Inhibitory effect of microalgae and cyanobacteria extracts on influenza virus replication and neuraminidase activity

Origin of microalgae and cyanobacteria strains

Microalgae and cyanobacteria strains used in this study belong to the Culture Collection of Microalgae at UFRJ (CCMR, Marine Phytoplankton Laboratory, Federal University of Rio de Janeiro). These strains were isolated by single-cell sorting in a flow cytometer (Fistarol et al., 2018) and are maintained in culture by successive transfers to fresh liquid medium at 3–6 week intervals, depending on the species. All 26 strains used here are originated from fresh (Paraguaçu River and Shallow Lake System of Lençois Maranhenses) and marine water (Guanabara Bay estuary and Abrolhos coral reefs) biomes in Brazil (Table 1). Taxonomic identification of the strains was based on morphology and, for most of them, also on DNA sequencing (SSU rDNA) (Fistarol et al., 2018). The field experiments were approved by Ministério do Meio Ambiente (MMA), Instituto Chico Mendes de Conservação da Biodiversidade (ICMBio) and Sistema de Autorização e Informação em Biodiversidade (SISBIO), approval number: 35854-2. The access to genetic heritage was approved by Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq), approval number: 010339/2014-0; and Sistema Nacional de Gestão do Patrimônio Genético e do Conhecimento Tradicional Associado (SisGen), approval number: A03D5C4.

Microalgae and cyanobacteria cultivation

Microalgae and cyanobacteria were grown in autoclaved sterile medium at 26 ± 1 °C, with a photon flux of ca. 300 µmoles photon m⁻² s⁻¹ and 16 h light: 8 h dark photoperiod. Freshwater strains were grown in ASM-1 medium whereas marine strains were grown in f/2 medium (Gorham et al., 1964; Guillard & Lorenzen, 1972; Reynolds & Jaworski, 1978). Cultures were grown in glass balloons with 300–500 mL of culture medium for 3–6 weeks, depending on the growth rate of each strain. The cultures were stored at −80 °C until extracts production.

Extracts production

All the microalgae and cyanobacteria cultures (cells and culture medium) were frozen, thawed and filtered through cheesecloth. The filtrate was extracted three times with ethyl acetate (added ratio 1:1 of volume) and three times with n-butanol (added ratio 1:1 of volume). The organic layers were concentrated under reduced pressure to dryness. A total of 38 extracts were produced, 26 in ethyl acetate (EtOAc) and 12 in n-butanol (n-BuOH). The crude extracts were stored at −20 °C until analysis. The extracts from one to 12 strains were obtained in the first batch with both solvents. After the screening test for anti-influenza activity, the second batch of extracts was produced from 13 to 26 strains, and they were solely obtained with ethyl acetate. Dried extracts were resuspended in dimethyl sulfoxide (DMSO) in a final concentration of 100 mg/mL for the in vitro tests. All the extracts were analyzed by TLC, HPLC, and Mass Spectrometry (MS). However, we could not identify any compounds, but the chemical profiles of the active extracts were preserved. Thus, the procedure must be repeated on a larger scale for the more active anti-influenza extracts. The chromatograms and mass spectra for extracts 5 and 21 are provided in the File S1 (S01).

Cells and virus

Madin-Darby canine kidney (MDCK) cells (London line) were kindly donated by the Centers for Disease Control and Prevention (CDC), Influenza Reagent Resources (IRR) (FR-58). These cells were used for cytotoxicity analysis, viral growth and experimental assays with influenza virus infection. These cells were cultured in Dulbecco’s modified Eagle’s medium (DMEM; Gibco, Waltham, MA, USA) supplemented with 10% of fetal bovine serum (FBS; Gibco) and 100 U/mL penicillin and 100 mg/mL streptomycin (Sigma-Aldrich, St. Louis, MO, USA). Cells were cultured at 37 °C in 5% CO₂ atmosphere (Szretter, Balish & Katz, 2006).

The viruses were grown and titrated according to WHO manual for the laboratory diagnosis and virological surveillance of influenza (WHO, 2011). The virus titration was performed by 50% Tissue Culture Infectious Dose (TCID₅₀) assay. We analyzed the inhibitory effect of the extracts against influenza A strains and B lineages, OST-sensitive (Wild type—WT) and resistant (mutant): A/California/04/2009 (A(H1N1)pdm09-WT); A/Perth/261/2009 (A(H1N1)pdm09-H275Y); A/Switzerland/9715293/2013 (A(H3N2)-WT); B/Phuket/3073/2013 Yam (WT); B/Brisbane/60/2008 Vic (WT); B/Perth/211/2001 Yam (WT) and B/Perth/211/2001 Yam (D197E). All the viruses were kindly provided by the International Society for Influenza and other Respiratory Viruses Diseases-Antiviral Group (isirv-AVG) and CDC –IRR.

Cytotoxicity assay

One hundred microliters of 2.0 × 10⁴ MDCK cells were seeded into the 96-well culture plates (flat bottom) and grown for 24 h at 37 °C in a 5% CO₂ atmosphere, then, the extracts were added with a final concentration of 1 mg/mL in DMEM. Control cells were treated with 1.0% DMSO, which did not affect the growth of the cells. The extracts were diluted in culture medium DMEM, the DMSO final concentration was equal to 1.0% (v/v). After 48 h of cells incubation with extracts or vehicle (DMSO), a freshly prepared XTT (Sigma) solution was added, as specified by the manufacturer’s instructions. The XTT is a colorless or slightly yellow compound that is reduced to formazan (bright orange) by mitochondrial dehydrogenases of viable cells, as described before (Roehm et al., 1991; Scudiero et al., 1988). The cell viability was calculated by the comparison between absorbance (A475 nm–A660 nm) from treated (A) and untreated (B) cells using the formula: A/B × 100. Tests were carried out performing three replicates (n = 3).

Screening of extracts on viral replication

Initially, we performed a screening to evaluate the inhibitory effect of the extracts. MDCK cells (2.0 × 10⁵ cells/well) were seeded in 24-well tissue culture plates and incubated overnight. Then, the monolayers of cells were infected with A(H1N1)pdm09-WT or A(H3N2)-WT at 400 TCID₅₀ per well during 1 h at 37 °C and 5% CO₂. After, the viral suspension was removed and the monolayers were treated with 200 µg/mL of each extract. There is no established protocol or concentration for the anti-influenza screening of crude extracts of natural products. There are papers that use 2 mg/ml as the maximum concentration, while others use 100 µg/ml (Ding et al., 2017; Ghoke et al., 2018; Rajasekaran et al., 2013; Shoji et al., 2017; Sood et al., 2012). At 48 h post-infection (hpi), the supernatant was harvested to quantify the influenza titer. For this purpose, the supernatants were diluted 1:5 in NA-Star buffer and the neuraminidase activity was analyzed in triplicate by the NA-StarTM assay kit (Life Technologies, Carlsbad, CA, USA), according to the manufacturer’s instructions. The antiviral activity of the extracts was calculated with respect to virus control only.

EC₅₀ determination

Half maximal effective concentration was determined for the extract that inhibited influenza replication in more than 80%. MDCK cells were seeded and infected as described above. After 1 hpi, the viral suspension was removed and the monolayers were treated with different extract concentrations (400 − 12.5 µg/mL). Oseltamivir carboxylate (100 − 3.2 µM) (OST-car; kindly donated by Hoffman-La Roche Inc., Basel, Switzerland) was used as a control. For influenza titration, the supernatants were harvested at 48 hpi, and the influenza virus was titrated by TCID50/mL using MDCK cells (5 × 104 cells/well in 96-well plates) through the Reed and Muench method (Reed & Mounch, 1938; Sacramento et al., 2015). Non-linear regression of the dose response curves was performed to determine the 50% inhibitory effect on viral replication (EC₅₀) for the extracts and reference compound (OST-car).

Functional antiviral assay—IC₅₀ determination

For extracts with high influenza inhibition (80%) in the screening test, we performed a functional antiviral assay to determine the concentration required to inhibit 50% of the NA enzymatic activity (IC₅₀). Wild-type and resistant strains of influenza A and B were tested against different concentrations of extracts (400 − 12.5 or 200 − 6.25 µg/mL), OST-car (1,000 − 0.01 nM) was used as control. This assay was done using the NA-Star™ assay kit (Life Technologies, USA), according to the manufacturer’s instructions (Lopes e Souza et al., 2015; Souza et al., 2011; Souza et al., 2013). The IC50 was calculated using a non-linear regression.

Statistical analysis

The dose–response curves used to calculate the pharmacological parameter values were generated using Excel 2010 for Windows software (Microsoft) (Souza et al., 2007). All of the experiments were performed at least three times, and the results are displayed as the mean or the mean ± standard error (SEM).

Low cytotoxicity of the extracts

The cells viability was the first analysis conducted, extracts with high cell toxicity in vitro were disregarded from the antiviral analysis. All the extracts produced were tested at 1 mg/mL in MDCK cell cultures. As control, we exposed MDCK to DMSO in the same proportion as we used in the extracts test, 1.0% V/V in culture medium, which did not affect the cells viability. The results showed that redox mitochondrial activity of the tested cells with extracts was inhibited at most by 20% after 48 h of exposure. Approximately 60% of extracts in ethyl acetate (EtOAc) and n-butanol (n-BuOH) solvent showed 90% of cell viability (Table 2).

The extracts inhibit influenza replication in more than 80%

After cytotoxicity analysis, we tested anti-influenza activity with all the extracts produced, EtOAc and n-BuOH. For this purpose, MDCK-infected cells were treated with 200 µg/mL of each extract. After 48 hpi, the supernatants were harvested and the influenza neuraminidase activity was analyzed. We observed that 17 extracts (16 in EtOAc and 1 in n-BuOH) inhibited the replication of influenza A strains, A(H1N1)pdm09-WT and A(H3N2)-WT. Seven of them (41%) resulted in more than 80% of inhibition (Table 2). Just one n-BuOH extract from filamentous green algae was capable of inhibiting influenza replication, but at low levels. Despite this, an EtOAc extract obtained from another filamentous green alga was able to produce satisfactory inhibitory effects (Table 2). All the EtOAc extract from Desmidiaceae family (Staurastrum sp.) and four out of six EtOAc extract from Scenedesmaceae family (Scenedesmus sp. and Desmodesmus sp.) were capable of inhibiting influenza replication with different magnitudes (Table 2). Besides these, extract 21, obtained from Chlorellaceae green microalgae presented the greatest capacity to inhibit both influenza replications. However, others from the same family had limited or null capacity against this virus (Table 2).

EC₅₀ determination and SI evaluation

We determined the EC₅₀ for seven marine extracts that inhibited in more than 80% the replication of both influenza A(H1N1)pdm09-WT and A(H3N2)-WT in the screening assay. The selective index (SI) is determined by the ratio between CC₅₀ and EC₅₀, and represents the relative effectiveness of the investigational product in inhibiting viral replication compared to inducing cell death. The EC₅₀ of five extracts was less than 100 µg/mL with SI higher than 10.0 (Table 3). The oseltamivir carboxylate (OST-car) was used as positive control with 177,667 and 66,625 SI for A(H1N1)pdm09-WT and A(H3N2)-WT, respectively. The SI values between a pool of molecules (extracts) and a molecule chemically defined (OST) are incomparable.

Neuraminidase inhibition and IC₅₀ determination

We performed a functional assay to analyze neuraminidase activity with all the seven extracts that inhibited influenza infection in more than 80%. IC50 is the concentration required to inhibit 50% of the influenza NA enzymatic activity. Thus, lower IC50 values imply in higher NA inhibition activity. Two of them, extracts from the cyanobacteria Leptolyngbya sp. (5) and microalga Chlorellaceae (21) inhibited influenza A NA activity with IC₅₀ lower than 210 and 20 µg/mL, respectively (Table 4 and Figs. 1A–1D). The OST-car was used as control (Table 4 and Figs. 1E and 1F). The HPLC and mass spectrum of extracts 5 and 21 are shown S01-1 and S01-2, respectively. The mass spectrum of extract 5 (S01-3) indicated the presence of chlorinated compounds m/z = 214.9850 and 232.9946. On the other hand, the mass spectrum of extract 21 (S01-4) exhibited signals that indicated the presence of peptides, terpenes and/or alkaloids.

EC₅₀ and IC₅₀ determination to influenza A and B OST-sensitive and resistant viruses

The OST is the main antiviral used in clinics against Influenza A and B. The pressure imposed by this drug has led to the selection of resistant mutants (Dixit et al., 2013; Lopes e Souza et al., 2015; Souza et al., 2011). Thus, the search for new anti-influenza compounds that inhibits both types of influenza and their resistant strains or lineages is extremely important. Therefore, we tested extracts from Leptolyngbya sp. (5) and Chlorellaceae (21) against replication and NA activity of influenza A OST-resistant (Table 4 and Figs. 1B and 1D) and B OST-sensitive and resistant (Table 4 and Figs. 1A–1D) viruses. The OST-car was used as control (Table 4 and Figs. 1E and 1F). Both extracts inhibited influenza replication and NA activity with low EC₅₀ and IC₅₀, including for OST-resistant strains and lineages (Table 4 and Fig. 1). The raw data of the IC₅₀ calculations are provided in S02.