Identifying SARS-CoV-2 antiviral compounds by screening for small molecule inhibitors of Nsp3 papain-like protease

Figures

• 

  Figure 1.. Design, expression and purification of PLpro enzyme. (A) Schematic of the COVID-19 genome showing the 29.9 kb single-strand (+) RNA. ORF 1ab is 21.5 kb and codes for a polyprotein which after processing, produces 16 proteins named nsp1–16. The nsp3 protein contains the catalytic core PLpro enzyme used in this study which comprises the Ubl2 and PLpro domains (highlighted in red). The ‘948 bp’ highlighted in red indicates the sequence that was the basis of the bacterial and insect cell expression constructs used for protein purification. (B) Purification of His-Sumo-PLpro. Lane 1, Digestion of the pulldown (PD) with Ulp1; lane 2, flow-through (FT) of second Ni-NTA; lane 3, FT from the MonoQ; lane 4, concentrated from lane 3 before applying to a gel filtration column; lanes 13–16 are the gel filtration fractions containing PLpro monomer from that were pooled for further assay.

• 

  Figure 2.. Enzyme assay design and enzyme characteristic. (A) Schematic of the Pro3 peptide designed for used in the FRET assay. The synthetic peptide contains 12 amino acids from the nsp2/3 junction in the natural polypeptide, in the centre of this is the FTLKGG//APTKVT sequence recognized by PLpro which cleaves between G and A. During synthesis the peptide had the fluorescent Anthranilate (2-aminobenzoyl-Abz) tag added to the N-terminus and the quencher nitro-L-tyrosine (Y(3-NO2)R) fused to the C-terminus. (B) Determination of the enzyme kinetics. The initial velocity of substrate hydrolysis over the titration of substrate is plotted. The velocity did not saturate at the tested concentration, so KM were estimated using Michaelis–Menten equation based on the incomplete dataset. Data was collected from three replicates. (C) Protease activity of PLpro. Titration of the purified enzyme (0.5–2 µM) incubated with the substrate. Fluorescent intensity was measured for 1 h. Data was collected from three replicates.

• 

  Figure 3.. High-throughput screening of the compound library. (A) (i) Flow diagram of the compound screen. Over 5000 compounds were dispensed into 24 of 384-well plates, enzyme was pre-incubated with compound before the addition of substrate and fluorescent intensity was read at an excitation 330 nm and emission at 420 nm with a Tecan Spark plate reader. (ii) Organization in 384-well plate. (B) Results of the screen. Two concentrations of the compound library were aliquoted- 1.25 µM and 3.75 µM. A single dot from the scatter plot represents each of the over 5000 compounds. There are 8 compounds and 29 compounds, in low and high concentration, respectively that reduce the PLpro activity more than 25%. Amongst those hits, four compounds were overlapping. In the low concentration (1.25 µM), one compound is auto-fluorescent and three have a Z score of more than −3.5.

• 

  Figure 4.. Validation of the hits with a gel-based protease assay. (A) Schematic of the substrate designed for the gel-based protease assay. A polypeptide with a N-terminal GST and a C-terminal MBP domain separated by 10 amino acids from the natural nsp2/3 junction (similar to that used in the fluorescence assay) was constructed. Cleavage between the amino acids G and A results in 25 kDa and 42 kDa products. (B) Flow diagram of the gel-based assay. (C) (i–iii) Two concentrations of each compound were used, 3.75 µM and 10 µM. Only Cdk4 inhibitor III, dihydrotanshinone I, GRL-0617, tanshinone IIA, cryptotanshinone, PDK1/Akt/Flt dual pathway inhibitor, Ro 08-2750 (2,3,4,10-Tetrahydro-7,10-dimethyl-2,4-dioxobenzo[g]pteridine-8-carboxaldehyde) and beta-lapachone show inhibition towards PLpro cleavage. GRL-0617, tanshinone IIA, and cryptotanshinone are the published compound inhibitors of PLpro.

• 

  Figure 5.. Determination of the IC50 of the validated hits with the fluorescent-based assay. (A) PDK1/AKT/Flt dual pathway inhibitor, (B) Ro 08-2750, (C) Cdk4 inhibitor III, (D) beta-lapachone, (E) dihydrotanshinone I, (F) GRL-0617, (G) tanshinone IIA, (H) cryptotanshinone.

• 

  Figure 6.. Inhibition of PLpro isopeptidase activities. An amount of 10 µM of each compounds were used. (A) PLpro DUB activities. PLpro cleaves K48-linked Ub3 to Ub2 and Ub1. These activities were inhibited strongly by beta-lapachone and dihydrotanshinone I, moderately by Ro 08-2750 and GRL-0617, but not by shikonin. (B) PLpro deISGylating activities. Pro-ISG15 is being cleaved to ISG15 by PLpro. Similar observation as in A.

• 

  Figure 7.. Determination of the EC50 of dihydrotanshinone I (Ai), Ro 08-2750 (Bi), GRL-0617 (Ci) in a cell culture-based assay and the synergistic effect of each compound with 0.5 µM remdesivir (Aii, Bii and Cii). Data was collected from three replicates.