Natural products isolated from Tetragonula carbonaria cerumen modulate free radical-scavenging and 5-lipoxygenase activities in vitro

Background Propolis and cerumen are plant-derived products found in honeybees and stingless bees, respectively. Although propolis is an ancient folk medicine, the bioactivities of cerumen obtained from Australian native stingless bees (Tetragonula carbonaria) have not been widely studied. Therefore, we investigated selected anti-oxidant and anti-inflammatory properties of T. carbonaria cerumen. Methods A methanolic extract was prepared from the combined cerumen of 40 T. carbonaria hives, and HPLC was used to screen for chemical constituents that scavenged 2,2-azobis(2-methylpropionamidine) dihydrochloride (AAPH). The ability of cerumen extracts to scavenge 1,1-diphenyl-2-picrylhydrazyl (DPPH) and to interfere with leukotriene B4 (LTB4) production in ionomycin-stimulated human neutrophils was also examined. Results The extract dose-dependently scavenged DPPH (EC50 = 27.0 ± 2.3 μg/mL); and inhibited the 5-lipoxygenase (5-LOX)-mediated oxidation of linoleic acid (IC50 = 67.1 ± 9.6 μg/mL). Pre-treatment of isolated human neutrophils with the methanolic cerumen extract additionally inhibited the ionomycin-stimulated production of LTB4 from these cells (IC50 = 13.3 ± 5.3 μg/mL). Following multi-solvent extraction, the free radical-scavenging and 5-LOX-inhibiting activities of the initial cerumen extract were retained in a polar, methanol-water extract, which contained gallic acid and a range of flavonone and phenolic natural products. Conclusions The findings identify free radical scavenging activity, and interference by extracts of T. carbonaria cerumen in 5-LOX–LTB4 signaling. Further investigation is needed to determine whether the extracts will provide therapeutic benefits for medical conditions in which oxidative stress and inflammation are implicated, including cardiovascular disease and impaired wound healing.


Background
Propolis is a resinous, plant-derived natural product of honeybees; made by foraging for plant resins and combining these with beeswax and salivary secretions [1]. Cerumen is a similar material produced by stingless bees of the Meliponini tribe [2]. Cerumen and propolis contain chemical constituents that protect the hive against bacterial infection and opportunistic pests [1,2]. Although propolis and cerumen typically comprise 50% plant resin, 30% beeswax, 10% essential and aromatic oils, 5% pollen and 5% organic debris [1,3], their exact chemical compositions may vary widely.
Propolis treatment has traditionally been indicated for a wide range of ailments, which has more recently been attributed to the broad anti-oxidant, anti-cancer, antibacterial, anti-viral, anti-inflammatory and woundhealing effects of its extracts (reviewed in [1,4]). These properties have often been correlated with a relatively small number of compounds in the sample or extract, including caffeic acid phenethyl ester (CAPE), artepillin C, kaempferol and galangin [5][6][7][8]. However, other studies have identified novel constituents within propolis extracts that are responsible for some of its observed bioactivities [9][10][11][12].
Tetragonula carbonaria is a stingless bee species native to Australia and commonly inhabits the Eastern coastline of southern Queensland and northern New South Wales [13]. T. carbonaria produce cerumen using the resins of the turpentine tree (Syncarpia glomulifera) [14] and Cadaghi gum (Corymbia torelliana) [15,16]. Our research group has shown that a methanolic extract of T. carbonaria cerumen comprises polar constituents including gallic acid, amyrins, C-methyl flavanones and phloroglucinols, amongst others [17][18][19]. This extract, which has a chemical profile distinct to typical honeybee propolis [17], elicited a vasorelaxant response in precontracted human and porcine artery preparations [20] and exerted antibacterial activity against Staphylococcus aureus [18,19]. An ethanolic extract of T. carbonaria cerumen inhibited 5-lipoxygenase (5-LOX) activity in a cell-free assay [17], although the kinetics for this response was not determined. Following on from this work, the aim of the present study was to investigate additional anti-oxidant and anti-inflammatory properties of T. carbonaria cerumen extracts. In particular, cellfree assays tested the potential of cerumen extracts to scavenge free radicals and inhibit the pro-inflammatory enzyme, 5-lipoxygenase (5-LOX). Stimulated human neutrophils additionally served as an in vitro model of human inflammation, to test the effects of a T. carbonaria cerumen extract on the 5-LOX mediatedproduction of the pro-inflammatory eicosanoid, leukotriene B 4 (LTB 4 ).

Cerumen collection and methanolic extraction
Cerumen collected from 40 T. carbonaria hives in the Brisbane region of South-East Queensland, Australia, was washed with water to remove debris and homogenised into one bulk sample. Raw T. carbonaria cerumen was extracted in 10 mL methanol and 5 mL hexane (per gram) with tumbling at 15 rpm and 22°C for 24 h. Following paper filtration, waxes contained in the upper hexane extract were discarded, and the remaining methanolic extract was evaporated under nitrogen gas (N 2 ) and freeze-dried overnight. Dried extract was reconstituted in dimethyl sulfoxide (DMSO; 1-500 μg/mL) for activity testing.
Multi-solvent extraction of the methanolic cerumen extract Hexane (15 mL) was added to the initial methanolic extract (30 mL), and the 'first' hexane extract was collected. Distilled water (15 mL) was then added to the remaining methanolic extract, which was extracted once more with hexane (20 mL). The 'second' hexane extract was separated from the methanol-water extract; both of which were collected. The two hexane extracts and the methanol-water extract obtained were evaporated under N 2 , freeze-dried and reconstituted in DMSO (1-5000 μg/mL) for activity testing (Fig. 1).

Collection and isolation of human neutrophils
Neutrophils were isolated from whole blood samples collected from healthy consenting adults, who were recruited to the study through internal advertisements at the University of the Sunshine Coast (USC). Persons who had recently taken medications known to affect neutrophil function (e.g. NSAIDs and, glucocorticoids), ingested alcohol or had undertaken strenuous exercise 48 h prior to blood collection, smoked regularly, or were pregnant, were excluded from the study. Whole blood was collected from healthy volunteers who provided their informed consent, and with ethics approval from the USC Human Research Ethics Committee (S/12/389). Blood collection followed the guidelines of the Declaration of Helsinki and Tokyo for humans. Venous blood (12 mL) was collected from the median cubital vein of four consenting individuals (24-62 years) into K 2 EDTA tubes. Neutrophils were obtained by layering 5 mL of whole blood onto 5 mL of Polymorphprep solution (Axis-Shield; Oslo, Norway) and centrifuging at 500×g for 30 min. The clear blood fraction containing neutrophils was collected into another centrifuge tube containing 6 mL of '20% media' (Media 199 containing 20% foetal bovine serum (FBS), 50 μg/mL penicillin/ streptomycin and 2 mM Glutamax-I) and centrifuged a second time at 500×g for 6 min. The cell pellets were resuspended in 1.3 mL of Dulbecco's Phosphate-Buffered Saline (PBS), with 10 μL of sample smeared onto a microscope slide and stained using Diff Quik differential dye to confirm successful isolation of neutrophils using brightfield microscopy.

High-performance liquid chromatography (HPLC) screening of free radical-scavenging constituents
Anti-oxidant compounds within the methanolic extract of T. carbonaria cerumen were identified using a modified HPLC screening method [21], using 2,2-azobis(2-methylpropionamidine) dihydrochloride (AAPH) as a free radical initiator. Dried methanolic T. carbonaria cerumen extracts (4 mg/mL) and AAPH (160 mg/mL) were reconstituted in 1:1 MilliQ water:acetonitrile, and equal volumes of each solution were incubated at 40°C. After 8 h, reversed-phase HPLC analysis of samples was performed using a PerkinElmer Series 200 HPLC pump and auto-sampler, with a Phenomenex Synergi 4 μm Fusion-RP 80 Å analytical column, 75 × 4.6 mm with 4 μm particles (Phenomenex, Inc.; Lane Cove, NSW, Australia). Mobile phase A (MPA) was 95:5 MilliQ water:acetonitrile and mobile phase B (MPB) was 10:90 MilliQ water:acetonitrile. Following 1 min equilibration at 100% MPA; 1.2 mL/min), samples were eluted with the following method: 100% MPA for 2 min, graded to 50:50 MPA:MPB over 10 min, 100% MPB for 20 min, 100% MPB for 10 min, graded back to 100% MPA over 5 min, 100 MPA for 3 min (total run time = 50 min). Photodiode array detection occurred at 205, 260, 290 and 340 nm. Constituents of the extract that scavenged AAPH-derived free radicals were detected by the reduction or disappearance of the peak intensity for the compound following HPLC analysis. An AAPHnegative control containing 2 mg/mL extract in 1:1 MilliQ water:acetonitrile was also included in the assay, and additional analyses confirmed that deterioration of the samples did not occur over 8 h (not shown).

Colorimetric 1,1-diphenyl-2-picrylhydrazyl (DPPH) assay
Reconstituted extracts of T. carbonaria cerumen were incubated with 100 μM DPPH (prepared in methanol) for 30 min at 22°C, then absorbance was measured at 518 nm. DPPH-scavenging activity of each sample was calculated by measuring the decline in absorbance after 30 min, and expressed as a percentage of a negative control.

Cell-free 5-LOX assay
The inhibitory effect of T. carbonaria cerumen on the 5-LOX-mediated oxidation of linoleic acid was examined using a modified colorimetric assay [22]. Briefly, 10 μL of each reconstituted cerumen extract was added to 0.5 mL of Solution A (containing 10 mM 3-(dimethylamino)benzoic acid, 0.05 M disodium phosphate (Na 2 HPO 4 ; pH 6.0), 500 μM linoleic acid) and 5-LOX enzyme (3.4 μg in 10 μL water), then incubated for 5 min at 22°C. Solution B (0.5 mL), containing 10 mM 3-methyl-2-benzothiazolinone and 0.1 mg/mL haemoglobin was added and incubated for a further 5 min at 22°C. Samples were centrifuged at 20,000×g. for 3 min at 22°C, and the absorbance of the supernatant was measured at 590 nm. Inhibition of 5-LOX by each sample was calculated by measuring the decline in absorbance after 30 min, and expressed as a percentage of a negative control.
The effect of the methanolic cerumen extract on 5-LOX enzyme kinetics was also determined. Reconstituted extract (10 μL; final concentration 100 μg/mL) was added to equal volumes of Solution A (containing 10-500 μM linoleic acid) and Solution B, then aliquotted into a 96-well microplate. Absorbance at 590 nm was measured immediately after the addition of 5-LOX (1.7 μg in 5 μL water), and periodically over one hour. The mean maximal reaction velocity (V max ) and Michaelis constant (K m ) were calculated for 5-LOX activity in the absence and presence of cerumen extract.

LTB 4 production in isolated human neutrophils
Resuspended neutrophils (35 μL) were made up to 70 μL with Dulbecco's PBS containing methanolic cerumen extract (final concentration 1-500 μg/mL). Cell suspensions were incubated at 37°C in a 5% CO 2 incubator for 20 min, then treated with 2 μM ionomycin for a further 5 min (final reaction volume 80 μL) to stimulate LTB 4 production [23]. Samples were centrifuged at 40×g for 6 min at 22°C, and LTB 4 concentration was determined spectrophotometrically at 405 nm using 50 μL aliquots of the supernatant in a LTB 4 ELISA, according to manufacturer's instructions (Cayman Chemical Company; Ann Arbor, USA). Background absorbance was measured in the absence of ionomycin and subtracted from all readings. Solvent (1% DMSO) and untreated controls were included in each assay, and were without effect (not shown).

Data analysis
Data are expressed as mean ± SEM. Data were compared using one way ANOVA with Tukey's post-hoc test, where differences were considered significant at P < 0.05.

Free radical-scavenging activities of T. carbonaria cerumen extracts
Polar extracts of T. carbonaria cerumen displayed free radical-scavenging properties in vitro. Reversed-phased HPLC screening of the methanolic cerumen extract identified several free radical-scavenging constituents, evidenced by reduced peak intensities after reacting with the free radical initiator, APPH (Fig. 2). The extract also scavenged DPPH, in a dose-dependent manner (EC 50 = 27.0 ± 2.3 μg/mL; Fig. 3). Following multisolvent partitioning of the methanolic extract, DPPHscavenging activity was identified in the resultant methanol-water and hexane extracts. However, the potency of this activity was significantly greater in the methanol-water extract (EC 50 = 31.1 ± 1.6 μg/mL) than the first and second hexane extracts (EC 50 = 352.2 ± 7.1 μg/mL and 128.1 ± 16.9 μg/mL respectively; P < 0.05).

Composition of a methanol-water extract of T. carbonaria cerumen
Analysis of the methanol-water extract using reversedphase HPLC revealed that it was a complex mixture of numerous compounds (Fig. 5). By comparison to known compounds, Peak 1 was identified as gallic acid, while the remaining constituents could only be assigned to compound classes ( Table 1). The majority of main compounds belong to the flavanone and phenolic classes of natural product.

Discussion
Propolis and cerumen are plant-derived bee products that exhibit a broad range of chemical and biological properties, regardless of their geographical origins. Although honeybee propolis is considered an ancient folk medicine, the recent popularity of alternative medicines and neutraceuticals has seen research and commercial interest in propolis and cerumen come to the fore. This study aimed to investigate selected anti-oxidant and anti-inflammatory properties of cerumen collected from Australian native stingless bees; a natural product that has not been widely studied. We subsequently found that polar extracts of T. carbonaria cerumen possessed potent free radical-scavenging properties, and exhibited inhibitory effects on the 5-LOX-LTB 4 signaling pathway in vitro. The polar methanol-water extracts were, on average, 10.6-and 4.9-fold more potent that the lower polarity first hexane and second hexane extracts, respectively, in our assays.
It is widely reported that propolis of diverse global origins exert anti-oxidant effects [1,4,24]. In particular, propolis extracts have previously been found to scavenge reactive oxygen species (ROS) and synthetic free radicals, inhibit lipid peroxidation, reduce ferric (Fe 3+ ) and cupric (Cu 2+ ) ions and elicit metal-chelating effects in vitro [5,11,[25][26][27][28][29][30][31]. In this study, we demonstrated that T. carbonaria cerumen also possessed free radicalscavenging properties, evidenced by the ability of its polar extracts to scavenge AAPH and DPPH in cell-free assays. However, whilst the anti-oxidant properties of propolis are often correlated with the phenolic acid and flavonoid content of its extracts [5,11,[25][26][27][28][29]31], T. carbonaria cerumen comprises a unique chemical profile. Previous gas chromatography-mass spectrometry (GC-MS) analysis of T. carbonaria cerumen extracts by our group showed that its chemical profile differed from New Zealand propolis and did not contain CAPE [17], a compound regarded to be largely responsible for the anti-oxidant and anti-inflammatory properties of temperate, honeybee propolis [32,33]. Further studies are required to investigate whether T. carbonaria cerumen may exert similar, and additional, anti-oxidant effects in cell-based systems. Gallic acid, one of the compounds identified in cerumen, was previously reported by our group to inhibit 5-LOX activity (pIC 50 = 5.62 ± 0.35 μg/mL) [17]. The methanolwater cerumen extract had 7.6-fold and 39.6-fold lower potency for inhibition of 5-LOX compared to gallic acid and kojic acid, respectively. Bioactivityguided fractionation of T. carbonaria polar extracts are ongoing to elucidate the remainder of its bioactive constituents.  Eicosanoids such as LTB 4 are pro-inflammatory signaling molecules produced from the enzyme-catalyzed metabolism of arachidonic acid (AA). Upon liberation from phospholipids by phospholipase A 2 , AA is oxidized to LTB 4 via an intermediate precursor, LTA 4 , in a pathway catalyzed by 5-LOX and LTA 4 hydrolase [34]. In the present study, we found that a methanolic extract of T. carbonaria cerumen inhibited the pro-inflammatory 5-LOX-LTB 4 signaling pathway in vitro. In cell-free assays, the extract dose-dependently inhibited the 5-LOXmediated oxidation of linoleic acid, by reducing the maximal reaction velocity and the affinity of 5-LOX to its substrate. These findings collectively suggest that the effects of the extract on 5-LOX resembled a mixture of competitive and non-competitive enzyme inhibition [35]. Using a cell-based model of human inflammation, the cerumen extract additionally suppressed LTB 4 production by ionomycin-stimulated neutrophils. Our results coincide with others who demonstrated that an ethanol extract of Brazilian green propolis inhibited the release of cysteinyl leukotrienes, LTC 4 , LTD 4 and LTE 4 , in peripheral leukocytes of patients with allergic rhinitis [36]. However, since the T. carbonaria cerumen extract inhibited neutrophilderived LTB 4 production with five-fold greater potency than cell-free 5-LOX activity (IC 50 = 13.3 ± 5.3 μg/mL versus IC 50 = 67.1 ± 9.6 μg/mL, respectively), we hypothesize that its mechanism of action may not be specific to 5-LOX inhibition. Although it is unclear whether propolis and cerumen may inhibit LTB 4 synthesis by exerting additional effects on phospholipase A 2 or LTA 4 hydrolase enzymes, it has been suggested that their anti-oxidant properties may assist in suppressing eicosanoid synthesis, by non-specifically scavenging the peroxy systems implicated in AA metabolism [37].

Conclusion
The aim of the present study was to investigate antioxidant and anti-inflammatory properties of polar extracts of cerumen from Australian native stingless bees, T. carbonaria. The study demonstrated that cerumen exerted potent free radical-scavenging effects, was a mixed enzyme inhibitor of 5-LOX, and reduced the Ca 2+ -ionophore-induced production of LTB 4 from human neutrophils in vitro. Polar constituents of cerumen belonged primarily to flavanone and phenolic classes of compound. Further investigation is needed to determine whether the extracts will provide therapeutic benefits for medical conditions in which oxidative stress and inflammation are implicated.