Targeting intracellular Neu1 for coronavirus infection treatment

Summary

There are currently no effective therapies for COVID-19 or antivirals against severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), and vaccines appear less effective against new SARS-CoV-2 variants; thus, there is an urgent need to understand better the virulence mechanisms of SARS-CoV-2 and the host response to develop therapeutic agents. Herein, we show that host Neu1 regulates coronavirus replication by controlling sialylation on coronavirus nucleocapsid protein. Coronavirus nucleocapsid proteins in COVID-19 patients and in coronavirus HCoV-OC43-infected cells were heavily sialylated; this sialylation controlled the RNA-binding activity and replication of coronavirus. Neu1 overexpression increased HCoV-OC43 replication, whereas Neu1 knockdown reduced HCoV-OC43 replication. Moreover, a newly developed Neu1 inhibitor, Neu5Ac2en-OAcOMe, selectively targeted intracellular sialidase, which dramatically reduced HCoV-OC43 and SARS-CoV-2 replication in vitro and rescued mice from HCoV-OC43 infection-induced death. Our findings suggest Neu1 inhibitors could be used to limit SARS-CoV-2 replication in patients with COVID-19, making Neu1 a potential therapeutic target for COVID-19 and future coronavirus pandemics.

Graphical abstract

Highlights

• •Coronavirus N protein is sialylated
• •Sialylation on N protein affects its RNA-binding activity
• •Neu1 regulates coronavirus replication
• •Sialidase inhibitor Neu5Ac2en-OAcOMe rescued mice from death from coronavirus infection

Classification Description: Virology

Introduction

The ongoing coronavirus disease 2019 (COVID-19) pandemic, which is caused by the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), currently affects millions of lives worldwide and poses an overwhelming burden on global health systems and economies. Although vaccines are available, they appear to be less effective against newly emerging variants of the virus,^1,2,3,4,5,6 and SARS-CoV-2 continues to ravage many communities worldwide. Furthermore, SARS-CoV-2 is likely to become endemic,⁷ possibly leading to the emergence of vaccine-resistant variants and reinforcing the need for the development of antiviral therapeutic agents. SARS-CoV-2 is a β-coronavirus with a large (30 kilobase [Kb]) positive-strand RNA genome.

Currently, no antiviral drugs exist that specifically target SARS-CoV-2, which puts health-care givers in a similar situation as the 2002 SARS outbreak and in many other human coronaviruses infections. A French study found that combination of hydroxychloroquine and azithromycin might be efficacious,⁸ although evidence on the safety and efficacy of these therapies is limited.⁹ In addition, remdesivir may modestly accelerate recovery time¹⁰ and has been approved by the United States Food and Drug Administration (U.S. FDA) for treatment of hospitalized COVID-19 patients over the age of 12 years. Many of the monoclonal antibody treatments that were effective against past variants have not demonstrated effectiveness with the COVID-19 Omicron variant, and evidence on the safety and efficacy of the novel COVID-19 oral antiviral molnupiravir is also limited.^{11,12,13,14,15,16} More recently, Pfizer’s co-administered antiviral treatment of nirmatrelvir and ritonavir (Paxlovid) was just approved by the FDA for use in children aged 12 years and above; however, its effectiveness against SARS-CoV-2 variants is still unclear. Thus, the development of novel therapeutics against SARS-CoV-2 remains a top priority for combating the current pandemic and future coronavirus outbreaks.

The coronavirus genome encodes four major structural proteins: the spike (S), nucleocapsid (N), membrane (M), and envelope (E) proteins. All four proteins are required to produce a structurally complete viral particle.^17,18 While our knowledge of SARS-CoV-2 pathogenesis is expanding rapidly, most studies have only focused on a few proteins, such as the S protein and how it interacts with host immune systems. Little is known about the mechanisms that globally control host and viral RNA synthesis during infection. N protein is one of the most abundant structural proteins, which plays key roles in the regulation of viral replication and virion assembly, and is a major immunogen in coronavirus infection-induced disease. Therefore, N protein is an attractive target for diagnosis and treatment strategies against coronaviruses including SARS-CoV-2.^18,19 Nevertheless, the mechanisms by which specific cellular genes are induced and regulated during SARS-CoV-2 infection are poorly understood. Knowledge of how host defense genes are controlled might be a key to understanding COVID-19 disease pathogenesis.

Sialylation is the most frequent modification of proteins and lipids, which entails the addition of sialic acids (a family of nine-carbon acidic monosaccharides) to the terminal residues of glycoproteins and glycolipids. This modification is important in self-nonself discrimination,²⁰ phagocytosis of cancer cells by macrophages,²¹ determination of immunoglobulin E allergic pathogenicity,²² and bacterial infection.^23,24 The sialylation level of a cell is largely dependent on the activity of two kinds of enzymes: sialyltransferases that add sialic acid residues to glycolipids or glycoproteins and sialidases that remove sialic acid residues from glycolipids or glycoproteins.²³ Mammals express four sialidases, Neu1 to Neu4. Neu1 is the most abundant of the four mammalian sialidases and is primarily localized in the lysosomes,^25,26 but it is relocalized from the lysosomes to the cell surface during the differentiation of monocytes and the monocytic cell line (THP-1) into macrophages²⁷ or after bacterial infection.²⁸ The enzyme works optimally at an acidic pH of 4–5 but can also hydrolyze glycoproteins and gangliosides at near-neutral pH.²⁶ Sialidase inhibitors are useful tools for studying sialidase function and serve as drugs for sialidase-related diseases, such as viral infection. Inhibition of viral neuraminidase activity has been successfully utilized as a therapeutic approach for influenza infection.²⁹ Tamiflu (oseltamivir) and Relenza (zanamivir), which are approved for treatment of influenza A and B, have almost no effect on human sialidases (neuraminidases) but are potent inhibitors of neuraminidase activity of the influenza neuraminidase (NA) protein.^29,30 Herein, we investigated whether N protein is sialylated and whether sialylation influenced the biological activity of N protein and developed novel sialidase inhibitors for treatment of coronavirus infection.

Results

Sialylation on coronavirus N protein

Virus glycans play an important role in virus infection, immune evasion, and immune modulation.³¹ Coronavirus S, E, and M proteins are glycosylated, although N- and O-link glycosylation were found on the N protein in vitro overexpressing system.³² It is not clear whether the virion N protein is glycosylated or not.³³ Human coronavirus OC43 (HCoV-OC43) is a β-coronavirus responsible for mostly mild respiratory symptoms; thus, we analyzed the N- and O-glycans on N protein from HCoV-OC43 virion by mass spectrometry. N protein from HCoV-OC43 virion is also heavily glycosylated, containing both N- and O-linked glycans (Figure S1). COVID-19 is caused by coronavirus SARS-CoV-2.³⁴ So, to determine whether sialylation occurred on N protein from both coronavirus HCoV-OC43 and SARS-CoV-2, we performed a lectin blot with samples immunoprecipitated from the serum of COVID-19 patients and healthy controls and cell lysates infected with HCoV-OC43 with anti-N protein antibodies.

The immunoprecipitated samples were treated with or without sialidase and then separated by SDS-PAGE. As shown in Figures 1A and 1B, N protein from both patients with COVID-19 and HCoV-OC43-infected cells was heavily sialylated. Sialic acid was mostly attached in α2, 6 linkage on the N protein (Figures 1A and 1B), and N protein sialylation was confirmed by sialidase treatment (Figure 1B). Sialylation was also observed on the SARS-CoV-2-N protein expressed in HEK293T cells (Figure 1C), the HCoV-OC43-N protein in THP-1 cells (Figure 1D), and the HCoV-OC43 virion (Figure 1E). Large bands seen below the molecular weight (MW) of the nucleocapsid are probably the degraded or truncated products of nucleocapsid. In addition, we detected HCoV-OC43-N protein in Sambucus nigra Lectin (SNA) pulldown assays with HCoV-OC43 virion but not in Maackia amurensis Lectin (MAA) pulldown assays (Figure S2).

Figure 1.

Sialylation on coronavirus nucleocapsid (N) protein is critical for its RNA-binding activity

(A–C) Immunoprecipitated concentration of nucleocapsid protein using the sera of COVID19 patients (A), cell lysates from HCoV-OC43-infected THP-1 cells (B), or HEK293T cell lysates overexpressing SARS-CoV-2 nucleocapsid (C), followed by immunoblot analysis of sialylation using biotin-MAA (α2,3-linkage), biotin-SNA (α2,6-linkage) lectins, biotin-anti-SARS-CoV-2-N (A), anti-HCoV-OC43-N, or anti-SARS-CoV-2-N antibodies (C). Cell lysates were treated with sialidase (250 U/mL) at 37°C for 2 h. IP, immunoprecipitation.

(D) Viral nucleocapsid is sialylated in host cell. Immunoblot analysis of proteins in THP-1 cells using biotin-MAA and biotin-SNA lectins and anti-HCoV-OC43-N (nucleocapsid) antibody. β-actin was used as the loading control. Nucleocapsid indicated with arrowhead.

(E) Viral nucleocapsid is sialylated in HCoV-OC43 virion. Immunoblot analysis of the sialylation of nucleocapsid (indicated with arrowhead) in HCoV-OC43 virion.

(F) Immunoblot analysis of the SARS-CoV-2 nucleocapsid protein in HEK293T cells. Cells were transfected with a construct expressing SARS-CoV-2 nucleocapsid (SARS-CoV-2-N) for 48 h, and cell lysates were detected with anti-SARS-CoV-2-N and β-actin antibodies.

(G and H) Gel mobility shift assay of the 32-mer ssRNA (G) or 32-mer ssDNA (H). The probe was incubated with no cell lysates (lane 1) or lysates with the treatment indicated (lanes 2–5). Cell lysates were treated with sialidase (250 U/mL) at 37°C for 2 h. Data are representative of three independent experiments.

(I) Antibodies shift assay: 32-mer ssDNA was incubated with cell lysates after transfection with SARS-CoV-2 N protein expression vector and then added IgG (lane 1) or anti-N protein antibodies (lane 2) and separated in a 7% acrylamide gel. Data are representative of three experiments.

The role of sialylation in coronavirus N protein

Because the primary role of N protein is to assemble with genomic RNA into the viral RNA-protein complex,³⁵ we investigated whether the sialylation on N protein affects its RNA-binding activity. To assess the nucleic acid-binding affinity of N protein, we conducted nucleic acid-binding assays in the presence of a 32-mer stem-loop II (32m) motif single-stranded RNA (ssRNA) and its 32-mer ssDNA mimic. The 32m ssRNA is a highly conserved sequence among coronaviruses and has been used in the past to map the putative RNA-binding domain of SARS-CoV N protein.^36,37

For nucleic acid-binding assays, we lysed HEK293T cells 48 h after transfection with SARS-CoV-2-N protein expression vector (Figure 1F) and then treated the cells with or without sialidase. SARS-CoV-2-N protein formed a strong complex with 32-mer ssRNA (Figure 1G) and 32-mer ssDNA (Figure 1H), which was supershifted in the presence of anti-N protein antibodies (Figure 1I), indicating that this complex is specific for N protein. As expected, HEK293T cell lysates transfected with empty vector did not form a complex with 32-mer ssRNA and ssDNA (Figures 1G and 1H). Furthermore, 32-mer ssDNA and ssRNA bound to sialidase-treated N protein dramatically increased (Figures 1G and 1H). These findings indicated a significant increase in N protein RNA-binding activity after sialidase treatment, supporting the critical role of N protein sialylation in RNA binding.

A critical role of Neu1 in coronavirus replication

The sialylation level of a cell is largely dependent on the activity of two kinds of enzymes: sialyltransferases and sialidase. Sialyltransferases are responsible for adding sialic acid residues to glycolipids or glycoproteins, while sialidases are responsible for removing sialic acid residues from glycolipids or glycoproteins.²³ We evaluated the contribution of endogenous sialidases to the sialylation of N protein using THP-1 cell lines. Real-time PCR (Figure 2A) and western blot analysis (Figure 2B) indicated that the expression of Neu1 was significantly increased but that Neu2, Neu3, or Neu4 were not significantly increased after infection with HCoV-OC43 for 72 h. Notably, NEU1 is also upregulated in COVID-19 patients.³⁸ In addition, N protein was associated with Neu1 in HCoV-OC43-infected cells (Figure 2C). The interaction between N protein and Neu1 was further confirmed by overexpression of SARS-CoV-2 N protein and Neu1 in HEK293T cells (Figure S3).

Figure 2.

A critical role of Neu1 in coronavirus replication

(A) The mRNA levels of Neu1, Neu2, Neu3, and Neu4 normalized by GAPDH in THP-1 cells with or without HCoV-OC43 infection for 72 h.

(B) Immunoblot analysis of Neu1 in naive and HCoV-OC43-infected THP-1 cells. β-actin was used as the loading control.

(C) N protein associated with endogenous Neu1 in HCoV-OC43-infected THP-1 cells (48 h postinfection).

(D) Immunoblot analysis of Neu1 and viral nucleocapsid in Neu1-overexpressing THP-1 cells.

(E) Overexpression of Neu1 dramatically reduced the levels of sialylation on viral nucleocapsid in HCoV-OC43-infected 293T cells. 293T cells were infected with HCoV-OC43 (MOI = 1) for 24 h, followed by transfection of empty vector or the construct expressing Neu1-Myc/Flag for 24 h. Immunoblot analysis of the indicated targets.

(F) The levels of intracellular viral RNA (upper) and extracellular viral titers (lower) in control and Neu1-overexpressing THP-1 cells.

(G) The efficiency of Neu1 knockdown in THP-1 cells. RT-qPCR analysis of Neu1 mRNA abundance in THP-1 cells transduced with either scrambled shRNA or shRNAs targeting Neu1.

(H) Immunoblot analysis of Neu1 and viral nucleocapsid in scrambled and Neu1-knockdown THP-1 cells.

(I) The levels of intracellular viral RNA (upper) and extracellular viral titers (lower) in scrambled and Neu1-knockdown THP-1 cells. Data are representative of three independent experiments and shown as mean ± SD ∗p < 0.05; ∗∗p < 0.01; ∗∗∗p < 0.001. Analysis was performed using two-way ANOVA. The densitometric analysis of the western blot (WB) bands in Figures 2B, 2D, 2E, and 2H can be found in Figure S9.

Coronavirus N protein is a multifunctional RNA-binding protein necessary for viral replication.³⁵ Because we determined that the sialylation on N protein affects its RNA-binding activity, we investigated whether this sialylation affects virus replication. Viral infection was quantified by real-time qPCR (RT-qPCR) with primers targeting the coding region of the viral N gene. RNA was collected from cells at indicated time points after viral challenge, and viral transcripts were quantified. Supernatants were also processed for quantification of viral titer via 50% tissue culture infective dose (TCID₅₀) assay.³⁹ We made THP-1 stable cell lines overexpressing Neu1 (Figure 2D).

The sialylation on N protein was significantly decreased in the cells overexpressing Neu1 compared with empty vector control cells (Figure 2E). The replication of HCoV-OC43 was more than 10 times higher at the level of viral transcripts and viral titers in cell culture supernatants (Figure 2F) of cells overexpressing Neu1 than in cells expressing empty vector 48 h after viral challenge. In contrast, the replication of HCoV-OC43 was more than 100 times lower at the level of viral transcripts and viral titers in the cell culture supernatants (Figure 2I) of cells overexpressing short hairpin RNA (shRNA) for Neu1 (Figures 2G and 2H) than in cells expressing scrambled shRNA. Compared with scrambled shRNA, Neu1sh3 significantly decreased HCoV-OC43 replication more than Neu1sh1 and Neu1sh2, consistent with the knockdown efficiency of shRNA (Figures 2G–2I). N protein levels were also significantly decreased in Neu1-knockdown cells (Figure 2H) but dramatically increased in Neu1-overexpressing cells (Figure 2D). These data indicated that host Neu1 is a regulator of HCoV-OC43 replication in THP-1 cells.

Sialidase inhibitors library screening

The significantly reduced HCoV-OC43 replication in Neu1-knockdown THP-1 cells suggested endogenous sialidase plays a key role in HCoV-OC43 replication and thus may be a valuable therapeutic target. To test this concept, we set up a sialidase inhibitor library, which included the following: (1) FDA-approved influenza sialidase inhibitors: zanamivir and oseltamivir; (2) cell surface sialidase inhibitors: Neu5Gc2en and Neu5Ac2en9N3; and (3) newly synthesized intracellular sialidase inhibitors that are expected to have higher cell membrane permeability and higher cellular uptake than those of the parent compound. The active form of the inhibitor is subsequently released via hydrolysis by esterases⁴⁰ in cytosol: Neu5Ac2en-OMe, Neu5Ac2en-OAcOMe, and Neu5Ac2en9N3-OAcOMe.

We first assessed these sialidase inhibitors for antiviral activity against HCoV-OC43 in vitro. THP-1 cells treated with these inhibitors were challenged with HCoV-OC43 for 2 h. RNA was collected from cells, and viral transcripts were quantified 72 h after viral challenge (Figures 3A and 3B upper). Supernatants were also processed at 72 h for quantification of viral titer by TCID₅₀ assay (Figures 3A and 3B lower). Three of the tested sialidase inhibitors significantly repressed viral replication (Figures 3A and 3B). Among them, Neu5Ac2en-OAcOMe showed the highest antiviral activities (Figure 3A), which were dose dependent.

Figure 3.

Screen for sialidase inhibitors suppressing coronavirus propagation

(A and B) The intracellular viral RNA levels (upper) and viral titers (lower) in the supernatant of HCoV-OC43 (MOI = 2)-infected THP-1 cells treated with sialidase inhibitors Neu5Gc2en, zanamivir, Neu5Ac2en9N3, Neu5Ac2en-OMe, Neu5Ac2en-OAcOMe, Neu5Ac2en9N3-OAcOMe (A), or oseltamivir (B) for 72 h.

(C) Vero 76 cells were treated with the sialidase inhibitors and infected with SARS-CoV-2 for 48 h; then cell viability was measured. Data are shown as mean ± SD and are representative of three independent experiments. ∗p < 0.05; ∗∗p < 0.01; ∗∗∗p < 0.001. NS: not significant. ND: not detected. Analysis was performed using one-way ANOVA.

Screening for sialidase inhibitors that block SARS-CoV-2 replication

We next used a high-throughput screen (HTS) assay to test whether these inhibitors had similar effects on the replication of SARS-CoV-2 and used cell viability 48 h post-SARS-CoV-2 infection as a readout. Neu5Ac2en-OAcOMe showed the best protection (Figure 3C), while viral neuraminidase inhibitors zanamivir and oseltamivir did not show inhibitory activity (Figures 3A–3C). These findings agreed with previous reports that zanamivir and oseltamivir have limited potency against all human sialidases.³⁰

Effects of Neu5Ac2en-OAcOMe on coronavirus life cycle

To further evaluate the antiviral activity of sialidase inhibitors, we chose Neu5Ac2en-OAcOMe, which showed the highest antiviral activities, and evaluated whether Neu5Ac2en-OAcOMe acts on the virus binding and entry steps of the viral life cycle. We treated THP-1 cells with Neu5Ac2en-OAcOMe for 2 h and then infected the cells with HCoV-OC43 at 4°C or 37°C. Cells incubated at 4°C were collected 1 h postinfection, and cells incubated at 37°C were collected 2 h postinfection. Intracellular viral RNA was quantified by RT-qPCR. As shown in Figure 4A, viral loads were similar among cells treated with different amounts of Neu5Ac2en-OAcOMe, indicating that Neu5Ac2en-OAcOMe treatment did not affect virus binding and entry steps of the viral life cycle.

Figure 4.

Neu5Ac2en-OAcOMe exerts antiviral effects in human cell models

(A) THP-1 cells were pre-treated with vehicle or Neu5Ac2en-OAcOMe and were inoculated with HCoV-OC43 (MOI = 2) at either 4°C (upper) or 37°C (lower). Intracellular viral RNA was analyzed by RT-qPCR.

(B–D) HCoV-OC43 (MOI = 2)-infected THP-1 cells were treated with a gradient concentration of Neu5Ac2en-OAcOMe for 72 h. Intracellular viral RNA (B, upper), extracellular progeny virus yields (B, lower and C), and intracellular nucleocapsid (D) were analyzed by RT-qPCR, TCID₅₀ assay, and immunoblot, respectively. The inhibitory and cytotoxic curves (C) were obtained using the data from the lower panel in B and cell viability measured by MTS assay. CellTiter 96 AQueous One Solution Cell Proliferation Assay (MTS) (catalog no. G3582; Promega, Madison, WI) was performed according to the kit manual.

(E) HCoV-OC43-infected THP-1 cells were treated with vehicle or Neu5Ac2en-OAcOMe for 24 h. Immunoprecipitation was performed to capture nucleocapsid using anti-HCoV-OC43-N antibody. Sialylation was detected with biotin-MAA and biotin-SNA lectins.

(F) CID-1067700 did not reverse the decrease in intracellular viral RNA and extracellular viral production triggered by Neu5Ac2en-OAcOMe. THP-1 (MOI = 1) cells were inoculated with HCoV-OC43 virus for 2 h, and then CID-1067700 and Neu5Ac2en-OAcOMe were added after removal of inoculum. The levels of intracellular viral RNA (upper) and virus release (lower) were analyzed by RT-qPCR and TCID₅₀ assay, respectively (72 h postinfection.

(G) Immunoprecipitation of nucleocapsid protein from HCoV-OC43-infected THP-1 cells or purified virions, followed by immunoblot analysis of sialylation using biotin-MAA and biotin-SNA lectins and anti-HCoV-OC43-N (nucleocapsid) antibody. Data are representative of three independent experiments.

(H) Neu1 overexpression reversed the inhibition of Neu5Ac2en-OAcOMe on HCoV-OC43 infection. HeLa cells transfected with empty vector or Neu1 expression vector were inoculated with HCoV-OC43 virus (MOI = 1) and then treated with or without Neu5Ac2en-OAcOMe for 48 h. The levels of intracellular viral RNA were analyzed by RT-qPCR.

(I and J) HEK293T cells were transfected with Neu1 expression vector for 48 h, and then sialidase activity was measured after incubation with Neu5Ac2en-OAcOMe (I) or oseltamivir (J). Data are representative of at least three independent experiments and shown as mean ± SD. Analysis was performed using one-way ANOVA(A-B) or unpaired Student’s t test (F, H-J). ∗p < 0.05; ∗∗p < 0.01; ∗∗∗p < 0.001. NS: not significant. ND: not detected. The densitometric analysis of the WB bands in Figure 4D can be found in Figure S9.

To determine whether Neu5Ac2en-OAcOMe affects post-entry steps of the viral life cycle, we infected THP-1 cells with HCoV-OC43 for 2 h and then treated the cells with Neu5Ac2en-OAcOMe. We quantified intracellular viral RNA and viral titers in the cell culture supernatants 72 h postinfection. Neu5Ac2en-OAcOMe treatment significantly decreased viral replication in THP-1 cells at the level of viral transcripts (Figure 4B, upper) and viral titers in cell culture supernatants (Figure 4B, lower) in a dose-dependent manner, with an IC₅₀ of 13.69 μM (Figure 4C). N protein levels were also significantly decreased in Neu5Ac2en-OAcOMe-treated cells (Figure 4D).

Compared with vehicle-treated cells, Neu5Ac2en-OAcOMe-treated cells showed increased sialylation of HCoV-OC43 N protein (Figure 4E). Importantly, Neu5Ac2en-OAcOMe was nontoxic at all tested concentrations in the One Solution Cell Proliferation (MTS) assay (Figure 4C) and did not induce cytokine production in THP-1 cells (Figure S4). Moreover, compound identification number (CID)-1067700, a competitive inhibitor of Rab7 activation that can block β-coronavirus egress,⁴¹ moderately inhibited HCoV-OC43 release, not viral RNA replication, at 24 h postinfection. CID-1067700 also did not reverse the decrease in intracellular viral RNA and extracellular viral production triggered by Neu5Ac2en-OAcOMe treatment (Figure 4F). In addition, similar levels of sialylation were observed on cellular N protein and N protein in HCoV-OC43 virion (Figure 4G), indicating that virus budding did not affect sialylation of N protein. Collectively, these results demonstrated that viral RNA synthesis is a target of Neu5Ac2en-OAcOMe.

To rule out the possibility that Neu5Ac2en-OAcOMe induced protein degradation, we treated the cells with MG132. As shown in Figure S5A, inhibiting proteasome activity with MG132 did not affect the decrease in N protein expression induced by Neu5Ac2en-OAcOMe. Neu5Ac2en-OAcOMe also did not alter the levels of ubiquitination on HCoV-OC43-N protein (Figure S5B). Thus, decreased N protein expression was not likely due to protein degradation.

Antiviral activity of Neu5Ac2en-OAcOMe

Interestingly, Neu1 overexpression reversed the inhibition of Neu5Ac2en-OAcOMe on HCoV-OC43 infection (Figure 4H), indicating Neu1 is the target of Neu5Ac2en-OAcOMe. Because Neu1 exists on the plasma membrane and in cells,²⁸ we investigated whether Neu5Ac2en-OAcOMe-sensitive Neu1 resides on the plasma membrane or within the cells. To test this, we incubated HEK293T cells transfected with Neu1 expression vector with Neu5Ac2en-OAcOMe at room temperature for 30 min. We then measured cell surface sialidase activity and total cell lysate sialidase activity. Neu5Ac2en-OAcOMe significantly decreased total sialidase activity but did not affect cell surface sialidase activity (Figure 4I), indicating that Neu5Ac2en-OAcOMe targets intracellular Neu1, where coronavirus replication takes place.⁴² By contrast, Neu5Gc2en targets sialidase on the cell surface,²⁸ which may explain why Neu5Gc2en did not inhibit HCoV-OC43 replication in THP-1 cells (Figure 3A) or protect against SARS-CoV-2 infection-induced cell death (Figure 3C). Consistent with previous research,³⁰ oseltamivir treatment only slightly decreased intracellular sialidase activity (Figure 4J) as it is a prodrug as an ethyl ester.

Next, we sought to ensure that the observed efficacy of Neu5Ac2en-OAcOMe was not restricted to THP-1 cells. We evaluated the efficacy of Neu5Ac2en-OAcOMe in three epithelial cell lines (HEK293T cells, HeLa cells, and epithelial cell line isolated from African green monkey kidney (BSC-1) cells). Just like THP-1 cells (Figure 2A), Neu1 is the most abundant of the four mammalian sialidases, and the expression of Neu1 but not Neu2, Neu3, or Neu4 was significantly increased after infection with HCoV-OC43 in HEK293T and HeLa cells (Figure S6). Neu5Ac2en-OAcOMe inhibited viral replication in all three cell lines (Figure 5A) and significantly decreased N protein levels (Figure 5B). Immunofluorescence staining for HCoV-OC43-N protein also showed Neu5Ac2en-OAcOMe treatment effectively suppressed viral replication in BSC-1 cells (Figure 5C). These results indicated that the tested sialidase inhibitors did not inhibit the entry step of viral replication but rather interfered in the subsequent steps of the viral life cycle of coronavirus. Therefore, sialidase inhibitors represent a potential treatment for coronavirus infections.

Figure 5.

Neu5Ac2en-OAcOMe exhibits robust antiviral activity for coronavirus in epithelial cell lines

(A and B) HEK293T (MOI = 1), HeLa (MOI = 1), and BSC-1 (MOI = 0.1) cells were inoculated with HCoV-OC43 virus for 2 h, and Neu5Ac2en-OAcOMe was added after removal of inoculum. The levels of intracellular viral RNA (A, upper), virus release (A, lower), and intracellular nucleocapsid (B) were analyzed by RT-qPCR (A, upper), TCID₅₀ assay (A, lower), and immunoblot (B), respectively (72 h postinfection, except BSC-1, 48 h postinfection).

(C) BSC-1 (MOI = 0.1) cells were inoculated with HCoV-OC43 virus for 2 h, and Neu5Ac2en-OAcOMe was added after removal of inoculum. The levels of intracellular nucleocapsid were analyzed by immunofluorescence (48 h postinfection). Red indicates nucleocapsid, and blue represents cell nuclei stained by DAPI. Scale bar, 100 μm. Data are representative of three independent experiments and data shown as mean ± SD. Analysis was performed using unpaired Student’s t test. ∗∗p < 0.01; ∗∗∗p < 0.001. ND: not detected. The densitometric analysis of the WB bands in Figure 5B can be found in Figure S9.

Therapeutic activity of Neu5Ac2en-OAcOMe

We further examined the antiviral efficacy of Neu5Ac2en-OAcOMe in vivo. No signs or symptoms of toxicity were observed during the treatment period (Figure S7), so we determined whether Neu5Ac2en-OAcOMe could prevent HCoV-OC43 infection-induced death in newborn mice. Seven-day-old C57BL/6 pups were injected (intraperitoneal, [IP]) with 30 μL virus dilution (1 x 10⁵ TCID₅₀ of HCoV-OC43). Based on our in vitro data, we chose to inject (IP) Neu5Ac2en-OAcOMe (20 mg/kg) 1 h before virus infection. As shown in Figure 6A, 100% of vehicle-treated mice succumbed to HCoV-OC43 challenge, while 50% of Neu5Ac2en-OAcOMe-treated mice survived throughout the observation period. Body weight was significantly lower in vehicle-treated mice than in Neu5Ac2en-OAcOMe-treated mice (Figure 6B). Neu5Ac2en-OAcOMe-treated mice also showed suppression of HCoV-OC43 viral replication in the lungs, blood, and brain (Figure 6C).

Figure 6.

Evaluation of the therapeutic effects of Neu5Ac2en-OAcOMe

(A–E) Seven-day-old mice were pre-treated with Neu5Ac2en-OAcOMe or vehicle and then IP infected with 30 μL virus dilution (1 x 10⁵ TCID₅₀ HCoV-OC43).

(A) Survival curves after HCoV-OC43 infection. (n = 11 for vehicle group, n = 10 for Neu5Ac2en-OAcOMe group).

(B) Body weight after HCoV-OC43 infection. (n = 11 for vehicle group, n = 10 for Neu5Ac2en-OAcOMe group).

(C) Viral RNA copies of HCoV-OC43 in the blood, brain, and lung on day 5 postinfection. (n = 6 for vehicle group, n = 7 for Neu5Ac2en-OAcOMe group).

(D) Cytokines in blood on day 5 postinfection. (n = 6 for vehicle group, n = 7 for Neu5Ac2en-OAcOMe group).

(E) Histological analysis of brain and lung tissues on day 5 postinfection. Tissue sections were stained with H&E. Scale bar, 100 μm. Data are representative of at least three independent experiments and shown as mean ± SD (A–E). Analysis was performed using Kaplan-Meier analysis (A), unpaired Student’s t test (B, C), or one-way ANOVA(D). ∗p < 0.05; ∗∗p < 0.01; ∗∗∗p < 0.001.

Both viral and host factors impact disease pathogenesis. Cytokine storm is a major cause of mortality^43,44 during infection with SARS-CoV, Middle East respiratory syndrome coronavirus (MERS), and SARS-CoV-2. In our previous study, sialidase inhibitors rescued mice from bacterial infection-induced death by inhibiting the activity of host cell surface Neu1 and suppressing the cytokine storm.²⁸ To determine whether Neu5Ac2en-OAcOMe exerts similar effects in coronavirus infection, we measured serum levels of interleukin 6 (IL-6) and tumor necrosis factor alpha (TNF-α), which contributed to the cytokine storm and correlated with respiratory failure and adverse clinical outcome in COVID-19.⁴⁵ We detected substantially decreased serum IL-6 and TNF-α levels in Neu5Ac2en-OAcOMe-treated mice (Figure 6D). Hematoxylin and eosin (H&E) staining also indicated less brain and lung tissue damage in Neu5Ac2en-OAcOMe-treated mice (Figure 6E). Taken together, these findings show Neu5Ac2en-OAcOMe conferred significant protection against HCoV-OC43 challenge by reducing viral replication in vivo and the associated inflammatory dysregulation.

Discussion

N protein has a mass of 50–60 kDa (Figures 1 and 2), indicating the presence of posttranslational modifications such as N- or O-linked glycosylation (Figure S1) and sialylation (Figure 1). Moreover, SARS-CoV-2-N protein is highly glycosylated, as demonstrated by glycomic and glycoproteomic analyses after expression in HEK293T cells.³² In this study, N protein from SARS-CoV-2 and HCoV-OC43 was significantly sialylated, and this sialylation was tightly regulated by host Neu1. Coronavirus replication occurs in the cytoplasm of infected host cells.⁴² The sialidase inhibitor Neu5Ac2en-OAcOMe targets intracellular sialidase but not cell surface sialidase, which is advantageous because β-coronaviruses traffic to lysosomes^41,46,47 where Neu1 is also known to be predominantly localized.²⁵ Notably, the newly developed sialidase inhibitor Neu5Ac2en-OAcOMe reduced HCoV-OC43 replication in vitro and in vivo by inhibiting host Neu1 activity and rescued mice from HCoV-OC43 infection-induced death. Moreover, Neu5Ac2en-OAcOMe reduced SARS-CoV-2 replication in vitro.

Our data suggest that the sialidase inhibitor Neu5Ac2enOAcOMe has no effect on HCoV-OC43 coronavirus binding and entry steps in their THP-1 models (Figures 4A and 4B) but targets intracellular sialidase Neu1 to control coronavirus replication (Figures 3 and 4). Although Neu2 is also cytosolic, Neu5Ac2enOAcOMe could interfere with trafficking of the virus to the lysosome by inhibiting activity of Neu2; however, Neu2 expression is very low and does not change upon HCoV-OC43 infection (Figures 2A and S6). A recent report suggested that gangliosides could serve as attachment receptors.⁴⁸ We believe that Neu5Ac2enOAcOMe does affect the sialylation levels of host glycolipids as Neu5Ac2enOAcOMe targets to intracellular sialidases. While Neu3 on the cell surface regulates the sialylation of glycolipids and Neu3 expression is low, it does not change upon HCoV-OC43 infection (Figure 2A).

Inhibition of viral neuraminidase activity has been developed as a therapeutic approach for influenza infection.²⁹ Tamiflu (oseltamivir) and Relenza (zanamivir), which are FDA approved for treatment of influenza A and B, have almost no effect on human neuraminidases.^29,30 Several clinical trials have assessed the efficacy of oseltamivir in treating SARS-CoV-2 infection, but no positive outcomes were observed.^{49,50,51,52,53} This lack of efficacy could be attributed to several reasons: (1) SARS-CoV-2 genomic RNA does not encode sialidase^54,55; (2) SARS-CoV-2 replication depends on the host Neu1 (Figure 2); and (3) oseltamivir has few inhibitory effects on host Neu1 (Figure 4J).³⁰ Together, these findings explain why oseltamivir has not shown efficacy in the treatment of COVID-19. Here, we demonstrated that the newly synthesized sialidase inhibitor Neu5Ac2en-OAcOMe specifically targets intracellular host sialidase Neu1 and inhibits coronavirus replication (Figure 7), accounting for the contributions of both the host and pathogen in the disease process. Based on our findings, sialidase inhibitors could be a generalizable and effective treatment in the current COVID-19 pandemic, as well as future coronavirus pandemics associated with the inflammatory response.

Figure 7.

Proposed inhibition of cytosolic desialylation (sialidase) by cytosolic sialidase inhibitor

Limitations of the study

There are some limitations in the study. First, although we found that sialylation of SARS-CoV-2-N protein affected the formation of a complex with 32-mer ssRNA and 32-mer ssDNA, it is not clear whether sialylation of SARS-CoV-2-N affects its binding to RNA in vivo. This limitation may be avoided in future studies through an immunoprecipitation of protein N from Neu1-wildtype and Neu1-knockout cells, followed by quantitating and sequencing of the associated RNA. Second, we have showed that a newly developed Neu1 inhibitor, Neu5Ac2en-OAcOMe, selectively targeted intracellular sialidase, which dramatically reduced HCoV-OC43 and SARS-CoV-2 replication. However, Neu5Ac2enOAcOMe would be a broad-spectrum sialidase inhibitor. We have made selective Neu1 inhibitor now and will study this in the future. Third, we indicated Neu5Ac2en-OAcOMe treatment rescued mice from HCoV-OC43 infection-induced death by using a mouse model in which the sialidase inhibitor was injected 1 h before virus infection. For a more clinically relevant effect of the drug, future studies will include a mouse model in which the sialidase inhibitor will be administered after virus infection.

STAR★Methods

Key resources table

REAGENT or RESOURCE	SOURCE	IDENTIFIER
Antibodies
Anti-SARS-CoV-2 N protein	Thermo Fisher Scientific	catalog no. MA5-36270; RRID: AB_2890570
Anti-HCoV-OC43 N protein	Sigma-Aldrich	catalog. no. mab9013; RRID: AB_95425
anti-human Neu1 antibodies	Sigma-Aldrich	catalog. no. PA5-42552; RRID: AB_2605856
Anti-ubiquitin mouse monoclonal antibody (FK2)	EMD Millipore	catalog. no. ST1200; RRID: AB_10681625
Anti-β-actin	Santa Cruz Biotechnology	catalog. no. sc-1615; RRID: AB_630835
Anti-Neu1	Santa Cruz Biotechnology	catalog. no. sc-32936; RRID: AB_2298197
Bacterial and virus strains
HCoV-OC43	ATCC	ATCC® VR-1558™
SARS-CoV-2	BEI resources	USA-WA1/2020
Biological samples
COVID-19 patient sera	Raybiotech	CoV-Neut-S-100
Chemicals, peptides, and recombinant proteins
Biotinylated Maackia Amurensis Lectin II	Vector Laboratories	catalog. no. B-1265
Biotinylated Sambucus Nigra Lectin	Vector Laboratories	catalog. no. B-1305
MG132	Peptide institute	catalog. no. 3175-v
Oseltamivir	Thermo Fisher Scientific	50,148,746
Zanamivir	Thermo Fisher Scientific	AC462960010
Neu5Gc2en	Thermo Fisher Scientific	362,000
Neu5Ac2en9N3	made in the lab	N/A
Neu5Ac2en-OMe	made in the lab	N/A
Neu5Ac2en-OAcOMe	made in the lab	N/A
Neu5Ac2en9N3-OAcOMe	made in the lab	N/A
Experimental models: Cell lines
Hela	ATCC	CCL-2
BSC-1	ATCC	CCL-26
HEK293T	ATCC	CRL-3216
THP-1	ATCC	TIB-202
Experimental models: Organisms/strains
C57BL/6J	The Jackson Laboratory	000,664
Oligonucleotides
ssDNA (5′-CGAGGCCACGCGG AGTACGATCGAGGGTACAG-3′)	Thermo Fisher Scientific	N/A
ssRNA (5′-CGAGGCCACGCG GAGUACGAUCGAGGGUACAG-3′)	Eurofins Genomics	N/A
Neu1: 5′-GGAGGCTGTAGGGTTTGGG-3′ (forward), 5′-CACCAGACCGAAGTCGTTCT-3′ (reverse)	Thermo Fisher Scientific	N/A
Neu2: 5′-CCATGCCTACAGAATCCCTGC-3′ (forward), 5′-CTCTGCGTGCTCATCCTTC-3′ (reverse);	Thermo Fisher Scientific	N/A
Neu3: 5′-AAGTGACAACATGCTCCTTCAA-3′ (forward), 5′-TCTCCTCGTAGAACGCTTCTC-3′ (reverse);	Thermo Fisher Scientific	N/A
Neu4: 5′-GGCCACGGGATGACAGTTG-3′ (forward), 5′-CAGGCGGATACCCATGTGAG-3′ (reverse);	Thermo Fisher Scientific	N/A
HCoV-OC43 N gene, 5′-CGATGAGGCTA TTCCGACTAGGT-3′ (forward) and 5′-CCTTC CTGAGCCTTCAATATAGTAACC-3′ (reverse).	Thermo Fisher Scientific	N/A
Recombinant DNA
human Neu1 expression vector	made in the lab	N/A
Software and algorithms
GraphPad Prism version 9	GraphPad	N/A

Resource availability

Lead contact

Further information and requests for reagents should be directed to the Lead Contact, Guo-Yun Chen (Gchen14@uthsc.edu).

Materials availability

Materials are available from the lead contact upon reasonable request, but a Material Transfer Agreement may be required.

Experimental model and subject details

Cell lines

HeLa, BSC-1, HEK293T and THP-1 cells were obtained from ATCC (Manassas, VA) and cultured in Dulbecco’s Modified Eagle Medium (DMEM) or Roswell Park Memorial Institute (RPMI) medium supplemented with 10% heat-inactivated fetal bovine serum, 2 mM glutamine, and 100 μg/mL penicillin/streptomycin.

Mouse experiments

WT C57BL/6J mice were obtained from The Jackson Laboratory (Bar Harbor, ME). All animal procedures were approved by the Animal Care and Use Committee of The University of Tennessee Health Science Center, Memphis, TN. Age- and sex-matched male and female mice were used in the experiments. The HCoV-OC43 infection mouse model was established as described previously.⁵⁶ Briefly, 7-day-old mice were separated randomly into two groups and injected intraperitoneally (IP) with either Neu5Ac2en-OAcOMe (20 mg/kg) or vehicle (0.5% dimethyl sulfoxide, DMSO). One hour later, mice were inoculated with 30 μL of virus dilution (1 x 10⁵ TCID₅₀ of HCoV-OC43) by IP injection. Neu5Ac2en-OAcOMe and vehicle were administered daily and mice were monitored up to 10 days for survival. To detect viral RNA loads in tissues and cytokine production, mice were euthanized at 5 days postinfection. Mouse brain, lung, and blood tissues were collected.

Method details

Reagents

COVID-19 male and female patients sera (both IgG and IgM antibodies to the N protein were negative) were purchased from Raybiotech (Peachtree, GA). Anti-SARS-CoV-2 N protein (HL5410, catalog no. MA5-36270) was obtained from Thermo Fisher Scientific (Waltham, MA). Anti-HCoV-OC43 N protein and anti-human Neu1 antibodies were purchased from Sigma-Aldrich (catalog. no. PA5-42552, St. Louis, MO). Anti-ubiquitin mouse monoclonal antibody (FK2) (catalog. no. ST1200, lot no. D00165221) was obtained from EMD Millipore (Merck KGaA, Darmstadt, Germany). MG132 (catalog. no. 3175-v, lot no. 640311) was purchased from Peptide institute (Osaka, Japan). Biotinylated Maackia Amurensis Lectin II (MAL II, MAA) (catalog. no. B-1265) and Biotinylated Sambucus Nigra Lectin (SNA, EBL) (catalog. no. B-1305) were purchased from Vector Laboratories (Burlingame, CA). Anti-β-actin, Anti-Neu1 (catalog. no. sc-32936, lot no. B2813, rabbit polyconal antibodies), Streptavidin-Horseradish peroxidase (HRP) and HRP-conjugated anti-mouse, anti-goat, or anti-rabbit secondary antibodies were purchased from Santa Cruz Biotechnology (Santa Cruz, CA). The lentiviral vectors expressing human Neu1 shRNAs and puromycin were purchased from Sigma. Stable clones were obtained after selection with puromycin (2.5 μg/mL) for 3 weeks after infection. Neuraminidase (sialidase) provided by Vibrio cholerae (catalog. no. 11080725001) was purchased from Sigma-Aldrich. The 32m ssDNA (5′-CGAGGCCACGCGGAGTACGATCGAGGGTACAG-3′) was purchased from Thermo Fisher Scientific. The 32m ssRNA (5′-CGAGGCCACGCGGAGUACGAUCGAGGGUACAG-3′) was purchased from Eurofins Genomics (Louisville, KY). SARS-CoV-2 nucleocapsid encoding plasmid was purchased from Sino Biological (cat. No. VG40588-UT, Beijing, China). Oseltamivir, Zanamivir and Neu5Gc2en were obtained from Thermo Fisher Scientific. Neu5Ac2en9N3, Neu5Ac2en-OMe, Neu5Ac2en-OAcOMe and Neu5Ac2en9N3-OAcOMe were synthesized as described.^57,58

Syntheses of sialidase inhibitors

Neu5Ac2en-OMe and Neu5Ac2en-OAcOMe were synthesized using our previously reported method.⁵⁷

Neu5Ac2en9N3 is synthesized by literature method.⁵⁸

Neu5Ac2en9N3-OAcOMe is newly made from Neu5Ac2en9N3-OMe.⁵⁸

Neu5Ac2en9N3-OMe (100 mg) was dissolved into a minimal amount of a 1:3 solution of acetic anhydride in anhydrous pyridine at 0 ºC, then warmed to room temperature and stirred for 18 hours. The solution was concentrated and the residue was purified by flash chromatography with 5:1 EtOAc-hexane as the mobile phase to give Neu5Ac2en9N3-OAc-OMe as a white solid (65 % yield,). ¹H NMR (400 MHz, CD₃OD): δ 6.05 ppm ( d, J=3 Hz, 1H), 5.53 ppm ( dd, J= 3 Hz, 1H), 5.51 ppm ( dd, J= 3 Hz, 1H), 5.23 ppm ( m, 1H), 4.39 ppm ( dd, J= 9, 3 Hz, 2H), 3.95 ppm ( dd, J=12, 2 Hz, 1 H), 3.86 ppm ( s, 3H), 3.54 ppm ( dd, J=8, 3 Hz, 1 H), 2.12 ppm ( s, 3H), 2.08 ppm ( s, 3H), 2.05 ppm ( s, 3H), 1.89 ppm ( s,3H). ¹³C NMR (101 MHz, CDCl3): δ 172.3, 171.5, 162.3, 147.3, 108.6, 73.4, 67.8, 52.3, 50.5, 46.9, 23.5, 21.8 (Figure S8).

Construction of plasmids

To generate the constructs expressing human Neu1, cDNA for Neu1 was amplified by RT-PCR and subcloned into expression vectors pcDNA6 and pLVX-puro (Life Technologies, Carlsbad, CA) as we previously reported.²⁸ All constructs were verified by restriction enzyme digestion and DNA sequencing. Stable clones were obtained after selection with puromycin (2.5 μg/mL) for 3 weeks after infection the lentiviral vectors expressing human Neu1.

Gel-mobility-shift assay

ssDNA or ssRNA in phosphate buffer (10 mM sodium phosphate, 50 mM NaCl, 1 mM EDTA, 0.01% NaN3, pH 7.4) was heated to 95°C and immediately put on ice to destroy its secondary structure. 1x 10⁷ HEK293T cells in a 10 cm dish transfected with empty vector or SARS-CoV-2 N protein expression vector for 48 hours were harvested, suspended in lysis buffer (20 mM Tris-HCl, 0.1% Triton X-100, 150 mM NaCl, pH 7.6) and separated equally. Half of the cell lysates were treated with sialidase for 2 hours at 37°C. The oligonucleotides were mixed with the cell lysates and incubated on ice for 10 min and then separated on 1% agarose gels. For antibody supershift assays, 1 μg of anti-N protein antibodies was added to the reaction mixture and then separated in a 7% acrylamide gel as described in previous publication.⁵⁹

Immunofluorescence

Cells were fixed with 4% paraformaldehyde in phosphate buffered saline (PBS) at room temperature for 15 min and then permeabilized with 1% Triton X-100 in PBS at room temperature for 15 min. Immunofluorescence staining was performed as described previously.⁶⁰ Images were acquired with an EVOS FL Auto Imaging System (Thermo Fisher Scientific).

HCoV-OC43 production and titration

HCoV-OC43 virus (ATCC VR-1558) was purchased from ATCC. The stock of HCoV-OC43 was produced and titrated using BSC-1 cells. Viral titers in cell-free culture supernatants were determined by endpoint dilution-based TCID₅₀ assays in 96-well plates.³⁹ Cytopathic effect was recorded and used for calculation of viral titers at 7 days postinfection.

SARS-CoV-2 high-throughput screen (HTS) cytopathic effect assay

Double-blinded SARS-CoV-2 high-throughput screen (HTS) cytopathic effect assay was performed with the fee service provided by The University of Tennessee Health Science Center Regional Biocontainment BSL3 Laboratory. The basic methods for the HTS for the identification of potential inhibitors of coronavirus have been previously described.⁶¹ Briefly, seven sialidase inhibitors (oseltamivir, zanamivir, Neu5Gc2en, Neu5Ac2en9N3, Neu5Ac2en-OMe, Neu5Ac2en-OAcOMe, and Neu5Ac2en9N3-OAcOMe) were plated in 384-well black wall plates containing 4,500 Vero 76 cells/well in single dose of indicated concentration in Eagle’s MEM with 5% heat inactivated FBS, 1% penicillin/streptomycin/L-glutamine, 1% HEPES, and 0.5% DMSO. The cells were infected with SARS-CoV-2 at an MOI of 0.1. Plates were then incubated at 37°C, 5% CO2, for 48 hours. The cell viability at the end of incubation period was measured as described elsewhere.⁶² After incubation, 100 μL of Promega CellTiter-GloR (Promega, Madison, WI) was added to each well using the BiomekR 2000. Plates were shaken for 2 min at speed 5 on a Labline Instruments (Kochi, India) plate shaker. Luminescence was then measured using a PerkinElmer Envision™ plate reader (PerkinElmer, Wellesley, MA).

Neuraminidase activity assay

Sialidase activity was measured using 2’-(4-methylumbelliferyl)-α-D-N-acetylneuraminic acid sodium salt hydrate (4-MU-NANA, catalog no. sc-222055, Santa Cruz Biotechnology) as the substrate. 1 x 10⁷ HEK293T cells in a 10 cm dish transfected with Neu1 expression vector were harvested after 48 hours, incubated with inhibitors for 30 min at room temperature, washed to remove the sialidase inhibitors, and separated equally. Half of the cells were used for detection of cell surface sialidase activity and half of the cells were suspended in lysis buffer (20 mM Tris-HCl, 0.1% Triton X-100, 150 mM NaCl, pH 7.6) for detection of whole cell sialidase activity. For the reaction, intact cells or lysed cells were incubated with 4-MU-NANA (final concentration, 15 μM) for 30 min at 37°C in 50 μL reaction buffer (50 mM Sodium phosphate, pH 5.0). The reaction was terminated by adding 600 μl stop buffer (0.25 M glycine-NaOH, pH 10.4). Fluorescence intensity was measured with a Synergy HTX Multi-Mode Reader (EMD Millipore, Merck KGaA) (excitation: 360 nm; emission: 460 nm).

Real-time quantitative PCR

Total RNA was extracted with TRIzol (Invitrogen, Carlsbad, CA) according to the manufacturer’s protocol and reverse transcribed with random primers and Superscript III (Life Technologies). The mRNA expression of human Neu1, Neu2, Neu3 and Neu4 was measured by real-time PCR. Samples were run in triplicate, and the relative expression was determined by normalizing the expression of each target to the endogenous reference, GAPDH. For RT-qPCR, copy numbers of HCoV-OC43 viral RNA were calculated based on a standard curve generated using pEF6-OC43-N-V5 His DNA. The following primers were used: Neu1: 5′-GGAGGCTGTAGGGTTTGGG-3′ (forward), 5′-CACCAGACCGAAGTCGTTCT-3′ (reverse); Neu2: 5′-CCATGCCTACAGAATCCCTGC-3′ (forward), 5′-CTCTGCGTGCTCATCCTTC-3′ (reverse); Neu3: 5′-AAGTGACAACATGCTCCTTCAA-3′ (forward), 5′-TCTCCTCGTAGAACGCTTCTC-3′ (reverse); Neu4: 5′-GGCCACGGGATGACAGTTG-3′ (forward), 5′-CAGGCGGATACCCATGTGAG-3′ (reverse); HCoV-OC43 N gene, 5′-CGATGAGGCTATTCCGACTAGGT-3′ (forward) and 5′-CCTTCCTGAGCCTTCAATATAGTAACC-3′ (reverse).

Immunoprecipitation and immunoblotting

Cell lysates were prepared in lysis buffer (20 mM Tris-HCl, 150 mM NaCl, 1% Triton X-100, pH 7.6, including protease inhibitors, 1 μg/mL leupeptin, 1 μg/mL aprotinin and 1 mM phenylmethylsulfonyl fluoride), sonicated, centrifuged at 13,000 rpm for 5 min, diluted in immunoprecipitation buffer (20 mM Tris-HCl, 150 mM NaCl, pH 7.6, supplemented with the protease inhibitors as described above) and pre-cleared with 50 μL of protein A/G-conjugated agarose beads (Santa Cruz Biotechnology) for 1 h at 4°C, before incubation with corresponding antibodies. Immunoprecipitates were washed four times with immunoprecipitation buffer and re-suspended in SDS sample buffer for Western blot analysis. The concentration of running gel was 10%. After blocking, the blots were incubated with the appropriate primary antibody (1: 1,000 dilution) or Biotin-MAA/SNA (1 μg/mL). After incubation with the secondary antibody (HRP-conjugated goat anti-rabbit IgG, goat anti-mouse IgG, 1: 5,000 dilution) or Streptavidin-HRP (1: 10,000 dilution), the signal was detected with an enhanced chemiluminescence (ECL) kit (Santa Cruz Biotechnology, Santa Cruz, CA). Original WB films for figures are depicted in the panels of Figure S10.

Lectin pulldown assay

HCoV-OC43 virion was lysed in Triton X-100 Lysis Buffer (pH 7.4, 200 mM Tris-HCl, 150 mM NaCl 1% Triton X-100 and 5 mM EDTA) and then sonicated for 30 second on ice. HCoV-OC43-N protein was pulled down with SNA or MAA from the virion lysates, separated by 10% SDS-PAGE and then detected with anti-N protein antibodies.

Measurement of inflammatory cytokines

Mouse blood samples were obtained at indicated time points, and cytokines in the serum were determined using a mouse cytokine bead array designed for inflammatory cytokines (552364, BD Biosciences, San Jose, CA). Human cytokines in cell culture-derived supernatants were determined using a human cytokine bead array designed for inflammatory cytokines (551811, BD Biosciences).

Quantification and statistical analysis

GraphPad Prism software (San Jose, CA) was used for data analysis. Data are shown as mean ± SD or mean ± SEM. Statistical significance was analyzed by two-tailed t-test for two groups or one-way analysis of variance (ANOVA) or two-way ANOVA for three or more groups. Differences in survival rates were analyzed by Kaplan-Meier plot, and statistical significance was determined using a log-rank (Mantel-Cox) test. ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001, n.s., not significant.

Published: January 24, 2023

Data and code availability

• •All relevant data are included in the manuscript.
• •This paper does not report original code.
• •Any additional information required to re-analyse the data reported in this paper is available from the lead contact upon request.