Prenylated flavonoid-enriched fraction from Maclura tinctoria shows biological activity against Staphylococcus aureus and protects Galleria mellonella larvae from bacterial infection

S. aureus ATCC 29213 was used as a test strain to evaluate the antibacterial activity of the plant extracts. Four field isolates of S. aureus (302/4, 469/4, 4130, and 4651), which originated from cows presenting mastitis, were also used to determine the MIC of the active fraction. The bacteria were maintained on Brain Heart Infusion agar (BHA, Fluka, Munich, GER) or Müeller-Hinton agar (MHA, Himedia, Mumbai, IND). For the antibacterial assays, bacteria were streaked onto MHA and cultured overnight at 37 °C before being used for inoculum preparation as described below. Stock cultures were stored at − 80 °C in BHI medium containing 20% glycerol.

Plant material from 23 species (Table 1) native to the Atlantic Forest were collected from a forest fragment at the coordinates 20°48′07″ S and 42°51′31″ W, which is part of the Universidade Federal de Viçosa (UFV), Minas Gerais, Brazil. The permission issued by the Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq) granted access to genetic heritage (permit No 010134/2014–0). A specimen voucher of all the samples was deposited in the VIC Herbarium of the UFV and given a voucher number (VIC 40.269). Identification was performed by Marcos Vinícius R.C. Simão with the aid of specialized bibliography, consultation with botanicals, and comparison with herborized material from the VIC Herbarium determined by specialists. The Brazilian Flora 2020 Project was used to check the species names and the botanical families. The leaves and stems were separated, dried in a ventilated oven at 40 °C for 48 h and then subjected to extraction by serial maceration using a dry vegetable/solvent ratio of 1:5 (w/v). Each sample was first extracted with a mixture of lower-polarity solvents (dichloromethane and methanol, 1: 1) followed by water, and for both extracts, the solvents were completely removed [10]. The tree species were enumerated randomly, and the extracts were identified by the abbreviations FO (organic extract of leaves), FA (aqueous extract of leaves), GO (organic extract of branches) and GA (aqueous extract of branches). Stocks of all the plant extracts were prepared at a 50 mg/mL concentration in dimethyl sulfoxide (DMSO, Vetec, Rio de Janeiro, BR) for use in the biological tests.

The antimicrobial activity was initially evaluated according to the well diffusion test. Inoculum was prepared as recommended by the Clinical & Laboratory Standards Institute (CLSI) [11]. A suspension of S. aureus ATCC 29213 corresponding to 0.5 McFarland was spread on MH agar medium. Holes of approximately 5 mm in diameter and 3 mm high were punched, and 20 μL of the extracts were added at a concentration of 50 mg/mL. DMSO and ampicillin (10 mg/mL) (A9518, Sigma-Aldrich, Dorset, UK) were used as controls. The plates were kept at 4 °C for 4 h and 37 °C for 24 h. The inhibition halos were measured in millimeters. The tests were performed on three independent biological replicates. The mean value ± SD for the replicates was calculated.

A 96-well microplate was filled with MH broth, with extract concentrations ranging from 0 to 10 mg/mL and 10⁵ CFU/mL of a suspension of S. aureus 291213. The control wells consisted of inocula added to 100 μL of the MH broth and MH broth plus DMSO in the volume corresponding to the highest concentration of tested extract. The microplates were incubated at 37 °C for 24 h. Subsequently, 5 μL of violet p-iodonitrotetrazolium (INT, I8377, Sigma-Aldrich, Dorset, UK) was added at 2 mg/mL. After 2 h of incubation, the MIC was defined as the lowest concentration of extract for which the purple color formation, which was indicative of cell viability, was not observed. The MIC of ampicillin for S. aureus 29213 was also determined. The MIC of the active fraction was estimated as described above using the reference strain and the isolates S. aureus 302/4, 469/4, 4130, and 4651. The assays were performed using three biological replicates.

The extract with the lowest MIC according to the preliminary screening was chosen for further assays. The phytochemical composition was investigated by liquid-liquid partitioning in which the bioactive extract (1.0 g) was resuspended in water (w) and partitioned into a separatory funnel in hexane (h), dichloromethane (d), ethyl acetate (ac), and n-butanol (b). A volume of 100 mL of each solvent was used in this step. The solvents were evaporated to give the 11 FO h (0.087 g), 11 FO d (0.333 g), 11 FO ac (0.017 g), 11 FO b (0.092 g), and 11 FO w (0.318 g) fractions. All the fractions were subjected to phytochemical analysis and an antibacterial activity assay. A phytochemical screening was performed by thin layer chromatography in glass plates covered with silica gel (Whatman, Kent, UK). The extract and fractions were investigated for the presence of cumarins, tannins, flavonoids, anthraquinones, essential oils, saponins, alkaloids, and triterpenes/steroids using standards and specific reagents for each class of secondary metabolites [12].

The most active fraction of the fractionated extract was subjected to analyses by LC-DAD-ESI/MS using UPLC (ACQUITY) system coupled with a photodiode array detector (Waters) and on Triple Quadrupole instrument (Waters ACQUITY® TQD) with an electrospray source (ESI) operating in the positive and negative mode. The ESI-MS/MS operation was performed using capillary voltage, 3500 V; capillary temperature, 320 °C; source voltage, 5 kV; vaporizer temperature, 320 °C; corona needle current, 5 mA; and sheath gas, nitrogen, 27 psi with argon as the collision gas, and the collision energy was set at 30 eV. Analyses were run in the full scan mode (100–1500 Da). Chromatographic separation was done on Acquity UPLC HSS C₁₈ column (50 mm × 2.1 mm i.d.; 1.7 μm; Waters, USA) in combination with an Acquity UPLC guard column (5 × 2.1 mm; 1.7 μm; Waters, USA) [13]. The mobile phase consisted of water 0.1% formic acid (solvent A) and acetonitrile 0.1% formic acid (solvent B). The elution protocol was 0–11 min, linear gradient from 5 to 95% B. The flow rate was 0.3 mL min − 1, and the sample injection volume was 4.0 μL. The UV spectra were registered from 190 to 450 nm. Phenolic compounds were identified on the basis of the typical UV absorption for each phenolic class analyzed here, in addition to the typical fragmentation patterns obtained by MS/MS in comparison to the literature data [14‐19].

The reference strain S. aureus 29213 was used to analyze the antibacterial activity of the 11FO d fraction [20]. A bacterial growth curve with an initial inoculum of 10⁶ CFU/mL was created at different MIC values, in the absence and presence of the 11FO d fraction. The aliquots were recovered at specific times, serially diluted and plated on BHI to determine the CFU/mL. The experiments were performed using two biological replicates.

S. aureus ATCC 29213 was grown in BHI medium at 37 °C for 12 h, centrifuged (2100 x g for 15 min), washed three times in phosphate buffer (5 mM, pH 6.5) and resuspended in the same buffer at approximately 10⁸ CFU/mL. The plant extract was then added at the MIC (0.04 mg/mL) for 4 h at 37 °C. Cells grown in culture media without the fraction were used as the control. After centrifugation, the pelleted cells were deposited on mica substrates for AFM (Ntegra Prima, NT-MDT, RUS). The AFM images were recorded using the intermittent contact mode and silicon probes with a 10 nm tip radius, 240 kHz resonant frequency and 11 N/m force constant [21].

Alterations in the cell permeability were evaluated by crystal violet assay [22]. In brief, a culture of S. aureus 29213 was centrifuged, and the cells were resuspended in 0.5 mM potassium phosphate buffer, pH 7.4. The cells were incubated at 37 °C for 120 min with CTAB (0.3, 30, and 120 μg/mL) or with the 11FO d fraction at different MIC values ​​(1/2 to 8X). CTAB is a cationic detergent that increases cell permeability and therefore was used as a positive control. The negative control consisted of cells incubated only in PBS. The samples were centrifuged at 9300 x g for 5 min, resuspended in PBS containing 10 μg/mL of crystal violet and incubated at 37 °C for 10 min. After further centrifugation, the supernatants were collected and read at an OD of 590 nm. The absorbance of the violet crystal solution was also checked. The percentage of crystal violet uptake was calculated as the (OD of samples / OD of crystal violet solution) × 100.

Galleria mellonella was reared on an artificial diet at 25 °C. The diet consisted of a mix of 400 g wheat bran, 200 g wheat germ, 120 g dried beer yeast, 80 mL glycerin, 200 g skim milk powder, and 300 mL liquid honey. Larvae weighing 250–350 mg were divided into six groups. Unless otherwise stated, all the experiments involved groups containing 10 larvae. Group 1 was injected with a suspension of S. aureus ATCC 29213 at a 10⁵ CFU concentration. Group 2 received the same inoculum as Group 1 but 2 h after bacterial injection 10 μL of the 11FO d fraction were injected into larvae’s last proleg. Before injection, this fraction was diluted in PBS at a dose corresponding to 200 mg/kg. Every 24 h, a new dose of the fraction was applied, and the larvae were monitored for 96 h after the last application. A total of four doses of the fraction were applied to each larva in this group. Group 3 resembled Group 2, but instead, the 11FO d fraction was replaced by gentamicin at 200 mg/kg (G1397, Sigma-Aldrich, Dorset, UK). The other groups (4–6) consisted of uninfected controls, which were treated with a single injection containing 10 μL of PBS, the 11FO d fraction, or DMSO. The larvae were kept in petri dishes in a dark environment at 37 °C. The experiments were performed using three biological replicates.

Survival data were collected using the Kaplan-Meier method, and the comparison between the control group and the treatment group was performed using the Log-rank test. A p ≤ 0.05 was considered significant. Data obtained from the agar well diffusion test and the cell permeability assay were subjected to one-way analysis of variance (ANOVA) followed by Dunnett’s test for comparison of more than two groups using PRISM 6 statistical software (San Diego, CA, USA). The results were considered statistically significant when p ≤ 0.05.

Twenty-six organic and aqueous extracts tested on S. aureus ATCC 29213 by the well diffusion test exhibited antimicrobial activity (Table 1). The largest inhibition halo was recorded for the organic extract of Miconia petropolitana leaves (24FO). The extracts from Anadenanthera peregrina led to the formation of inhibition zones ranging from 9 to 13 mm. Inhibition halos were observed when the leaf extracts from Maclura tinctoria (11FA and 11FO) were tested. Different extracts from Bathysa nicholsonii, Piptadenia gonoacantha, Tabernaemontana hystrix, Xylopia sericea, and Myrciaria glazioviana also showed activity against S. aureus. One extract from Luehea grandiflora, Ceiba speciosa, Siparuna guianensis, and Alchornea glandulosa produced inhibition halos. MIC assays were performed with all the extracts that resulted in inhibition zones and values ranged from 0.08 to 10 mg/mL (Table 1). The 11FO organic extract prepared from Maclura tinctoria leaves had the smallest MIC, and therefore, it was chosen for later assays.

Coumarins, flavonoids, essential oils, saponins, tannins and triterpenes/steroids were detected in the 11FO extract. The MICs of the fractions obtained with the solvents hexane (11FO h), ethyl acetate (11FO ac), n-butanol (11FO b), and water (11FO a) ranged from 0.04 to 10 mg/mL. The MIC obtained for the dichloromethane fraction of 11FO d was reduced to half that obtained for the crude extract (Fig. 1). The MIC of fraction 11FO d for bovine S. aureus isolates was also determined, and an antimicrobial activity between 0.01 and 0.04 mg/mL was recorded (Fig. 1). TLC showed that the fraction 11FO d consisted of a mixture of flavonoids.

The phytochemical profile of active fraction 11FO d was determined by UPLC-DAD (Fig. 2), which showed the presence of five major compounds. All compounds showed UV spectra with maximum absorption (λ_max.) near 264 and 270 nm, which are typical wavelength values for isoflavone or flavanone derivatives. Full scan mass spectra (100–1500 Da) were obtained for 11FO d by LC-DAD-ESI-MS in positive and negative modes, and their molecular ion and fragmentation analyses allowed us to confirm the structures of five prenylated flavonoids (Fig. 3). Compound 1 showed the MS spectrum in negative mode presented an [M-H]⁻ at m/z 353. Positive mode MS² focused on m/z 355, m/z 337 [M + H-18]⁺ and m/z 299, which could be attributed due to the loss of a prenyl unit (56 u). Data analysis along with literature reports on the Moraceae family suggests that compound 1 is luteone. Compound 2 was the major constituent of 11FO d and presented an [M + H]⁺ ion at m/z 339 with an abundant product ion at m/z 283 that was attributed to a loss of 56 Da, which is characteristic of isoflavone compounds. The molecular formula of this ion (C₂₀H₁₈0₅) and its fragmentation and absorption maximum (λ_max.) at 266 suggest that this compound is the prenylated isoflavone known as wightenone. Spectra obtained for peaks 3 and 4 showed the UV absorption maxima with two bands at 270 nm, which are characteristic of the flavanone derivatives. The ESI-MS for both compounds showed the same [M + H]⁺-protonated molecule at m/z 423 and [M-H]⁻-deprotonated molecule at m/z 421 with very similar retention times, suggesting the presence of isomers. These protonated molecules refer to isomers of flavanones that are attached to two prenyl substituents on their aglycone. In the negative mode, compound 3 presented fragmentation ions at m/z 367 which indicates the loss of a ring-closed prenyl group, C₄H₆. The fragment at m/z 293 [M-H-54-56-18] indicates loss of a second neutral prenyl group, C₄H₈, plus a water molecule. These data indicated that peaks 3 and 4 refer to isomers of flavanones with two prenyl substituents, one in an open chain form and the other in a closed ring form, confirming the presence of euchrestaflavanone C and cudraflavanone A (C₂₅H₂₆O₆). Compound 5 also showed spectra suggestive of a flavanone (λ_max 270 nm). The MS spectrum in positive mode was focused on a m/z 407 with an abundant product ion at m/z 351 that was attributed to a loss of 56 Da, which is characteristic of the presence of prenyl group. The spectral data suggest that compound 5 is a diprenylated flavanone.

After 4, 8, and 12 h of exposure to 2X MIC, MIC and 1/2 MIC, there was no further bacterial growth, which revealed the bactericidal effect of the fraction (Fig. 4). To explore the possible antibacterial mechanism of fraction 11FO d, the morphological changes in S. aureus were observed by AFM. No remarkable changes in the bacterial surface or cellular dimensions (1 μm × 500 nm) were apparent, although a decrease in the number of bacteria was observed after 4 h of treatment, as were changes in the cellular arrangement (Fig. 5). The alterations in cellular permeability were tested by incubating the cells with crystal violet, a hydrophobic dye that is capable of crossing the lipid bilayer (Fig. 6). In the presence of increasing concentrations of CTAB, a cationic detergent, there was greater dye entry, which is consistent with the increased permeability of cell membrane. The same trend was not observed when the cells were incubated in the presence of the 11FO d fraction, suggesting that the membrane properties were not altered by the compounds found in the fraction. Cells treated with the fraction had a lower uptake of crystal violet in relation to the cells without the added fraction (control).

The 11FO d fraction used at the 200 mg/kg concentration increased the survival of insects that were infected with S. aureus ATCC 29213 when compared to those in the control groups. After 24 h, 50 % of the larvae were dead compared to the high survival seen in the group that received the fraction. After 48 h, there were still live larvae in the group treated with 11FO d. All the larvae injected with PBS, PBS and DMSO, or PBS and extract survived after 96 h of monitoring (Fig. 7).