Physicochemical, bioactive properties and antioxidant of Apis mellifera L. honey from western Paraná, Southern Brazil

Honey cannot be used as a nutritionally complete option of food for humans, but it has a good potential to serve as food supplement. It has been used in medicine against various diseases since ancient times in Egypt, Greece and Rome, and also in western, eastern, Chinese and Ayurveda medicines (Kuropatnicki et al., 2018). In traditional medicine, it has effects on the treatment of burns, diarrhea and ulcers (Ahmed et al., 2018). In addition, it boosts the immune system by measuring the effects of antiinflammatory, antibacterial, antifungal and antiviral activities. Also, its positive effects against breast, colorectal, renal, prostate, cervical and oral cancer have already been proven (Ahmed et al., 2018). The medicinal effects of honey are due to its antioxidant action as it has intrinsic compounds, such as glucose oxidase enzymes, catalase and peroxidase. Other compounds are carotenoids, organic acids, ascorbic acid, amino acids and proteins. (Machado De-Melo et al., 2018). Phenolic acids and flavonoids are listed as the main reason why honey has an antioxidant capacity (Baglio, 2018), due to their mechanism of action and reactions based on the transference of hydrogen atoms or simple electrons (EterafOskouei & Najafi, 2013). Honey antioxidant capacity depends on its botanical origin, environmental conditions and seasonal changes (Machado De-Melo et al., 2018).


Introduction
Honey cannot be used as a nutritionally complete option of food for humans, but it has a good potential to serve as food supplement. It has been used in medicine against various diseases since ancient times in Egypt, Greece and Rome, and also in western, eastern, Chinese and Ayurveda medicines (Kuropatnicki et al., 2018). In traditional medicine, it has effects on the treatment of burns, diarrhea and ulcers (Ahmed et al., 2018). In addition, it boosts the immune system by measuring the effects of antiinflammatory, antibacterial, antifungal and antiviral activities. Also, its positive effects against breast, colorectal, renal, prostate, cervical and oral cancer have already been proven (Ahmed et al., 2018). The medicinal effects of honey are due to its antioxidant action as it has intrinsic compounds, such as glucose oxidase enzymes, catalase and peroxidase. Other compounds are carotenoids, organic acids, ascorbic acid, amino acids and proteins. (Machado De-Melo et al., 2018). Phenolic acids and flavonoids are listed as the main reason why honey has an antioxidant capacity (Baglio, 2018), due to their mechanism of action and reactions based on the transference of hydrogen atoms or simple electrons (Eteraf-Oskouei & Najafi, 2013). Honey antioxidant capacity depends on its botanical origin, environmental conditions and seasonal changes (Machado De-Melo et al., 2018).
Brazilian honey has been recognized for its quality that also depends on the aforementioned factors and the region where it is produced (Braga et al., 2019). This quality presents variations in quantity and type of phenolic compounds, as well as in flavor and aroma. The country has a vast geographical area, with great diversity of honey bee species for apiculture products, which reflects on the wide variety of honey (Moraes et al., 2019;Sekine et al., 2019;Nascimento et al., 2018). Furthermore, its quality is also attributed to the resistance of Africanized bees to diseases and pests, so there is no need to use chemicals that contaminate honey to maintain hives. Moreover, the increase in studies on the quality of honey has been showing the incidence of trace elements (Altunatmaz et al., 2018). These trace elements and pesticides had been monitoring, to seek an alternative and methods to improve the quality of Physicochemical, bioactive properties and antioxidant of Apis mellifera L. honey from western Paraná, Southern Brazil the honey produced (Yaqub et al., 2020). Demonstrating, even more, the importance of the georeferencing of production (Camargo et al., 2014) and denomination of origin of honey.
Honey from the Western region of Paraná has been granted the "Indication of Origin" seal (Instituto Nacional da Propriedade Industrial, 2017), which gives it higher added value. For that reason, producers follow international quality standards. Being awarded the seal requires information on the relationship between honey characteristics and edaphoclimatic factors in the region, which is the object of this study, since the honey produced in western Paraná has not been a topic of research regarding physicochemical characteristics, phenolic compounds or antioxidant activity. Thus, the aim of this study was to characterize the honey produced by beekeepers from the aforementioned region, with regard to physicochemical parameters, phenolic compounds and antioxidant activity, as well as the degree of similarity between the samples from different municipalities.

Physicochemical analysis
Physicochemical analyses were performed according to Marchini et al. (2004), in triplicate, which allows higher reliability to the measured data. Moisture was determined following the Chataway refractometric method, which uses the refractive index measurement of the sample to be converted to moisture content by using an Abbe refractometer (Atago, Abbe refractometer, Tokyo, Japan). The pH was performed by diluting 10 g of honey in 75 mL of distilled water, determining the concentration of hydrogen ions contained in the honey solution used. Total acidity was determined based on neutralization of the honey acid solution, titrated with 0.05 N sodium hydroxide to the equivalence point. The ash content was determined by the gravimetric method after incineration in muffle furnace at 550ºC for 5 hours. For honey color determination, honey samples' absorbance was read with a spectrophotometer at 560 nm (UV-1800, Shimadzu, Columbia, USA). The reading was transformed into color by using the Pfund scale. Electrical conductivity was measured in a solution with 20% honey dry matter, diluted in distilled water and quantified by using a conductivity meter (Tec-4MP, Tecnal, Piracicaba, Brazil). The total nitrogen content was determined by the Micro-Kjeldahl method using factor 6.25 to convert nitrogen into proteins. Determination of HMF content was based on White's reaction by spectrophotometry with readings at 284 nm and 336 nm. Determination of total sugars (%), reducing sugars (%) and apparent sucrose (%) was based on the procedure described by Sereia et al. (2017). The method was based on the ability of sugars, such as glucose and fructose, to reduce the copper found in the cuproalcaline (Fehling liquor) solution, changing from Cu 2+ to Cu + (reduction of cupric ions in cuprous), and the sugars were oxidized into organic acids.

Total phenolic content (TPC) and Total flavonoid content (TFC)
Honey extract was obtained by diluting the honey in methanol (1:11, w / v) and stirring in Vortex for one minute. That was followed by an ultrasonic bath for 15 minutes. Then the extract was filtered through a qualitative paper filter for retention of larger particles. The extract remained refrigerated until the analyses were performed in triplicate. The total phenol content was determined according to Meda et al. (2005). Gallic acid (0.05 -0.08 mg mL -1 ) was used as standard for constructing the calibration curve (R 2 : 0.9999), expressed as mg GAE 100 g -1 of honey. Total flavonoid content was determined according to the method described by Meda et al. (2005), with some modifications. Quercetin (0.01to 0.07 mg mL -1 ) was used as standard for constructing the calibration curve (R 2 : 0.9972), expressed as mg EQ100 g -1 of honey.

Statistical analysis
Low and high values, medians, mean and quartiles were calculated for the physicochemical parameters, bioactive compounds and antioxidant activity. The data of the variables were analyzed by the multivariate factor analysis. Non-Metric Multidimensional Scaling (NMDS) was used to test the parameter differences between the different honey samples. Later, the Euclidean distance was used with the normalized data, using the "metaMDS" command to generate random and interactive processes, aiming to find the best possible solution. The measured NMDS adjustment value was assessed by the stress value. To estimate the adjustment between the dissimilarity matrix and the generated dendrogram, we calculated the cophenetic correlation coefficient (Estevinho et al., 2016). All statistical analyzes were performed using the R statistical software, version 3.0.2.

Physicochemical parameters
The results of the physicochemical evaluations of the samples are presented in Table 1. The values found were compared with Brazilian (Brasil, 2000) and international (Food and Agriculture Organization, 2001) standards. The honey samples evaluated were characterized as acidic, because of the average pH value (3.26), which was lower than the minimum standard pH (3.30). In addition, the mean total acidity of the samples was 34.54 meq kg -1 , with only three samples with values above 60 meq kg -1 , which is the limit established by legislation (Brasil, 2000;Food and Agriculture Organization, 2001). Values above this may indicate fermentation of the product. The pH and free acidity of honey are also important in determining taste and stability against microbial development.
The moisture content of the honey samples analyzed presented an average of 18.75% (Table 1), but 7% of them presented moisture content above the allowed limit (20%) (Brasil, 2000). Several factors may influence moisture content, such as the degree of maturity, environmental conditions during the harvest, processing techniques and conditions of storage (Silva et al., 2016;Guo et al., 2019).
As for color, the honey analyzed presented an average value of 0.26 ± 0.11 inc (Table 1) and, according to a classification using the Pfund scale, 67.16% of the samples presented light amber color; 1.49%, white; 20.9% extra light amber; and 10.45%, amber. This information corroborates studies on honey samples from previous harvests in this region, which presented colors ranging from white to dark amber, with prevalence of lighter colors (Camargo et al., 2014). That is the reason why the honey in question has advantages, since light colors have higher acceptability by consumers (Silva et al., 2016;Kortesniemi et al., 2018). For this reason, warehouses from other regions of Paraná and Brazil purchase honey from this region to mix it with darker types, favoring commercialization. The low HMF values indicate the good quality of the honey produced, since all samples presented values below 13.67 mg kg -1 (Table 1). According to Machado De-Melo et al. (2018), the honey evaluated is considered fresh because the lower the HMF value, the higher the indicative of honey that was stored for a short time or did not undergo heating and fermentation process.
The ash contents of the samples presented an average value of 0.14% (Table 1). Variations of this parameter may be due to floral and geographical origin (Nascimento et al., 2018), honey management techniques applied and / or the type of extraction used (Nascimento et al., 2018;Baglio, 2018). All samples comply with current regulations (Brasil, 2000), which consider 0.6% of ash content for floral honey, whereas the limit tolerated for melate or melate honey and its mixtures with honey is 1.2%. Silva et al. (2016) state that electrical conductivity in honey is directly related to ash content. The authors reported that this parameter was incorporated by Codex Alimentarius Standards, replacing the parameter of ash content. The legislation (Brasil, 2000) does not present low and high values for electrical conductivity, but Codex (Food and Agriculture Organization, 2001) considers the maximum value of 800.00 µS cm -1 . Thus, the values obtained, which ranged from 181.20 to 565.8 µS cm -1 (Table 1), were well below the limit. Electrical conductivity of honey is a parameter that, together with evaluations of insoluble solids and minerals, is used to determine honey purity (Silva et al., 2016).
Protein percentage of honey samples presented an average value of 0.28% (Table 1). Honey protein levels vary due to different factors, such as bee species and nectar botanical origin (Alvarez-Suarez et al, 2018), as well as the proteins from secretions of the mandibular and hypopharyngeal glands that are incorporated by bees in the process of transformation and dehydration of nectar into honey (Baglio, 2018).
The total sugar content is not standardized and, in this study, it presented an average value of 69.09% (Table 1). As for reducing sugars, there was a variation between 56.4 to 72.78% (Table 1), which is within the limits (Brasil, 2000;Food and Agriculture Organization, 2001), as well as the sucrose content that presented mean values of 4.28% (Table 1).

Phenolic compounds
The total phenolic compound content of the honey samples in this study ranged from 11.39 to 61.27 mg GAE 100 g -1 ( Table 2). These values were similar to honey samples from Portugal (Ferreira et al., 2009) in a way that an increase in the total phenol content of darker samples was detected: the white, amber and dark amber samples presented increasing values of 13.2, 16.8 and 20.4 mg GAE/100 g -1 , respectively. It may be the explanation for the amplitude of the low and high values observed in this study.
Flavonoid contents of the honey samples analyzed (Table 2) ranged from 7.97 to 44.99 mg EQ/100 g -1 . We also observed that all honey samples had higher total phenol content than flavonoids. This is due to the fact that flavonoids represented more than 50% of total phenols in plants and are transferred from plants to honey by bees (Silva et al., 2016). Flavonoid levels are related to the variation in honey color, as there is a significant increase of 12.6, 34.2 and 58.7 mg EQ/100 g -1 honey, for white, amber and dark amber (Ferreira et al., 2009). The values show that the honey analyzed has bioactive compounds and can be  Brasil (2000). Q1: 25% of the samples are below the median. Q3: 75% of the samples are above the median; c Mean ± standard deviation. 0.04 0.16 0.12 0.11 0.13 0.12 ± 0.02 Q1: 25% of the samples are below the median. Q3: 75% of the samples are above the a standard deviation. ABTS (2,2-azinobis (3-etilbenzotiazolina-6-sulfonic acid); DPPH (2,2-diphenyl-1picryl-hydrazyl); (FRAP) ferric reducing antioxidant power (2,4,6-tripyridyl-s-triazine).
used as functional food or as a source of food antioxidants with nutraceutical properties (Ahmed et al., 2018).

Antioxidant activity
The antioxidant activity FRAP presented a variation from 0.03 to 11.12 µmol FeSO 4 /g -1 (Table 2). It showed higher values than ABTS methods (0.43 to 1.54 µmol ET/ g -1 ) and DPPH (0.04 to 0.16 µmol ET/g -1 ), which may be due to the fact that FRAP is expressed as iron sulfate, while ABTS and DPPH are expressed as Trolox equivalent. FRAP antioxidant activity is in accordance with what was observed in studies with honey samples from Minas Gerais, Rio de Janeiro and Rio Grande do Sul, in Brazil (Sant' Ana et al., 2012;Nascimento et al., 2018). The observed gap is compatible with the FRAP, ABTS and the DPPH values of honey samples from countries such as Algeria (Khalil et al., 2012) and Cuba (Alvarez-Suarez et al., 2018). The use of different equivalents as ferric (II) and quercetin for FRAP makes it difficult to compare the values of work with data from other studies (Kuś et al., 2014;Duarte et al., 2012). The same is observed with ABTS and DPPH, whose results are expressed as percent inhibition of DPPH free radical (Sant' Ana et al., 2012;Nascimento et al., 2018) or in equivalent quercetin and gallic acid (Duarte et al., 2012).
Studies show that samples from the same region have different values, showing variations due to botanical, geographical origin and seasonality (Sant' Ana et al., 2012;Salgueiro et al., 2014). Honey antioxidant activity may increase with industrial honey processing (Machado De-Melo et al., 2018). Another factor that influences the antioxidant activity is storage time, because honey degradation processes occur by enzymatic or Maillarde reaction, which releases intermediate chemical groups with reducing power.
During the collection period there were changes in the color intensity of honey, and the samples collected in October and November were characterized by lighter shades. During this period, there are flowering grapes from Japan (Hovenia dulcis), abundant species in the study area, mainly in the lakeside forests of Itaipu bi-national hydroelectric (Camargo et al., 2014). Moraes et al. (2019) verified that in honey samples from the lake area, there is a predominance of pollen type H. dulcis in most of the honey sampled, characterized as monofloral of light shades. Nascimento et al. (2018) found out that monofloral honey of this species has light shades, as well as a total phenol content of 30.5 mg GAE/100 g -1 and 0.2 mg EQ/100 g -1 of flavonoids. Between December and March, the honey got darker, which may be linked to the Eucalyptus flowering, quite prevalent in the region (Sekine et al., 2019;Camargo et al., 2014) and significantly present in the honey samples of the region (Sekine et al., 2019;Moraes et al., 2019). It causes an increase in the bioactive compound contents and antioxidant activity. (Nascimento et al., 2018).
The similarity relationship between the evaluated parameters and honey samples from the 14 municipalities was assessed by the Non-Metric Multidimensional Scaling (NMDS) (Figure 2A). The ordering produced a two-dimensional solution with a final stress value of 0.033. According to Estevinho et al. (2016), the ordering is representative because there was a low stress value.
The nine groups were formed by using the criterion of the mean silhouette widths, from the result of the cluster analysis of the honey samples ( Figure 3). The correlation coefficient (CCC) was 0.73, a value considered reasonable as a factor of representativeness (Estevinho et al., 2016). NMDS showed that The analyses show similarity between the honey samples of the region in relation to the parameters that were evaluated. This study demonstrates that the types of honey analyzed have quality, commercial maturity, and bioactive properties that may be related to the phytogeographic origin of the product. the studied samples were grouped according to their similarities in relation to the evaluated parameters ( Figure 2B). The blue and purple groups have higher values for electrical conductivity. On the other hand, the samples of the yellow group presented the highest values for total phenols. The group in green presented the highest values for color and flavonoids, parameters that have a correlation of 0.96% according to Khalil et al. (2012).
The best partition by this criterion is the one with the highest average silhouette width. The group in red presented the highest values for pH, HMF, humidity, ash content, total sugars and reducing sugars. The gray group presented the highest values for FRAP, ABTS, acidity and protein. Khalil et al. (2012) reported that the antioxidant activity FRAP and Proline showed a high correlation (0.98%). The group with two red samples showed similarity for the vast majority of compounds and may indicate a relationship of bioactive compounds despite their different geographical origin. Finally, the two groups formed by only one sample each did not show similarity relationship with other samples. These results are important because they indicate that, within the nine groups, there was a homogeneous distribution of samples from different locations, showing that, in general, the honey samples evaluated present great similarity. As there is a differentiation of the samples according to their botanical origin, it would be necessary to analyze these honey samples coming from different municipalities to observe if they were grouped according to their botanical origins (Marchini et al., 2005), or if some pollen types interfere with the parameters evaluated (Estevinho et al., 2016).

Conclusion
The honey samples evaluated complied with national and international normative instructions. In addition, this first work quantifying bioactive compounds and antioxidant activity in honey samples from Western Paraná found relevant concentrations of phenolic acids and flavonoids, as well as antioxidant activity, which means that the product has nutraceutical properties.