Eugenia aurata and Eugenia punicifolia HBK inhibit inflammatory response by reducing neutrophil adhesion, degranulation and NET release

Background Eugenia spp. are used in popular medicine in the treatment of pain, diabetes, intestinal disorders and cough. The aim of the work is to evaluate, ex vivo and in vivo, the anti-inflammatory activity of the hydroethanolic extracts of the leaves of Eugenia aurata (EA) and Eugenia punicifolia HBK (EP) upon neutrophils. Methods Ex vivo, isolated human neutrophils were sensitized by Eugenia extracts (0.1–1000 μg/mL) and stimulated by PMA. In these conditions, different neutrophil activities related to inflammatory process were measured: adhesion, degranulation and NET release. Neutrophil viability and tumor line cells were monitored. In vivo, neutrophil influx was evaluated by peritonitis model performed in mice pretreated with different concentrations of Eugenia extracts. Phytochemical profile was assessed by mass spectrometry. Results Ex vivo, EA and EP (1000 μg/mL) reduced cell adhesion and degranulation, respectively. NET release was inhibited by EA and EP. Anti-inflammatory activities occurred in the absence of cytotoxicity. In vivo, both EA as EP inhibited neutrophil migration. The phytochemical profile revealed that EA contains myricitrin, rutin, quinic acid and quercetin derivatives. EP presents gallic acid, quercetin derivatives, syringic acid, ellagic acid, monogalloyl-glucose, glycosyringic acid, mudanoside B, HHDP glucose isomer and digalloylglucose isomer. EA and EP inhibit neutrophil migration by different pathways. Conclusion Different chemical compositions may explain the anti-inflammatory effects described herein for EA and EP. Both extracts inhibit NET release but only EA reduces cell adhesion whereas EP decreases elastase secretion. This work contributes to the elucidation of cellular mechanisms related to the anti-inflammatory activity for leaves of E. aurata and E. punicifolia HBK.


Background
Inflammation is a process that includes a complex immune response, which occurs in several steps and may be caused by chemical, physical, microbiological and immunological stimuli. It involves leukocyte recruitment where the first leukocytes to be recruited and act on the inflamed tissue are neutrophils. Neutrophils have been considered a target for pharmacological intervention given their abilities to kill microorganisms, to begin and amplify the inflammatory process. Neutrophil recruitment and inflammatory activities require a complex sequence of events, including cell adhesion, degranulation, and more recently, neutrophil extracellular traps (NET) release [1]. The control of the inflammatory process is critical because of the associated risks: tissue damage, loss of organ performance and failure.
Eugenia genus with over 500 species, of which about 400 are in Brazil, assumes prominence in popular medicine, mainly for their anti-inflammatory activities in the treatment of wounds and infections [2,3]. Flavonoids, tannins, terpenoids and essentials oils were isolated from this genus [4,5]. Different crude extracts of Eugenia show several medicinal effects, such as anti-inflammatory [6], antifungal [7], neurological [8], antimicrobial [9], among others. Leaves of Eugenia punicifolia are popularly used to treat inflammation [10], diabetes [6,10], fever and flu [11,12]. E. aurata is an endangered species [13] with low studies in the literature and, by analogy, there is a need of registering its effects on inflammatory processes.
Although leaves of Eugenia species are widely used in popular medicine for inflammatory diseases, efficacy of cellular and molecular mechanisms remains elusive. Our aim was to evaluate the cellular mechanisms involved in the anti-inflammatory activity of Eugenia aurata and Eugenia punicifolia. For this purpose, ex vivo assays were performed and the anti-inflammatory activity was confirmed using in vivo assays.

Animals
Swiss male mice, weighing on average 40 grams, were kept in controlled temperature rooms (23-25°C) with access to food and water. All animal care and experimental procedures were performed in accordance with the internationally accepted principles for laboratory animal use [14].

Human neutrophils isolation
Human neutrophils were isolated and prepared according to previous methods described [15,16]. Cells were suspended in Hank´s balanced salt solution (HANKS) (Sigma, St. Louis, MO, USA) containing 0.1 % gelatin (w/v) (HANKS-gel) with over 90 % viability as determined by the Trypan blue (Sigma) exclusion test, and 90-95 % of cells were found to be neutrophils.

Cell viability (MTT assay)
Cytotoxicity was evaluated by the colorimetric method of MTT (3-(4,5-dimethylthiazol-2-yl) 2,5-Diphenyl Tetrazolium bromide) (Sigma), which consists of indirectly measuring of cell viability by mitochondrial enzyme activity of living cells. Human neutrophils (2 x 10 5 / well) seeded into 96 well culture plates were incubated with different concentrations of Eugenia spp. during 1 h at 37°C. Subsequently, MTT (1 mg/mL) was added to each well and incubated at 37°C for 4 h. After incubation, formazan crystals were diluted by addition of Dimethyl Sulfoxide (DMSO, Sigma) and the optical density (O.D.) of samples measured in a spectrophotometer at 570 nm. Neutrophils incubated either with RPMI-1640 (Sigma) or 50 μM H 2 O 2 [17] were used as negative and positive control (100 % viable) to cell death, respectively.

Cell adhesion
Cell adhesion assays were performed in 96 well micro plates. Human neutrophils (4 x 10 5 ) suspended in RPMI medium (Sigma) plus 5 % Fetal Bovine Serum (FBS) (Vitrocell, Campinas, SP, Brazil) were added to wells of a micro plate containing different concentrations of Eugenia spp. After 15 min, cells were then stimulated by Phorbol Myristate Acetate (PMA 25nM) (Sigma) for 1 h at 37°C. Non-adherent cells were removed and adherent cells were made evident via a colorimetric test with Bicinchoninic Acid (BCA; Pierce). Sample absorbance was measured in a Multiskan FC (Thermo Scientific, Waltham, MA, USA) reader at 560 nm.

Elastase activity
Elastase enzyme activity upon degranulation was assessed as follows: Neutrophils (2 x 10 5 ) suspended in Hank's solution were incubated for 30 min in the presence of different Eugenia spp. concentrations then stimulated with PMA (25nM) for 3 h at 37°C. Succeeding incubation, neutrophils were centrifuged (437 x g, 5 min) and the resulting supernatants incubated in 1 mM elastase substrate (N-Methoxysuccinyl-Ala-Ala-Pro-Val p-nitroanilide) (Sigma) for 30 min. After incubation, color reaction was measured at 405 nm on microplate reader Multiskan FC (Thermo Scientific). Several concentrations of purified elastase enzyme from human neutrophils (EMD Chemicals Inc., Billerica, MA, USA) were used as standards.

Neutrophil Extracellular Traps (NETs) release
Human neutrophils (2 x 10 5 ) were incubated with different concentrations of Eugenia spp. during 30 min and then stimulated with PMA (50 nM) for 4 h at 37°C. NETs generated by activated neutrophils were digested with 500 mU/mL micrococcal nuclease (MNase, Worthington Biochemical Corp.) [18]. The nuclease activity was ceased by 5 mM Ethylene Diamine Tetra Acetic Acid (EDTA) and the culture supernatant collected and stored at 4°C until the moment of quantification. NETs were quantified using the PicoGreen dsDNA kit (Invitrogen) according to the manufacturer's recommendations.

Peritonitis model (in vivo)
Mice received 1 mL of 3 % Thioglycolate injected intraperitoneally one hour after plant extract subcutaneous administration (3-300 mg/kg) [19,20]. Six hours later, mice were euthanized by cervical dislocation. The cells were immediately harvested with 5 mL PBS containing heparin (5 IU/mL). Total counts of harvested cells were performed in a Neubauer chamber. Differential counts were made on smears stained using Panoptic Fast Stain (LaborClin, Siqueira Campos, PR, Brazil). The results were reported as the number of neutrophils per mL of cavity wash. The control groups animals received: (1) subcutaneous and intraperitoneal PBS injection; (2) subcutaneous PBS and intraperitoneal Thioglycolate injection; (3) subcutaneous Dexamethasone (0.5 mg/kg) and intraperitoneal Thioglycolate injection.

Phenolic content
Phenolic content was performed as previously described [23] with minor adaptations. Briefly, 2.5 mL Folin-Ciocalteau 10 % (v/v) and 2.0 mL 4 % (m/v) sodium carbonate were added to a 0.5 mL extract in ethanol solution (1 mg/mL). After a 2 h incubation in the dark, at room temperature, absorbance was measured at 750 nm and results were expressed as equivalent milligrams of gallic acid per gram of sample. All tests were performed in triplicate.

Flavonoid content
Flavonoid content was performed as previously described [23] with minor adaptations. Briefly, 1.5 mL ethanol, 0.1 mL potassium acetate (1 M) and 2.8 mL distilled water were added to a 0.5 mL extract in ethanol solution (1 mg/mL). After 30 min incubating in the dark at room temperature, absorbance was measured at 425 nm and results were expressed as equivalent milligrams of quercetin acid per gram of sample. All tests were performed in triplicate.

ESI-MS n analysis
The mass spectrometry experiments were performed on a 6550 iFunnel Q-TOF (Agilent Technologies). The studied matrix was analyzed by Dual Agilent Jet Stream ESI (Dual-AJS-ESI) (ESI) and fragmented in the MS/MS collision cell. The negative mode was selected for the generation and analysis of first order mass spectra (MS) and the remaining multistage experiments under the following conditions: Gas Temp at 290°C, Drying Gas flow at 11 Lmin −1 , Nebulizer at 45 psi, Sheath gas temp at 350°C, Sheath gas flow 12 Lmin −1 , VCap 3000, Nozzle voltage 320 V, Fragmentor 100 V, OCT 1 RFVpp 750 V, and collision energy 35 V Agilent MassHunter Qualitative Analysis software version B.06.00 used for data acquisition and processing.

Statistical analysis
Experimental data was evaluated by variance analysis (one-way ANOVA) followed by Bonferroni test. A significance level of 5 % was adopted. All assays were performed in triplicate at least in three independent assays.

Results and discussion
Inflammatory process involves cellular and molecular events that begin with neutrophil recruitment. This process is commonly separated in four steps: rolling, adhesion, transmigration and degranulation. The rolling is mediated by the interaction between neutrophil selectins (selectin-L) and endothelium selectin (selectins P and E). Sequentially, the adhesion happens by the interaction between endothelium and neutrophil integrins as well as by mobilization of neutrophil secretory vesicles. After that, the transmigration -or diapedesisalso occurs under effect of integrins interactions and counting on the help of release of neutrophil tertiary granules to digest endothelium basal membrane. Finally, the proteins released from neutrophil secondary/primary granules can be associated with DNA and oxidative enzymes, also called NETs. The release of NETs improves microbe capture, as well as increases phagocytosis efficiency [1,24].
The cellular mechanisms related to the anti-inflammatory activity of E. punicifolia (EP) and E. aurata (EA) were evaluated. Therefore, some neutrophil functions were analysed ex vivo and others in vivo because neutrophils are the first inflammatory cells to be recruited to the damaged tissue. Both Eugenia extracts were able to inhibit neutrophils responses, by different pathways, under clear evidence of no toxicity for the cells.
Ex vivo assays aimed the evaluation of the role of EA and EP in adhesion, elastase secretion and NET release. In the first step, adhesion, neutrophils incubated with 25 nM PMA and adhered to the culture plate were considered as 100 % adhesion (Fig. 1). Neutrophils incubated only with culture medium RPMI-1640 showed basal adhesion rate (~50 %). Pretreatment of neutrophils with several EP concentrations did not alter their ability to adhere under PMA stimulus. On the other hand, pretreatment with EA 1000 μg/mL significantly reduced cell adhesion. EP (1000 μg/mL) was the only concentration able to induce a weak neutrophil adhesion when neutrophils were incubated with extracts alone (data not shown), although this effect has not been statistically significant.
The second step was to evaluate elastase secretion, indirectly assessed through elastase enzyme activity. Supernatants of cultures of neutrophils stimulated by PMA (25nM) and previously sensitized by plant extracts were analyzed. The release of elastase induced by PMA was considered as100%. Neutrophils incubated with medium alone showed significant reduction (Fig. 2). The presence of EA did not alter elastase secretion at any tested concentration . However, EP 1000 μg/mL significantly reduced PMA induced elastase secretion. None of the extract concentrations in test was able to induce elastase secretion (data not shown).
Next inflammatory event evaluated was the release of NETs. The presence of both plant extracts studied here inhibited significantly the release of NETs induced by PMA (50nM) for all tested concentrations (0.1-1000 μg/ mL) (Fig. 3). The positive control (50nM PMA) induced DNA release of 230 ng/ml while negative control (Medium) showed~50 ng/mL. The inhibition of this event may have a suppressive effect on inflammation, activation of neutrophils and capture/elimination of pathogens by decreasing the inflammatory stimulus that comes from genetic material released. Furthermore, the  presence of extracellular DNA (NETs) has been appointed as a direct source of stimulus to inflammatory and autoimmune diseases [1].
Neutrophil viability was monitored during all ex vivo assays. This is an important factor once some inhibitory effects could be related to cytotoxicity. The viability was assessed by MTT assay. The data revealed that the extracts, evaluated in different concentrations (0.1-1000 ug/mL), did not reduce neutrophil viability when these cells were sensitized either with EP or with EA (Fig. 4). As positive and negative controls, H 2 O 2 and RPMI medium were used. H 2 O 2 (50 μM) reduced 70 % of cell viability.
Since no cytotoxic effects were observed in leukocytes, other cytotoxicity assays were performed with E. punicifolia hydroethanolic extract solutions (HEEP) as well as with E. aurata hydroethanolic extract solutions (HEEA), using tumor and not tumor cell lines. As others [25], HEEP and HEEA (mean TGI > 141 μg/mL and 198 μg/mL, respectively) did not exhibit a cytotoxicity against all tested cell lines. HEEP presented a moderate selective activity effect against cell line K562 (leukemia, 12.9 ± 7.19 μg/mL) and weak for MCF-7 (mammary, 39.0 ± 5.80 μg/mL). For all the analyzed lineages, including normal cell VERO (Green monkey kidney), HEEP was inactive (Table 1). HEEA was inactive for the tested lineages, including normal cell VERO (Green monkey kidney).
According to ex vivo assays, E.aurata and E.punicifolia inhibit neutrophil functions in the absence of cell death. Based on these results, in vivo tests were performed. The in vivo experiment comprises a greater complexity of events when compared to in vitro and ex vivo experiments, in which experimental conditions are better monitored. Therefore, the results obtained using in vivo analyses are closer to the real scenario.
Acute peritonitis model induced by Thioglycolate in mice was assessed in order to evaluate Eugenia anti-inflammatory activity. Both extracts showed anti-inflammatory effect by inhibiting neutrophil influx. No animal showed symptoms of toxicity or even death. Mice received a subcutaneous injection of extract, and 1 h later were administered with 3 % Thioglycolate (TG) intraperitoneally. When mice are assayed in a peritonitis model, a 6 h period is necessary to reach maximum acute neutrophil recruitment. In vivo cell migration analysis showed that subcutaneous injection of different extracts concentrations significantly reduced cell  influx into the peritoneal cavity (Fig. 5). EP showed antiinflammatory activity at concentrations of 30 and 300 mg/ mL (Fig. 5a). As for EA, the anti-inflammatory activity was observed for all the tested concentrations (Fig. 5b). As negative control, extracts were s.c. injected and PBS was i.p. injected. As a result, neutrophil migration did not occur. Either for EA as for EP, the anti-inflammatory activity was comparable to the effect of dexamethasone (DEX), a potent anti-inflammatory drug in clinical use.
Differences in chemical composition between E. aurata and E. punicifolia HBK may explain the anti-inflammatory effects described herein. Both extracts inhibit NET release but only EA reduces cell adhesion whereas EP decreases elastase secretion.
The control of accute/chronic inflammatory processes as rheumatoid arthrite, ashtma, vaculitis among others like diabetes is relevant, once such processes may be related to the damage caused by the release of free radicals [26]. Moreover, neutrophil activation is largely dependent on the generation of reactive oxygen species (ROS) that are known to be inhibited by antioxidant compounds, as catechin and rutin, found abundantly in plant extracts [27]. Selected phenolic compounds, as diosmin and hesperidin, decrease the adhesion of inflammatory cells to the endothelium [28], whereas others can inhibit degranulation of neutrophils without affecting superoxide production [29].
In addition, after assessing the phenolic and flavonoid contents, an ESI-MS was performed, in order to correlate the main secondary metabolites found in HEEP and in HEEA with their biological activity without further chromatographic separations, since a preliminary HPLC HEEP analysis was already published by our group [31]. Gallic acid derivatives, flavonols, glycosides and procyanidins were the most common phenolic compounds in fruits and leaves of the Myrtaceae family [32]. Database search showed no prior HEEA research published to the moment.
The most characteristic corresponding molecular formulas of HEEP and HEEA, their fractions, MS/MS fragments are shown in Table 2. All peaks found were   [33]. The ion of m/z 191 represents quinic acid, frequently found in higher plant as major compounds in the leaves [34] showed fragments of m/z 173 after H 2 O loss [35]. The peak at m/z 359 was characterized as glycosyringic acid [36].  [42]. Dissociation of fragment m/z 593 showed a loss of 308 units (corresponding to a rhamnose plus glucose group) and yielded directly a fragment ion at m/z 285 (assigned as kaempferol).
Some phenolic compounds found in HEEP/HEEA extracts belong to the family of ellagitannins. They are hydrolyzable tannins, a class of polyphenols whose structure consists of ellagic acid units linked to a polyol, usually glucose or quinic acid. These compounds are also characterized by their hexahydroxydiphenoyl (HHDP) group which is released on acid hydrolysis and spontaneously lactonizes to ellagic acid. Ellagic acid was characterized by diagnostic mass ion of m/z 301 and ma ss fragments at m/z 257 and 229 [40].  [33]. However, ellagitannins had lower efficacy in the inhibition of cell proliferation compared to ellagic acid, the breakdown product [44].
At m/z 783, the MS 2 peak fragmentation produced an ion at m/z 481 [M-H-302] − , loss of HHDP), and after losing a HHDP-glucose [M-H-481] − , an ion at m/z 301 which corresponds to ellagic acid. This fragmentation pattern was assigned to a bis-HHDP-glucose isomer. Additionally there were peaks at m/z 481 [M-H-469] − , loss of a trisgalloyl group) and m/z 301, corresponding to ellagic acid. These results suggest a HDDP-glucose and a trigalloyl group. All those results are consistent with data reported for other Eugenia [42,45]. Table 2 shows that EA contains the phenolic compounds myricitrin, rutin, quinic acid and quercetin derivatives. Among those compounds, some present a role on neutrophil activity. Aqueous extract containing quinic acid or its molecule derivate are described to inhibit neutrophil migration [46] and elastase secretion [47]. Despite inhibitory roles demonstrated for myricetin in elastase secretion [48], its derivate myricitrin does not present any inhibitory effect [49]. The suppressor character of rutin on neutrophil functions were better studied than the compounds cited above. Isolated rutin or the one found in plant extracts decreases in vitro and in vivo neutrophil migration [50], adhesion [51], elastase secretion [52] and NET release [27]. Quercetin derivatives were commonly found in E.aurata and E.punicifolia extract fractions. There are no evidences that quercetin-3-O-β-rhamanose or quercetin-galloyl-hexoside promotes effects on neutrophil activities. In contrast, quercetin suppresses elastase secretion [48] and in vivo and in vitro neutrophil migration [20]. The role of quercetin in the adhesion events is still unclear [53].
ESI-MS n analyses also revealed the presence of gallic acid, quercetin-3-O-α-arabinopyranoside, syringic acid, ellagic acid, monogalloyl-glucose, glycosyringic acid, mudanoside B, HHDP glucose isomer and digalloylglucose isomer. The participation of these compounds in the neutrophil biology is poorly studied. There were reported downregulation of elastase secretion, NET release and neutrophil migration for gallic and ellagic acids present in plant extracts or in isolated form [53]. Evidences were not found for neutrophil adhesion.

Conclusions
Based in our results, we are able to propose a model for the anti-inflammatory properties exhibited by both hydroethanolic extracts of E. aurata and E. punicifolia (Fig. 6). The extracts in this model have a different chemical composition in terms of phenolic compounds, but both present in vivo and ex vivo anti-inflammatory activity, besides not being toxic to neutrophils. E. aurata reduces neutrophil adhesion and E. punicifolia decreases elastase degranulation. NET release is inhibited by both extracts ex-vivo. Together, these effects result on a reduced inflammatory response and provide support to their use in popular medicine. Furthermore, these results show a potential of these extracts for the development of phytomedicines with anti-inflammatory properties, including the treatment of rheumatology, neoplastic, selfinflammatory, autoimmune or infectious disorders.