Combinations of Favipiravir with Doxycycline, Azithromycin or Ivermectin Exert Synergistic Effects Against Influenza A H3N2 Virus Replication

Abstract

Influenza A viruses constantly threaten the global population, with seasonal outbreaks occurring in different parts of the world, including avian influenza. Severe influenza A virus infections are strongly associated with the cytokine storm, which can contribute significantly to morbidity and even mortality. The virulence and high mutability of these viruses necessitate more effective treatment strategies and regimens to manage patients, especially those with a severe disease. Favipiravir is an antiviral agent approved in Japan for treating influenza virus strains resistant to the current antivirals. The objective of this study is to investigate the combination treatment of Favipiravir paired with selected repurposed drugs to determine the effectiveness of these combinations against influenza A virus replication as well as their effects on cytokine expression. Specific combinations of Favipiravir with Doxycycline, Azithromycin or Ivermectin were identified to be highly synergistic and effective in inhibiting live virus titers of an influenza H3N2 clinical strain by 4 log₁₀. Furthermore, combinations of Favipiravir with Doxycycline or Azithromycin also exhibited immunomodulatory effects on pro-inflammatory cytokines by strongly reducing the relative mRNA expression of IFN-γ, IL-6, TNF-α and IL-1β. Notably, monotherapy with Andrographolide also completely inhibited influenza virus titers by 4 log₁₀. Specific combinations of Favipiravir with Artesunate or Andrographolide revealed additive effects by inhibiting influenza virus titers by about 2 or 1.5 log₁₀, respectively. Our findings indicate that specific drug combinations show promising efficacy and potential in the treatment of influenza and warrant further studies using influenza models of human cell, tissue and animal infection.

1. Introduction

Influenza viruses belong to the Orthomyxoviridae family and are single-stranded ribonucleic acid (RNA-enveloped viruses that cause acute respiratory illnesses ranging from mild fatigue to potential death due to respiratory failure. Influenza virus infections cause high morbidity and mortality rates around the world, with an estimated three to five million severe cases, and 290,000 to 650,000 deaths every year [1]. Moreover, the ongoing highly pathogenic avian influenza A/H5N1 virus circulating in the United States, and its transmission from birds to cattle poses a global health concern of a possible spillover to humans [2].

Influenza viruses are categorized into four types, i.e., A, B, C and D. Types A and B are the most common cause of disease in humans [1]. Influenza A viruses (IAVs) are further classified into subtypes based on their hemagglutinin (HA) and neuraminidase (NA) glycoproteins, while influenza B viruses are categorized based on their lineages (Yamagata and Victoria).

While antiviral vaccines may be useful as initial protection against emerging pandemic viruses, the options available for both prophylactic and therapeutic treatment of severe IAV infections remain limited, and the rise of antiviral drug resistance poses a significant challenge [3]. While current antivirals such as Oseltamivir and Baloxavir Marboxil are still being used at present, there have been reported instances of IAV gaining resistance to them, such as a naturally occurring Oseltamivir-resistant IAV detected in a seasonal H1N1 virus in Norway in 2007 and Baloxavir-resistant IAV found in children or immunocompromised individuals infected with the H3N2 or H1N1pdm09 strains [4,5].

Combination therapy involves the use of two or more drugs to treat pathogenic infections, with an ideal combination consisting of an antiviral and a drug that targets the host cell pathway [6,7]. Combination therapy has been successfully used to treat diseases such as cancer and human immunodeficiency virus [8,9]. Combination therapy has also been investigated in the context of influenza, e.g., combining Oseltamivir with Baloxavir [7]. One major advantage of combination therapy is that it limits susceptibility to drug resistance as compared with monotherapy. Given that influenza is able to develop resistance easily due to antigenic drift, deploying multiple drugs simultaneously can significantly increase the duration for resistance to develop against them [10]. Since most influenza complications arise due to the cytokine storm, modulation of the host’s pro-inflammatory responses by using host-targeting drugs in combination therapy also acts as a complementary therapeutic strategy along with antivirals [11]. Host-targeting drugs could also potentially downregulate or temporarily inhibit some host-cell pathways that are essential for viral replication, further diminishing both viral replication and the inflammatory response [11]. Combination therapy could potentially give rise to synergistic drug combinations that could enhance their therapeutic effects over and above a single drug. This allows for lower concentrations of drugs to be administered, thus minimizing the risk of toxicity and adverse effects while exerting enhanced clinical efficacy [12]. Additionally, dual-drug use can also suppress the amount of pro-inflammatory cytokines produced by macrophages, thereby mitigating excessive inflammatory reactions [13].

Drug repurposing refers to the use of existing drugs for a new treatment that has not been indicated before [14]. The major advantage is that it lessens the time, risks and costs of developing new treatments as compared with the established way of de novo drug discovery which often suffers from high failure rates. Drugs already on the market have also passed safety and toxicity studies, which renders drug applications more cost-effective [15].

Since 2014, Favipiravir (T-705, Avigan) has been administered to patients in Japan against new and re-emerging pandemic influenza virus infections [16,17]. As a nucleoside analog, the active form of Favipiravir, ribofuranosyl-5′-triphosphate, acts as a selective and potent inhibitor of RNA-dependent RNA polymerase (RdRp) by incorporating itself into the nascent viral RNA, leading to chain termination and viral mutagenesis [18]. This drug also exhibits broad-spectrum antiviral activity against other RNA viruses such as Zika virus and Ebola virus [19,20]. However, resistance towards Favipiravir has been recorded, with a K229R mutation in the PB1 subunit of IAV RNA polymerase [21]. Fortunately, such resistance is only limited to viruses that were cultured in vitro but has not been detected in a clinical setting. In a study on severe influenza patients, combination therapy with Favipiravir and Oseltamivir did not prevent the emergence of Oseltamivir-resistant mutants [22]. There is thus a research gap to discover novel treatment methods coupled with different types of drugs to mitigate the risk of viral resistance and to improve clinical outcomes [23].

Studies have indicated that certain antibiotics and antiparasitic drugs exhibit some antiviral and anti-inflammatory activities against influenza and other virus infections [24,25,26,27,28,29,30]. Furthermore, when an antiviral is combined with such repurposed drugs, e.g., Remdesivir with Ivermectin, such combination therapy can exert highly synergistic and promising antiviral and immunomodulatory activities against coronavirus infection [13].

On the basis of the antiviral effects of Favipiravir, we hypothesized that its dual-drug combination with the repurposed drugs Andrographolide, Artesunate, Azithromycin, Doxycycline and Ivermectin could produce synergistic effects in lowering both live viral load and the production of pro-inflammatory cytokines. Hence, this study aims to determine the efficacy of monotherapy on IAV; the efficacy of dual-drug combinations against IAV; the cytotoxicity of both monotherapy and combination therapy; and the effects of combination therapy on the pro-inflammatory cytokine expression of IL-1β, IL-6, TNF-α and IFN-γ.

2. Materials and Methods

2.1. Cell Line

Madin–Darby canine kidney cells containing cDNA of human 2,6-sialyltransferase (MDCK-SIAT1, ATCC, Manassas, VA, USA) were maintained for 3 to 4 days in Dulbecco’s modified Eagle’s medium (DMEM) (HyClone, SH30022.01, Logan, UT, USA) containing 10% heat-inactivated fetal bovine serum (FBS) (BioWest, S1810, Bradenton, FL, USA) as well as 1 mg/mL of Geneticin G418 selective antibiotic (Gibco, 10131035, Waltham, MA, USA) at 37 °C with 5% CO₂. MDCK-SIAT1 cells (1.3 × 10⁵ per well) were seeded in 24-well plates for virus plaque assays and half-maximal inhibitory concentration experiments, while 6.5 × 10⁴ cells per well were seeded in 96-well plates for the half-maximal cytotoxic concentration experiments. These cell seeding densities were determined to be optimal since they produced ~90% confluency in each well after 24 h of incubation.

2.2. Influenza Virus

The influenza A/Singapore/G2-31.1/2014(H3N2) strain is a clinical isolate from an influenza patient in Singapore [31]. The initial live virus titer was quantified via a plaque assay to be 7.5 × 10⁷ plaque-forming units (PFU) per mL in order to determine the volume needed to achieve a multiplicity of infection (MOI) of 0.1.

2.3. Drugs

The direct-acting anti-influenza drug tested was Favipiravir (MedChemExpress, Cat# HY-14768, Monmouth Junction, NJ, USA), while the repurposed drugs included Andrographolide (MedChemExpress, HY-N0191), Artesunate (MedChemExpress, HY-N0193), Azithromycin (MedChemExpress, HY-17506), Doxycycline (Sigma-Aldrich, D989, St Louis, MO, USA) and Ivermectin (Sigma-Aldrich, I8898). The drugs were prepared as 10 mM solutions using dimethyl sulfoxide (DMSO), while Doxycycline was prepared using sterile water. Drug dilutions were subsequently prepared in DMEM containing 0.5% DMSO and 2% FBS.

2.4. Half-Maximal Cytotoxic Concentration (CC₅₀)

MDCK-SIAT1 cells were first incubated in a 96-well plate for 24 h at 37 °C with 5% CO₂. The wells were washed twice with sterile phosphate-buffered saline (PBS) before adding drugs at varying concentrations. Each well was topped up to a final volume of 100 μL with DMEM containing 0.5% DMSO and 2% FBS before being maintained for 48 h at 37 °C with 5% CO₂. Cell viability was determined by the MTS tetrazolium assay using the CellTiter 96 AQueous One Solution cell proliferation assay (Promega, G3580, Madison, WI, USA) following the manufacturer’s protocol. Absorbance values were read at 490 nm using the Spark 10M multimode microplate reader (Tecan, Mannedorf, Switzerland). Final absorbance values were obtained after subtracting the background absorbance from the cells and drug, and normalizing with untreated cells. Uninfected cells cultured in a medium containing 0.5% DMSO (without any drug) served as the background reference control. Biological replicates were performed (n = 3). The optical density readings at 490 nm of the untreated uninfected control cells (with 0.5% DMSO) were extremely high, thus indicating living, viable cells. In contrast, the absorbance reading for the medium only with 0.5% DMSO (without cells) was extremely low (Supplementary Table S1), i.e., consistent with the principle of the MTS assay, whereby higher absorbance indicates greater cell viability.

2.5. Quantification of Virus Titer via Plaque Assays

MDCK-SIAT1 cells were incubated in a 24-well plate for 24 h at 37 °C with 5% CO₂. Ten-fold serial dilutions from 10⁻¹ to 10⁻⁶ of the virus supernatant obtained from the checkerboard assays were prepared using plain DMEM. The medium in the seeded 24-well plate was then removed, and the plate was rinsed twice with PBS. A virus inoculum of each dilution (100 μL) was then added into each well for infection, and the plate was incubated at 35 °C with 5% CO₂ for 1 h with gentle rocking every 15 min. An overlay concentration was prepared, comprising 1.2% Avicel CL-611 (FMC, Philadelphia, PA, USA), 1× MEM (Gibco, 41500, Waltham, MA, USA) and 0.5 μg/mL N-tosyl-L-phenylalanine chloromethyl ketone (TPCK)-trypsin (Sigma-Aldrich, 4352157). The overlay (500 μL) was added into each well, and the plate was incubated for 72 h at 35 °C with 5% CO₂. The overlay was subsequently removed, and cells were fixed for 1 h in 4% formaldehyde before staining with 1% crystal violet [32].

2.6. Half-Maximal Inhibitory Concentration (IC₅₀)

MDCK-SIAT1 cells were first incubated in a 24-well plate for 24 h at 37 °C with 5% CO₂. The wells were then washed twice with sterile PBS, then IAV (at MOI of 0.1) was added to each well and incubated for 1 h at 35 °C with 5% CO₂. TPCK-trypsin (2 μg/mL) was added to each well and further incubated for 2 h at 35 °C with 5% CO₂. The inoculum was subsequently removed, and varying concentrations of drugs (chosen on the basis of their CC₅₀ values) were added to each well. The cells were then incubated for 48 h at 35 °C with 5% CO₂. The supernatant in each well was then harvested, and the virus plaque assay was performed to determine the live virus titer. Biological replicates were performed for treatments with Favipiravir, Doxycycline, Azithromycin and Ivermectin (n = 2). Technical replicates were performed for treatments with Artesunate and Andrographolide (n = 2).

2.7. Effects of Combination Therapy Versus Monotherapy via Checkerboard Assays

Checkerboard assays (7 × 6 and 6 × 6) were performed to determine the effects of drug combinations on cell viability, live viral titers and viral RNA levels in comparison with their corresponding monotherapies. Varying drug concentrations were evaluated on the basis of the IC₅₀ values obtained for the monotherapy. Similar procedures were performed as for the CC₅₀ and IC₅₀ experiments. Infected cells cultured in the medium containing 0.5% DMSO (without any drug) served as the background reference control. The virus titers of the infected control cells mock-treated with 0.5% DMSO were high, ranging from 1 to 7.5 × 10⁵ PFU/mL, thus verifying the lack of virus inhibition by 0.5% DMSO (Supplementary Table S2).

2.8. RNA Extraction and Reverse Transcription-Quantitative Polymerase Chain Reaction (RT-qPCR)

For viral RNA (vRNA) qPCR analysis, RNA was extracted from virus-infected culture supernatants using the QIAmp viral RNA Mini kit (Qiagen, 52906, Venlo, The Netherlands) according to the manufacturer’s protocol. The concentration and purity of RNA was evaluated using the NanoDrop One Microvolume UV–Vis spectrophotometer (Thermo Fisher, Waltham, MA, USA). Each RNA sample (150 ng) was topped up to 6.25 μL with nuclease-free water, and 0.5 μL of random primers (Promega, C1181) was added. The mixture was first placed in the thermocycler at 70 °C for 5 min and then placed in ice for 1 min. A master mix containing 2 μL of M-MLV RT 5× buffer (Promega, M531A), 0.5 μL of 10 mM dNTP mix (Promega, U1511), 0.5 μL of M-MLV reverse transcriptase or RT (Promega, M170A) and 0.625 μL of recombinant RNasin ribonuclease inhibitor (Promega, N2511) was subsequently added to the RNA mixture and incubated at 37 °C for 1 h. The mixture was diluted 2.5 times with nuclease-free water after RT.

The qPCR assays were performed using the LightCycler 96 real-time PCR system (Roche Molecular Systems, Pleasanton, CA, USA). Each sample contained 1 μL of the cDNA sample obtained from RT, 5 μL of FastStart Essential DNA Green Master (Roche Molecular Systems, 06402712001), 0.5 μL each of both forward and reverse primers (5 μM each) and 3 μL of nuclease-free water. The IAV primers used and the thermocycling protocol are shown in

Table A1. Nucleotide sequences of RT-qPCR primers for RNA detection of IAV, cytokines and the GAPDH housekeeping gene.

Primer	Nucleotide Sequence (5′ to 3′)
IAV M1-M2-Forward	GGACTGCAGCGTAGACGCTT
IAV M1-M2-Reverse	CATCCTGTTGTATATGAGGCCCAT
IL1-β-Forward	TGCAAAACAGATGCGGATAA
IL1-β-Reverse	GTAACTTGCAGTCCACCGATT
IL-6-Forward	TCCAGAACAACTATGAGGGTGA
IL-6-Reverse	TCCTGATTCTTTACCTTGCTCTT
TNF-α-Forward	CGTCCATTCTTGCCCAAAC
TNF-α-Reverse	AGCCCTGAGCCCTTAATTC
IFN-γ-Forward	CCAGATCATTCAAAGGAGCA
IFN-γ-Reverse	CGTTCACAGGAATTTGAATCAG
GAPDH-Forward	TTCCACGGCACAGTCAAG
GAPDH-Reverse	ACTCAGCACCAGCATCAC

and

Table A2. Protocol for qPCR conditions to quantify viral RNA and cytokine mRNA expression.

Stage		Temperature	Duration	Number of Cycles
Pre-incubation		95 °C	10 min	1
3-step amplification	Denaturation	95 °C	10 s	55
	Annealing	40 °C (viral) 51 °C (cytokine)	5 s
	Extension	72 °C	8 s
Melt curve analysis		95 °C	10 s	1
		65 °C	60 s
		97 °C	1 s

of Appendix A [33]. Quantification of the absolute viral RNA level was performed (n = 1) via a standard curve generated from cDNA of the virus stock of known titer using dilutions ranging from 10⁶ to 10¹. The Pearson correlation R² value (coefficient of determination) was then calculated.

2.9. Cytokine mRNA Expression Assays

For cytokine qPCR analysis, RNA was extracted from cell lysates harvested from both the monotherapy and checkerboard assays using the RNeasy Mini kit (Qiagen, 74106) as recommended by the manufacturer. The protocols for RT-qPCR (n = 1) remain the same as stated in Section 2.8, except for the annealing temperature of 51 °C. The cytokine primers used and the thermocycling profiles are shown in Table A1 and Table A2 [34].

The relative mRNA expression of pro-inflammatory cytokines (IL-1β, IL-6, TNF-α and IFN-γ) in MDCK-SIAT1 cells was quantified with reference to that of the housekeeping gene glyceraldehyde-3-phosphate dehydrogenase (GAPDH). The fold change in cytokine mRNA levels was determined by the formula 2^−∆∆Ct, where ∆∆Ct represents the difference in ∆Ct values between the control and the treated sample. ∆Ct is the difference in the Ct values of the target gene and the GAPDH reference gene [35].

2.10. Statistical Analyses

Dose–response curves, checkerboard heatmaps, bar graphs and a standard curve of quantitation were plotted using GraphPad Prism (version 9.5.0). Nonlinear regression analysis was used. For the bar graphs, total vRNA copies and the relative mRNA expression of cytokines are log₁₀-transformed values. To interpret the results of each dual-drug combination, the Bliss independence model was used to score synergistic effects [36]. Bliss independence assumes that both drugs act independently when administered at the same time [37]. The average Bliss synergy score of the checkerboard assays was calculated using SynergyFinder+ 3.0 [38]. According to the SynergyFinder program, synergy scores below −10 suggest antagonism, while scores between −10 and 10 indicate additive effects. Synergy scores above 10 suggest synergism for the dual-drug combination.

3. Results

3.1. Antiviral Efficacy of Monotherapy with Andrographolide or Favipiravir

Monotherapy with Andrographolide showed minimal cytotoxicity in MDCK-SIAT1 cells with a CC₅₀ value of 144 μM (Figure 1A), while exhibiting complete viral inhibition with a 4 log₁₀ reduction in live viral load at higher concentrations (Figure 1B). The total vRNA also showed a decrease from 10⁷ to 10⁵ with increasing concentrations of Andrographolide (Figure 1C).

Not surprisingly, Favipiravir monotherapy could completely inhibit live virus replication at higher concentrations, e.g., 4 μM (Figure A3, panel A). In Figure A2 and Figure A3, the effects of monotherapies with the six tested drugs on cell viability and live viral loads are compared (including statistical analyses).

3.2. Combination Treatment with Favipiravir and Doxycycline

The combination of Favipiravir and Doxycycline (FVP-DOX) showed minimal cytotoxicity in MDCK-SIAT1 cells, with all 30 combinations in the FVP-DOX matrix exhibiting cell viability levels of 80% and above (Figure 2A). In drug testing, a cell viability of 80% is generally accepted to be the threshold for a drug to be considered as non-cytotoxic according to international standards (e.g., ISO 10993-5).

Complete inhibition of the live viral load (4 log₁₀ reduction) was observed with 0.5 μM Favipiravir paired with 10 μM Doxycycline, as well as with combinations using higher concentrations of Favipiravir or Doxycycline (Figure 2B). This reduction in the live viral load was significantly greater than the 1 log₁₀ reduction with Doxycycline monotherapy at 40 μM and the <0.5 log₁₀ reduction with Favipiravir monotherapy at 0.5 μM. However, the FVP-DOX combination did not significantly decrease total vRNA levels compared with Favipiravir monotherapy (Figure 2C). It should be noted that the total vRNA levels are generally higher than the live viral titers, since total vRNA detection includes vRNA from defective viral particles as well as vRNA that is not incorporated into infectious virions [39].

Although the predicted Bliss synergy score was −1.58 (Figure 2E), the additional reductions in the live viral load by the FVP-DOX combination over monotherapy treatments indicate synergistic drug effects.

3.3. Combination Treatment with Favipiravir and Azithromycin

The combination of Favipiravir and Azithromycin (FVP-AZM) showed minimal cytotoxicity in MDCK-SIAT1 cells, with 26 out of 30 FVP-AZM combinations exhibiting a cell viability of at least 80% (Figure 3A). Complete inhibition of the live viral load (4 log₁₀ reduction) was achieved by combining 0.5 μM Favipiravir and 60 μM Azithromycin (below its IC₅₀ of 93.80 μM), as well as by combinations at higher concentrations of Favipiravir and Azithromycin (Figure 3B). This reduction in the live viral load was greater than the 1 log₁₀ reduction each for Azithromycin monotherapy at 60 μM and for Favipiravir monotherapy at 0.5 μM. However, the FVP-AZM combinations showed no reduction in total vRNA levels compared with the respective monotherapies (Figure 3C). Compared with the monotherapies, the additional reduction in the live viral load by certain combinations of Favipiravir and Azithromycin thus indicates the synergistic effects of the FVP-AZM combination despite the predicted Bliss synergy score of 0.31 (Figure 3E).

3.4. Combination of Favipiravir and Ivermectin

The combination of Favipiravir and Ivermectin (FVP-IVM) showed minimal cytotoxicity in MDCK-SIAT1 cells for lower concentrations of Ivermectin, with 23 out of 30 combinations exhibiting a cell viability of 80% and above. Combinations with low cell viability were observed when the highest concentration of Ivermectin at 8 μM was tested (Figure 4A). Combinations of Favipiravir with 8 μM Ivermectin were thus not considered in the interpretation of synergistic effects in view of the observed cytotoxicity.

Complete inhibition of the live viral load (4 log₁₀ reduction) was achieved by combining 2 μM Favipiravir with 0.5, 1, 2 or 4 μM Ivermectin (below its IC₅₀ of 4.97 μM), as well as by combinations of higher concentrations of Favipiravir with Ivermectin (Figure 4B). This reduction in the live viral load was greater than the <1 log₁₀ reduction with Ivermectin monotherapy at 4 μM and the <2 log₁₀ reduction with Favipiravir monotherapy at 2 μM. With the FVP-IVM combinations, total vRNA showed no observable decreases compared with the monotherapies and remained at above 10⁶ copies (Figure 4C). The dual FVP-IVM drug combination could exert greater virus-inhibitory effects than the monotherapy treatments, thus indicating drug synergism despite the predicted Bliss synergy score of −4.75 (Figure 4E).

3.5. Combination of Favipiravir and Artesunate

The combination of Favipiravir and Artesunate (FVP-ART) did not cause any significant cytotoxicity, with cell viability above 85% even at the highest tested concentrations of both drugs (Figure 5A). The combinations of 0.24 μM Favipiravir and 16.0 μM Artesunate as well as 0.30 μM Favipiravir and 16.0 μM Artesunate yielded a 2.30 log₁₀ reduction in live viral titers compared with the untreated infected control (Figure 5B). In addition, the FVP-ART combination could further reduce live virus titers compared with Favipiravir monotherapy and also led to a ~2 log₁₀ reduction in the live virus titer compared with Artesunate monotherapy. However, the viral RNA loads for the FVP-ART combinations were generally similar to those of Favipiravir and Artesunate monotherapies (Figure 5C). Overall, the predicted Bliss synergy score for this combination was 22.34, indicating synergism between the two drugs (Figure 5E).

3.6. Combination of Favipiravir and Andrographolide

The combination of Favipiravir and Andrographolide (FVP-ADG) did not show any cytotoxicity in MDCK-SIAT1 cells, with all of the cells remaining completely viable (Figure 6A). Compared with the untreated control, a reduction in the live virus titer of ~1.5 log₁₀ was achieved by combining 0.18 μM of Favipiravir with 24 μM of Andrographolide (Figure 6B). This FVP-ADG combination resulted in live virus titers lower than monotherapy with Favipiravir or Andrographolide. However, the FVP-ADG combination led to total viral RNA levels that were similar to the respective monotherapies (Figure 6C). Overall, the Bliss synergy score of 0.089 (Figure 6E) predicted additive effects for this combination at the tested drug concentrations.

3.7. Relative mRNA Expression of Pro-Inflammatory Cytokines of Treatments Versus Controls

For cytokine mRNA expression analysis, infected and untreated MDCK-SIAT1 cells served as the control for the tested monotherapies and combination therapies—each control cytokine mRNA expression was recorded as 1. The relative cytokine mRNA expression values for selected drug concentrations of monotherapy and combination therapy were determined. Values less than 1 indicate decreased relative expression of the cytokine, and vice versa. The relative cytokine mRNA expression (fold change) of each combination treatment compared against its corresponding monotherapies was then calculated.

For single and combination treatments with Favipiravir, Artesunate and Andrographolide at selected concentrations, uninfected and untreated MDCK-SIAT1 cells served as an additional control. With this control, the fold changes in cytokine mRNA expression were similar to those with the first control, thus ensuring reproducibility.

3.8. Relative mRNA Expression of IL-1β

Table 1. Relative IL-1β mRNA expression of combination treatments versus their corresponding monotherapies.

Combination of Drugs X + Y	Fold Change in mRNA Expression (Versus Drug X Monotherapy)	Fold Change in mRNA Expression (Versus Drug Y Monotherapy)
FVP0.25 + DOX20	↓ 10.06 (FVP0.25)	↓ 3.27 (DOX20)
FVP0.25 + DOX40	↓ 10.13 (FVP0.25)	↓ 1.66 (DOX40)
FVP0.5 + DOX5	↓ 5.06 (FVP0.5)	↓ 8.17 (DOX5)
FVP0.25 + AZM15	↓ 19.43 (FVP0.25)	↓ 2.33 (AZM15)
FVP0.25 + AZM30	↓ 10.13 (FVP0.25)	↓ 1.03 (AZM30)
FVP2 + IVM0.5	↓ 3.18 (FVP2)	↓ 5.58 (IVM0.5)
FVP2 + IVM1	↓ 1.66 (FVP2)	↓ 1.27 (IVM1)
FVP0.24 + ART16	↓ 1.23 (FVP0.24)	↓ 1.001 (ART16)
FVP0.12 + ADG24	↑ 1.47 (FVP0.12)	↑ 2.20 (ADG24)
FVP0.18 + ADG18	↑ 2.03 (FVP0.18)	↓ 1.03 (ADG18)

Shown are the fold changes in the IL-1β mRNA expression of each combination treatment relative to their corresponding monotherapies. ↓ indicates reduced while ↑ indicates increased IL-1β mRNA expression of the combination treatment versus its monotherapy. Infected and untreated MDCK-SIAT1 cells served as the reference control. FVP: Favipiravir; DOX: Doxycycline; AZM: Azithromycin; IVM: Ivermectin; ART: Artesunate; ADG: Andrographolide. The number after each drug indicates the drug concentration in μM.

shows the relative IL-1β mRNA expression (fold change) of the various combination treatments when compared against their corresponding monotherapies. The two tested combinations of Favipiravir with Azithromycin resulted in the greatest reduction in IL-1β mRNA expression relative to Favipiravir monotherapy, i.e., a ~10-fold or 19-fold reduction. This was followed by the three tested combinations of Favipiravir with Doxycycline, i.e., about a 5- to 10-fold reduction compared with Favipiravir monotherapy. The combination of 2 μM Favipiravir with 0.5 μM Ivermectin resulted in a modest reduction in the relative IL-1β expression, i.e., ~3 to 5.5-fold compared with both monotherapies. However, the fold changes in the relative IL-1β expression for the Favipiravir–Artesunate and Favipiravir–Andrographolide combinations were quite unremarkable compared with their monotherapies.

3.9. Relative mRNA Expression of IL-6

Table 2. Relative IL-6 mRNA expression of combination treatments versus their corresponding monotherapies.

Combination of Drugs X + Y	Fold Change in mRNA Expression (Versus Drug X Monotherapy)	Fold Change in mRNA Expression (Versus Drug Y Monotherapy)
FVP0.25 + DOX20	↓ 8.36 (FVP0.25)	↓ 2.95 (DOX20)
FVP0.25 + DOX40	↓ 15.82 (FVP0.25)	↓ 1.55 (DOX40)
FVP0.5 + DOX5	↓ 29.24 (FVP0.5)	↓ 7.06 (DOX5)
FVP0.25 + AZM15	↓ 10.22 (FVP0.25)	↓ 1.78 (AZM15)
FVP0.25 + AZM30	↓ 362.98 (FVP0.25)	↓ 19.43 (AZM30)
FVP2 + IVM0.5	↓ 6.73 (FVP2)	↓ 5.31 (IVM0.5)
FVP2 + IVM1	↓ 4.23 (FVP2)	↓ 2.38 (IVM1)
FVP0.24 + ART16	↓ 1.64 (FVP0.24)	↓ 1.53 (ART16)
FVP0.12 + ADG24	↑ 1.93 (FVP0.12)	↑ 1.83 (ADG24)
FVP0.18 + ADG18	↑ 2.97 (FVP0.18)	↑ 4.86 (ADG18)

Shown are the fold changes in the IL-6 mRNA expression of each combination treatment relative to their corresponding monotherapies. ↓ indicates reduced while ↑ indicates increased IL-6 mRNA expression of the combination treatment versus its monotherapy. Infected and untreated MDCK-SIAT1 cells served as the reference control. FVP: Favipiravir; DOX: Doxycycline; AZM: Azithromycin; IVM: Ivermectin; ART: Artesunate; ADG: Andrographolide. The number after each drug indicates the drug concentration in μM.

shows the relative IL-6 mRNA expression (fold change) of the various combination treatments when compared against their corresponding monotherapies. The two tested combinations of Favipiravir with Azithromycin resulted in the most pronounced reduction in IL-6 mRNA expression relative to Favipiravir monotherapy, i.e., about 10-fold or 360-fold reductions. This was followed by the three tested combinations of Favipiravir with Doxycycline, i.e., about 8- to 30-fold reductions compared with Favipiravir monotherapy. The two combinations of Favipiravir with Ivermectin culminated in a modest reduction in relative IL-6 expression, i.e., 2.4 to 6.7-fold compared with both monotherapies. However, the fold changes in relative IL-6 expression for the Favipiravir–Artesunate and Favipiravir–Andrographolide combinations were less remarkable in comparison with their monotherapies.

3.10. Relative mRNA Expression of TNF-α

Table 3. Relative TNF-α mRNA expression of combination treatments versus their corresponding monotherapies.

Combination of Drugs X + Y	Fold Change in mRNA Expression (Versus Drug X Monotherapy)	Fold Change in mRNA Expression (Versus Drug Y Monotherapy)
FVP0.25 + DOX20	↓ 1.08 (FVP0.25)	↓ 2.17 (DOX20)
FVP0.25 + DOX40	↓ 6.45 (FVP0.25)	↓ 10.7 (DOX40)
FVP0.5 + DOX5	↓ 15.46 (FVP0.5)	↓ 18.00 (DOX5)
FVP0.25 + AZM15	↓ 2.14 (FVP0.25)	↓ 2.25 (AZM15)
FVP0.25 + AZM30	↓ 9.19 (FVP0.25)	↓ 16.11 (AZM30)
FVP2 + IVM0.5	↓ 3.03 (FVP2)	↓ 1.89 (IVM0.5)
FVP2 + IVM1	↑ 1.02 (FVP2)	↑ 1.19 (IVM1)
FVP0.24 + ART16	↑ 1.03 (FVP0.24)	↓ 1.69 (ART16)
FVP0.12 + ADG24	↓ 1.28 (FVP0.12)	↑ 1.19 (ADG24)
FVP0.18 + ADG18	↑ 3.71 (FVP0.18)	↓ 1.40 (ADG18)

Shown are the fold changes in the TNF-α mRNA expression of each combination treatment relative to their corresponding monotherapies. ↓ indicates reduced while ↑ indicates increased TNF-α mRNA expression of the combination treatment versus its monotherapy. Infected and untreated MDCK-SIAT1 cells served as the reference control. FVP: Favipiravir; DOX: Doxycycline; AZM: Azithromycin; IVM: Ivermectin; ART: Artesunate; ADG: Andrographolide. The number after each drug indicates the drug concentration in μM.

shows the relative TNF-α mRNA expression (fold change) of the various combination treatments when compared against their corresponding monotherapies. The highest reduction in TNF-α mRNA expression was observed for the combination of 0.5 μM Favipiravir with 5 μM Doxycycline, i.e., about 15- and 18-fold reductions compared with Favipiravir and Doxycycline monotherapies, respectively. Additionally, another combination of 0.25 μM Favipiravir with 40 μM Doxycycline yielded reductions in the relative TNF-α mRNA expression of about 6.5- and 11-fold compared with their monotherapies. The combination of 0.25 μM Favipiravir with 30 μM Azithromycin was also notable in decreasing the relative expression of TNF-α, i.e., about 9- and 16-fold reductions against their monotherapies. A modest reduction in relative TNF-α expression was observed for the combination of 2 μM Favipiravir with 0.5 μM Ivermectin, i.e., ~3-fold compared with Favipiravir monotherapy. However, the fold changes in the relative TNF-α expression for the Favipiravir–Artesunate and Favipiravir–Andrographolide combinations were quite unremarkable compared with their monotherapies.

3.11. Relative mRNA Expression of IFN-γ

Table 4. Relative IFN-γ mRNA expression of combination treatments versus their corresponding monotherapies.

Combination of Drugs X + Y	Fold Change in mRNA Expression (Versus Drug X Monotherapy)	Fold Change in mRNA Expression (Versus Drug Y Monotherapy)
FVP0.25 + DOX20	↓ 162.04 (FVP0.25)	↓ 7.62 (DOX20)
FVP0.25 + DOX40	↓ 50.91 (FVP0.25)	↑ 1.67 (DOX40)
FVP0.5 + DOX5	↓ 1008.40 (FVP0.5)	↓ 240.16 (DOX5)
FVP0.25 + AZM15	↓ 359.65 (FVP0.25)	↓ 2.75 (AZM15)
FVP0.25 + AZM30	↓ 256.03 (FVP0.25)	↓ 1.99 (AZM30)
FVP2 + IVM0.5	↓ 5.46 (FVP2)	↑ 1.70 (IVM0.5)
FVP2 + IVM1	↓ 73.17 (FVP2)	↓ 16.60 (IVM1)
FVP0.24 + ART16	↓ 2.53 (FVP0.24)	↑ 1.13 (ART16)
FVP0.12 + ADG24	↑ 1.06 (FVP0.12)	↑ 1.37 (ADG24)
FVP0.18 + ADG18	↑ 1.50 (FVP0.18)	↓ 1.38 (ADG18)

Shown are the fold changes in IFN-γ mRNA expression of each combination treatment relative to their corresponding monotherapies. ↓ indicates reduced while ↑ indicates increased IFN-γ mRNA expression of the combination treatment versus its monotherapy. Infected and untreated MDCK-SIAT1 cells served as the reference control. FVP: Favipiravir; DOX: Doxycycline; AZM: Azithromycin; IVM: Ivermectin; ART: Artesunate; ADG: Andrographolide. The number after each drug indicates the drug concentration in μM.

shows the relative IFN-γ mRNA expression (fold change) of the various combination treatments when compared against their corresponding monotherapies. The most prominent reduction in IFN-γ mRNA expression was notable for the combination of 0.5 μM Favipiravir with 5 μM Doxycycline, i.e., about 1000- and 240-fold reductions compared with Favipiravir and Doxycycline monotherapies, respectively. Furthermore, two other combinations of 0.25 μM Favipiravir paired with 20 μM or 40 μM Doxycycline could decrease the relative IFN-γ mRNA expression by about 160- or 50-fold respectively compared with Favipiravir monotherapy. Also noteworthy are the two combinations of 0.25 μM Favipiravir paired with 15 μM or 30 μM Azithromycin, which strongly decreased the relative expression of IFN-γ, i.e., about 360- or 256-fold reductions respectively compared with Favipiravir monotherapy. A marked reduction in relative IFN-γ expression was noted for the combination of 2 μM Favipiravir with 1 μM Ivermectin, i.e., about 73-and 16-fold against their monotherapies. However, the fold changes in relative IFN-γ expression for the Favipiravir–Artesunate and Favipiravir–Andrographolide combinations were quite unremarkable compared with their monotherapies.

4. Discussion

The combinations of Favipiravir with Doxycycline performed extremely well, with overall cell viability of over 80%, while 18 out of 30 tested combinations completely inhibited the live viral load (by 4 log₁₀). Although the Bliss synergy score was predicted to be −1.58, there were multiple synergistic combinations (such as 0.5 µM Favipiravir with 10 µM Doxycycline, and 1 µM Favipiravir with 2.5 µM Doxycycline) which displayed highly efficacious antiviral activity. Furthermore, the relative mRNA expression of cytokines was strongly reduced by certain FVP-DOX combinations compared with their monotherapies, especially IFN-γ, IL-6 and TNF-α. IAV-infected MDCK cells show elevated levels of matrix metalloproteinases (MMPs), which play crucial roles in lung injury, especially in severe influenza pneumonia [25]. While its exact antiviral mechanisms are still unclear, Doxycycline acts as an inhibitor of MMPs and shows good potential in ameliorating infections such as influenza and tuberculosis [40,41,42]. Therefore, it may be promising to further explore this FVP-DOX combination for treating IAV infections by pairing Favipiravir as an RdRp inhibitor that targets viral RNA replication together with the anti-inflammatory and other properties of Doxycycline.

Another encouraging combination was Favipiravir with Azithromycin, with 26 out of 30 tested combinations showing a cell viability of 80% and above. However, there was no significant change in total vRNA levels across FVP-AZM combinations. Although the Bliss synergy score was predicted to be only 0.31, 16 out of 30 combinations (such as 60 µM Azithromycin paired with 0.5 µM or 1 µM Favipiravir) could achieve complete inhibition of the live viral load, suggesting potent drug synergism. It is noteworthy that the two tested FVP-AZM combinations could strongly decrease the relative mRNA expression of cytokines compared with their monotherapies, particularly IFN-γ, IL-6 and IL-1β. A proposed mechanism of action for Azithromycin is the inhibition of endosomal acidification during the early phase of viral invasion, which can impact IAV replication [29]. Furthermore, Azithromycin has anti-inflammatory activity such as regulating T-helper functions and decreasing production of the pro-inflammatory cytokines IL-8, IL-6 and TNF-α [27,43]. Hence, Azithromycin could complement the RdRp inhibition by Favipiravir in targeting different viral pathways and components of IAV infections to augment robust treatment.

For the combination of Favipiravir with Ivermectin, the concentration of 8 µM Ivermectin exhibited significant cytotoxic effects on MDCK-SIAT1 cells. Out of 23 FVP-IVM combinations, seven could attain complete inhibition of the live viral loads, but the total vRNA levels were generally unchanged at above 10⁶ copies. Despite the Bliss synergy score predicted to be −4.75, several FVP-IVM combinations were strongly synergistic by completely inhibiting the live virus, e.g., 2 µM Favipiravir with Ivermectin at either 1 µM, 2 µM or 4 µM. Certain FVP-IVM combinations could reduce the relative mRNA expression of cytokines (especially IFN-γ) compared with their monotherapies. Ivermectin is known to efficiently inhibit the import of viral ribonucleoprotein (vRNP) into the nucleus for IAV replication. The inhibition of vRNP import by Ivermectin coupled with RdRp inhibition by Favipiravir makes the FVP-IVM combination a potentially promising strategy by targeting different pathways in the IAV replication cycle [44].

All tested combinations of Favipiravir with Artesunate showed minimal cytotoxicity effects, with a cell viability above 80%. Although none of the FVP-ART combinations exhibited complete live viral inhibition, the Bliss synergy score of 22.34 suggested that this FVP-ART combination was synergistic. This was evident in the combinations of 0.24 μM Favipiravir with 16 μM Artesunate and of 0.30 μM Favipiravir with 16 μM Artesunate, which culminated in a 2.30 log reduction in the live viral titers. While the exact mechanisms of the FVP-ART combination are still unknown, Artesunate itself is known to disrupt the mitogen-activated protein kinase (MAPK) signaling pathway, which prevents IAV vRNP nuclear export [45]. Hence, the FVP-ART combination could further impede IAV replication by hindering its replicative cycle at the transcriptional and vRNP stages, thus rendering it as a potential novel treatment against IAV infections. Future studies should investigate combinations of Favipiravir with Artesunate at higher drug concentrations to determine whether they can achieve more pronounced antiviral efficacy.

All the tested combinations of Favipiravir with Andrographolide (FVP-ADG) showed no cytotoxicity effects, with cell viability at 100%. However, none of these combinations exhibited complete live viral inhibition. Despite the low Bliss synergy score of 0.089, some reduction in the live virus titers was observed for 0.18 μM Favipiravir with 24 μM Andrographolide, and for 0.3 μM Favipiravir with 18 μM Andrographolide. Andrographolide is reported to hamper the activation of NF-κB, which subsequently prevents the production of pro-inflammatory cytokines and lowers the efficiency of IAV replication [46]. Andrographolide also confers protective actions against pulmonary inflammation in mice exposed to cigarette smoke and infected with non-typeable Haemophilus influenzae [47]. This suggests that Andrographolide does not exert direct anti-influenza activity but rather indirectly targets IAV via its immunomodulatory activity [48]. This may also explain why its combination with Favipiravir did not result in a significant reduction in IAV RNA levels. It is noteworthy that Andrographolide monotherapy performed well by achieving complete viral inhibition at higher concentrations (84 μM and above). Monotherapy of infected cells with 24 μM Andrographolide could also reduce the relative mRNA expression of the pro-inflammatory cytokines IFN-γ, IL-6 and IL-1β by 10.4-fold, 5.2-fold and 4.6-fold, respectively (compared with infected and untreated control cells). Therefore, applying higher concentrations of both Andrographolide and Favipiravir in combination may potentially accomplish even greater antiviral and immunomodulatory effects.

One limitation of this study is that it was conducted only in the MDCK-SIAT1 cells commonly used for influenza virus propagation. However, this canine kidney cell line differs from human respiratory cells in terms of viral entry, replication, and immune response pathways. This may limit extrapolating the conclusions to humans. Future studies should thus evaluate testing on human cell models such as A549 human lung carcinoma, primary human nasal and bronchial epithelial cells [49], and subsequently on animal models which better mimic in vivo conditions [25,50,51]. This can provide useful insights into the host responses to IAV as well as the effects of combination drug therapy (and potential adverse reactions) on animals, including non-human primates. Future experiments could also include known and effective anti-influenza drug combinations (e.g., Favipiravir and Oseltamivir) as positive controls to validate the sensitivity and reliability of the checkerboard assays [22].

The challenges of applying the findings of this study in clinical practice should also be considered. Examples of some further questions and challenges include the feasibility of achieving effective synergistic concentrations of Azithromycin and Doxycycline in the human body (such as 60 μM and 10 μM), the potential risks of antibiotic misuse and the development of resistance. Another crucial issue to be addressed is drug toxicity, e.g., the relationship between the known neurotoxicity of Ivermectin and its effective concentration window [52].

5. Conclusions

Favipiravir is a suitable drug to be used as the main antiviral in combination therapy, due to its wide safety profile. The repurposed drugs, while unsuitable to be used in monotherapy, have exhibited synergistic results when paired with Favipiravir. Their anti-inflammatory properties and the downregulation of the production of pro-inflammatory cytokines are highly useful when paired with Favipiravir in attenuating the cytokine storm associated with severe IAV infection. These repurposed drugs also target different mechanisms from Favipiravir, which renders combination therapy a more robust treatment strategy.