Bioactive compounds and antibacterial properties of monofloral Ulmo honey

Monofloral Ulmo honey is a very appreciated product in the international market but more information is needed in order to support their health properties. Total phenolic and flavonoid contents, antioxidant capacity and antibacterial activity of monofloral Ulmo (Eucryphia cordifolia) honey samples were determined. The samples contained between 176 and 208 mg gallic acid eq (GAE)/100 g for total phenolic content (TPC) and 43–90mg Quercetin eq/100 g for total flavonoid content (TFC). The antioxidant activity ranged between 91 and 152 mM eq Trolox/g and 28–49 mM eq Trolox/g for DPPH and ABTS assays, respectively. All phenolic extracts from honey samples inhibited Staphylococcus aureus, Streptococcus pyogenes, Pseudomonas aeruginosa and Escherichia coli by plate assays. The chromatographic profile showed that Ulmo honey contained the polyphenolic gallic, caffeic, and coumaric acids, the flavonoids pinocembrin, chrysin, quercetin, luteolin and apigenin, and abscisic acid. Compuestos bioactivos y propiedades antibacterianas de la miel monofloral de Ulmo RESUMEN La miel monofloral de Ulmo es un producto muy apreciado en el mercado internacional, pero se necesita más información para respaldar sus propiedades saludables. Se determinaron los contenidos fenólicos y flavonoides totales, la capacidad antioxidante y la actividad antibacteriana de muestras de miel monofloral de Ulmo (Eucryphia cordifolia). Las muestras contenían entre 176 – 208 mg eq de ácido gálico (GAE)/100 g para el contenido fenólico total (TPC) y 43 – 90 mg eq de quercetina/100 g para el contenido total de flavonoides (TFC). La capacidad antioxidante osciló entre 91 – 152 mM eq Trolox/g y 28 – 49 mM eq Trolox/g, para los ensayos de DPPH y ABTS, respectivamente. Todos los extractos fenólicos de muestras de miel de ulmo inhibieron Staphylococcus aureus, Streptococcus pyogenes, Pseudomonas aeruginosa y Escherichia coli mediante ensayos en placa. El perfil cromatográfico mostró que la miel de ulmo contenía como compuestos bioactivos los ácidos, gálico, cafeico y cumárico, los flavonoides pinocembrina, crisina, quercetina, luteolina y apigenina, y ácido abscísico. ARTICLE HISTORY Received 21 August 2019 Accepted 3 December 2019


Introduction
Honey is the natural product obtained after flora nectar is collected, transformed, and stored by honey bees (Apis mellifera L.). There are different types of honey according to the floral source from which the bees have collected nectar. Honey is usually classified according to the floral pollen present in the sample (Anklam, 1998).
There is a close relationship between the botanical origin of honey and its nutritional, antioxidant, and antimicrobial properties. For example, black forest honey has a lower antioxidant content than buckwheat honey (USDA, 2016), while monofloral Manuka honey from the tree Leptospermum scoparium is recognized worldwide for its nutraceutical properties (Alzahrani et al., 2012). Identification of the flora source of honey associates the presence of certain attributes to a particular honey.
In Chile, different types of honeys, with particular qualities and characteristics, have been discriminated based on the different climate zones, which affect the floral profile and floral abundance (Montenegro, Santander, Jara, Nuñez, & Fredes, 2013). One species with flowers that are very attractive to the bee is Ulmo (Eucryphia cordifolia Cav.). Recently, Ulmo honey has gained attention for its antioxidant and antimicrobial properties. In vivo studies have shown that Ulmo honey is able to activate the monocytic response through regulation of antiinflammatory factors (Jull, Rodgers, & Walker, 2008;Tonks et al., 2003) that allow the regeneration of epithelial cells in burn wounds (Schencke, Salvo, Vasconcellos, & Del Sol, 2013). Ulmo honey was reported to have a greater inhibitory activity against E. coli, Pseudomonas aeruginosa and five types of methicillinresistant Staphylococcus aureus (MRSA) than Manuka honey (Sherlock et al., 2010). More recently, Acevedo et al. (2017) reported activity of Ulmo honey on these bacteria strains as well as Aeromonas hydrophilia, Klebsiella pneumoniae and Salmonella enteric. However, both studies used limited honey samples and did not report the botanical origin of the honey.
These regenerative and antimicrobial activities are mainly attributed to the phenolic compounds present in honey, such as flavonoids, benzoic acid derivatives and volatile compounds (Pita-Calvo & Vázquez, 2017). P-coumaric, ferulic, chlorogenic, and caffeic acids have been identified in Ulmo honey, as well as the flavonoids kaempferol and luteolin (Velásquez et al., 2017) along with some volatile ketones (Montenegro et al., 2009). However, there is a knowledge gap, since no studies have shown a relationship between the botanical origin, the polyphenol profile, and the antioxidant and antimicrobial properties of Ulmo honey.
The current study aimed to analyze monofloral Ulmo honey samples, those with high pollen content from Eucryphia cordifolia, and to identify the potential molecules responsible for its reported medicinal properties.

Samples
Samples were purchased from local beekeepers from the Mediterranean climate area of Southern Chile (X Región de Los Lagos). Only honeys with ≥45% of Eucryphia cordifolia pollen determined by a melissopalinological analysis were selected.

Melissopalinological analysis
The botanical origin of the honeys was determined according to Chilean Standard Normative 2981 (Montenegro, Gómez, Díaz-Forestier, & Pizarro, 2008). Briefly, 10 g of honey were diluted in 10 mL of distilled water, centrifuged at 2500 rpm for 5 min, and the sediment was re-suspended in 0.1 mL of distilled water. Optical microscopy was used to observe the pollen grains and were identified using a palinoteque. Only monofloral honeys, with a composition equal or higher than 45% of a unique plant species, were considered.

Ulmo honey extracts
Samples were diluted in acidic water (pH 2 with HCl). The dissolution was put into an open chromatographic column (250 mm x 20 mm) filled with amberlite and washed with acidic water, distilled water and ethanol. Collected extracts were concentrated with a vacuum rotary evaporator until dryness (Buchi, Flawil, Switzerland), dissolved in distilled water, filtrated and stored at −20ºC until use. Controls were carried out to ensure the innocuity of the extracts before doing the tests.

Total phenolic content (TPC)
The phenolic content of the samples was evaluated by Folin-Ciocalteu (FC) assay based on Bridi et al. (2014). Five hundred microliters of honey samples diluted in water (0.2 g honey to 4 mL of water) were mixed with 2.0 mL of the FC reagent 1:10 (v/v) and 1.2 mL of a solution of Na 2 CO 3 . The absorbance at 760 nm was measured after 60 min at room temperature in the dark, using an Agilent 8453 UV-visible spectrophotometer (Palo Alto, Calif., USA). A calibration curve constructed with gallic acid was used to quantify and results are expressed as mg GAE/100 g honey.

Total flavonoid content (TFC)
Total flavonoids were determined using an aluminum chloride method based in Woisky and Salatino (1998) procedure. Five hundred microliters of honey sample diluted in water (0.1 g honey to 1.6 mL of water) were mixed with 500 µL of 2% AlCl 3 ethanol solution. After 60 min at room temperature, the absorbance at 420 nm was measured using an Agilent 8453 UV-visible spectrophotometer (Palo Alto, Calif., USA). Total flavonoids content was calculated as mg QE/100 g honey, from a calibration curve.
2.6. 2,2-diphenyl-1-picrylhydrazyl radical scavenging assay (DPPH) The scavenging activity was determined based on Thaipong, Boonprakob, Crosby, Cisneros-Zevallos, and Byrne (2006) procedure. Twenty-four milligrams of DPPH were dissolved in 100 mL of methanol. Ten milliliters of the solution were diluted with methanol until an absorbance of 1.1 ± 0.02 at 515 nm is obtained. Five hundred microliters of aqueous honey dilution reacted with 4,5 mL of final DPPH solution in the dark for 30 min, and the absorbance at 517 nm was measured. A calibration curve with TROLOX was employed and the results were expressed in mg eq TROLOX/g honey. 2.7. 2,2ʹ-azino-bis (3-ethylbenzothiazoline-6-sulphonic acid) radical scavenging activity (ABTS· + ) The scavenging activity of the samples measured on ABTS· + radical was based on the method described by Van den Berg, Haenen, van den Berg, and Bast (1999), with some modifications. Twenty milligrams of potassium persulfate were added to 20 mL of 7 mM ABTS aqueous solution. The mixture was incubated at darkness between 12 and 16 h and then diluted with ethanol until final absorbance of 0.70 ± 0.02 at 732 nm. Fifty microliters of aqueous sample dilution were mixed with 1.45 mL of reagent solution and absorbance at 732 nm was measured after 30 min in the dark. Calibration curve was constructed with and the results were expressed in equivalent mg Trolox/g honey.

Antibacterial activity
Bacteria strains E. coli ATCC-25922, S. aureus ATCC-25923, Pseudomonas aeruginosa ATCC-15692, Streptococcus pyogenes were used. The strains were propagated on Mueller Hinton agar 24 h in advance. Colonies were selected and diluted in saline solution to ensure a concentration of viable 10 8 CFU/mL by visual comparison with a 0.5 McFarland standard (Becton Dickinson and Company, USA). With a swab, strains were placed on petri dishes with Mueller Hinton agar, and 6 mm diameter holes were made and filled with 100 µL of honey dilution. The Petri plates were incubated for 24 h at 35ºC. Then, the inhibition diameter that appeared around each hole was measured (CLSI, 2006).

Statistical analysis
Statistical analyses of the results were performed using a variance analysis by Tukey's HSD method with 95% of confidence computed by STATGRAPHICS Centurion XV software 15.02.05. All tests were carried out in triplicate, and the results are reported as the means and standard deviation. Correlations between the analyses performed were made using the Pearson procedure (p < 0.05). In order to evaluate the general effect of the determined parameters, the principal component analysis, PCA (The Unscrambler 9.7 software, CAMO AS, Oslo, Norway) methods was chosen.

Botanical origin
The antioxidant and antimicrobial activities of Chilean honey have been studied over the last few years. These properties have been mainly attributed to the phenolic compounds present in the samples (Bridi et al., 2015;Giordano et al., 2018;Montenegro et al., 2013). However, these honey samples not included the polyphenolic profile of several Ulmo honey samples and its relation with the antibacterial activity.
Eight monofloral Ulmo honey samples were collected in the X Los Lagos region of southern Chile. The botanical origin was determined by analysis of pollen grain present in the honeys, according to the official policy established by the Standards Division of the National Institute for Standardization (Montenegro et al., 2008), which classifies the honeys as monofloral, bifloral, or polyfloral (Montenegro et al., 2009). Monofloral honeys are those in which at least 45% or more of the pollen grains belong to the same species. The eight evaluated samples contained between 86.6% and 96.8% Eucryphia cordifolia pollen grain (Table 1). Pollen analysis showed a low diversity of secondary species, with contributions from Arrayán (Luma apiculata), Tineo (Weinmannia trichosperma) and alfalfa chilota (Lotus pedunculatus).
Among the Ulmo honey samples, samples 4 and 5 had significantly higher TPC compared to samples 3 and 6. The TP content was higher when the secondary species of nectar origin was Tineo, and lower when it was alfalfa chilota or Arrayán.
The total flavonoid (TF) content of the Ulmo honey samples ranged from 40 to 90 mg QE/100 g of honey, with an average of 75.50 ± 14.45 mg QE/100 g (Figure 1(b)). The average TF content of Ulmo honey was about 10 times higher than that of Sesame honey (5.50 ± 0.02 mg QE/100 g of honey), and four to five times higher than Manuka honey (16.29 ± 0.12 mg quercetin/100 g of honey) (Jantakee & Tragoolpua, 2015). Sample 5 showed a statistically different TF content from the rest of the samples, presenting the lowest flavonoid content, however, it seems unrelated to the botanical origin.
The antioxidant activities of the Ulmo honey samples were evaluated using two different antioxidant assays, the ABTS and DPPH assays, with Trolox as the standard (Figure 1(c)). Honey samples 1, 2 and 8 showed a significantly higher antioxidant capacity when evaluated using the DPPH radical. The samples with the highest DPPH scavenging activity were also sourced from Arrayán, while the others did not. The presence of Arrayán nectar could indicate that these species increases the antioxidant activity of the Ulmo honeys. In the ABTS assay, samples 1 and 5 showed the highest antioxidant activities. Overall, the samples showed a three-fold higher value when evaluated using DPPH (up to 152 mM eq Trolox/g honey) compared to when using ABTS radicals (up to 49 mM eq Trolox/g honey).
These results are contrary to those reported by Attanzio, Tesoriere, Allegra, and Livrea (2016), who found higher values using ABTS compared to DPPH when evaluating Sicilian monofloral honeys.
For a better characterization of the Ulmo honey samples (Pita-Calvo & Vázquez, 2017) liquid chromatography was used to analyze the polyphenol content. All samples exhibited  (Ferreres, Andrade, & Tomás-Barberán, 1996), indicating that the phenolic profile of honey would depend on its botanical origin, and also the isoprenoid abscisic acid, that has been used as floral marker also (Silva et al., 2013). It is noteworthy that the polyphenols identified in the Ulmo honey samples are known to have biological activity. For example, it has been reported that gallic acid has antimicrobial, anticancer and antiviral activities (Khatkar, Nanda, Kumar, & Narasimhan, 2017), while rutin, also called vitamin P, chrysin and pinocembrin have demonstrated anti-inflammatory, antiasthmatic, antitumor and antimicrobial activities. These health benefits are mainly attributed to the powerful antioxidant activities of these molecules as scavengers of free radicals (   A correlation study was conducted between the polyphenolic composition and the antioxidant properties of the Ulmo honey samples using a Pearson multivariate analysis (Table 3). High positive correlations were found between the concentration of chlorogenic acid and the antioxidant activity of the samples as measured by the ABTS method (0.824) and between the antioxidant activity as measured by the DDPH assay and the nectar source Arrayán (0.844). There were no statistically significant correlations between ABTS/ TPC, DPPH/TPC, TFC/DPPH, DPPH/ABTS and TFC/TPC. Similar conclusions were found in other monofloral honeys [34]. These results suggested that the antioxidant capacity of Ulmo honey is influenced by the content of some of the phenolics in the samples.
On the other hand, Principal Component Analysis (PCA) showed that there were no distinctive groups according to the difference in the contribution of the secondary species (Figure 2(a)). However, there was a relationship between Tineo and TP content, which implies that the highest concentration of Tineo generates the highest concentration of TP content. There is also an inverse correlation between TP content and Arrayán, with sourcing from Arrayán species generating a lower TP content. There were no strong relationships between the TF content and any of the other compounds, with the Ulmo, nor with the accompanying species (Figure 2(b)). In regards to the antioxidant activity, no direct relationships were observed between ABTS radical scavenging activity and either the botanical origin or the polyphenol content or molecules (Figure 2(c)). A direct relationship between the antioxidant activity, the Arrayán source of nectar, and the flavonoid luteolin was observed, as was one between the DPPH antioxidant activity and luteolin (Figure 2(d)).

Phenolic and flavonoid contents and antibacterial properties
The antibacterial activity of the eight Ulmo honey samples were tested in inhibition zone assays. Most of the honey samples inhibited growth of S. aureus, St. pyogenes, Ps. aeruginosa and E. coli (Figure 3). Samples 2, 4, 5 and 7 showed the greatest antibacterial activity against S. aureus. Samples 1 to 7 showed similar inhibition activity against St. pyogenes, while sample 8 showed slightly less activity. The highest activity against E. coli was observed in samples 1 and 8, both with Arrayán as secondary species of origin. Sample 5 exhibited the highest antibacterial activity against Ps. aeruginosa, while sample 2 showed the lowest. Of the microbes assayed, Ulmo honey samples showed greater inhibition of Gram-positive bacteria compared to Gramnegative bacteria.
Previous studies reveal similar inhibition patterns. Sherlock et al. (2010) reported that Ulmo honey had inhibitory activity against E. coli, Ps. aeruginosa and five strains of S. aureus. Acevedo et al. (2017) also reported activity against Aeromonas hydrophilia, Klebsiella pneumoniae and Salmonella enterica.
Having a greater number of Ulmo honey samples from different geographical origins and harvest seasons could strengthen the relevance of Eucryphia cordifolia in the local beekeeping industry.
Correlation analysis (Table 3) showed statistically significant correlations between the inhibition of Ps. aeruginosa and chlorogenic acid levels (0.733) and between inhibition of E. coli and the percentage of Arrayán as secondary species Antiradical capacity against 2,2ʹ-azino-bis (3-ethylbenzothiazoline-6-sulphonic acid) (ABTS); Antiradical activity against 2,2-diphenyl-1-picrylhydrazyl (DPPH); total phenolic content (TPC); total flavonoid content (TFC); gallic acid (GalA); Cinnamic acid (  Antiradical capacity against 2,2ʹ-azino-bis (3-ethylbenzothiazoline-6-sulphonic acid) (ABTS); Antiradical activity against 2,2-diphenyl-1-picrylhydrazyl (DPPH); total phenolic content (TPC); total flavonoid content (TFC); Gallic acid (GalA); Cinnamic acid ( (0.832). No correlation between antibacterial activity and botanical origin was observed, likely since the dominance of Eucryphia cordifolia is not significate different among the samples (range 92.2 ± 3.9%, as shown in Table 1) The PCA indicated that there were no distinctive groups according to the contribution of secondary species (Figure 4(a)). Despite this, relationships were established between the botanical origin, polyphenols and antibacterial activity (Figure 4(b)). The antibacterial activity against Ps. aeruginosa showed a direct relationship with alfalfa sourcing, Ulmo and chlorogenic acid, abscisic acid and quercetin, which are mostly linked to the alfalfa content. The antibacterial activity against St. pyogenes and S. aureus showed direct relationships with the amount of Ulmo and Tineo sourced nectar. The inhibition of Ps. aeruginosa had a greater relationship with Tineo content than with Ulmo content. Both antibacterial activities showed indirect correlations with alfalfa sourcing, myricetin, quercetin and coumaric acid. The inhibition of E. coli was directly related to Arrayán sourcing and the presence of apigenin and chrysin, and to a lesser extent, with pinocembrin, caffeic acid and gallic acid.

Conclusions
In this study, several Ulmo honey samples were compared for phenolic composition, antioxidant and antimicrobial activities for the first time. The dominance of Eucryphia cordifolia in the samples range from 86.6% to 96.8%. Ulmo honey contained  gallic, caffeic, coumaric, abscisic and chlorogenic acids as well as pinocembrin, chrysin, quercetin, luteolin, apigenin, and rutin. These polyphenols content was contrast to the high antioxidant activity and strong antibacterial activity against S. aureus, St. pyogenes, Ps. aeruginosa and E. coli. According to PCA analysis, these activities could be related to the polyphenols present in the samples, especially chlorogenic acid. On the other hand, this study highlights how phenolic profiles are significant in the biological activity of honey samples and that further research should focus on identifying a broad set of molecules in a larger set of honey samples.

Disclosure statement
No potential conflict of interest was reported by the authors.