A screen of the NIH Clinical Collection small molecule library identifies potential anti-coronavirus drugs

Abstract

With the recent emergence of Middle East Respiratory Syndrome coronavirus in humans and the outbreak of devastating porcine epidemic diarrhea coronavirus in swine, therapeutic intervention is urgently needed. However, anti-coronavirus drugs currently are not available. In an effort to assist rapid development of anti-coronavirus drugs, here we screened the NIH Clinical Collection in cell culture using a luciferase reporter-expressing recombinant murine coronavirus. Of the 727 compounds screened, 84 were found to have a significant anti-coronavirus effect. Further experiments revealed that 51 compounds blocked virus entry while 19 others inhibited viral replication. Additional validation studies with the top 3 inhibitors (hexachlorophene, nitazoxanide and homoharringtonine) demonstrated robust anti-coronavirus activities (a reduction of 6 to 8log10 in virus titer) with an IC50 ranging from 11nM to 1.2μM. Furthermore, homoharringtonine and hexachlorophene exhibited broad antiviral activity against diverse species of human and animal coronaviruses. Since the NIH Clinical Collection consists of compounds that have already been through clinical trials, these small molecule inhibitors have a great potential for rapid development as anti-coronavirus drugs.

Keywords: Coronavirus; NCC library; Screening; Small molecule.

Graphical abstract

Fig. 1

Effect of candidate drugs on MHV infection at different MOIs. DBT cells were treated with drugs at 10 μM for 1 h and then infected with MHV-2aFLS at MOI of 1 or 0.1 in the presence of the indicated drugs for 8 h. Cells were then lysed for determining luciferase activity. Cells treated with 1% DMSO were served as control. The antiviral effectiveness of the drugs is expressed as percent luciferase activity to the control, which is 100%. Data represent mean of 3 independent treatments and standard deviation of the means.

Fig. 2

Correlation of inhibition on luciferase reporter expression and virus titer. DBT cells were treated with various drugs as indicated (10 μM) or DMSO (1%) as a control for 1 h and then infected with MHV-2aFLS at MOI of 1 for 8 h. The medium was harvested for determining viral titer (TCID₅₀) (B) and cells were lysed for determining luciferase activity (A). Data indicate the mean of 3 replicates (independent treatments) and standard deviation of the mean. Statistical significance of the inhibitory effect of the drugs on luciferase activity (A) or virus titer (B) as compared to those of DMSO control is indicated by the number of asterisks (^∗p < 0.05; ^∗∗p < 0.01; ^∗∗∗p < 0.001).

Fig. 3

Inhibitory effect of hexachlorophene on MHV infection. (A) Chemical structure of hexachlorophene. (B) Determination of IC₅₀. DBT cells were treated with hexachlorophene at various concentrations as indicated or 1% DMSO (vehicle control) for 1 h, and were infected with MHV-2aFLS at MOI of 1 in the presence of the drug for 8 h. Cells were then lysed for luciferase assay. Inhibition of MHV infection was expressed as percent reduction in luciferase activity following drug treatment compared to the control, and the IC₅₀ was then calculated as indicated by the solid lines. (C) Inhibition of viral titer. DBT cells were treated with hexachlorophene (10 μM) or DMSO (1%) as a control for 1 h and then infected with MHV-2aFLS at MOI of 1 for 12 h. The medium was harvested for determining viral titer (TCID₅₀). Data indicate the mean of 3 replicates and standard deviation of the mean. (D) Inhibition of viral N protein expression. The experiments were performed identically to (C), except that different concentrations of the drugs were used. Following drug treatment and viral infection, cells were lysed to evaluate viral N protein expression levels by Western blotting. Beta-actin serves as a loading control.

Fig. 4

Inhibitory effect of nitazoxanide on MHV infection. (A) Chemical structure of nitazoxanide. (B) Determination of IC₅₀. DBT cells were treated with nitazoxanide at various concentrations or 1% DMSO (vehicle control) for 1 h, and were infected with MHV-2aFLS at MOI of 1 in the presence of the drug for 8 h. Cells were then lysed for luciferase assay. Inhibition of drug on MHV infection was expressed as percent reduction on luciferase activity to the control and the IC₅₀ was then calculated as indicated by the solid lines. (C) Inhibition of viral titer. DBT cells were treated with nitazoxanide (10 μM) or DMSO (1%) as a control for 1 h and then infected with MHV-2aFLS at MOI of 1 for 12 h. The medium was harvested for determining viral titer (TCID₅₀). Data indicate the mean of 3 replicates and standard deviation of the mean. (D) Inhibition of viral N protein expression. The experiments were performed identically to (C), except that different concentrations of the drugs were used. Following drug treatment and viral infection, cells were lysed to evaluate viral N protein expression levels by Western blotting. Beta-actin serves as a loading control.

Fig. 5

Inhibitory effect of homoharringtonine on MHV infection. (A) Chemical structure of homoharringtonine. (B) Determination of IC₅₀. DBT cells were treated with homoharringtonine at various concentrations or 1% DMSO (vehicle control) for 1 h, and were infected with MHV-2aFLS at MOI of 1 in the presence of the drug for 8 h. Cells were then lysed for luciferase assay. Inhibition of drug on MHV infection was expressed as percent reduction on luciferase activity to the control and the IC₅₀ was then calculated as indicated by the solid lines. (C–E) Cells were pretreated with homoharringtonine at various concentrations as indicated or 1% DMSO for 1 h. Cells were then infected with MHV-A59 for 12 h (C and D in DBT cells) or MHV-A59GFP for 16 h (E in 17Cl-1 cells) at MOI of 1 in the presence of the drug. (C) Virus titer in the medium was determined by TCID₅₀. Data represent the mean of 3 replicates and standard deviation of the mean. (D) Viral N protein expression in cell lysates was detected by Western blotting using mAb J.3.3. Beta-actin serves as a loading control. (E) Expression of GFP was directly observed using a fluorescence microscope (Olympus IX-70), and images were captured using a digital camera (Zeiss).

Fig. 6

Identification of candidate drugs that inhibit MHV infection during viral replication. (A) DBT cells were either treated with drugs (10 μm) as indicated or DMSO (1%) at 1 h before or 3 h after infection with MHV-2aFLS at MOI of 1. The cells were lysed at 8 h p.i. and luciferase activity was measured and expressed as a percentage of DMSO control. Data represent the mean of triplicate experiments and standard deviation from the mean. (B) 17Cl-1 cells were either treated with drugs (10 μm) as indicated or DMSO (1%) at 1 h before or 3 h after infection with MHV-A59GFP at MOI of 1. At 16 h p.i. EGFP expression was observed using a fluorescence microscope (Olympus IX-70), and images were captured using a digital camera (Zeiss). Representative images for 3 h post infection are shown.

Fig. 7

Identification of candidate drugs that inhibit MHV infection during cell entry. (A) DBT cells were either treated with drugs (10 μM) as indicated or DMSO (1%) at 1 h before or 3 h after infection with MHV-2aFLS at MOI of 1. The cells were lysed at 8 h p.i. and luciferase activity was measured and expressed as a percentage of DMSO control. Data represent the mean of triplicate experiments and standard deviation from the mean. (B) 17Cl-1 cells were either treated with drugs (10 μM) as indicated or DMSO (1%) at 1 h before or 3 h after infection with MHV-A59GFP at MOI of 1. At 16 h p.i. EGFP expression was observed using a fluorescence microscope (Olympus IX-70), and images were captured using a digital camera (Zeiss).

Fig. 8

Candidate drugs are capable of inhibiting infection with diverse coronaviruses. (A) DBT cells were infected with various MHV strains (MHV-1, MHV-2, and MHV-JHM) in the presence of the indicated drugs for 12 h, and the viral N protein was detected by Western blot using mAb J.3.3. Beta-actin serves as loading control. (B and C) HRT-18 cells were infected with either BCoV-L9 (B) or HECoV-4408 (C) in the presence of the drugs for 36 h, and the viral N protein was detected by IFA using specific mAb#46. The concentration of homoharringtonine is 1 μM and hexachlorophene is 5 μM. The control is 1% DMSO.