Aloin isoforms (A and B) selectively inhibits proteolytic and deubiquitinating activity of papain like protease (PLpro) of SARS-CoV-2 in vitro

The most common host entry point of human adapted coronaviruses (CoV) including SARS-CoV-2 is through the initial colonization in the nostril and mouth region which is responsible for spread of the infection. Most recent studies suggest that the commercially available oral and nasal rinse products are effective in inhibiting the viral replication. However, the anti-viral mechanism of the active ingredients present in the oral rinses have not been studied. In the present study, we have assessed in vitro enzymatic inhibitory activity of active ingredients in the oral mouth rinse products: aloin A and B, chlorhexidine, eucalyptol, hexetidine, menthol, triclosan, methyl salicylate, sodium fluoride and povidone, against two important proteases of SARS-CoV-2 PLpro and 3CLpro. Our results indicate only aloin A and B effectively inhibited proteolytic activity of PLpro with an IC50 of 13.16 and 16.08 μM. Interestingly, neither of the aloin isoforms inhibited 3CLpro enzymatic activity. Computational structural modelling of aloin A and B interaction with PLpro revealed that, both aloin isoforms form hydrogen bond with Tyr268 of PLpro, which is critical for their proteolytic activity. Furthermore, 100 ns molecular dynamics (MD) simulation studies predicted that both aloin isoforms have strong interaction with Glu167, which is required for PLpro deubiquitination activity. Our results from the in vitro deubiquitinase inhibition assay show that aloin A and B isomers exhibit deubiquitination inhibitory activity with an IC50 value of 15.68 and 17.51 µM, respectively. In conclusion, the isoforms of aloin inhibit both proteolytic and the deubiquitinating activity of SARS-CoV-2 PLpro, suggesting potential in inhibiting the replication of SARS-CoV-2 virus.


Scientific Reports
| (2022) 12:2145 | https://doi.org/10.1038/s41598-022-06104-y www.nature.com/scientificreports/ mediated through the internalization of the virus-ACE2 complex via clathrin-mediated endocytosis 3 . The viral genome (ssRNA) uses the host ribosomes to translate the viral RNA into a long polypeptide chain (PP) of about 800 kDa. Two proteases encoded by the viral genome, papain like proteases (PLpro) and 3-chymotrypsin like protease (3CLpro), auto-cleave the newly formed PP chain to generate several non-structural proteins (NSPs) required for the viral replication. The PP chain is cleaved into 16 NSPs by PLpro and 3CLpro, with the 3CLpro generating 11 of the 16 NSPs making this protease the main target for developing anti-SARS-CoV-2 drugs 4 . In addition to the protease activity, SARS-CoV-2 PLpro exhibits deubiquitination (DUB) activity 5 . Ubiquitination, a process of attachment of ubiquitin (UB) and ubiquitin like proteins (UBL) to the cellular proteins that needs to be degraded by the host proteasomal complex in cytosol, is an essential process required to maintain the host protein turn over. Ubiquitination also plays an important role in degrading the foreign proteins such as viral proteins upon infection to prevent the viral propagation 6 . Thus, the deubiquitination activity of SARS-CoV-2 PLpro leads to the disruption of host's anti-viral immune response. During viral infection, innate immune cells such as dendritic cells produce Type-I interferon (IFN-α/β), which in turn activates interferon-sensitive gene-15 (ISG-15). The upregulated ISG-15 protein conjugates with various signaling molecules including JAK, STAT, IRF-3 through a process called ISGylation to mediate the Type-I IFN induced anti-viral function [7][8][9] . It has been shown that SARS-CoV-2 PLpro mediates deISGylation of ISG-15 to the host signaling molecules that leads to the inhibition of the host anti-viral innate immune response 7,10 . Therefore, the DUB activity of SARS-CoV-2 dysregulates the primary interferon mediated anti-viral response, which is the hallmark of COVID-19 11 . Plethora of reports suggest that the SARS-CoV-2 mediated mortality is by the pro-inflammatory cytokine storm 12 . The possible mechanism of pro-inflammatory cytokine storm upon SARS-CoV-2 infection might be due to the dysregulated interferon-mediated anti-viral response (Fig. 1). Thus, PLpro serves as a drug target not only to inhibit the viral replication but also to suppress the cytokine storm during SARS-CoV-2 infection. Taken together, these studies suggest that the drug candidates, which specifically inhibits the enzymatic activity of 3CLpro and PLpro will control the replication of SARS-CoV-2. However, the drugs that can specifically target DUB activity of SARS-CoV-2 PLpro may prevent the cytokine storm mediated tissue damage.
Although global vaccination is being currently undertaken, with new variants emerging, there are no SARS-CoV-2 specific FDA approved drugs to prevent the viral transmission from person to person. Several studies have shown that the major mode of transmission for SARS-CoV-2 is through respiratory droplets expelled from the infected person 13 . Additionally, the air droplet produced from asymptomatic individuals is one of the major silent source for the spread of the SARS-CoV-2 virus 14 . Importantly, the high initial viral loads are concentrated in mouth and nasopharyngeal region 15 . Therefore, it has become a huge risk factor for healthcare workers such as dentists, physicians, nurses who come in close contact with asymptomatic or infected patients. This warrants an urgent need to develop strategies to prevent the transmission of the virus. It has been shown that commercially available anti-septic mouthwash products exhibited anti-bacterial and anti-viral activity in the oral cavity 16 . After rinsing the mouth, Listerine and chlorhexidine reduced the herpes simplex virus-1 load in the saliva 17,18 .

Figure 1.
Schematic representation of proteolytic and DUB activity of PLpro. SARS-CoV-2 PLpro proteolyticaly cleaves the viral protein at three sites to generate the mature NSPs. Additionally, PLpro inhibits the binding of UBLs such as ISG-15 to signaling molecules that are required to elicit the type-I interferons mediated anti-viral immune response that leads to survival of the virus in the host and ultimately leading to viral-mediated pro-inflammatory cytokine response in host. ISG Interferon sensitive gene, UBL Ubiquitin like protein, NSP Nonstructural protein.  [19][20][21] . The mechanism could be due to the membrane-disruption or viral protein inactivation by the reagents used in the mouthwash 16 . However, the specific target of the active components used in the mouthwash to prevent the viral replication is still not clear. Herein, we have investigated the inhibitory effect of active ingredients used in several commercially available mouthwash products through the 3CLpro and PLpro enzymatic activity assay. Most commonly found active ingredients in mouth rinses such as aloin A and B, chlorhexidine, eucalyptol, hexetidine, menthol, triclosan, methyl salicylate, sodium fluoride, and povidone iodide were studied for their inhibitory effects. We observed that none of these compounds were able to inhibit the 3CLpro enzymatic activity. However, both aloin A and B were able to inhibit more than 70% PLpro proteolytic and DUB activity. Our data suggest that aloin A and B might be potential drug candidates not only to inhibit the SARS-CoV-2 replication, but also to control the cytokine storm in COVID-19 patients. However, the SARS-CoV-2 anti-viral effect of aloin isomers need to be validated by preclinical and clinical studies.

Results
Inhibition of SARS-CoV-2 proteases by the active ingredients present in the mouthwash products. Since commercially available mouth rinses are effective in inactivation of SARS-CoV-2 19-21 , we performed an in vitro enzymatic assay for the most important SARS-CoV-2 proteases such as 3CLpro and PLpro with active ingredients of mouth rinses. As shown in the Fig. 2, none of the active ingredients of the mouthwashes were able to inhibit the 3CLpro enzymatic activity at 50 µM concentration. Interestingly, as shown in Fig. 3, out of the 9 compounds tested, only aloin isomers (A and B from) were able to inhibit more than 70% of PLPro proteolytic activity at 50 µM concentration. These studies suggest that aloin isomers exhibit specific inhibitory activity towards SARS-CoV-2 PLpro.

Structural interaction and 100 ns MD Simulation of aloin A and B with SARS-CoV-2 PLpro.
The structure of SARS-CoV-2 PLpro is divided into four sub-domains, the N-terminal Ubiquitin-like domain, the α-helical Thumb domain, the β-stranded Finger domain and the Palm domain ( Figure S1). This arrangement is similar to ubiquitin specific proteases deubiquitinating enzyme (DUB) with a very low homology (10%) 22 . The thumb comprises of six α helices and a small β hairpin. The fingers subdomain is made of six β strands and two α helices. The palm subdomain comprised of six β strands. It suggests that the proteolytic and DUB sites are independent of each other suggesting two possible activates of PLpro. The conventional catalytic triad Cys 111 -His 272 -Asp 286 is located between the interface of palm and thumb subdomains. In addition to catalytic triad three additional residues play an important role in the enzymatic activity of SARS-CoV-2 PLpro: 1) An important β turn/loop (Glu 266 -Gly 271 ) which closes upon substrate and/or inhibitor binding is found adjacent to the active site. 2) Tyr 268 part of the (Glu 266 -Gly 271 ) plays a critical role in proteolytic activity of SARS-CoV-2 PLpro. The mutation of Tyr 268 has shown to interfere with the proteolytic activity of SARS-CoV-2 PLpro and The interaction of aloin isomers to the ligand site of GRL0617, in the SARS-CoV-2 PLpro (PDBID: pbd7cmd) was analyzed using MOE software. Orientations that showed strong structural interaction with Tyr268 were considered for 100 ns MD simulation, as any hydrogen bonding interaction with Tyr 268 will interfere with the proteolytic activity. Molecular docking studies of aloin A and B with PLpro resulted in 20 orientations for each of them. Four best orientations (two for each) with hydrogen bonding interaction at Tyr 268 were chosen and then simulated with MD to evaluate the stability of these orientations at the 100ns time interval (Fig. 4A-D). The MD simulation data suggests that all four orientations were stable, and the molecules remained bound to the enzyme throughout the 100ns simulated time (Movies S1-S4). The analysis of protein-ligand interaction fingerprint between the SARS-CoV-2 PLpro enzyme and aloin A showed that the orientation-1 had a very weak Try 268 interaction throughout the duration of simulation (Fig. 5A, panel-I). Orientation-1 for aloin A also showed significant interaction with Gln 269 (Fig. 5A, panel-I). In contrast, orientation-2 for aloin A (Fig. 5B, Panel-I) showed significant interaction with Try 268 , Gln 269 , and Glu 167 . As mentioned earlier, Glu 167 plays an important role in the deubiquitination of the enzyme and aloin A, molecular modeling predicted that the orientation-2 of aloin A significantly impairs the DUB activity of SARS-CoV-2 PLpro. The fingerprint region of aloin A with PLpro over 100ns time showed that the S-score for orientation-1 fluctuated from − 6.25 to − 4.25 kcal/mol but was stable for the period of computation (Fig. 5A, panel-II). The S-score for orientation-2 (− 6.5 to − 4.25 kcal/ mol) of aloin A was stable for the first 80ns but fluctuated over the last 20ns of the calculation, however, the molecule always stayed bound to the enzyme over the period of study (Fig. 5B, panel-II). Hence based on the interaction of aloin A with Glu 167 , Tyr 268 , and Glu 269 , the orientation-2 of aloin A seems to be predominant for its interaction with the SARS-CoV-2 PLpro. Figure 5C, D shows the fingerprint region and S-score of the two different orientations of aloin B with the SARS-CoV-2 PLpro enzyme. Figure 5C, panel-I shows that the orientation-1 of aloin B has a very strong interaction with Try 268 , Gln 269 , and Glu 167 during the period of simulation similar to orientation-2 of aloin A (Fig. 5B, panel-I). These interactions of orientation-1 of aloin B with Try 268 , Gln 269 , and Glu 167 can explain the strong inhibition of proteolytic as well as DUB activity of PLpro. The S-score for orientation-1 of aloin B ranged from − 6.75 to − 5.65 kcal/mol and was very stable over the period of 100ns (Fig. 5C, panel-II). Figure 5D, Panel-I shows that the orientation-2 of aloin B has a very strong interaction with Try 268 and Gln 269 but did not show any interaction with Glu 167 during the period of simulation. The S-score for orientation-2 of aloin B was also very unstable throughout the period of simulation (Fig. 5D, panel-II) and varied from − 6.75 to − 3.85 kcal/mol. Hence based on the interaction of aloin B with Glu 167 , Tyr 268 , and Glu 269 orientation-1 of aloin B seems to be predominant during its interaction with the SARS-CoV-2 PLpro. Thus, orientation-2 of aloin A and orientation-1 of aloin B with triplicate values were presented graphically. P value < 0.001 considered as statistically significant. One-way ANOVA with Bonferroni's Multiple Comparison post-hoc test was used to calculate the statistical significance.

Inhibition of deubiquitination (DUB) activity of SARS-CoV-2 PLpro by aloin isomers. Our MD
Simulation data revealed that both aloin isomers may demonstrate DUB inhibitory activity of PLpro, therefore, we investigated the in vitro DUB activity of PLpro in the presence of aloin A and B. Previous data suggests that aloin isoforms inhibit the proteolytic activity of SARS-CoV-2 PLpro enzyme (Fig. 3). We performed the in vitro DUB activity using the Papain-like Protease (SARS-CoV-2) Deubiquitinase Assay Kit (BPS biosciences). As shown in Fig. 6, both the isoforms were able to inhibit more than 70% of DUB activity of the SARS-CoV-2 PLpro enzyme at 50µM concentration. These data aligns with the predicted interaction from MD simulation studies data suggesting that both aloin isomers A and B not only inhibits the proteolytic activity of SARS-CoV-2 PLpro but also its DUB activity.
Dose and time dependent inhibition of both proteolytic and DUB activity of SARS-CoV-2 by aloin isomers. Next, we subjected both aloin A and B for further dose-dependent studies to calculate the concentration required to inhibit the 50% of PLpro enzymatic activity (IC 50 ). IC 50 is the most widely used measure of antagonist drug potency in pharmacological research. In this study, IC 50 represents the concentration of aloin compounds required for 50% inhibition of PLpro and DUB enzymatic activity in vitro. As we observed in Fig. 7, aloin isomers inhibited around 80% proteolytic and DUB activity of PLpro at 100 µM concentration. The concentration of aloin isomers against the percent activity of PLpro was used to determine the IC 50 with nonlinear curve fit model as described in Methods section. The IC 50 value for aloin A and B was found to be 13.16 and 16.08µM for proteolytic activity and 15.68 and 17.51µM for DUB activity, respectively. Further, the time dependent data (Fig. 8) suggests that the aloin isomers started exhibiting their inhibitory effect towards both proteolytic and DUB activity of PLpro as early as 1h and attained their maximum inhibitory effect by 4h under our assay conditions and the inhibition continued till 18h. Additionally, aloin A and B did not exhibit cytotoxic effect on African green monkey kidney epithelial cells Vero-E6 (C1008) for 24 and 48h at 50 and 100µM concentration (Supplementary figure S2) respectively. Vero-E6 cells are known to be sensitive to SARS-CoV-2 23,24 therefore we selected Vero-E6 cells for cell viability assay. Although aloin isomers did not alter the cell viability at the tested dose for up to 48h, it is possible that the aloin isomers cytotoxic effect may alter up on viral infection, therefore future studies are warranted for possible changes in the effective concentrations of aloin isomers.
Taken together, our data suggest that both aloin A and B are specific inhibitors of proteolytic and DUB activity of PLpro but not 3CLpro enzyme of SARS-CoV-2 virus and thus may help in the inhibition of SARS-CoV-2 viral replication.

Discussion
SARS-CoV-2 virus is responsible for causing COVID-19, which is now a global threat especially with the emerging variants because of their high infectivity and mortality rate [25][26][27][28] . Though vaccines have been developed and are still in pursuit, there are no SARS-CoV-2 specific drugs available to stop the spread of this virus. Wearing mask is one of the means to prevent the spread of the virus; however, it is a challenge to control the spread of the virus in a setting where mask needs to be removed such as dental health care or encounters with asymptomatic individuals. SARS-CoV-2 mainly spreads through respiratory droplets 14 , therefore there is an immediate need to identify potential molecules that can reduce the viral load at the first point of entry such as the mouth and nasal region to prevent the spread of the virus . Recent studies have shown that most of the commercially available mouth rinses eliminate the SARS-CoV-2 virus in the mouth cavity by disrupting the outer envelope of the virus mainly because of the presence of peroxides or alcohol but these molecules also disrupt the host cells in the mouth cavity 16 . Therefore, it would be ideal to prepare mouthwashes with active ingredients that directly target the virus replication. SARS-CoV-2 PLpro enzyme is a protease required to generate the NSPs essential for viral replication through its protease activity 29 . In addition, PLpro also exhibits de-ubiquitination and de-SIGylation in order to prevent the INF-α/β mediated anti-viral activity, thus may be responsible for pro-inflammatory cytokine storm in COVID-19 patients 5,30-32 . Therefore, PLpro serves as an excellent drug target not only to control the viral replication but also to prevent cytokine tsunami in COVID-19 patients. In this report, we have established that the aloin A and B inhibits the PLpro enzymatic activity. Aloin is an anthraquinone abundantly found in aloe vera plant. Anthraquinones including aloin A and B exhibit analgesic, antimicrobial [33][34][35] and antiviral properties [36][37][38] . Recent reports suggest that aloin B inhibits hepatitis B virus replication in an in vitro setting 39 . Additionally, aloin exhibits anti-influenza activity by inhibiting the neuraminidase enzymatic activity including the oseltamivir-resistant influenza strain 40 . Further, pre-clinical studies suggest that aloin also promotes host immunity by enhancing the hemagglutinin specific T cells during PR8H1N1influenza infection in mice 40 . To the best of our knowledge, this is the first study to report that both aloin A and B specifically inhibit the SARS-CoV-2 PLpro proteolytic and DUB activity but not 3CLpro proteolytic activity. Both aloin A and B were able to inhibit more than 70% PLpro proteolytic activity. Structural analysis of aloin A and B with PLpro using computational studies revealed that both aloin isoforms form hydrogen bond with Tyr 268 of PLpro, which is critical for their proteolytic activity. In addition, the 100ns MD simulation fingerprint analysis predicted a strong interaction of aloin A and B with Glu 167 , which is required for deubiquitination activity of PLpro. The in vitro experiments confirmed that both aloin A and B inhibited viral deubiquitination activity of PLpro enzyme suggesting their potential benefit in preserving anti-viral immune response, as well as, preventing the replication of SARS-CoV-2 virus. Mechanistically, our computational modeling studies suggest that aloin A and B interacts with the Glu 167 , Tyr 268 and Glu 269 residue of PLpro, which are essential for proteolytic and DUB activity, however, these studies need to be further validated by crystallographic data. Since PLpro enzyme is known to suppress the IFN-α/β mediated anti-viral response though DUB activity and responsible for cytokine storm in COVID-19 patients, use of aloin isoforms as an anti-SARS-CoV-2 drug may promote the IFN-α/β mediated anti-viral response and limit the cytokine storm in COVID-19 patients. In addition, it has also been shown that 1 ml of the aloe vera juice has approximately 10 μg/mL of aloin, and no cytotoxic effects were observed in vitro at the concentration of 120 μM of aloin. Furthermore, the recommended concentration of aloin for human consumption is 11mM 41 . Taken together, these studies suggest anti-viral effects of aloin from aloe vera with suitable safety profile.
In conclusion, several active ingredients present in most of the commercially available mouth rinses were selected as targets to screen the enzymatic inhibitory effect of SARS-CoV-2 specific 3CLpro and PLpro enzymes. Among the 10 active ingredients tested, only aloin A and B inhibited both PLpro proteolytic and DUB activities but not 3CLpro, suggesting the specificity of aloin A and B towards PLpro. Taken together, our data suggest that aloin A and B might be potential drug candidates not only to inhibit the SARS-CoV-2 replication, but also to control the cytokine storm in COVID-19 patients. However, the therapeutic potential of aloin A and B to prevent the spread of SARS-CoV-2 infection needs to be further validated by viral challenge and clinical studies.

Materials and methods
Reagents and drugs. Molecular  In vitro enzymatic assay. SARS-CoV-2 specific 3CLpro enzymatic assay was carried out as we previously reported 4 and PLpro proteolytic assay was performed according to the manufacturer protocol. Briefly, 1ng/µl of 3CLpro and 0.4ng/µl of PLpro in 30µl of assay buffer was pre-incubated with the 10µl of 250µM compounds for  42 . In all, 20 binding conformations were collected for each aloin A and B. Two binding conformations for each aloin were retained for further analysis based on the highest binding affinity and orientation in the binding pocket.
MD simulation studies. MD simulations were used to test the stability of the inhibitor in presence of docked aloin A and B. The simulation cell and NAMD 2.14 43 input files were generated using MOE. The protein/ ligand complexes were embedded in a TIP3P water box with cubic periodic boundary conditions, keeping a distance of 10Å between the boundaries and the protein. The net charge of the protein was neutralized with 100mM NaCl. For energy minimization and MD simulations, the AMBER10:EHT force field was used and the electrostatic interactions were evaluated by the particle-mesh Ewald method. Each system was energy-minimized for 5000 steps using the Steepest Descent and Conjugate Gradient method. For equilibration the system was subjected to a 100ps simulation to gradually heating the system from 10 to 300K. Next, a 100ps NVT ensemble was generated at 300K followed by an NPT ensemble for 200 ps at 300K and 1 bar. Then, for each complex, a 100ns production trajectory was generated for further analysis. The trajectory analysis was done using scripts shared by the CCG support group.

Identification of ligand-binding mode. The protein-ligand interaction fingerprints (PLIF) descriptors
implemented in the MOE were used. Interactions are classified as hydrogen bonds, ionic interactions, and surface contacts according to the residues. The PLIF descriptors for all protein-bound aloin were generated with the default parameter set in MOE over the recorded MD trajectories.
Statistical analysis and reproducibility. Statistical analysis was carried out using one-way analysis of variance (ANOVA) with Bonferroni's Multiple Comparison test with 99.9% confidence intervals and represented as the mean ± SEM. Two-way ANOVA was carried out with Bonferroni's post-test to compare the grouped data. P values < 0.001 and P < 0.05 were considered statistically significant. Non-linear regression (curve fit) with four variable dose vs inhibition was performed to calculate the IC 50 values. GraphPad Prism (version 6.07; La Jolla, CA, USA) was used for statistical analysis. The experiments were performed a minimum of three times with triplicates for reproducibility. Authors performing the assay were blinded for the drugs being tested in the assay.

Data availability
Data information can be obtained from the corresponding author upon reasonable request.