The Effect of Famotidine on SARS-CoV-2 Proteases and Virus Replication

The lack of coronavirus-specific antiviral drugs has instigated multiple drug repurposing studies to redirect previously approved medicines for the treatment of SARS-CoV-2, the coronavirus behind the ongoing COVID-19 pandemic. A recent, large-scale, retrospective clinical study showed that famotidine, when administered at a high dose to hospitalized COVID-19 patients, reduced the rates of intubation and mortality. A separate, patient-reported study associated famotidine use with improvements in mild to moderate symptoms such as cough and shortness of breath. While a prospective, multi-center clinical study is ongoing, two parallel in silico studies have proposed one of the two SARS-CoV-2 proteases, 3CLpro or PLpro, as potential molecular targets of famotidine activity; however, this remains to be experimentally validated. In this report, we systematically analyzed the effect of famotidine on viral proteases and virus replication. Leveraging a series of biophysical and enzymatic assays, we show that famotidine neither binds with nor inhibits the functions of 3CLpro and PLpro. Similarly, no direct antiviral activity of famotidine was observed at concentrations of up to 200 μM, when tested against SARS-CoV-2 in two different cell lines, including a human cell line originating from lungs, a primary target of COVID-19. These results rule out famotidine as a direct-acting inhibitor of SARS-CoV-2 replication and warrant further investigation of its molecular mechanism of action in the context of COVID-19.


A B S T R A C T
The lack of coronavirus-specific antiviral drugs has instigated multiple drug repurposing studies to redirect previously approved medicines for the treatment of SARS-CoV-2, the coronavirus behind the ongoing COVID-19 pandemic. A recent, large-scale, retrospective clinical study showed that famotidine, when administered at a high dose to hospitalized COVID-19 patients, reduced the rates of intubation and mortality. A separate, patient-reported study associated famotidine use with improvements in mild to moderate symptoms such as cough and shortness of breath. While a prospective, multi-center clinical study is ongoing, two parallel in silico studies have proposed one of the two SARS-CoV-2 proteases, 3CL pro or PL pro , as potential molecular targets of famotidine activity; however, this remains to be experimentally validated. In this report, we systematically analyzed the effect of famotidine on viral proteases and virus replication. Leveraging a series of biophysical and enzymatic assays, we show that famotidine neither binds with nor inhibits the functions of 3CL pro and PL pro . Similarly, no direct antiviral activity of famotidine was observed at concentrations of up to 200 µM, when tested against SARS-CoV-2 in two different cell lines, including a human cell line originating from lungs, a primary target of COVID-19. These results rule out famotidine as a direct-acting inhibitor of SARS-CoV-2 replication and warrant further investigation of its molecular mechanism of action in the context of COVID-19.

I N T R O D U C T I O N
A large part of the current therapeutic discovery effort against the severe acute respiratory syndrome coronavirus 2 (SARS-CoV)-2 is focused on drug repurposing 1 . Of such agents, only remdesivir has thus far shown clinical evidence of antiviral effect 2 , while several others have not met their primary endpoints in various clinical studies 3,4 . Recently, famotidine has gained attention as a therapeutic option against SARS-CoV-2, initially based on anecdotal evidence of its positive effects in COVID-19 patients in China. Famotidine (PEPCID ® ), a histamine-2 receptor (H2R) antagonist, is an FDA approved drug for the treatment of gastroesophageal reflux disease (GERD) and gastric ulcers 5 .
Earlier reports of the beneficial effect of famotidine in China were recently supported by a retrospective clinical study involving 1620 patients in the U.S., which noted that hospitalized COVID-19 patients receiving a total median dose of 136 mg famotidine, in oral or IV formulation once daily, for 6 days had a reduced risk of death or intubation 6 . Another study involving 10 non-hospitalized patients linked the use of high-dose oral famotidine (240 mg per day for a median of 11 days) with patient-reported improvements in symptoms such as shortness of breath and cough 7 . These two reports conclude that the use of high-dose famotidine may be associated with improvements in both mild and severe symptoms of COVID-19. While a large, multi-center clinical trial to confirm these observations is in progress, the mechanism by which famotidine purportedly improves the clinical outcomes in COVID-19 patients is unknown. In silico modeling and molecular docking studies have separately suggested either of the two SARS-CoV-2 proteases as potential targets of famotidine activity 8,9 . In one computational study, Wu et. al. docked a library of approved drugs on to the available X-ray crystal structure of the 3chymotrypsin-like protease (3CL pro ) of SARS-CoV-2, identifying famotidine as one of the drugs likely to act on the protease 8 . Other computational reports have predicted famotidine as an inhibitor of the Papain-like protease (PL pro ), a second SARS-CoV-2 protease 9 . Together, these studies have raised the prospect of a direct antiviral effect of famotidine on SARS-CoV-2 replication. While both proteins are attractive targets for SARS-CoV-2 drug development 10-19 , there are at present no clinical-stage or approved drugs targeting either protein. The possibility of famotidine, an approved drug, acting on SARS-CoV-2 proteases is of significant clinical interest.
In this study, we performed an array of biochemical, biophysical, and antiviral experiments to test if famotidine is an effector of SARS-CoV-2 proteases and whether it inhibits virus replication in cultured cells.

M A T E R I A L S A N D M E T H OD S Compounds
Famotidine was acquired from Sigma Aldrich (Missouri, USA; cat. No. F6889). Compound 6, a previously reported inhibitor of SARS-CoV-2 PL pro function 20  Similarly, remdesivir. (cat. No HY-104077) an inhibitor of SARS-CoV-2 replication 2 was purchased from the same vendor. All compounds were dissolved in 100% DMSO at 100mM.

Cloning, expression, and protein purification
The complete sequences encoding 3CL pro and residues 746-1060 of PL pro (Wuhan-Hu-1 isolate, GenBank accession NC_045512) were cloned into a charge modified SUMO fusion expression vector, generated in-house. The fusion protein was expressed for 24 hours in Rosetta-2 (DE3) pLysS at 18°C in ZYP-5052 autoinducing media. Harvested cells were resuspended in 50 mM Hepes pH 7.5, containing 150 mM NaCl and lysed by sonication. The clarified supernatant was loaded onto a HiTrap HP SP column (Cytiva, Massachusetts, USA; cat no. 17115201) and the target fusion protein was captured in a cation-exchange chromatography step and eluted using a NaCl gradient. SUMO hydrolase was added to the pooled fractions to liberate the target protein and the sample dialyzed against 20 mM Tris, 10 % v/v glycerol, 5 mM DTT pH 7.0 overnight at 4°C. The protein was reloaded on the HiTrap HP SP column to remove the SUMO protein and hydrolase in a subtractive step. The flow-through, containing 3CL pro or PL pro was further purified by anion exchange chromatography using a HiTrap HP Q column (Cytiva; cat. no. 17115401) employing a NaCl gradient to elute the protein. Pooled fractions were further purified by size exclusion chromatography in 20 mM Tris pH 7.4, 150 mM NaCl and 5 mM DTT. The final protein was concentrated to 4 mg/mL for PL pro and 5 mg/mL for 3CL pro and flash frozen in aliquots.

R E S U L T S Famotidine is not an inhibitor of SARS-CoV-2 proteases
Processing of the SARS-CoV-2 polyprotein is critical to the generation of a functional virus replication complex 11,18,24 . To carry out this essential proteolytic function, the SARS-CoV-2 genome encodes two cysteine proteases, called PL pro and 3CL pro18 . Due to their critical roles in viral polyprotein processing and virus proliferation, both proteases are considered attractive targets for drug discovery 10,11,[13][14][15][16][17]20 . Since in silico docking studies have predicted these proteases as putative molecular targets of famotidine 6,8,9 , we methodically investigated the effect of famotidine on the catalytic functions of each protease.
First, we developed an in vitro activity assay of PL pro . PL pro is a protease domain found within the large multi-domain nsp3 protein encoded by SARS-CoV-2. While many coronaviruses encode two papain-like proteases, SARS-CoV, MERS-CoV and SARS-CoV-2 possess only one PL pro , which processes the amino-terminal end of the viral polyprotein liberating nsp1, nsp2 and nsp3 19,20 . Additionally, PL pro deubiquitinates host cell proteins by cleaving the consensus motif of LXGG 18,19 and is known to efficiently hydrolyze both diubiquitin and synthetic peptide substrates 19 . We leveraged the deubiquitinating property of PL pro to set up a functional activity assay using ubiquitin-AMC, a fluorogenic substrate cleavable by PL pro . Upon incubation with PL pro , the ubiquitin is recognized and cleaved at the C-terminus  Figure 1A).
We next tested whether famotidine can inhibit the enzymatic activity of 3CL pro , the second protease encoded by the SARS-CoV-2 genome. This protein, also referred to as the main protease (M pro ) or nsp5, cleaves the viral polyprotein at 11 unique sites 11 . This proteolytic activity generates multiple individual functional proteins required for the assembly of the SARS-CoV-2 replication/transcription complex, which drives viral genome replication 24 . Owing to its central role in the coronavirus life cycle, 3CL pro has received significant attention as a drug target resulting in the discovery of several potent inhibitors 10,14,15,17 . Native 3CL pro exists as a homodimer and requires dimerization for its proteolytic activity 11 . The catalytic mechanism of 3CL pro activity is typical of cysteine proteases, where the Cys-His catalytic dyad drives site-specific cleavage of substrates. We evaluated the enzymatic activity of 3CL pro using a FRETpeptide substrate that quenched fluorescence in its intact form, however, cleavage of the peptide substrate by 3CL pro produced fluorescence that could be measured at the excitation/emission wavelengths of 490/535 nm. The inclusion of ML188, a previously reported 3CL pro inhibitor served as a control, also aiding assay setup and optimization. Results of the FRET assay for various ML188 and famotidine concentrations are shown in Figure 1B.

Famotidine does not directly engage PL pro or 3CL pro of SARS-CoV-2
The function of many enzymes, such as proteases and kinases, can extend beyond their catalytic roles and includes a wide spectrum of non-catalytic activities such as allosteric regulation, scaffolding, protein-protein interactions, and protein-DNA interactions 25 . To rule out whether famotidine could bind away from the active site of the two viral proteases, and exert an effect through interference with non-proteolytic functions, we asked if famotidine is able to bind directly with either of the two SARS-CoV-2 proteases. For this, we employed two distinct biophysical techniques i.e. surface plasmon resonance (SPR) and differential scanning fluorimetry (DSF), that are routinely used to probe drug-protein engagement.
For our SPR studies, the biotinylated viral proteases were captured to a high density on In agreement with the SPR data, the control inhibitors produced a quantitative increase in observed T m (Figure 3). While compound 6, the known PL pro inhibitor, stabilized PL pro by a T m of 5.5 C (Figure 3A), and ML188, the 3CL pro inhibitor, produced a T m shift of 4.8 C ( Figure   3B), famotidine did not alter the T m of either of the two viral proteases. Taken together, the biophysical data decisively rules out the possibility of famotidine exerting its effect on PL pro or 3CL pro through interference with catalytic or non-catalytic protein functions as it is unable to bind with either of the two proteases.

Famotidine does not inhibit SARS-CoV-2 replication in cultured cells
Having  Figure 4B). Our results are consistent with previously reported studies in which remdesivir exerted a greater antiviral effect in human lung A549 cells than in Vero E6 cells 26 .
In-parallel cytotoxicity assays, carried out in both Vero E6 and A549 cells, showed that famotidine was not toxic up to the highest tested concentrations of 200 μ M (Figure 4 A and B).
Remdesivir, on the other hand, exhibited dose-dependent cytotoxicity at higher concentrations, well above its IC 50 . Together, these results show that famotidine does not inhibit SARS-CoV-2 replication in cultured cells and that its purported clinical benefit may be due to an alternative mechanism of action.

D I S C U S S I O N
Two in silico studies have separately predicted the 3CL pro or PL pro of SARS-CoV-2 as potential molecular targets of famotidine 8,9 , implying that famotidine associated improvement in COVID-19 patients may be due to a direct antiviral mechanism of action 6  It is therefore conceivable that famotidine-related benefit in managing respiratory symptoms may be due to an anti-inflammatory mechanism of action.
It is noteworthy that H2R, the established molecular target of famotidine, is involved in the activation of several mediators of the adaptive immune response, such as Th1 lymphocytes, which are implicated in pro-inflammatory cytokine production 34 . Histamine, the H2R ligand, also regulates bronchoconstriction, airway inflammation, and vasodilation 34 . Mast cells are a major source of histamine and their activation has been reported following viral infections of the respiratory tract [35][36][37] . Therefore, Mast cells may represent an underappreciated source of pro-inflammatory cytokine release in COVID-19 patients 35 . A better understanding of the role of the H2R pathway in COVID-19 will help elucidate the molecular details of how famotidine reduces the disease severity.
Our study redirects the mechanism behind the potential beneficial effect of famotidine, away from an antiviral effect to likely an anti-inflammatory action in COVID-19 patients. Given that there is an ongoing randomized clinical trial (NCT04370262), our results may assist the investigators in reshaping their interventional study to include inflammation-related outcomes.
Also, it should be noted that while famotidine is one of the relatively safer drugs, its use is not without risk [38][39][40] , especially in elderly patients (a high-risk population for COVID- 19), in which famotidine use has been associated with CNS complications 41 . Provided the ongoing clinical trial yields promising results, further investigation of famotidine and its safety profile in different age brackets will be needed before the drug can be used, most likely as part of a combination therapy, for COVID-19 disease management.

Conflicts: None.
Author contributions:          Quantitative real-time PCR was used to determine SARS-CoV-2 RNA with primers for the viral envelope (E) gene. Viral mRNA copies/mL of culture medium are shown when infected cells were grown in the presence of a range of famotidine (green) and remdesivir (blue) concentrations. While remdesivir exerts a dose-dependent effect on suppressing virus replication as seen by reduction in viral RNA copies, famotidine showed no effect in either Vero E6 or human lung A549 cells.