D,L-Lysine-Acetylsalicylate + Glycine (LASAG) Reduces SARS-CoV-2 Replication and Shows an Additive Effect with Remdesivir

The severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) causing the coronavirus disease-19 (COVID-19) is still challenging healthcare systems and societies worldwide. While vaccines are available, therapeutic strategies are developing and need to be adapted to each patient. Many clinical approaches focus on the repurposing of approved therapeutics against other diseases. However, the efficacy of these compounds on viral infection or even harmful secondary effects in the context of SARS-CoV-2 infection are sparsely investigated. Similarly, adverse effects of commonly used therapeutics against lifestyle diseases have not been studied in detail. Using mono cell culture systems and a more complex chip model, we investigated the effects of the acetylsalicylic acid (ASA) salt D,L-lysine-acetylsalicylate + glycine (LASAG) on SARS-CoV-2 infection in vitro. ASA is commonly known as Aspirin® and is one of the most frequently used medications worldwide. Our data indicate an inhibitory effect of LASAG on SARS-CoV-2 replication and SARS-CoV-2-induced expression of pro-inflammatory cytokines and coagulation factors. Remarkably, our data point to an additive effect of the combination of LASAG and the antiviral acting drug remdesivir on SARS-CoV-2 replication in vitro.


Introduction
The severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), that emerged in late 2019, is the pathogen causing COVID-19 [1]. The virus has rapidly spread around the world, leading to a pandemic with more than 500 million cases worldwide and more than 6 million deaths (updated information from the World Health Organization, as of 9 June 2022). In the meantime, frequent mutations have resulted in multiple variants of SARS-CoV-2, whose specific sequences cause changes in the viral properties [2].
Although infection is mild or asymptomatic in many cases, SARS-CoV-2 can cause severe courses including life-threatening pneumonias with a high mortality rate [1]. In particular, older patients with comorbidities, such as lung pathologies, cardiovascular diseases, diabetes predisposition and obesity often develop severe illnesses and complications [3][4][5]. The early stages of SARS-CoV-2 infection are characterized by a high virus load and milder symptoms-most commonly fever and dry cough [6]. With the progression of well as thrombotic factors. When LASAG was used in combination with remdesivir, an additive effect of the antiviral activity of both substances could be demonstrated in vitro.

LASAG Inhibits SARS-CoV-2 Replication at Non-Toxic Concentrations In Vitro
The antiviral potential of LASAG was already demonstrated on human coronaviruses and MERS coronavirus in vitro [34].
Here, the effective concentration of 50% (EC 50 ) of LASAG was determined by standard plaque assay 24 h post infection (p.i.) with a SARS-CoV-2 Alpha variant (isolate 5159) and a Delta variant (isolate 4749). LASAG treatment resulted in an efficient inhibition of the virus replication at millimolar concentrations (Alpha variant: EC 50 of 6.7 mM; Delta variant: EC 50 of 5.4 mM) (Figure 1a,b). To exclude the notion that the described effects of LASAG on the viral replication are due to impaired cell viability, cell propagation was analyzed. For that, Calu-3 and Vero-76 cells were counted 24 h after treatment with LASAG (Figure 1c,d). The results clearly show that the used concentrations of LASAG do not affect cellular proliferation and viability at all in Calu-3 cells and only slightly in Vero-76 cells. Shown are the means (±SD) of viable cells of three independent experiments including two biological samples. Statistical significance was analyzed by ordinary one-way ANOVA followed by Dunnett's multiple comparisons test (*, p < 0.0332; ****, p < 0.0001).

In the Presence of LASAG, the Replication of Different SARS-CoV-2 Variants Is Reduced In Vitro
To further elucidate the effect of LASAG on the replication of three different SARS- CoV-2  Alpha variants and one Delta variant, we used the human lung carcinoma cell line Calu-3 and  primate kidney Vero-76   In the presence of LASAG, SARS-CoV-2 replication of different variants is significantly reduced in vitro. Calu-3 (a-d,f) and Vero-76 (e) cells were left untreated or were incubated with the indicated concentrations of LASAG for 1 h. Subsequently, cells were left uninfected (mock) or were infected with the SARS-CoV-2 variants (Alpha variants: isolate 5159 (a,e,f), isolate 5587 (b), isolate 5588 (c); Delta variant: isolate 4749 (d)) at 0.5 MOI in the absence and presence of LASAG for 2 h. After a medium change, cells were further incubated in the absence and presence of LASAG. At 24 h p.i., the analyses were performed. (a-e; left panel) Progeny virus titers were determined in the supernatant by standard plaque assay. Shown are the means (±SD) of plaque forming units (pfu) mL −1 of three independent experiments including two biological samples. (a-e; middle panel) Total cell lysates were used to investigate the viral spike protein expression by Western blot analysis. ERK2 served as the loading control. Shown are representative example blots of three independent experiments. (a-e; right panel) The spike protein signals were analyzed with ImageJ and normalized to the ERK2 signals. Shown are the means (±SD) of three independent experiments. Untreated, infected samples were arbitrarily set to 100%. (f) In immunofluorescence microscopy, SARS-CoV-2 spike protein was detected by a spike-specific antibody and an Alexa Fluor 488-conjugated goat anti-mouse IgG antibody (green). The nuclei were stained with Hoechst 33342 (blue). Immunofluorescence microscopy was acquired by use of the Axio Observer.Z1 instrument (Zeiss) with a ×200 magnification. Scale bars represent 100 µm. Shown are the representative example pictures of three independent experiments. (a-e) Statistical significance was analyzed by ordinary one-way ANOVA followed by Dunnett's multiple comparisons test (ns, non-significant; *, p < 0.0332; **, p < 0.0021; ***, p < 0.0002; ****, p < 0.0001).
Thus, the treatment of SARS-CoV-2-infected epithelial cells with LASAG at non-toxic concentrations efficiently reduced viral replication and propagation.

LASAG Reduces SARS-CoV-2-Induced Cytokine and Chemokine Production In Vitro
The cytokine storm that is induced in severe cases of COVID-19 is known to be a key cause of morbidity in SARS-CoV-2 infection [40,41]. One major goal of this work was to investigate adverse side effects of LASAG during SARS-CoV-2 infection, especially since LASAG is commonly used to combat inflammation [42]. Initially, the effect of LASAG on viral mRNA expression was determined by qRT-PCR analyses 24 h p.i. (Figure 3a). Consistent with the data on viral titers and viral protein expression (Figure 2a-d), a significantly reduced mRNA synthesis of SARS-CoV-2 (N1) was observed in Calu-3 cells treated with 10 mM LASAG. Concomitantly, the SARS-CoV-2-mediated interferon-gamma-induced protein 10 kDa (IP-10), the interferons (IFNs) IFNβ, IFNλ1 and IFNλ2/3, the tumor necrosis factor-related apoptosis-inducing ligand (TRAIL) and the 2 -5'-oligoadenylate synthetase 1 (OAS1) mRNA synthesis were significantly reduced upon LASAG treatment ( Figure 3a) probably due to reduced viral infection.

LASAG Reduces Cytokine and Chemokine Production of SARS-CoV-2-Infected Epithelial Cells within the Chip Model
To investigate the effects of LASAG on SARS-CoV-2 infection in a more complex cell culture model, the previously described chip model [44] containing epithelial Calu-3 cells, endothelial human umbilical vein endothelial cells (HUVEC) and integrated peripheral blood mononuclear cells (PBMC) was used. Analogue to the experiments described above, both cellular layers were pre-treated with LASAG for 1 h. Furthermore, LASAG was added during and after the infection.   Since the endothelial tissue plays a pivotal role in hemostasis, we focused on the effect of LASAG on coagulation factors. Although the endothelial layer was not infected in the chip model, a slight activation of IL-6, IL-8 and tissue factor (TF) (Figure 4c) was demonstrated, probably due to virus-induced factors of the epithelial layer [44]. Of note, TF and tPA were significantly reduced in infected, LASAG-treated endothelial cells (Figure 4c). Other cytokines and thrombotic factors (IL-6, IL-8 and PAI-1) were slightly but not significantly downregulated in the presence of LASAG. All these factors are known to be aberrantly changed during SARS-CoV-2 infection.
It has to be mentioned that LASAG treatment of the endothelial layer in the chip model showed slight disturbance of the membrane function independent of SARS-CoV-2 infection (Supplementary Figure S3a), which was not observed in endothelial cells cultured in cell culture plates (Supplementary Figure S3b).  The highest concentrations of both substances used in combination show no effect on cell growth, which was determined by cell counting after 24 h of incubation. Statistical significance was analyzed by ordinary one-way ANOVA followed by Dunnett's multiple comparisons test (*, p < 0.0332).

Discussion
ASA is one of the most commonly used drugs worldwide. It is used for many indications, mainly due to its impact against inflammation and pain of a general nature. In addition to these analgesic and anti-inflammatory properties, ASA has anti-platelet, anti-thrombotic and neuroprotective effects [46]. However, adverse effects of ASA (e.g., bleeding, Reye's Syndrome in childcare) have to be taken into account, especially when high doses are administered [47,48].
In particular, several studies on the administration of ASA in COVID-19 patients show a variable picture on the outcome [38,39,[48][49][50][51][52][53][54]. On the one hand, an improvement in mechanical ventilation and in-hospital deaths [38], and on the other hand, no effect of ASA on mortality [39,48,50,51,53,54] were reported. While ASA administration was associated with a decreased risk of thrombosis, an increased risk of bleeding occurred [48]. Although some patients who took ASA prior to hospital admission had less need for upgrading ventilatory support [39], other patients had an increased risk of death using ASA before COVID-19 [52].
Based on the fact that severe courses of COVID-19 are often associated with complications caused by inflammatory processes and blood clotting and that LASAG is used for many indications, the present study was intended to investigate beneficial and adverse effects of LASAG during SARS-CoV-2 infections in vitro.
Here, we were able to show that the replication of SARS-CoV-2 in vitro is reduced in the presence of 10 mM LASAG. Similar observations have been made previously for the coronaviruses, including MERS coronavirus but also IAV in vitro [34,36]. These studies indicated that the reduction in viral replication relied on the inhibition of virus-supportive functions of NFκB-mediated signaling during infection. Especially in the case of IAV infection, it was demonstrated that IAV-induced detrimental hyperinflammation was reduced upon NFκB inhibition [55]. Likewise, NFκB is implicated in the development of acute respiratory distress syndrome in COVID-19 patients due to the hyper-activation of the immune system [56][57][58]. Within the present study, a reduction in virus-induced mRNA synthesis of IP-10, IFNβ, IFNλ1, IFNλ2/3 and TRAIL of Calu-3 mono cell culture was observed, which was verified on the level of IP-10, IFNβ, IFNλ1 and IFNλ2/3 protein secretion. In the more complex chip model, a LASAG-mediated reduction in IP-10 and IFNλ1 protein secretion was observed. The SARS-CoV-2-induced IL-6 protein secretion was not affected in the presence of LASAG.
The difference in the LASAG-mediated effect on SARS-CoV-2-induced cytokine and chemokine expression might be due to the close proximity of epithelial cells, endothelial cells and PBMCs. Thus, it might be that the interplay of these cells provokes differences in the cytokine expression. Furthermore, the prolonged growth of cells in this model system might be a reason for lower infection and replication compared to monocultured cells and probably leads to changes in infection-related effects.
The different effects of LASAG in vitro and in vivo are probably due to the difference in concentration. Millimolar amounts of LASAG, which were used in in vitro experiments to study coronavirus but also IAV infection [34,36], cannot be achieved in vivo. More than 3 g/L would be required to reach a concentration of 10 mM in the blood, but this is toxic [59]. After intravenous administration of 500 mg ASA i.v. the highest plasma concentration is 54.25 mg/L and after oral administration 4.84 mg/L [34,60]. Indeed, we encountered some toxicity of LASAG on the endothelial layer of the chip model, which was not observed in mono cell culture. Nonetheless, the application directly to the lung via inhalation may be considered, since a study on the treatment of asthma with LASAG showed that higher concentrations can be achieved this way without causing toxicity [61]. Further investigations are required to determine if patients might benefit from the anti-viral and anti-inflammatory effects of LASAG.
In addition, it might be considered to increase the antiviral effect of LASAG by combinatory treatment with antiviral substances, such as remdesivir. Combination therapies are standard for the treatment of viral infections due to a better efficacy, decreased toxicity and the prevention of resistance emergence compared to single antivirals [45]. With the help of the webtool Syngeryfinder 2.0, several pharmacological studies were carried out to find useful drug combinations that could be repurposed to treat COVID-19 [62][63][64]. Our data clearly indicate an additive effect on SARS-CoV-2 inhibition by combining LASAG with remdesivir in vitro.
In severe COVID-19 cases, a dysfunction of the lungs and other organs is linked to vascular injuries and thrombosis and several blood parameters are activated during this process [65,66]. Among those, TF is the primary initiator of blood coagulation and is activated after vessel injury. Aberrant TF expression can induce intravascular thrombosis [67]. Interestingly, TF was shown to be activated during infections with Herpes simplex virus, HIV and Ebola as well as SARS-CoV-2 infection [68][69][70][71]. In primates, ASA-mediated TF inhibition did not result in increased bleeding [72], and in pigs, TF inhibition with recombinant TF-pathway inhibitor and ASA prevented acute thrombosis by decreasing thrombus areas without further bleeding complications [73]. Clinical studies showed that TF pro-coagulant activity declined after combined treatment with a coagulation inhibitor (clopidogrel) and ASA [74] and that a combined treatment with ASA and other antiplatelet-acting drugs reduced TF levels and thrombin generation in patients with angina pectoris [75]. Similarly, TF inhibition has been discussed as a potential therapeutic strategy in COVID-19 patients in several studies [76][77][78]. Our data showed a significant increase in TF expression in the endothelial layer of the chip model after SARS-CoV-2 infection of the epithelial layer. Since the endothelial layer was not infected, these results indicated an indirect deregulation of hemostasis probably due to SARS-CoV-2-induced cell responses.
The two coagulation factors tPA and PAI-1 are part of the fibrinolytic system and-in balance-dissolve blood clots and prevent blood clotting [79]. Elevated tPA and PAI-1 levels were found in patients hospitalized with COVID-19 and these high levels were associated with a worse respiratory status [80]. In our studies, we showed a strong induction of PAI-1 after SARS-CoV-2 infection in mono cell culture followed by a LASAG-dependent significant reduction. The tPA was not affected in mono cell culture studies. In the chip model, both tPA and PAI-1 were not induced but the significant reduction in tPA of the LASAG-treated endothelial cells correlates with very early studies on ASA which show an inhibition of tPA activity by ASA [81].
Taken together, numerous beneficial mechanisms of the action of ASA have been reported during COVID-19, although the negative side effects cannot be ignored. Our own in vitro data show a LASAG-mediated reduction in viral titers, SARS-CoV-2-induced cytokines, chemokines and coagulation factors, when used at high concentrations. Admittedly, this treatment also resulted in the disturbance of the endothelial layer in the chip model, which was independent of the SARS-CoV-2 infection, and maybe due to limited growth areas or medium exchange. However, LASAG-mediated cell damage was neither observed in mono cell culture for the endothelial cells nor for epithelial cells. Nevertheless, such millimolar concentrations of LASAG cannot be achieved in vivo, but were always used in cell culture experiments to demonstrate the antiviral and anti-inflammatory effects of ASA and LASAG during viral infections [34,36].
Since aerosolized LASAG was already used for the treatment of influenza in hospitalized patients [31] and based on the fact that LASAG was able to reduce SARS-CoV-2 load, SARS-CoV-2-induced pro-inflammatory cytokines and thrombotic factors, this substance seems to be promising for future studies against COVID-19. In particular, combinations with antiviral substances, such as remdesivir that could increase the effect and reduce toxicity, might be considered. For determining the cytotoxicity of drugs, Calu-3 and Vero-76 cells were seeded in 24-well plates and treated with different concentrations for 24 h at 37 • C. After that, the viable cell number was counted on a Countess TM II (Invitrogen, Dreieich, Germany).

Cell Culture, Cytotoxicity and Virus Infection
Mono cell culture infection was performed as follows: cells were washed with phosphate-buffered saline (PBS, Biozym, Hessisch Oldendorf, Germany) and pre-treated for 1 h with LASAG, remdesivir or solvent controls at the indicated concentrations in DMEM/FCS. The cells were either left uninfected or were infected with SARS-CoV-2 at the indicated MOI for 2 h in DMEM/FCS supplemented with the indicated pharmaceutical substances and concentrations. After infection, the medium was removed, cells were washed with PBS and further cultivated in fresh DMEM/FCS supplemented with the substances at 37 • C and 5% CO 2 .
The BioChips (BC-002) were purchased commercially from the Dynamic42 GmbH (Jena, Germany). The BioChips were equipped with epithelial and endothelial cells as well as PBMCs as described previously [44,82]. For use in the chip model, the Calu-3 cells were cultured in RPMI 1640 (Sigma-Aldrich, Taufkirchen, Germany) supplemented with 10% FCS and the HUVECs in endothelial cell growth medium (EC, Promocell, Heidelberg, Germany) including supplement mix (Promocell, Heidelberg, Germany). HUVECs were isolated from anonymously acquired human umbilical cords according to the Declaration of Helsinki, "Ethical Principles for Medical Research Involving Human Subjects" (1964). LASAG was solved in RPMI1640 to obtain a 200 mM stock solution. Furthermore, LASAG was diluted in RPMI1640/FCS or EC/supplements for incubation in the epithelial chamber or the endothelial chamber, respectively. For the infection of the chip model, the membranes were washed with PBS and pre-treated for 1 h with 10 mM LASAG in both chambers. The epithelial chamber was left uninfected or infected with SARS-CoV-2 (MOI 1) in RPMI1640 supplemented with 0.2% autologous human serum, 1 mM MgCl 2 , 0.9 mM CaCl 2 . After 3 h incubation, the membranes were washed and supplemented with fresh medium containing LASAG at 37 • C and 5% CO 2 . For plaque assay, Vero-76 cells were seeded in 6-well plates and infected with serial dilutions of the supernatants in PBS supplemented with 1 mM MgCl 2 , 0.9 mM CaCl 2 , 0.2% BSA and 100 U mL −1 Pen/Strep (Sigma-Aldrich, Taufkirchen, Germany) for 90 min at 37 • C. After aspiration, the cells were incubated with 2 mL MEM supplemented with 0.9% agar (Oxoid, Wesel, Germany), 0.01% DEAE-Dextran (Pharmacia Biotech, Freiburg im Breisgau, Germany), 0.2% BSA and 0.2% NaHCO 3 (Sigma-Aldrich, Taufkirchen, Germany) at 37 • C and 5% CO 2 for 3 days. To visualize the plaques, a staining with neutral red solution (Sigma-Aldrich, Taufkirchen, Germany) in PBS was performed and the number of infectious particles (pfu mL −1 ) was determined by counting.

Immunfluorescence Microscopy
For immunofluorescence microscopy studies, Calu-3 cells were cultivated in 24-well plates supplemented with glass slides and infected as described above. At 8 h or 24 h p.i., the slides were fixed for 30 min in 4% paraformaldehyde (PFA, Sigma-Aldrich, Germany) at 37 • C and permeabilized with 0.1% Triton-X 100 (Sigma-Aldrich, Taufkirchen, Germany) for 15 min.
The human chip model membranes were fixed for 30 min with 4% PFA at 37 • C. The membrane was removed from the chip and cut in half to analyze either the epithelial or the endothelial side and permeabilized in PBS supplemented with 0.1% Saponin (Sigma-Aldrich, Taufkirchen, Germany) and 3% goat serum (Invitrogen, Dreieich, Germany) for 1 h at room temperature.
A The fluorescent images were acquired using an Axio Observer.Z1 microscope (Zeiss, Jena, Germany) with Plan Apochromat 20×/0.8 objective (Zeiss, Jena, Germany), Apo-Tome.2 (Zeiss, Jena, Germany) and Axiocam 503 mono (Zeiss, Jena, Germany) and the software Zen 2.6 (blue edition; Zeiss, Jena, Germany). Apotome defolding with phase error correction and deconvolution and Z-stack merging with maximum intensity projection was done carried out with the software Zen 2.6 as well.

Detection of mRNA Expression by Using qRT-PCR
RNA isolation was performed using the RNeasy mini kit (Qiagen, Hilden, Germany) according to the manufacturer's protocol. The cells were lysed and scraped with 350 µL RLT lysis buffer and the RNA was eluted at the end in 30 µL RNase-free water. A NanoDrop spectrophotometer ND-1000 (Peqlab, Erlangen, Germany) was used for measuring the RNA concentration.
QuantiNova reverse transcription kit (Qiagen, Hilden, Germany) was used for cDNA synthesis. In total, 500 ng RNA was diluted to a total volume of 13 µL with RNase-free water and together with 2 µL gDNA removal mix incubated at 45 • C for 2 min. After cooling, 5 µL RT master mix (4 µL reverse transcription mix and 1 µL reverse transcription enzyme) was added to each sample and incubated at 25 • C for 3 min, at 45 • C for 10 min and inactivated at 85 • C for 5 min.
For primer sequences, see Table 1.

Flow Cytometry Analyses
The flow cytometry analyses were performed using the LEGENDplex TM assays (Biolegend, Amsterdam, Netherlands) according to the manufacturer's protocol. In total, 25 µL of supernatant of Calu-3 cells or the different chambers of the chip model were incubated with the premixed beads (Human Anti-Virus Response Panel, Cat. 740349; Human Thrombosis Panel, Cat. 740891) overnight at room temperature on a plate shaker. The detection antibodies were incubated for 1 h and afterwards the SA-PE was added and further incubated for 30 min at room temperature on a plate shaker. The samples were fixed for 30 min in 4% PFA and after washing directly measured on an Accuri C6 Plus flow cytometer (BD Bioscience, Heidelberg, Germany). The data analysis was performed using the LEGENDplex TM webtool powered by QOGNIT (San Carlos, CA, USA) (https://legendplex.qognit.com/, access dates: Figure 3b

Statistical Analysis
Statistical analyses were performed in GraphPad Prism 8, the methods used are described in the figure legends. Informed Consent Statement: Informed consent was obtained from all subjects involved in the study.

Data Availability Statement:
The data presented in this study are available in the present article and supplementary material.