Famotidine increases cellular phospho-tyrosine levels

Abstract

While vaccines were being developed, the SARS-CoV-2 pandemic triggered a race to find known drugs that could be quickly repurposed to treat patients. One such candidate was famotidine, which retrospective cohort studies had shown increased survival in hospitalized patients. Computational studies had suggested that famotidine may target early viral proteases; however, ultimately, famotidine was shown not to function as a viral inhibitor. In contrast, we have observed a change in the cellular levels of phospho-tyrosine in A549 lung epithelial cells following treatment with famotidine. This quick change in phosphorylation was due mainly to a dose-dependent increase in cellular production of H₂O₂. Notably, these changes in phospho-tyrosine levels were able to affect cell signaling; we detected an increased short- and long-term response to IFNα stimulation. Our results suggest that famotidine can increase the anti-viral state of non-infected cells thereby potentially increasing viral resistance.

Introduction

In early 2020, the novel COVID-19 disease, caused by severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), started an initial rush to repurpose drugs for the emergency treatment of patients[1,2] while effective vaccines were approved and produced. One of the proposed drugs was famotidine, sold under the brand name PEPCID, which is a well-tolerated and safe over-the-counter drug used to treat peptic ulcers and gastroesophageal reflux. Famotidine acts as a Histamine 2 receptor (H2R) antagonist in gastric parietal cells and thereby reduces the secretion of acid into the stomach[3]. Histamine 2 receptors are widely expressed in a range of different organs, including stomach, brain, and lung [3,4]. Histamine receptors belong to the G-protein-coupled receptor (GPCR) family and can affect different pathways via downstream signaling; in particular, H2R induces activation of PKA, which can lead to changes in phosphorylation of downstream targets and changes in cellular calcium levels[3].

The US Department of Veterans Affairs, which provides healthcare to military veterans, conducted an observational study of hospitalized patients with COVID-19 infections. In this initial review of patient records, it was suggested that there was a survival benefit for those patients undergoing famotidine treatment at the time of infection[5]. In addition, several computational studies to examine the structure of viral proteases suggested that famotidine could be an inhibitor of the early viral 3-chymotrypsin-like protease (3CL^PRO/Main protease) and the Papain-Like protease (PL^Pro)[6,7] enzymes that governed processing of the initial polypeptide produced at the beginning of the viral replicative cycle[8]. These initial results, together with older studies suggesting that famotidine could inhibit Human Immunodeficiency Virus (HIV) replication[9], prompted the initiation of clinical trials to explore further the potential benefits of famotidine treatment for COVID-19 patients, either administered on hospital admission[10] or undergoing self-administration[10,11]. These initial results suggested that famotidine use reduced the risk of clinical deterioration and requirement for patient intubation in a hospital setting[10,11] and improved patient-reported outcomes in subjects that were not hospitalized[10,11]. Mechanistically, work from other groups indicated that famotidine had no effect on viral proteins or viral replication when compared to known inhibitors[12,13], suggesting that the observed effect was most likely due to an influence on the host cell or organ[12,13].

In response to a viral infection, host cells release interferons (IFN) to combat infection and prepare neighboring cells. Interferons induce the synthesis of hundreds of IFN-stimulated genes (ISGs), including ISG15, OAS1, and Mx1, which target different stages of the viral life cycle in order to protect the cell and inhibit viral replication[14,15]. Initial transmission of the IFN signal is relayed via phosphorylation of tyrosine residues in proteins through the JAK-STAT pathway[14,15], and is regulated by protein tyrosine phosphatases (PTPs), such as PTP1B and TCPTP among others[16]. Recent work from our lab had shown that use of PTP1B inhibitors could protect from acute lung injury, suggesting that modulation of PTPs might be a beneficial treatment option in cases of acute respiratory distress syndrome[17]. To examine further potential mechanisms by which famotidine may be exerting its observed effects in the context of COVID, we examined its potential role modifying cellular tyrosine phosphorylation and anti-viral signaling and have revealed new links between the drug and redox regulation of PTP function.

Materials and methods

Reagents and Antibodies:

All common reagents were obtained from ThermoFisher Scientific (Waltham, MA) or Sigma-Aldrich (St. Louis, MO). The ZRLRGGAMC Acetate peptide was from Bachem (#4027158.003, Torrence, CA). Famotidine and Ranitidine were from Sigma-Aldrich (PHR1055 and R101). Amthamine was from Abcam (ab120778, Cambridge, UK). Catalase was from Calbiochem (219261, San Diego, CA). Human IFNα was from Stemcell Technologies (78076, Cambridge, MA). The following antibodies were used in this study: Anti global phospho-tyrosine p-Y-1000 (#8954), pY701-Stat1 (#7649), Stat1 (#14994), pY690-Stat2 (#88410), Stat2 (#72604), Mx-1 (#37849), Irf-7 (#4920), IFIT1 (#14769), and OAS1 (#14498) were purchased from Cell Signaling Technology (Danvers, MA). ISG-15 was from Sigma-Aldrich (#HPA004627). Loading control GAPDH was from Abcam (ab125247). Antibodies were used at a 1/1000 dilution except GAPDH (1/10,000).

PL^pro and Cathepsin L activity assays:

Purified recombinant PL^pro (aa1564–1880, 125nM) purchased from SinoBiological (#40593-V07E, Beijing, China) was mixed with increasing concentrations of famotidine, ZnCl₂ and E-64 (0.05–100mM) in assay buffer (50 mM HEPES, pH 7.5, 5mM DTT, 2% DMSO, 0.1mg/ml BSA) at 37°C for 10min. The activity of PL^pro was determined by its ability to release amino methyl coumarin (AMC) from a Z-Arg-Leu-Arg-Gly-Gly-AMC peptide (25μM). The Cathepsin L activity assay kit was purchased from Abcam (#ab197012) and was performed following the manufacturer’s instructions using the same concentrations. PL^pro assay was measured at 360/460nm and the Cathespsin L assay was measured at 405/505nm in a Spectramax Gemini XPS microplate reader (Molecular Devices, San Jose, CA).

pTyr measurements and STAT signaling.

A549 and NCI-H1975 lung epithelial cells (ATCC, Manassas, VA) cultured in DMEM containing 10% FBS and 1% Penicillin/Streptomycin solution. Cells (A549: 100,000 cells/well, NCI_H1975: 65,000 cells/well) were serum starved overnight and then treated with famotidine (25 to 300 μM), ranitidine (50 to 300μM) for 1 h or amthamine (50 to 300μM) for 16 h. To test the effect of suppressing H₂O₂ levels on the effect of famotidine, cells were treated with N-acetyl cysteine (NAC). NAC was added before famotidine treatment for 30min. To measure STAT1/2 activation, A549 cells were plated as before and serum starved overnight then treated with famotidine for 15 min before adding hIFNα (25ng/ml) for 5, 15 and 30mins. To determine ISG levels, cells were plated and serum starved as previously, then treated with famotidine 100μM for 15min before adding hIFNα (10ng/ml) for 30mins. Afterwards, media was replaced with fresh starvation media and cells were collected at 4, 16, 24 and 48hrs. Cells were lysed in RIPA buffer (25mM HEPES pH 7.5, 150mM NaCl, 0.25% Deoxychloate, 10% Glycerol, 25mM NaF, 10mM MgCl2, 1mM EDTA, 1% Triton X-100, 0.5mM PMSF, 10mM Benzamidine, cOmplete protease inhibitor cocktail (Roche)). Soluble proteins were harvested by centrifugation and quantitated using the BCA method. Total proteins in the cell lysates were separated by SDS-PAGE and the specific proteins were monitored with their corresponding antibodies. GAPDH was used as a loading control.

Amplex UltraRed H₂O₂ detection:

To detect ROS, we used Amplex UltraRed reagent (A36006, ThermoFisher). Cells (22,000 cells/well) were plated in black 96 well plates (Corning, Corning, NY) and serum starved overnight and then treated with 20mM NAC for 30mins, after which the media was replaced with media containing either famotidine (25 to 300 μM) or vehicle along with the Amplex UltraRed reagent (100μM) and Horseradish peroxidase (10U/ml, Sigma-Aldrich) for 1 h. Fluorescence was measured at 530/590nm in a Spectramax Gemini XPS microplate reader (Molecular Devices).

Human IFNα ELISA:

The media at 16, 24, and 48 h from the cells used in the ISG expression experiments was collected and tested using the hIFNalpha ELISA (BMS216, ThermoFisher), following the manufacturer’s instructions.

Statistics:

All data are presented as mean ± SEM and represent a minimum of three replicates. Significance was determined by two-tailed Student’s T-test. *p<0.05, **p<0.01, ***p<0.001.

Results

Famotidine did not inhibit proteases implicated in SARS infection

The SARS CoV-2 PL^pro papain-like protease plays an essential role in the cleavage and maturation of viral proteins and, therefore, was considered as a potential target for anti-viral agents[6]. To test whether famotidine could inhibit SARS CoV-2 PL^pro protease, we used a fluorescence-based in vitro assay incorporating purified enzyme exposed to different concentrations of famotidine. We used an aminomethylcoumarin fluorophore bound to an Arg-Leu-Arg-Gly-Gly peptide as substrate, to test the deISGylation activity of PL^pro and test whether this activity was affected by famotidine. We were unable to detect any effect of various concentrations of famotidine on the enzyme, whereas, in parallel, there was inhibition of protease activity by zinc, starting at a concentration of 5μM (Fig. 1A); this agrees with results of other groups[12,13]. In addition, we considered possible targets for famotidine in the viral entry cycle. Cathepsins are also cysteine-based proteases that are required for release of the viral genome from the early endosomes into the cytoplasm[8,18]. Due to the structural similarities between the cathepsin cysteine-protease inhibitor E-64 and famotidine, we tested whether famotidine might inhibit Cathepsin L and thus inhibit the SARS-CoV-2 life cycle; however, we were unable to detect inhibition of Cathepsin L when incubated with famotidine, until concentrations in the higher millimolar range were tested (Fig. 1B). As expected, both E-64 and the cathepsin inhibitor zFF-FMK displayed strong inhibition, whereas zinc only inhibited at the highest dose (Fig. 1B). These initial results suggest that famotidine does not affect the virus itself or host enzymes required for viral entry, so we turned our attention to the cell response to virus infection.

Fig. 1. Famotidine does not inhibit viral proteases.

(A) %relative inhibition of PL^pro was tested at different concentrations (0.05–100mM) of famotidine, ZnCl₂, and E-64 using an AMC-fluorophore bound to an Arg-Leu-Arg-Gly-Gly peptide. (B) The %relative inhibition of Cathepsin L at different concentrations (0.05–100mM) of famotidine, ZnCl₂, E64 and FF-FMK was tested using a commercial assay. Small error bars not shown due to graph scale.

Famotidine alters phospho-tyrosine levels

Recent work by our lab demonstrated that inhibition of the protein tyrosine phosphatase PTP1B could protect mice from acute lung injury by regulating CXCR4 signaling in neutrophils[17]. Additionally, tyrosine phosphorylation is an important controlling element in many different signaling pathways, including anti-viral interferon signaling[15]. To test for a possible effect on cellular phospho-tyrosine (pTyr) levels, we treated A549 lung epithelial cells with different concentrations of famotidine (0-300μM) and probed for total pTyr. The concentrations of famotidine used were consistent wth those achieved in the serum of patients following oral administration of the drug[19]. A549 cells displayed a dose-dependent increase in cellular pTyr levels after one hour of treatment with famotidine (Fig. 2A and B). A similar, but less pronounced, effect was obtained after treatment with a different H2R antagonist, Ranitidine[3] (Fig. 2A and C). In addition, we observed similar results in the NCI-H1975 non-small cell lung cancer cell line (Fig. 2D–E) Since famotidine is an H2R antagonist, we treated A549 cells with the H2R specific agonist amthamine and observed a reduction in cellular pTyr levels following overnight treatment (Fig. 2A and 2F).

Fig. 2. Famotidine increases cellular pTyr levels.

(A) Structures of Famotidine, Ranitidine and Amthamine. (B) A549 cells were treated with the indicated concentrations of famotidine for 1 hour and the lysates were probed for cellular pTyr levels using Western blot. (C) Cells were treated with the indicated concentrations of ranitidine for 1 hour and lysates were probed for cellular pTyr levels using Western blot. (D-E) Equivalent experiments to 2B-C performed in NCI-H1975 cells. (F) A549 cells were treated with the indicated concentrations of amthamine for 16 hours and the lysates were probed for cellular pTyr levels using Western blot. Band quantifications shown to the right.

Changes in phosphorylation usually occur rapidly after ligand-induced receptor activation[20]. One mechanism for fine-tuning protein tyrosine phosphorylation and transmission of a signal is by localized release of a small burst of reactive oxygen species (ROS), which can reversibly oxidize and transiently inhibit PTP function[21]. To examine whether famotidine could trigger a rapid increase in pTyr levels, we used the AmplexRed reagent[22] to measure changes in H₂O₂ release. One hour of famotidine treatment increased ROS levels in a dose-dependent manner (Fig. 3A); however, at higher doses (300μM) the effect was less pronounced and may be due to a slight reduction in cell viability (Suppl. Fig 1). Additionally, pretreating cells with the antioxidant N-acetyl cysteine (NAC) attenuated the increase in pTyr but not to the extent of the control group (Fig. 3B). These results suggest that famotidine can increase cellular pTyr levels by activating ROS production in treated cells.

Fig. 3. Famotidine increases ROS production.

(A) A549 cells were treated with famotidine at the indicated concentrations in the presence of AmplexRed reagent for 1 hour. (B) Cells were pretreated with N-Acetylcysteine (NAC) for 20min and then treated with famotidine (100μM) for 1 hour. Lysates were probed for cellular pTyr levels by Western blot. Band quantifications shown to the right. ns=p>0.05.

Famotidine augments Interferon signaling

Interferon signaling is a cellular response mechanism to viral infection. Infected and immune cells release IFNs, which can prepare neighboring cells to resist the virus by expressing antiviral proteins or interferon stimulated genes (ISGs)[14,15]. Canonical IFN signaling is based on the JAK-STAT axis, which is triggered and controlled in part by changes in tyrosine phosphorylation. Based on our previous results, we tested whether cells treated with famotidine displayed increased STAT phosphorylation when stimulated with IFN alpha. Pretreatment of cells briefly with famotidine, promoted the response to IFNα stimulation, leading to higher STAT1 phosphorylation across both early and late time-points; in contrast, enhanced phosphorylation of STAT2 was detected only at later time points, up to 1 h (Fig. 4A, B).

Fig. 4. Famotidine promotes IFN signaling.

(A) A549 cells were pre-treated with famotidine (50μM, 15min) and then stimulated with IFNα for the indicated timepoints after which cells were lysed and the pTyr status of STAT1/2 was probed by Western blot. (B) Cells were pre-treated with famotidine (100μM, 15min) and then stimulated with IFNα for 30mins (0.5hrs), after which, the media was replaced with fresh media and the 1- and 2-hour timepoints were collected. As in A, lysates were collected and probed for pSTAT1/2 via Western blot. (C) Similar to B, A549 cells were cultured for the indicated longer timepoints then lysate was probed for various ISGs via Western blot. Band quantification on the right. (D) The media of the 0-, 16-, 24-, and 48-hour timepoints from C was collected before lysing and IFNα levels were quantified via a commercial ELISA kit. Band quantifications shown to the right.

Long-term IFN/STAT signaling leads to changes in protein transcription, increasing anti-viral and immune signaling proteins. To test whether increased signaling induced by famotidine treatment led to increased IFN gene transcription, we probed cell lysates for differences in several well-known anti-viral ISGs[14,23]. Following famotidine treatment, levels of the ubiquitin-like modifier protein ISG15 were dramatically decreased; this effect was transient and returned to the control state approximately 48 hours after treatment (Fig. 4C). Interestingly, different proteins followed distinct time-courses of expression following famotidine treatment. We were unable to detect a difference in IFIT1 levels and cells displayed a trend towards modest increases in the levels of MX-1, OAS1 and IRF7, a result that was also observed by others[13], although this did not reach statistical significance (Fig. 4C). Media collected from these experiments was also probed for IFNα levels by ELISA (Fig. 4D). We observed that famotidine-treated cells displayed a modest increase in IFNα levels in the media when compared to control group. Altogether, these results suggest that famotidine can indirectly affect IFN signaling by increasing ROS levels in A549 cells.

Discussion

Early in the COVID-19 pandemic, before vaccines were readily available, it was proposed, based on preliminary clinical trials, that famotidine, an over-the-counter treatment for acid reflux and stomach ulcers, could attenuate the deterioration of virus-infected patients. These early studies focused on the potential for famotidine to bind to the active site of the early viral proteases 3CL^PRO and PL^PRO[6,10,11] and attenuate the effects of viral infection leading to an increased survival of hospitalized patients[5,10]. Subsequently, several groups reported that famotidine had neither an inhibitory effect on viral enzymes or on the viral life cycle[12,13], however, the mechanisms underlying its protective effects remained unclear.

Here, we report that famotidine treatment increased global tyrosine phosphorylation in non-infected A549 lung carcinoma cells. This was achieved by a dose-dependent induction of ROS in cells, which would result in transient inhibition of protein tyrosine phosphatase function[20,21]. Considering the crucial role played by PTPs in the regulation of cell signaling[16,20], this would be expected to alter pathways that are dependent on protein tyrosine phosphorylation, such as interferon signaling, which is transmitted mainly through the JAK-STAT axis[15]. We were able to detect increased phosphorylation of STAT1, which led to increased downstream transcription and would contribute to a heightened anti-viral state (Fig. 4). Older studies had demonstrated a link between famotidine and viral inhibition[9], suggesting that famotidine could be useful in fighting different viruses. Carrau and colleagues discovered that SARS-CoV-2 infection is initially localized to the respiratory tract and that this localized infection was able to generate an antiviral response in other organs due to increased circulation of interferons[24]. Any variant that could evade the initial localized response was able to induce viremia and viral tropism[24]. Our results suggest that famotidine might be able to help patients in two different ways. On the one hand, famotidine increases IFN signaling by facilitating STAT phosphorylation, which leads to an increased anti-viral state in non-infected cells; on the other hand, the heightened anti-viral state might help with preparing a localized immune response and ultimately preparing the rest of the body.

The Dikic group studied the effect of famotidine on infected cells and found that it reduced the histamine-induced expression of TLR3, leading to a lower inflammatory response and a reduced potential for a cytokine release syndrome[13]. They focused on longer term transcription changes caused by extended famotidine treatment in cells undergoing infection. Although they saw differences in transcription, it is still unclear how famotidine was able to affect the different transcriptional profiles. Importantly, they also saw a reduction in ISG15 mRNA and protein after famotidine treatment, consistent with our data (Fig. 4C). ISG15 is up-regulated by IFN signaling[14,15] and is considered an important modulator of the IFN response. Research in children with ISG15 mutations has shown that they are highly resistant to viral infections and have higher levels of phospho-STATs and circulating IFNs[25]. ISG15 exerts negative feedback on IFN signaling through USP18 by ubiquitinating and degrading STATs, thus attenuating signaling[26]. Although the acute, transient changes in ROS-induced tyrosine phosphorylation elicited by famotidine may culminate in reduced levels of ISG15 at later time points (16-24 h), in turn promoting an anti-viral response, the differential expression of ISGs (fig. 4C) and TLRs[13] suggests that famotidine acts also through additional pathways that have yet to be determined.

Although we did not investigate the source of ROS, a potential candidate is DUOX1, a NADPH oxidase that is highly expressed in epithelial tissues and uses Ca²⁺ for optimal activity[27]. We observed that the addition of Ca²⁺ to the medium increased pTyr levels, whereas BAPTA, a Ca²⁺ chelator, decreased pTyr levels (data not shown). Interestingly, DUOX1 is associated with an epithelial defense role, as the H₂O₂ it produces can fight microorganisms in the lung lumen[28] and is also used to create microbicidal oxidants in the gastrointestinal tract[29]. Additionally, H2R acts mainly through the Gs subtype GPCR which leads to calcium sequestration[3]. Famotidine acting as an inverse agonist could dampen H2Rs ability to sequester calcium making more calcium available for DUOX1 to activate. This opens the possibility that famotidine could also protect the lungs from infection by increasing H₂O₂ production in the lung epithelium, degrading viral particles and reducing infective potential.

In summary, our results shed light on one potential mechanism by which famotidine may ameliorate symptoms of viral infection, such as COVID-19. We found that in cells treated with famotidine, increased ROS production triggered acute changes in protein tyrosine phosphorylation, which enhanced IFN signaling to promote heightened anti-viral state. Further studies will be required to define the full complement of the effects of ROS oxidation on other sensitive proteins.

Highlights.

• Famotidine does not inhibit viral proteases.

• Famotidine increases phospho-tyrosine levels via ROS production and inhibition of protein tyrosine phosphatases.

• Increased reactive oxygen species production can augment Interferon signaling leading to an enhanced anti-viral state.