Coronaviruses, which are enveloped positive-single-stranded RNA viruses that belong to the Coronaviridae
family, cause mild to severe respiratory, enteric, and neurological diseases in humans and animals (1
). At the end of 2019, a pneumonia outbreak caused by severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) was reported in Wuhan, China, and this novel coronavirus gave rise to the current coronavirus disease 2019 (COVID-19) pandemic (2
). By cross-species transmission to humans, as of December 2021, over 280 million COVID-19 confirmed cases and 5 million deaths globally have been reported, according to COVID-19 situation reports from WHO (https://www.who.int/emergencies/diseases/novel-coronavirus-2019/situation-reports
Although coronavirus disease 2019 (COVID-19) is primarily characterized as a respiratory disease, multiple organ dysfunction syndromes may occur in several organs, including the brain, which contributes to neurological manifestations (3
). Acute neurological and psychiatric complications of COVID-19 often occurred even in persons younger than 60 years of age (4
). Moreover, cortical signal alteration (5
), loss of white matter, and axonal injury (6
) have been reported, as well as increasing observations of neurological issues, including headache, ischemic stroke, seizures, delirium, anosmia, ageusia, encephalopathy, and total paralysis (7–14
). Patients with more severe infections are more likely to have neurological manifestations and impairment and are at a higher risk of mortality (15
Microglia are macrophage-like brain immune cells in the central nervous system (CNS). They have key functions in maintaining brain homeostasis and in the rapid response to injury and inflammation (16
). When microglia respond to immunological stimuli, they become activated and transform from a ramified into an amoeboid morphology, releasing interleukin (IL)-1β, IL-6, and tumor necrosis factor-α (TNF-α) (17
). Activated microglia consist of a dual phenotype, wherein M1, or the classically activated state, is neurotoxic and involved in neuroinflammation, and M2, or the alternatively activated state, is neuroprotective (18–20
). Increasing evidence suggests that the overactivation and dysregulation of microglia might result in disastrous and progressive neurotoxic consequences (21–24
). In the brains of deceased COVID-19 patients, microgliosis and immune cell accumulation were observed (25
), as well as microglial nodules caused by massive microglial activation in the medulla oblongata (26
) and cerebellar dentate nuclei (27
). The neuroinvasive capacity (28
) and olfactory transmucosal invasion of SARS-CoV-2 in patients with COVID-19 (29
) have also been reported. Additionally, human microglia express SARS-CoV-2 entry factors, such as angiotensin-converting enzyme 2 (ACE2) and transmembrane protease serine subtype 2 (TMPRSS2) (30
). Thus, we hypothesized that microglial activation by direct SARS-CoV-2 infection could be one of the major mechanisms driving the neuroinflammation and neurological complications.
Despite accumulating evidence, little is known regarding the mechanisms involved in the neuroinflammation of SARS-CoV-2 infection. In this study, we demonstrated that SARS-CoV-2 can directly infect human microglia and induce proinflammatory responses, reflecting polarization toward the M1 phenotype. We further showed that the SARS-CoV-2 infection led to apoptosis as a cytopathic effect (CPE) through both intrinsic and extrinsic apoptotic pathways. Moreover, we found that murine microglia expressing human ACE2 (hACE2) were infected by intranasally administered SARS-CoV-2, followed by microglial proinflammatory activation and loss of their population.
Although COVID-19 is primarily known as a respiratory disease, symptoms related to changes in the peripheral nervous system have recently been reported in patients (15
). There is evidence of viral invasion into the brain through the olfactory or vagal nerve and/or the oral and ophthalmic routes, with transsynaptic neuronal spread to other brain regions in COVID-19 patients’ autopsies and a SARS-CoV-2-infected mouse model (25
). Microglia, expressing SARS-CoV-2 entry factors, including ACE2 and TMPRSS2 (30
), can be infected by the invading SARS-CoV-2 and, subsequently, activated (49
). Recently, emerging data have shown the presence of NP expression of Iba1-positive cells in the brains of SARS-CoV-2-infected hamsters and even COVID-19 patients with microgliosis (32
). Activation of microglia could render individuals more at risk of developing neurological and psychiatric complications, even after complete clinical recovery from the infection (52
). In addition, dysfunctional or aberrant microglial activity could severely impair cognitive functions, including judgment, decision making, learning, and memory (53
). Hence, proinflammatory activation by SARS-CoV-2 infection of microglia might have critical outcomes on the short-, moderate-, or long-term neurological and psychiatric consequences of SARS-CoV-2 infection.
In this study, we found that SARS-CoV-2 directly infects a human microglial cell line. Regarding the RNA-seq analysis, it was observed that the viral infection-induced ER stress and immune responses in the early phase and apoptosis in the late phase. The microglia infected with SARS-CoV-2 were activated and polarized toward the M1 phenotype, the mediator of proinflammatory responses. In addition, SARS-CoV-2 infection triggered apoptosis as one of the CPEs in human microglia through both the intrinsic and extrinsic pathways, further provoking cell death. Indeed, SARS-CoV-2 has been reported to induce both intrinsic and extrinsic apoptosis in lung tissues and cell lines (54
). More recently, SARS-CoV-2-encoded membrane glycoprotein (M) and NP have been demonstrated to trigger apoptosis (55
), suggesting a mechanism of apoptosis in SARS-CoV-2-infected microglia. Moreover, we observed SARS-CoV-2 infection of microglia in K18-hACE2 mice, followed by an increase in the number of activated microglia and immune cell accumulation in the brain, and a decrease in the total number of microglia.
Given the colocalization of viral proteins and Iba1 in brains of patients with COVID-19 by postmortem studies (32
), we herein suggest a possibility that SARS-CoV-2 directly infects human microglia, inducing a CPE, which is cell death. In addition, human microglia could be one model to study viral pathogenesis. Microglia are important not only for the innate but also for the adaptive immune responses to pathogen infection of the brain (56
). Depletion of microglia can lead to ineffective T cell responses by reduction of the total number and percentage of CD4+
and regulatory T cells. Consequently, the depletion of microglia causes increased viral replication and increased neurological manifestations in the brain (57
). Therefore, microglial cell death by SARS-CoV-2 infection might be linked to the lack of immune response and consequent increased viral replication, leading to increased neurological manifestations.
The activation of microglia and complement-mediated pathways, which leads to the synthesis of inflammation mediators, is the key element of main inflammatory neurological diseases (59
). Many viruses, including HIV-1, herpes simplex virus (HSV), and ZIKV, infect microglia and cause neuroinflammation via microglial activation. C3 and its cleavage products attract microglia to gather around the neurons to exert phagocytotic activity and clear the presynaptic ends (60
). The increase in the number of activated microglia via viral infection can have detrimental effects, indirectly by activating astrocytes (61
) and T lymphocytes (62
), and directly by inducing neuronal damage and death, further contributing to neuronal degeneration (56
). In addition, the cytokine storm produced by microglia can lead to increased blood-brain barrier (BBB) breakdown and might be responsible for several neurological symptoms of COVID-19 (49
). Thus, microglia could be a potential target for SARS-CoV-2 and could help the spread of the virus in the CNS. In future studies, to gain more insights into the SARS-COV-2 neuropathology in patients with COVID-19, it will be important to consider the relationship between SARS-CoV-2 and microglia. Taken together, our findings indicate that microglia are potential mediators of neurological diseases by SARS-CoV-2 and, consequently, can be targets of therapeutic strategies against neurological disorders in patients with COVID-19.
Our study has some limitations. Viral RNA detection and proinflammatory responses and CPE in iPSC-Microglia should have been evaluated to further evaluate susceptibility to SARS-CoV-2 infection of microglia; however, the long incubation time was a limitation in this investigation because iPSC-Microglia undergo the cell death after prolonged incubation post the microglial maturation. Another limitation is that we used K18-hACE2 mice to assess microglial susceptibility to SARS-CoV-2, rather than the more naturally sensitive to viral infection model, such as wild-type Syrian hamsters. While the expression and distribution of hACE2 are under the cytokeratin-18 gene promoter in K18-hACE2 mice, these mice have been widely used and have the advantage of investigating extrapulmonary replication of SARS-CoV-2.
MATERIALS AND METHODS
Cells and viruses.
Human microglial clone 3 (HMC3) (CRL-3304), Caco-2 (HTB-37), and Vero E6 (CRL-1586) cell lines were purchased from ATCC (Manassas, VA, USA). These cells were maintained in Eagle’s Minimum Essential Medium (EMEM; Welgene, Gyeongsangbuk-do, South Korea) containing 10% fetal bovine serum (FBS) (Gibco, Waltham, MA, USA) and 1% Pen/Strep (Gibco). iPSC-Microglia were derived from human induced pluripotent stem cells using a published protocol (65
) with modifications. Primary human brain microglial cells (PHM) were purchased from Alphabioregen (PHM001, Boston, MA, USA). The SARS-CoV-2 Korean strain (GISAID accession no. MW466791
), isolated from a patient in South Korea, was obtained from Korea Centers for Disease Control and Prevention (KCDC) and propagated in Vero cells (CCL-81, ATCC).
Cells (1 × 105 cells per well) were plated into six-well plates and inoculated with one multiplicity of infection (MOI) SARS-CoV-2 in EMEM containing 2% FBS on the next day. After incubation for 1 h, the inoculum was removed, and the cells were washed two times. The medium was changed to EMEM containing 10% FBS. At 2, 4, or 6 dpi, the total cellular RNA was extracted using the RNeasy Minikit (Qiagen, Hilden, Germany). For iPSC-Microglia and PHM cells, 2 × 104 cells per well were plated into 24-well plates and infected with one MOI SARS-CoV-2 as described before, followed by detection of viral RNA after 24 h of incubation.
For the virus neutralization, the inoculum was further incubated with a CR3022 neutralizing antibody (ab273073, Abcam, Cambridge, UK) for 1 h. For crystal violet staining, cells were stained with 0.5% crystal violet staining solution in 25% methanol for 30 min. The plates were then washed three times with water and dried. The cell-covered areas were measured by the ImmunoSpot reader (CTL, Shaker Heights, OH, USA). The caspase inhibitors used in this study, including Z-DEVD-FMK (S7312), Z-VAD-FMK (S7023), Z-IETD-FMK (S7314), and Belnacasan (VX-765, S2228), were purchased from Selleckchem (Houston, TX, USA).
All procedures were performed in a biosafety level 3 (BSL3) or animal BSL3 facility for SARS-CoV-2-related experiments, with approval from the Korea Research Institute of Chemical Technology (KRICT) and by personnel equipped with powered air-purifying respirators.
Eight-week-old male B6.Cg-Tg (K18-hACE2) 2Prlmn/J mice were purchased from the Jackson Laboratory and maintained in a biosafety level 2 (BSL2) animal facility in the Korea Research Institute of Chemical Technology (KRICT). All protocols were approved by the Institutional Animal Care and Use Committee (IACUC; protocol no. 8A-M6, 2021-8A-02-01, and 2021-8A-03-03).
Virus inoculations (SARS-CoV-2; 2 × 104 PFU) were performed by the intranasal route (I.N.) under anesthesia using isoflurane in a BSL3 animal facility, and all efforts were made to minimize animal suffering. Body weights were measured every day postinfection.
Isolation of microglia in mice was performed following the protocols previously described (66
), with minimal modifications. Mock or SARS-CoV-2 infected mice were anesthetized by isoflurane, followed by perfusion with 10 or 20 mL of cold 1× DPBS (Gibco) into the left ventricle to remove blood from the tissues. Brains were transferred to a six-well plate containing cold Hanks’ Balanced Salt Solution (HBSS) (Gibco) and the plates were kept on ice. The generation of brain cell suspension by a 70-μm pore sized-cell strainer (SPL, Gyeonggi-do, South Korea) was made in 10 mL per brain of digestion cocktail containing 0.5 mg/mL DNase I (Roche, Basel, Switzerland) and 1 mg/mL Collagenase A (Roche) in HBSS. The suspension was incubated at 24°C for 30 min, followed by centrifugation for 7 min at 300 × g
, 18°C. The cell pellet was resuspended with 30% Percoll (Sigma-Aldrich, St. Louis, MO, USA) in HBSS, and then was slowly layered over 70% Percoll in HBSS in a 15 mL-conical tube. About 2 mL of interphase volume was collected to a new tube after gradient centrifugation for 40 min at 200 × g
, 18°C. Isolated mononuclear cells were washed three times in a volume of 500 μL of HBSS containing 0.01 M HEPES (Gibco), using a microcentrifuge for 7 min at 600 × g
Flow cytometry analysis.
Isolated brain mononuclear cells from the mock or SARS-CoV-2-infected mice in cell staining buffer (phosphate-buffered saline [PBS] with 1% FBS and 0.09% NaN3) were stained for 30 min with fluorescence-conjugated antibodies, namely, brilliant violet 421 anti-mouse/human CD11b antibody (101236, BioLegend, San Diego, CA, USA), PE/Cyanine7 anti-mouse CD45 antibody (103114, BioLegend), APC anti-mouse TNF-α antibody (506307, BioLegend), FITC anti-mouse IL-6 monoclonal antibody (MP5-20F3) (11-7061-82, eBioscience, San Diego, CA, USA), FITC anti-human ACE2 antibody (NBP2-7211F, Novus Biologicals, Centennial, CO, USA), and Alexa 647 anti-Iba1 antibody (78060S, Cell Signaling Technology, Danvers, MA, USA). Cells were then analyzed by FACSAria III sorter (BD Biosciences, San Jose, CA, USA), and data were analyzed by FlowJo software (BD Biosciences). All fluorochromes were compensated. The total leukocyte population was gated for microglia (CD11b+, CD45low). For annexin V staining, mock or SARS-CoV-2-infected HMC3 were stained with FITC-recombinant human annexin V, following the protocols of Annexin V-FITC Apoptosis Detection kit (BMS500FI-20, eBioscience).
Quantitative RT-PCR (QuantStudio 3, Applied Biosystems, Foster City, CA, USA) was performed with a one-step Prime script III RT-qPCR mix (RR600A, TaKaRa, Kyoto, Japan). The viral RNA of NP was detected by the 2019-nCoV-N1 probe (catalog number 10006770, Integrated DNA Technologies, Coralville, IA, USA). The IL-1β, IL-6, IL-12, TNF-α, IFN-β, IFN-λ1, NOS2, and Arginase-1 genes were detected by individually customized probes (Integrated DNA Technologies).
For RT-PCR analysis of GPR34, MERTK, and P2RY12, reverse transcription was carried out using GoScript Reverse Transcription System (Promega) according to the manufacturer’s instructions. The PCR amplifications were performed in SYBR green PCR master mix and CXR (Applied Biosystems, Warrington, U.K.) including primers using an ABI 7500 (Applied Biosystems). Average threshold cycle (Ct) values of GPR34, MERTK, and P2RY12 from triplicate PCRs were normalized from average Ct values of GAPDH.
Focus forming assay.
The cell culture medium was serially diluted in EMEM containing 2% FBS and was added to 4 × 104 Vero E6 cells plated on 96-well plates. After incubation for 8 h at 37°C, cells were washed and fixed with a 4% formaldehyde solution. Cells were stained with the anti-SARS-CoV-2 NP antibody (40143-R001, Sino Biological, Beijing, China) and a secondary horseradish peroxidase-conjugated goat anti-rabbit IgG (Bio-Rad, Hercules, CA). The signal was developed using an insoluble tetramethylbenzidine (TMB) substrate (Promega, Madison, WI, USA), and the number of infected cells was counted using an ImmunoSpot reader (CTL, Shaker Heights, OH).
RNA-seq and analysis.
A sequencing library was prepared with TruSeq Stranded mRNA Sample Prep kit and sequenced on NovaSeq 6000 (Illumina, San Diego, CA, USA), yielding more than 6G bases of sequences for each sample. From the sequenced reads, adaptor sequences were removed using Cutadapt (version 3.1) (68
) and aligned to the hybrid reference genomes of humans (GRCh38.p13_ENS100) and SARS-CoV-2 (ASM985889_v3) with STAR aligner (version 2.7.6a) (69
). Aligned reads were quantified at the gene level by HTSeq (version 0.13.5) (70
) with “intersection-nonempty” mode. Genes with lower than five counts for the total count per gene were removed for further analyses. DEG analysis was processed with DESeq2 (version 1.30.1) (71
) using abs (log2
change) > 1 and adjusted P
value (Benjamini-Hochberg) < 0.01 as the cutoff. Multidimensional scaling analysis was performed with the clustermap function in Python seaborn package (version 0.11.1) using genes with mean FPKM > 1 among the samples and transformed to log2
(FPKM + 1). Overrepresentation analysis of the DEGs enriched to GO Biological Process 2018 with EnrichR (72
) with adjusted P
value (Benjamini-Hochberg) <0.05 as the cutoff. Network analysis was presented by Metascape (73
) using the GO Biological Process gene sets.
After transcardial perfusion with cold 4% paraformaldehyde in PBS, brain tissues of SARS-CoV-2 infected K18-hACE2 mice were dissected and fixed by immersion in 4% paraformaldehyde in PBS overnight at 4°C. The brain sections (30 μm thickness) were permeabilized with 0.2% Triton X-100 in 1% BSA/PBS for 30 min, washed in PBS, and blocked with 0.5% BSA in PBS for 15 min, followed by incubation overnight at 4°C with primary antibodies, namely, anti-SARS-CoV-2 S (40150-T62-COV2, Sino Biological), and anti-Iba1/AIF1 (MABN92, Merck Millipore, Burlington, MA, USA). After washing twice, further incubation was carried out with Alexa Fluor 488-conjugated anti-rabbit antibody (A32731, Thermo Fisher Scientific, Waltham, MA, USA) and Alexa Fluor 594-conjugated anti-mouse antibody (A32744, Thermo Fisher Scientific). Immunofluorescence was observed by confocal microscopy (LSM700, Carl Zeiss, Oberkochen, Germany).
For the cell lines, the SARS-CoV-2-infected cells were fixed with 4% paraformaldehyde in PBS overnight at 4°C and then permeabilized with 0.5% Triton X-100 in PBS for 10 min, followed by washing thrice with PBS. Blocking buffer (0.1% Tween 20, 1% bovine serum albumin [BSA] in PBS) was added to remove the nonspecific binding. Cells were immunostained overnight at room temperature with primary antibodies, namely, anti-dsRNA J2 (MABE1134, Sigma-Aldrich), anti-SARS-CoV-2 NP (40143-R019, Sino Biological), anti-SARS-CoV-2 S (40150-T62-COV2, Sino Biological), and anti-CD68 (sc-17832, Santa Cruz Biotechnology, Dallas, TX, USA). After washing thrice, further incubation was carried out with Alexa Fluor 488-conjugated anti-rabbit antibody (A32731, Thermo Fisher Scientific) and Alexa Fluor 594-conjugated anti-mouse antibody (A32744, Thermo Fisher Scientific). Immunofluorescence was observed by confocal microscopy.
The culture supernatants were collected from infected cells and used for the detection of IL-1β, IL-6, IL-10, and TNF-α. Each cytokine was determined by the corresponding ELISA kit (IL-1β, K0331800; IL-6, K0331194; IL-10, K0331123; TNF-α, K0331131; Komabiotech, Seoul, South Korea), following the manufacturer’s instructions.
Cells were lysed in radioimmunoprecipitation (RIPA) buffer (Thermo Fisher Scientific), and proteins in the lysate were separated in a denaturing polyacrylamide gel and transferred to a polyvinylidene fluoride (PVDF) membrane (Merck Millipore, Burlington, MA, USA). The membrane was incubated with 5% skim milk (BD Biosciences) in tris-buffered saline with 0.1% Tween 20 (TBST) buffer and the primary antibodies, namely, anti-SARS-CoV-2 NP (40143-R001, Sino biological), anti-CD68 (sc-17832, Santa Cruz Biotechnology) anti-GSDMDC1 (sc-81868, Santa Cruz Biotechnology), anti-Actin (sc-47778, Santa Cruz Biotechnology), anti-Hsp70 (sc-24, Santa Cruz Biotechnology), anti-CX3CL1 (ab25088, Abcam), anti-CX3CR1 (ab8021, Abcam), anti-CD16 (80006S, Cell Signaling Technology), anti-phospho-Stat1 (9167S, Cell Signaling Technology), anti-Stat1 (14994S, Cell Signaling Technology), anti-Fas (4233T, Cell Signaling Technology), anti-DR4 (42533T, Cell Signaling Technology), anti-DR5 (8074T, Cell Signaling Technology), anti-TNFR2 (3727T, Cell Signaling Technology), anti-Bcl-2 (4223T, Cell Signaling Technology), anti-Bim (2933T, Cell Signaling Technology), anti-Bid (2002T, Cell Signaling Technology), anti-Bax (5023T, Cell Signaling Technology), anti-caspase-9 (9502S, Cell Signaling Technology), anti-caspase-8 (9746S, Cell Signaling Technology), anti-caspase-3 (9665S, Cell Signaling Technology), anti-caspase-1 (3866S, Cell Signaling Technology), anti-NLRP3 (15101S, Cell Signaling Technology), and anti-PARP (9542S, Cell Signaling Technology). Horseradish peroxidase (HRP)-conjugated secondary antibodies from Bio-Rad and ECL reagents (Thermo Fisher Scientific) were used for protein detection.
All experiments were performed at least three times. All data were analyzed using GraphPad Prism 8.0 software (GraphPad Software, San Diego, CA, USA). P < 0.05 was considered statistically significant. Specific analysis methods are described in the figure legends.
The SARS-CoV-2 (NCCP43326) was kindly provided by the National Culture Collection for Pathogens at the Korea Centers for Disease Control and Prevention. This work was supported by the National Research Foundation of Korea (NRF) grant, funded by the Ministry of Education, Science, and Technology (MIST) of the Korean government (2020R1C1C1003379 to Y.-C.K.) and the National Research Council of Science & Technology (NST) grant, funded by the Korean government (MSIP) (CRC-16-01-KRICT to Y.-C.K.). W.-H.S. and Y.C.C. are supported by the Korea Institute of Toxicology (1711133844). J.-Y.L. are supported by a faculty research grant of Yonsei University College of Medicine (6-2021-0155), a grant of the Korea Health Technology R&D Project through the Korea Health Industry Development Institute (KHIDI), funded by the Ministry of Health & Welfare, Republic of Korea (HV21C0050), and NRF grant funded by MSIT of Korean government (2021R1C1C1006912).
Conceptualization: G.U.J., and Y.-C.K.; methodology: G.U.J., K.-D.K., J.K., W.H.S., and Y.-C.K.; investigation: G.U.J., J.R., K.-D.K., Y.C.C., G.Y.Y., S.L., and I.H.; writing: G.U.J. and J.R.; review and editing: G.U.J., K.-D.K., W.H.S., J.K., J-Y.L., and Y-C.K.; funding acquisition: W.-H.S., J.-Y.L., and Y.-C.K.; and supervision: Y.-C.K.
We declare no conflict of interest.