (Cat# 07922; StemCell Technologies, Vancouver, CA) and passaged for maintenance or macrophage differentiation.

Human epithelial lung carcinoma cells stably overexpressing human ACE2 (A549-ACE2; a549d-cov2r; InvivoGen, San Diego, CA, USA), human hepatocellular carcinoma cells (Huh-7; ATCC PTA-4583), and human embryonic kidney epithelial cells (293T/17; ATCC CRL-11268) stably overexpressing human ACE2 (293T/17-ACE2) were cultured in DMEM (Cat# SH30243.02; Cytiva, Marlborough, MA, USA) with 10% fetal bovine serum (FBS; Cat# F7524; Sigma, St. Louis, MO, USA). Monkey kidney epithelial cells (VeroE6; ATCC CRL-1586) were cultured in Opti-MEM (Cat# 22600-050; Invitrogen, Waltham, MA, USA) with 10% FBS.

Generation of hiPSC-derived macrophages

The protocol for generating hiPSC-derived macrophages (iMΦ) was adapted from previously published protocols (Wilgenburg et al., 2013; Gutbier et al., 2020), with minor modifications. On day 0, 4 × 10⁶ cells of hiPSCs were initially stimulated to form embryoid bodies (EBs) by culturing them in a single well of an AggreWell 800 plate (Cat# 34850; StemCell Technologies, Vancouver, CA) with EB medium and incubating at 37 °C/5% CO₂. EB medium consisted of mTeSR-1 medium (Cat# 85850; StemCell Technologies, Vancouver, CA) supplemented with 10 µM Rock inhibitor (Y27632) (Cat# 72302; StemCell Technologies, Vancouver, CA), 50 ng/mL rhBMP4 (Cat# 120-05ET-10UG; PeproTech, Rocky Hil, NJ, USA), 20 ng/mL rhSCF (Cat# 78062.1; StemCell Technologies, Vancouver, CA), and 50 ng/mL rhVEGF (Cat# 100-20-10UG; PeproTech, Rocky Hill, NJ, USA). From days 1–3, the differentiated hiPSCs were replenished daily with the EB medium. On day 4, approximately 300 EBs were harvested into a 50-mL centrifuge tube containing differentiation medium. Differentiation medium consisted of X-VIVO 15 medium (Cat# BE02-053Q; Lonza, Basel, Switzerland); 100 ng/mL rhM-CSF (Cat# 216-MC; R&D Systems, Minneapolis, MN, USA); 25 ng/mL rhIL-3 (Cat# 203-IL; R&D Systems, Minneapolis, MN, USA); 2 mM Glutamax (Cat# 35050061; Gibco, Waltham, MA, USA); 1% Antibiotic-Antimycotic (Cat# 15240062; Gibco, Waltham, MA, USA); and 0.055 mM 2-mercaptoethanol (Cat# 21985023; Gibco, Waltham, MA, USA). Approximately 150 EBs could be cultured in a T175 flask pre-coated with Matrigel (Cat# 354234; Corning, Corning, NY, USA) using at least 15 mL of differentiation medium. Then, the cultured EBs were incubated at 37 °C/5% CO₂ without disturbing the flasks during the first week of differentiation. From week 2–3, at least 10 mL of fresh differentiation medium was added to each flask once per week. From week 4, half of differentiation medium for each flask was replaced with fresh differentiation medium once per week until macrophage precursors appeared in the supernatant. Then, the entire differentiation medium was changed twice per week.

The floating macrophage progenitors were harvested and resuspended in macrophage medium. Macrophage medium consisted of X-VIVO 15 medium; 100 ng/mL rhM-CSF; 2 mM Glutamax; and 1% antibiotic-antimycotic. Macrophage precursors were cultured in macrophage medium at 37 °C/5% CO₂ for 7 days to obtain mature macrophages (iMΦ) ready for use. For M1 (iM1Φ) polarization, iMΦ were cultured for 24 h in X-VIVO 15 medium supplemented with 100 ng/mL LPS (Cat# L4391-1MG; Sigma, St. Louis, MO, USA); 20 ng/mL rhIFN γ (Cat# 285-IF, R&D Systems; Minneapolis, MN, USA); 2 mM Glutamax; and 1% Antibiotic-Antimycotic. For M2 (iM2Φ) polarization, macrophages were cultured for 24 h in X-VIVO 15 medium supplemented with 50 ng/mL rhIL-4 (Cat# 204-IL; R&D Systems, Minneapolis, MN, USA).

Viruses

SARS-CoV-2 isolates were obtained from saliva samples of COVID-19 patients in Thailand and verified by nucleotide sequencing as the Delta (B.1.617.2; GISAID accession number EPI_ISL_11327790) and Omicron (B.1.1.529; EPI_ISL_11327789) variants. All experiments with SARS-CoV-2 were performed in a biosafety level 3 (BSL-3) laboratory at the National Center for Genetic Engineering and Biotechnology (BIOTEC), Thailand. Viruses were propagated in A549-ACE2 cells at 37 °C. Viral titers were measured using the median tissue culture infectious dose (TCID₅₀) assay on A549-ACE2 cells. For the TCID₅₀ assay, virus samples were serially diluted ten-fold (from 1:10 to 1:10⁶) in Opti-MEM (Cat# 22600-050; Invitrogen, Waltham, MA, USA) with 0.1% TrypLE Select (Cat# 0040090DG; Gibco, Waltham, MA, USA) and then incubated with confluent A549-ACE2 cells in 96-well plates for 72 h at 37 °C/5% CO₂. After incubation, cells were examined for cytopathic effects (CPEs) and confirmed by dot blots on nitrocellulose membranes (Bio-Rad, Hercules, CA, USA) with the following antibodies: primary anti-SARS-CoV-2 nucleocapsid (N) antibody (1:1000 dilution; in-house) and secondary goat anti-mouse IgG conjugated with HRP (1:4000 dilution; Cat# 405306; BioLegend, San Diego, CA, USA). Viral titers were calculated by the Spearman-Karber method and expressed as TCID₅₀/mL.

Infection of viruses

iMΦ were infected with SARS-CoV-2 Delta or Omicron variant at a multiplicity of infection (MOI) of 0.1 in RPMI-1640 (Cat# 31800022; Gibco, Waltham, MA, USA) without serum. Mock-infected cells were used as controls. Cells were adsorbed with virus for 2 h, washed with PBS, replenished with RPMI-1640 supplemented with 0.05% TrypLE Select, and incubated at 37 °C/5% CO₂. Cell supernatants were harvested at 0, 24, 48 and 72 h post-infection (hpi) for virus titration. Cells were subjected to immunofluorescence assay to determine the expression of SARS-CoV-2 N protein.

Immunofluorescence assay of viral proteins

Cells were fixed and permeabilized with 80% acetone for 15 min and blocked with 1% bovine serum albumin (BSA; Cat# A8412; Sigma, St. Louis, MO, USA) in PBST (PBS + 0.1%Tween 20) for at least 30 min. For SARS-CoV-2 detection, cells were stained with primary mouse anti-SARS-CoV-2 N antibodies (1:1000 dilution; in-house) overnight at 4 °C and washed three times with PBS. Cells were then stained with secondary anti-mouse IgG antibodies conjugated with Alexa Fluor 568 (1:1000 dilution; Cat# A11004; Invitrogen, Waltham, MA, USA) and phalloidin conjugated with Alexa Fluor 488 (1:400 dilution; Cat# A12379; Invitrogen, Waltham, MA, USA) in the dark for at least 1 h at room temperature and washed three times with PBS.

Cell nuclei were counterstained with DAPI (Cat# H-1200; Vector Laboratories, Newark, CA, USA). Cells were imaged with an IX73 fluorescence microscope (Olympus, Tokyo, Japan) and analyzed with cellSens Standard Imaging Software (version 2.1; Olympus, Tokyo, Japan) and ImageJ (version 1.53c; Bethesda, MD).

Flow cytometry

For surface marker analysis, macrophages were harvested with enzyme-free cell dissociation buffer (Cat# 13151014; Gibco, Waltham, MA, USA). Cells were resuspended in ice-cold FACS buffer (consisting of PBS and 0.5–1% BSA) at a concentration of 1-5 × 10⁶ cells/mL. To block non-specific binding of the primary antibodies, Fc receptor blocking antibody (Cat# 422301; BioLegend, San Diego, CA, USA) was added to the cell suspension. Cells were incubated with the conjugated primary antibodies or the isotype-matched control in the dark for at least 30 min at 4 °C. Cells were washed three times with ice-cold FACS buffer before analyzing with a FACSAria Fusion flow cytometer (Becton Dickinson, Franklin Lakes, NJ, USA). Data were analyzed using FlowJo (version 10, Ashland, OR).

The following antibodies (all purchased from Invitrogen, Waltham, MA, USA) were used for staining surface markers: CD11b-APC (1:20 dilution; Cat# 17-0118-41), CD14-PE (1:20 dilution; Cat# 12-0149-42), CD16-FITC (1:10 dilution; Cat# 11-0168-41), CD86-PE (1:20 dilution; Cat# 12-0869-41), CD163-APC (1:20 dilution; Cat# 17-1639-41) and isotype controls: mouse-IgG1 kappa-APC (1:40 dilution; Cat# 17-47148-1), mouse-IgG1 kappa-PE (1:40 dilution; Cat# 12-4714-81), mouse-IgG1 kappa-FITC (1:200 dilution; Cat# 11-4714-81), mouse-IgG2b kappa-PE (1:40 dilution; Cat# 12-4732-81).

Phagocytosis assay

Macrophage progenitors were cultured for 7 days in macrophage medium in a 96-well plate at a density of 1 × 10⁵ cells/well. Mature iMΦ were incubated with or without reconstituted pHrodo Green Zymosan Bioparticles (Cat# P35365; Invitrogen, Waltham, MA, USA) at 37 °C for 2 h. Cells were harvested and analyzed by flow cytometry. Zymosan-free cells were used as negative controls to set a threshold for measuring the percentage of positive cells (Wilgenburg et al., 2013).

Real-time quantitative reverse transcription PCR (RT-qPCR)

For human gene expression, total RNA was extracted from cell lysates using the GeneJET RNA Purification Kit (Cat# K0732; Thermo Fisher, Waltham, MA, USA). RT-qPCR was performed using the iTaq Universal SYBR Green One-Step Kit (Cat# 10032048; Bio-Rad, Hercules, CA, USA) on a CFX Opus 96 Real-Time PCR System (Bio-Rad, Hercules, CA, USA). Primers for ACE2, CD86, IL-1β, IL-6, IL-8, IL-18, TNF-α, CCL2, IFN-α, and GAPDH are described in Table S1. Cycling conditions were as follows: 1 cycle of 50 °C for 20 min, 1 cycle of 95 °C for 5 min; 40 cycles of 95 °C for 10 s and 58 °C for 30 s. Fluorescence signals were detected at the end of each 58 °C step. The dissociation curve was as follows: 1 cycle of 95 °C for 10 s, 65 °C for 5 s with a temperature increment of 0.5 °C to 95 °C. Ct values and dissociation curves were analyzed using CFX Maestro Software (Bio-Rad; Hercules, CA, USA, version 2.0). Gene expressions were normalized against GAPDH and compared with 293T/17 cells for the ACE2 gene or mock infection for cytokine genes using the 2^−ΔΔCT method (Schmittgen & Livak, 2008). With the exception of ACE2 and CD86 gene expression, all other samples were measured in triplicate.

Statistical analysis

All statistical tests were performed with GraphPad Prism 9 (San Diego, CA). One-way ANOVA with Tukey’s post-hoc tests were used for multiple comparisons between experimental groups. Data are presented as mean ± SD. The P value < 0.05 was considered statistically significant.

Characterization of hiPSC-derived macrophages (iMΦ)

We generated iMΦ using previously published protocols (Figs. 1A–1B) (Wilgenburg et al., 2013; Gutbier et al., 2020). Differentiated iMΦ displayed round, oval, and spindle shapes (Fig. 1B), which are typical of macrophage morphology (Rey-Giraud, Hafner & Ries, 2012). Flow cytometry analysis of cell surface marker expression revealed that iMΦ expressed the monocyte/macrophage lineage markers CD11b, CD14, CD16, CD86 and CD163 (Fig. 2A). To assess phagocytic activity, iMΦ were incubated with pH-sensitive zymosan particles that fluoresce at an acidic pH typically found in phagosomes. After 2 h of zymosan incubation, the majority (>90%) of iMΦ engulfed the particles (Figs. 2B–2C), confirming that the cells were functional with their phagocytic activity. Taken together, the iMΦ generated in this study exhibit key characteristics of macrophages in terms of their morphology, surface marker expression, and phagocytic activity.

iMΦ are susceptible to SARS-CoV-2 infection

Given that SARS-CoV-2 can infect mouse and human macrophages via ACE2-independent pathways (Lv et al., 2021; Jalloh et al., 2022), we first sought to investigate the susceptibility of iMΦ to SARS-CoV-2 infection. Under bright-field microscopy, we infected iMΦ with SARS-CoV-2 (Delta or Omicron variants) at an MOI of 0.1 and found no apparent signs of CPE at 72 hpi. In contrast, an immunofluorescence experiment revealed the presence of SARS-CoV-2 N protein in the cytoplasm of iMΦ infected with both Delta and Omicron (Fig. 3A; Fig. S1). SARS-CoV-2 N protein was detectable in 11.5% of Delta-infected iMΦ but only in 3.6% of cells infected with Omicron (Fig. 3B). Of note, we observed clear signs of cell–cell fusion or syncytia formation only in Delta-infected iMΦ (Fig. 3A; Fig. S1). It is worth noting that, despite the signs of viral entry, RT-qPCR and immunoblot analysis revealed undetectable mRNA and protein expression of ACE2 in iMΦ (Figs. S2A– Figs. S2B; Table S2). Viral replication was further examined by titrating infectious virions in iMΦ supernatants collected up to 72 hpi on A549-ACE2 cells using the TCID₅₀ assay (Fig. S3; Table S3). Replication of the Delta variant in iMΦ was productive within the first 48 hpi (Fig. 3C). Replication kinetics gradually increased from 8.73 × 10 TCID₅₀/mL at 0 hpi to 3.58 ×10² and 8.16 ×10² TCID₅₀/mL at 24 and 48 hpi, respectively (Fig. 3C). Later, however, viral titers dropped to 3.68 × 10 TCID₅₀/mL at 72 hpi (Fig. 3C). In contrast, Omicron replication in iMΦ was abortive, as we found no viral titer of the Omicron variant at any time point (Fig. 3C). Taken together, these results suggest that SARS-CoV-2 Delta variant replicates in iMΦ in an ACE2-independent manner and induces syncytia formation.

iMΦ show moderate expression of proinflammatory cytokine genes in response to SARS-CoV-2 infection

We further investigated whether SARS-CoV-2 infection could activate iMΦ to overexpress proinflammatory cytokine genes, leading to overproduction of proinflammatory cytokines or cytokine storms as observed in severe COVID-19 patients (Blanco-Melo et al., 2020; Hadjadj et al., 2020; Grant et al., 2021). To this end, we examined mRNA expression of proinflammatory and antiviral cytokines in iMΦ infected with SARS-CoV-2 (Delta or Omicron) at 24, 48, and 72 hpi as well as in iMΦ polarized to the proinflammatory M1 phenotype by LPS and IFN-γ. After normalization to GAPDH using the 2^−ΔΔCT method (Schmittgen & Livak, 2008), the results are expressed as a fold change in cytokine gene expression compared to mock infection (Fig. 4; Table S4). Although the infection rates of SARS-CoV-2 in iMΦ were approximately 11% and 4% for Delta and Omicron, respectively (Fig. 3B), the infection did activate iMΦ to upregulate mRNA expression of the M1 phenotype marker CD86 (Fig. S4; Table S5).

Compared to mock controls, SARS-CoV-2 infection of iMΦ did not substantially upregulate cytokine mRNA expression (Figs. 4A–4G). Delta and Omicron infection only marginally altered mRNA expression of IL-1β (Delta vs Mock: p = 0.6114; Omicron vs Mock: p = 0.7534), IL-6 (p = 0.1910; p = 0.2619), IL-18 (p = 0.7755; p = 0.6092), TNF-α (p = 0.9080; p = 0.9405) and IFN-α (p = 0.0612; p = 0.2138) (Figs. 4A–4B, 4D–4E, 4G). Moreover, the infection suppressed mRNA expression of IL-8 (∼3-fold; p < 0.0001; p < 0.0001) and CCL2 (∼3-fold; p < 0.0001; p = 0.0002) (Figs. 4C, 4F).

However, among groups of SARS-CoV-2-infected iMΦ, there was an overall trend toward a time-dependent increase in mRNA expression of most cytokines during infection. At 72 hpi, Delta infection upregulated at least two- to seven-fold mRNA expression of IL-1β (Fig. 4A), IL-6 (Fig. 4B), IL-18 (Fig. 4D), CCL2 (Fig. 4F) and IFN-α (Fig. 4G), compared to those at 24 hpi (IL-1β: p = 0.0005; IL-18: p = 0.0008; CCL2: p < 0.0001; IFN-α: p = 0.0007) and 48 hpi (IL-1β: p = 0.0063; IL-6: p = 0.0048; IL-18: p = 0.0289; CCL2: p = 0.0001). When infected with Omicron at 72 hpi, mRNA expression of the same group of cytokines also increased by similar magnitudes compared to those at 24 hpi (IL-1β: p = 0.0259; IL-18: p = 0.0002; CCL2: p = 0.0452; IFN-α: p = 0.0293) and 48 hpi (IL-1β: p = 0.0006; IL-6: p = 0.0043; IL-18: p = 0.0011) (Figs. 4A–4B, 4D, 4F–4G). Interestingly, only at 24 hpi did Omicron infection induce significantly higher mRNA expression of IL-8 (∼4-fold; p < 0.0001) and IFN-α (∼2-fold; p = 0.0132) than Delta infection (Figs. 4C, 4G).

Compared to LPS/IFN-γ polarization, SARS-CoV-2 infection resulted in significantly reduced mRNA expression of IL-1β (∼15-fold; Delta vs LPS/IFN-γ: p < 0.0001; Omicron vs LPS/IFN-γ: p < 0.0001), IL-6 (∼108-fold; p < 0.0001; p < 0.0001), IL-8 (∼4-fold; p < 0.0001; p < 0.0001), TNF-α (∼10-fold; p = 0.0052; p = 0.0272) and CCL2 (∼7-fold; p < 0.0001; p < 0.0001) (Figs. 4A–4C, 4E–4F). However, the infection generated significantly higher IFN-α mRNA expression (∼4-fold; p < 0.0001; p < 0.0001) than non-viral polarization by bacterial LPS and type II IFN-γ (Fig. 4G).

These results suggest that SARS-CoV-2 infection triggers moderate gene expression of proinflammatory cytokines in iMΦ at the transcriptional level. However, LPS/IFN-γ polarization causes significant transcriptional activation of proinflammatory cytokine genes in iMΦ. Thus, it is unlikely that a direct interaction between SARS-CoV-2 and macrophages is the main cause of increased proinflammatory cytokine production or cytokine storm.