Schinus terebinthifolia leaf lectin (SteLL) has anti-infective action and modulates the response of Staphylococcus aureus-infected macrophages

Staphylococcus aureus is recognized as an important pathogen causing a wide spectrum of diseases. Here we examined the antimicrobial effects of the lectin isolated from leaves of Schinus terebinthifolia Raddi (SteLL) against S. aureus using in vitro assays and an infection model based on Galleria mellonella larvae. The actions of SteLL on mice macrophages and S. aureus-infected macrophages were also evaluated. SteLL at 16 µg/mL (8 × MIC) increased cell mass and DNA content of S. aureus in relation to untreated bacteria, suggesting that SteLL impairs cell division. Unlike ciprofloxacin, SteLL did not induce the expression of recA, crucial for DNA repair through SOS response. The antimicrobial action of SteLL was partially inhibited by 50 mM N-acetylglucosamine. SteLL reduced staphyloxathin production and increased ciprofloxacin activity towards S. aureus. This lectin also improved the survival of G. mellonella larvae infected with S. aureus. Furthermore, SteLL induced the release of cytokines (IL-6, IL-10, IL-17A, and TNF-α), nitric oxide and superoxide anion by macrophagens. The lectin improved the bactericidal action of macrophages towards S. aureus; while the expression of IL-17A and IFN-γ was downregulated in infected macrophages. These evidences suggest SteLL as important lead molecule in the development of anti-infective agents against S. aureus.

SteLL is a 14 kDa glycoprotein and the antimicrobial activity of this protein was reported towards both Gram-positive and Gram-negative bacteria and Candida albicans 25 . SteLL also affected the survival and nutrition of the beetle Sitophilus zeamais adults 28 . Recently, the antitumoral activity of SteLL was shown in sarcoma 180-bearing mice. The authors reported that the treatment with SteLL did not induce hematological changes nor genotoxic effects in mice, advocating for the safety of in vivo use of this lectin 29 .
This present work provides insights into the in vitro effects of SteLL on S. aureus and evaluates the phenotypic response induced by this lectin in macrophages uninfected and/or infected by S. aureus. In addition, the in vivo activity of this lectin against S. aureus is reported using Galleria mellonella larvae (Lepidoptera: Pyralidae) as infection model.

Results
SteLL induced changes in the cell size/DnA content ratio of S. aureus. As previously reported 25 , SteLL was able to inhibit the growth of all S. aureus strains tested (S. aureus 8325-4, S. aureus ATCC 6538, S. aureus ATCC 29312) with a MIC (minimum inhibitory concentration) of 2 µg/mL. A flow cytometry-based assay was performed to assess the effects of SteLL on cell size (seen by mean light scattering detected in LS1) and DNA content (seen by the fluorescence intensity in FL2 channel) of S. aureus. Ciprofloxacin (inhibitor of DNA replication) and chloramphenicol (inhibitor of protein synthesis) were used as controls in this assay (Table 1 and Fig. 1).
The exposure of S. aureus to ciprofloxacin at MIC resulted in significantly increase in cell mass, while the DNA content decreased when compared with untreated cells. This resulted in a decrease in the cellular DNA concentration (DNA/mass ratio = 0.41) compared to untreated bacteria (DNA/mass ratio = 0.71). Bacteria treated with chloramphenicol did not alter cell size as cell division requires protein synthesis and hence was blocked. In these bacteria, the cellular DNA content increased (DNA/mass ratio = 1.20), which is consistent with a DNA replication arrest specifically at the level of initiation. Consequently, runout chromosome synthesis was observed and the bacterial cells end up with integral numbers of fully replicated chromosomes (Fig. 1C).
Cells treated with SteLL did not follow any of these profiles (Fig. 1D,E). A significant increase (31%) on cell size was observed in bacteria treated with SteLL at 8 × MIC (16 µg/mL), this was accompanied by a similar increase in DNA content (39%), and consequently the DNA concentration remained almost unchanged in these cells (DNA/mass ratio = 0.75) ( Table 1). The effects of SteLL on S. aureus cell size were also confirmed by fluorescence microscopy ( Fig. 2), where cells treated with the lectin (Fig. 2C,D) appeared with increased size when compared with control cells (Fig. 2A).
We also evaluated whether sub-inhibitory concentration of SteLL (0.5 × MIC) had any effect on the SOS response, a pathway associated with acquisition of drug resistance and virulence phenotypes. After 3 h of treatment, recA transcription determined from a recA-lacZ transcriptional fusion was not altered by SteLL (Fig. 3). On the other hand, ciprofloxacin (a DNA damaging agent) induced more than 5-fold increase in recA transcription. These results revealed that SteLL, unlike ciprofloxacin, did not affect DNA integrity and thus did not induce any SOS-related mutagenic pathways.
SteLL increased ciprofloxacin activity against S. aureus. Checkerboard assays were performed to evaluate the interaction of SteLL with two selected antibiotics: ciprofloxacin and ampicillin. The MIC values for these two drugs towards S. aureus 8325-4 were 0.78 µg/mL and 25 µg/mL, respectively. A synergistic activity was observed when combining SteLL with ciprofloxacin (ΣFIC: 0.47); while an additive effect was found between SteLL and ampicillin (ΣFIC: 0.53).
When SteLL was combined at 2 × MIC, it did not induce any effect on ciprofloxacin (2 × MIC) bactericidal action (ΔLC were less than 1 log CFU/mL when compared with ciprofloxacin-treated cells). On the other SteLL inhibits staphyloxanthin production. Next, we evaluated whether sub-inhibitory SteLL concentrations (0.065 × MIC, 0.125 × MIC, 0.25 × MIC and 0.5 × MIC) could inhibit the production of staphyloxanthin, the golden pigment of S. aureus ATCC 29312. SteLL induced a dose-dependent reduction in staphyloxanthin production relative to untreated bacteria (Fig. 5A). The reductions ranged from 48.78% to 82.88% (Fig. 5B). The treatment with the highest tested concentration of SteLL (0.5 × MIC) reduced the staphyloxanthin content to around 18% (in comparation to untreated bacteria), resulting in almost colorless S. aureus cells.
SteLL increased the release of nitric oxide and superoxide by mice macrophages. The macrophage viability was not affected by any tested SteLL concentrations (2-16 µg/mL) (data not shown). On the other hand, the treatment of macrophages with different concentrations of this lectin resulted in a significant increase in NO production in relation to untreated cells (Fig. 6A). Maximum NO production was observed in the presence of SteLL at 8 µg/mL and 16 µg/mL. At these concentrations, the levels of NO were similar (p > 0.05) to those produced by M1 macrophages.
The induction of mitochondrial superoxide by SteLL was evaluated using the MitoSOX fluorescent probe. The macrophages were incubated for 30 min with SteLL at 8 µg/mL and 16 µg/mL. As shown in Fig. 6B, only the treatment with SteLL at 16 µg/mL significantly enhanced superoxide production by the macrophages (increase of 40%).    www.nature.com/scientificreports www.nature.com/scientificreports/ indicated that the reduction of bacteria in supernatant of SteLL-treated macrophages is related with the increased release of reactive species.

SteLL modulated cytokine release by uninfected macrophages or Staphylococcus aureus-uninfected macrophages.
We selected the most active concentration of SteLL (16 µg/mL) to evaluate its influence on cytokine release pattern of macrophages uninfected and infected with S. aureus. For uninfected macrophages, the treatment with this dose resulted in a significant enhancement of IL-6, IL-10, IL-17A, and TNF-α compared to untreated cells (p > 0.05) (Fig. 7A,B,C,E). The levels of IFN-γ were also higher in the supernant of SteLL-treated cells, although no statitical differences were found when compared with the values obtained for control cells (Fig. 7D). The lectin did not influence the levels of IL-4 and IL-12 (data not shown).
The macrophages infected with S. aureus expressed high levels of IL-6, IL-10, IL-17A, and TNF-α (Fig. 7). The roles of theses cytokines in pathogenesis of S. aureus have been described previously [30][31][32] . The level of IFN-γ was also increased by S. aureus infection (about 2-fold). SteLL was able to downregulate the expression of IL-17A (Fig. 7B) and IFN-γ (Fig. 7D) by infected macrophages, while the levels of the other tested cytokines were also  www.nature.com/scientificreports www.nature.com/scientificreports/ reduced albeit not in a significant manner. It is important to highlight that SteLL did not totally inhibit the release of IL-17A and IFN-γ; in fact, their levels were reduced to approximately what was found in the supernatant of control cells.

SteLL protected Galleria mellonella larvae against Staphylococcus aureus infection. To finally
show the treatment efficacy of SteLL, we employed an infection assay using G. mellonella larvae. This model has been widely used to study microbial pathogenesis 33 and to assess the in vivo activity of antimicrobial agents 34,35 . Uninfected larvae inoculated with PBS or SteLL exhibited similar survival curves (p > 0.05). On the other hand, infection with S. aureus 8325-4 reduced the larval viability by 30%, 80% and 100%, on day 1, 2 and 3, respectively. The median survival of this group was 2 days.
The single-dose treatment using 0.2 mg/kg SteLL (corresponding to the administration of 10 µL of a SteLL solution at 2 × MIC) increased the survival of S. aureus-infected insects. For this group, the median survival was not possible to be defined, and only a 30% reduction in survival was recorded after 3 days (Fig. 8A). The survival curves of SteLL-treated larvae and S. aureus-infected group were significantly different (p < 0.05).
Next, we evaluated the number of S. aureus colonies in the hemolymph of larvae. The untreated larvae infected with S. aureus exhibited increased levels of bacteria in hemolymph during the experiment (Fig. 8B). The www.nature.com/scientificreports www.nature.com/scientificreports/ treatment with SteLL was able to significantly inhibit the bacteria growth (p < 0.05). During the first two days the bacterial load in SteLL-treated animals remained the same as at time of inoculation and only a small increase was observed in the third day (about 1 log CFU/mL). This effect is in accordance to the effect of SteLL in the time kill assay (Fig. 3). In this sense, SteLL treatment reduced the S. aureus proliferation in larvae hemolymph resulting in increased animal survival.

Discussion
In this study, we evaluated the in vitro effects of SteLL on S. aureus and the responses induced by this protein in two models of infections (using macrophages and G. mellonella). SteLL is purified from leaves of S. terebinthifolia, a medicinal plant widely used in Northeastern Brazilian (where it is popularly known as "Aroeira da praia") to treat skin wounds and inflammation 36 . Besides SteLL, other products derived from S. terebinthifolia have exhibited antimicrobial activity such as essential oils and some purified compounds [37][38][39] .
We employed a flow cytometry-based method to analyze whether SteLL could induce any effect on cell size and DNA content 40 . Proper cell cycle control is essential to ensure the generation of two identical daughter cells as result of cell division [41][42][43] . Perturbations in cell cycle regulation are therefore deleterious for bacterial proliferation; and thus the proteins involved in this pathway constitute potential targets for drug action 44 . We observed that SteLL induced significant increases in both the cell size and DNA content after 3 hours of incubation. These effects were confirmed by fluorescence microscopy and may indicate impairment of cell division (since both DNA and cell mass are increased), a kind of response that has been described for other antimicrobial agents such as targocil, a cell wall stressor 45 .
The ability to bind N-acetylglucosamine residues has been associated with the bacteriostatic properties of chitin-binding lectins (such as SteLL) 25,26,46 , since high amount of peptidoglycan in S. aureus cell wall provides multiples targets for interactions. We demonstrated that 50 mM NAG partially inhibited the antimicrobial action of SteLL. Thus, it is possible to hypothesize that SteLL binds N-acetylglucosamine residues present in cell wall and disturbs the process of cell division.
In addition, SteLL treatment did not induce the expression of recA which is instrumental in triggering the SOS response. RecA detects ssDNA generated by DNA degradation or inhibition of DNA replication, stimulates LexA autocleavage which in turn leads to derepression of a number of LexA regulated genes. These genes encode enzymes involved in DNA repair and mutagenesis (reviewed by Simmons et al. 47 ). Taken together, these findings indicated that SteLL may inhibit bacterial growth by impairing division without affecting DNA structure.
Since SteLL inhibits the growth of S. aureus, we evaluated whether this lectin could affect the activity of antibiotics in clinical use. The combinatory effects of SteLL and the antibiotics ciprofloxacin or ampicillin were assessed using checkboard and time-kill experiments. Initially, we found that this lectin improved the action of ampicillin (β-lactam) and ciprofloxacin (quinolone) through additive and synergistic effects, respectively. SteLL at 16 µg/mL (8 × MIC) could also increase the bactericidal properties of ciprofloxacin, however, this action was only observed during the first hours. The synergistic interactions with antibiotics have been reported only for few plant lectins, including the lectins extracted from Alpinia purpurata (ApuL) 24 and Vatairea macrocarpa (VML) 48 .
Another effect of SteLL on S. aureus physiology is the inhibition of staphyloxanthin production, a carotenoid pigment encoded by the crtOPQMN operon. Staphyloxanthin has been associated with the protection against oxidant attack promoted by immune cells 49 , which brought a new light in the use of this pigment as target for drug development 34,50 . Some plant derived compounds have inhibitory effects on staphyloxanthin 34,51,52 , however, this is the first report of a similar action for a plant lectin.
Given the well known ability of plant lectins to alter the phenotic responses of immune cells, we examined the effects of SteLL on macrophages and S. aureus-infected macrophages. The results showed that SteLL induced the the release of cytokines (IL-6, IL-10, IL-17A, and TNF-α) and reactive species (nitric oxide, superoxide anion) by uninfected macrophagens. The ability of plant lectins to alter the macrophage responses have been demonstrated www.nature.com/scientificreports www.nature.com/scientificreports/ by other authors 18,19,53 . The effects of SteLL on nitric oxide and superoxide anion production may be related to the improvement of macrophages bactericidal action, since these reactive species play essential roles in the defense against infectious diseases 54,55 .
Importantly, SteLL modulated the expression of two proinflammatory cytokines (IL-17A and IFN-γ) in infected macrophages. In the context of S. aureus infectious, IL-17A is essential for antimicrobial peptides production and bacterial clearance 56 , while IFN-γ increase the macrophages response against S. aureus 57 . Intriguingly, the overproduction of both cytokines can also exacerbate the severity of some infections 14,58 . For example, S. aureus phenol-soluble modulins that induce high levels of IL-17 leading to skin inflammatory response 13 .
The excess of IFN-γ is also related to the harmful inflammation state associated with the damage of essential organs (such as liver and kidney) 59 . Recent evidences suggested that IFN-γ favored the outgrowth of S. aureus 58 . Thus, inhibition of the expression of both IFN-γ and IL-17A has been reported as a beneficial effect in several models of S. aureus-induced infection 59,60 .
Although the molecular mechanism underlying the suppression of IL-17A and IFN-γ induced by lectin treatment in macrophages infected with S. aureus remains to be elucidated, it is possible that decreased levels of these cytokines help to attenuate the deleterious effects of persistent inflammatory responses at the site of infection. Furthermore, because the non-infected cells treated with SteLL presented different cytokine profiles than those found in the S. aureus-infected cells treated with SteLL, it is likely that bacterial components also contribute to modulation of the macrophage response.
These paradoxical effects on the production of inflammatory mediators when comparing uninfected and infected hosts have been observed for other plant lectins 16,18,19 . For instance, the lectin isolated from Cratylia mollis (Cramoll) has been described as pro-inflammatory agent in in vitro and in vivo models 18,61-63 ; however the treatment with this protein was shown to reduce the release of cytokines (such as TNF-α, IL-6) in experimental models of infection induced by S. aureus (using peritoneal cells) 18 and Cryptococcus gatti (using mice) 64 .
Similarly, the lectin from Canavalia brasiliensis (ConBr) induced different responses in Salmonella enteritidis-infected and uninfected macrophages. The exposition of uninfected macrophages to ConBr resulted in high levels of mRNA transcripts for IL-6 (in relation to untreated and uninfected macrophages). However, S. enteritidis-infected macrophages treated with this lectin exhibited lower levels of IL-6 gene transcription when compared to untreated S. enteritidis-infected cells. ConBr treatment also suppressed the transcription of IL-10 gene in macrophages infected with S. enteritidis 19 . For SteLL purification, the powder from dried leaves (20 g) was suspended in a saline solution (0.15 M NaCl) and submitted to agitation at 4 °C. After 16 h, the filtered extract was centrifuged (3000 × g for 15 min) and then submitted to chitin column (Sigma-Aldrich, MO, USA). The elution was performed using acetic acid (1 M) and SteLL was obtained after dialysis (10 kDa cut-off membrane; Sigma-Aldrich) against distilled water (4 h, 4 °C) and in sequence against 0.15 M NaCl (4 h, 4 °C). The protein concentration was determined according to Lowry et al. using a standard curve of bovine serum albumin (31.25-500 μg/mL) 67 .

Antibacterial activity and combinatory effects with antibiotics. MIC determination.
The antimicrobial activity of SteLL was confirmed by determination of the Minimum Inhibitory Concentration (MIC) against S. aureus strains (S. aureus 8325-4, S. aureus ATCC 6538 and S. aureus ATCC 25923) using broth microdilution assay 25 . Briefly, serial dilutions of SteLL were prepared in 96-wells plates containing Luria-Bertani (LB) broth to obtain concentrations ranging from 128 to 0.25 µg/mL. Following, each well received 10 μL of a microbial suspension (resulting in a bacterial load of approximately 1.0 × 10 7 CFU/mL for each well). Bacterial growth was detected measuring the optical density at 600 nm (OD 600 ).
The antimicrobial action of SteLL was also evaluated in the presence of NAG, to evaluate the participation of carbohydrate-binding domain. For this, SteLL (16 µg/mL) was pre-incubated with 50 mM NAG. After 1 h, the bacteria were added as described for MIC determination. The bacterial growth was determined after 3 h, 6 h and 9 h of incubation. The inhibition index (IN%) was calculated using the following equation: where "L" is the concentration (µg/mL) of SteLL in a given well, and MIC L represents the control MIC of SteLL alone. "D" is the concentration of the tested drug in a given well, and MIC D represents the control MIC of the tested drug alone. FICI mean (ΣFIC) is derived by averaging the FICI values along the growth-no growth interface. Data interpretation: ΣFIC ≤ 0.5: synergism (syn); 0.5 < ΣFIC ≤ 1: addition (add); 1 < ΣFIC < 4: noninteraction (non); ΣFIC ≥ 4: antagonism (ant) 68 . After overnight incubation at 37 °C, the tubes were centrifuged (9000 × g for 10 min), suspended with 1 mL of phosphate-buffered saline (PBS) and re-centrifuged. Bacteria cells were then photographed. Next, an assay to quantify carotenoid pigments (including staphyloxanthin) was performed. For this, each pellet was resuspended in methanol (0.2 mL) and incubated for 3 min at 55 °C. The methanol phase (supernatant) and cell debris were separated by centrifugation (9000 × g for 10 min) and the pellets were submitted to entire pigment extraction procedure three more times. Finally, the absorbance of methanol extract was determined at 465 nm 34 . Assays with mice peritoneal macrophages. Isolation of mice peritoneal macrophages. Peritoneal macrophages were obtained from inbred strains of C57BL/6 mice of both sexes at 8-10 weeks of age. Exudate cells were harvested by peritoneal lavage using 10 mL of ice-cold sterile phosphate-buffered saline (PBS) (pH 7.2). After centrifugation at 120 × g for 5 min, the cell pellets were suspended in RPMI-1640 medium supplemented with bovine calf serum (10%; v/v), penicillin and streptomycin (100 U/mL) (all from Sigma-Aldrich). For all assays, macrophages (1 × 10 6 cells/mL) were cultured in 24-or 96-well plates, and non-adherent cells were removed. All animal experiments were performed according to the ethical standards of the CEUMA University and were approved by the ethics committee for animal experimentation of this institution (CEUA-CEUMA) (Protocol of Approval N° 107/14), which follows the principles of care with laboratory animals.
Determination of nitric oxide (NO) production and cell viability. For both assays, macrophages (1 × 10 6 cells/mL) were seeded in 96-well plates for 24 h at 37 °C and 5% CO 2 . The cells were then treated with SteLL (2, 4, 8 and 16 µg/mL) for another 24 h. Next, the supernatant was used for determination of NO production, and the adherent cells were assessed by the MTT assay (below). Untreated cells were used as negative control. LPS (Escherichia coli; 2000 ng/mL; Sigma-Aldrich) + INF-γ (100 ng/mL; BD Pharmingen) were used as inductors of macrophages activation (M1 macrophages), while IL-4 (400 ng/mL; BD Pharmingen) + IL-13 (400 ng/mL; BD Pharmingen) were used for induction of alternative activation (M2 macrophages). The assays were performed following the protocols described below in quadruplicate in two independent experiments. The results are expressed as the mean ± standard deviation (S.D.). www.nature.com/scientificreports www.nature.com/scientificreports/ • NO production: The measurement of NO production by peritoneal macrophages was determined using the Griess assay. Briefly, a 50 μL sample from the supernatant of each well was mixed with 50 μL of Griess reagent in a 96-wells plate. After incubation for 15 min at room temperature, the optical density was determined at 540 nm with a microplate reader (Benchmark Plus, Bio-Rad, CA, US). The nitrite concentration (μmol/10 6 cells) was quantified by extrapolation from a sodium nitrite standard curve for each experiment. • MTT assay: Cell viability was evaluated using the MTT assay, which measures the metabolic conversion of 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) salt to colored formazan dye. At the end of the incubation period, the medium was removed, and fresh RPMI medium containing 5 mg/mL MTT solution was added, and the sample was incubated for 3 h. Subsequently, the medium was removed, and the intracellular formazan product was dissolved in DMSO. The optical density (OD) was measured at 595 nm. Cell viability was expressed as the % of viable cells compared to the control.

Scientific RepoRtS
Mitochondrial superoxide production. Mitochondrial superoxide anion production by the macrophages was evaluated using a MitoSOX ™ Red Mitochondrial Superoxide Indicator (Molecular Probes Bacterial killing assay. The effects of SteLL on the bactericidal effect of peritoneal macrophages were evaluated towards intracellular and extracellular bacteria. To quantify extracellular bacteria, aliquots of 4 µL of 10-fold-diluted suspensions from cell supernatants were added to agar plates. For the measurement of intracellular bacteria, the supernatants were removed and each well was washed 5× with ice-cold PBS containing trypan blue in order to remove extracellular bacteria. Following, the cells lysed after washing in 0.1 mL sterile water 71 . The cell lysates were 10-fold-diluted and added in agar plates. All plates were incubated at 37 °C for 24 h. After this period, the colonies were counted for the calculation of CFU/mL. In vivo infection model with Galleria mellonella. Survival assay. G. mellonella larvae (~200 mg) were randomly distributed in three experimental groups (n = 10). Two groups were infected by injection of 10 μL of a fresh S. aureus 8325-4 suspension (1 × 10 5 CFU/mL) into the last left proleg, followed by incubation at 37 °C. After 2 h, one group of animals received 10 μL of 4 μg/mL SteLL (2 × MIC) dissolved in PBS (resulting in a dose of 0.2 mg/kg). The second group of animals was treated with PBS. The larvae were incubated at 37 °C, and the larval viability was determined daily for 4 days.
Quantification of S. aureus in G. mellonella hemolymph. G. mellonella larvae were infected with S. aureus and treated as described above. Each day, a total of 5 larvae were cut in the cephalocaudal direction with a scalpel blade and squeezed to remove the hemolymph. Each sample was 10-fold-diluted in PBS and 4 µL was plated on LB agar. After 24 h-incubation at 37 °C, the colonies were enumerated, and the results were expressed as CFU/mL.

Statistics analysis.
All experiments were performed in quadruplicates and in at least two independent assays. Plotting of data was performed using GraphPad Prism 5. Experiments with p < 0.05 were considered significant and are stated in the results section. The survival plots for in vivo infection were performed using Kaplan-Meier analysis on pooled data for repetitive experiments. Statistical analysis was carried out with log-rank (Mantel-Cox) test for comparison of survival curves. Experiments with p < 0.05 were considered significant and are stated in the results section.
conclusion Altogether, the findings of this study suggest that the SteLL impairs cell division of S. aureus without provoking DNA damage. The lectin also alters bacteria metabolism resulting in a reduced staphyloxanthin production. These effects may be responsible for the anti-infective activity of SteLL in G. mellonella. Furthermore, we demonstrate that SteLL increase the production of cytokines by uninfected macrophages and modulate the release of IL-17A and IFN-γ in S. aureus-infected macrophages. Taken together, these findings from both in vitro and in vivo studies suggest SteLL as a promising lead for the development of new anti-infective agents against S. aureus.