Antibacterial properties of Tualang, Acacia and Yemeni Sumur honey against selected food spoilage bacteria and foodborne pathogens

The antibacterial activity of honey is mainly credited to its acidity, osmolarity and enzymatic generation of hydrogen peroxide via glucose oxidase. Additional honey components, such as aromatic acids or phenolic compounds, also contribute to the overall antibacterial activity. The level of antibacterial activities found in honey varies with different types of honey, due to mainly the composition, percentage as well as the nature of the sugars present in the honey. This study aimed to evaluate the antibacterial activity of four types of honey, namely Tualang honey (TH1), Tualang honey (TH2), Acacia honey (AH) and Yemeni Sumur honey (YSH). Nine bacterial strains were used. Disc diffusion, well diffusion, minimum inhibitory concentration (MIC), Minimum bactericidal concentration (MBC), and time-kill methods were performed to reveal the antibacterial potential of the selected honey. The MIC values ranged between 12.5 to 50% for both TH1 and YSH while for TH2, and AH it ranged between 25 to 50%. For MBC, it ranged from 25 to 50%. The time-kill in TH1 Staphylococcus aureus (food isolate) showed total inhibition at 6 hrs in 2 X MIC, and for Staphylococcus aureus ATCC 29737 was 3.84 log CFU/g at the 6 hrs. Physicochemical quality of honey resulted as follows: the pH of the honey samples was acidic in nature ranging between 3.69 to 3.94, and the aw of the honey samples were between 0.53 to 0.69. For colour analysis, YSH was observed to has the maximum lightness and yellowness, and TH1 has the maximum redness. While, AH had a minimum lightness, redness, and yellowness.


Introduction
Honey is characterised as a vicious, aromatic and sweet food substance that many people around the world enjoy its consumption. Since ancient times, honey has been used in medical treatment as an antiseptic until the antibiotics were invented. However, the treatment of diseases has become more difficult than ever due to the emergence of antibiotic-resistant bacteria. As a result, humans seek a natural alternative such as honey to treat diseases. Honey contains antibacterial compounds that are effective in killing or inhibiting a broad spectrum of bacteria (Boukraâ, 2013). Various studies have shown that honey is effective against a wide range of microorganisms, including methicillin-resistant Staphylococcus aureus (MRSA) and vancomycinresistant enterococci (VRE).
For all antibiotic classes, including the major last-resort drugs, resistance is increasing worldwide and even more alarming, very few new antibiotics are being developed. The potent activity of honey against antibiotic-resistant bacteria resulted in renewed interest for its application. Several types of honey have been approved for clinical applications. The incomplete knowledge of the antibacterial compounds involved and the variability of antibacterial activity are, however, major obstacles for applicability of honey. Therefore, the objective of this work is to determine the in vitro antibacterial activity of Tualang Honey 1 (TH 1 ), Tualang Honey 2 (TH 2 ), Acacia Honey (AH) and Yemeni Sumur Honey (YSH) against foodborne pathogens and spoilage bacteria.

Test organisms
Nine bacterial strains were assayed; four were Grampositive Staphylococcus aureus ATCC 29737, Bacillus cereus (food isolate), Staphylococcus aureus (food isolate) and Listeria monocytogenes ATCC 19112, and five Gram-negative Escherichia coli (food isolate), Campylobacter jejuni ATCC 29428, Pseudomonas aeruginosa ATCC 27853, Escherichia coli ATCC 10536 and Salmonella enterica ser. Typhimurium ATCC 13311. Stock cultures of the ATCC were obtained from Thermo Fisher Scientific (US) while the remaining microorganisms were obtained from the culture collection of the Laboratory of Food Microbiology, Faculty of Food Science and Technology, University of Putra (UPM), Malaysia.

Preparation of test organisms
Stock cultures of the bacterial pathogens were subcultured in Nutrient agar (Oxoid, UK) at 37°C for 24 hrs. Colonies of fresh cultures of the different microorganisms from overnight growth were picked with a sterile inoculating loop and suspended in 3 to 4 mL of Nutrient broth (Oxoid, UK) in sterile test tubes and incubated for 2 to 3 hrs at 37°C. This was diluted with sterile distilled water to standardise the inoculum density used in this study.
2.5 Antibacterial activity of honey 2.5.1 Disc diffusion assay Disc diffusion assay was performed to demonstrate the sensitivity of microorganisms to the honey samples with a slight modification in which an overnight bacterial culture was suspended into 5 mL of 0.1% normal saline and adjusted with 0.5 Mc Farland standards (Reller et al., 2009). A 50 µL of the bacterial suspension was spread well using sterile swabs on Mueller Hinton agar (Oxoid, UK). Sterile paper discs (Whatman AA discs, 6 mm in diameter) were used to impregnate nine different dilutions of each of the honey samples as follows (0.195, 0.390, 0.781, 1.56, 3.125, 6.25, 12.5, 25, and 50% w/v). Twenty microliters of each dilution of honey were soaked into the paper discs and allowed to dry at room temperature and placed aseptically on the surface of inoculated agar using sterile forceps and finally incubated for 16 to 24 hrs at 37°C. The assessment of antibacterial activity was based on the measurement of the diameter of the inhibition zone appearing around the disc. The sterile blank disc was used as a negative control. As a positive control, ciprofloxacin and gentamicin (Oxoid, UK) were used for Gram-negative bacteria, and for Gram-positive bacteria, ampicillin and penicillin G (Oxoid, UK) were used (Tumin et al., 2005).

Well diffusion assay
The procedure was similar to that of disc diffusion assay. Mueller Hinton plates were inoculated with 50 µL of 0.5 Mc Farland standard bacterial suspension. A hole with a diameter of 6 to 8 mm was punched aseptically with a sterile cork borer or a pipette tip, and 100 to 200 µL of honey of different concentrations (0.195, 0.390, 0.781, 1.56, 3.125, 6.25, 12.5, 25, and 50% w/v) was introduced into the well. Plates were incubated for 16 to 24 hrs at 37°C. The assessment of antibacterial activity was based on the measurement of the diameter of the inhibition zone appearing around the hole. Sterile distilled water was used as a negative control. As a positive control, ciprofloxacin and gentamicin (Oxoid, UK) were used for Gram-negative bacteria, and for Gram-positive bacteria, ampicillin and penicillin G (Oxoid, UK) were used (Tumin et al., 2005).

Minimum Inhibitory Concentration (MIC)
The MIC of different honey samples was determined using the two-fold dilution method (Balouiri et al., 2016). Honey batches were investigated for their MIC against all the nine bacterial strains. About 1 mL of honey was transferred into each test tube followed by the addition of 1 mL of Mueller Hinton (Oxoid, UK) broth that contained standardised inoculum of 0.  3.125, 6.25, 12.5, 25, and 50% (v/v). The inoculated tubes were incubated at 37°C for 24 hra. The highest dilution of the tested honey to inhibit growth (no turbidity in the tube) was considered as the MIC value of the honey against the tested bacterial species.

Minimum Bactericidal Concentration (MBC)
MBC test was performed following the MIC test via a streak plate method. Each honey dilution with no bacterial growth (turbidity) from the MIC test was assayed. A loopful of bacterial strain was inoculated into sterile Mueller Hinton plates. The plates were then incubated at 37°C for 24 hrs. The least concentration that did not show any growth of tested organisms was considered as the MBC value of the tested honey against the tested bacterial species (Balouiri et al., 2016).

Time-Kill
Time-kill assay was performed to determine the time taken for the microorganisms to be killed by the honey, and it was performed following the method of Rukayadi et al. (2010). Mueller Hinton broth was used, and bacterial inoculum was adjusted between 6 to 8 log CFU/ mL. Final concentrations of honey were 0 MIC, ½× MIC, 1× MIC, and 2× MIC. Four bijou bottles were added with the right amount of the inoculum and the honey respectively to produce the final concentrations of 0 MIC, ½× MIC, 1× MIC, and 2× MIC aliquots. These steps were performed for all the strains. The cultures were then incubated at 37°C with the agitation of 200 rpm. At predetermined intervals 0, ½, 1, 2, 4 and 6 hrs, 1 mL aliquots were serially diluted in 1% phosphatebuffered saline (PBS) and plated onto Mueller Hinton (Oxoid, UK). The plates were incubated at 37°C for 24 hrs, and the number of colonies was counted. The assays were carried out in duplicate. The graph of log 10 CFU/ mL was plotted against time.

Determination of pH
The pH values of the honey samples (10% aqueous honey solution) were measured at 28±2°C using a digital pH meter (Mettler Toledo) calibrated at pH 4.0, 7.0 and 9.0 using standard buffer solutions (Saxena et al., 2010).

Determination of water activity
The water activity of honey solutions was determined at 25±0.2°C using an electronic dew-point water activity meter (Aqualab Series 3 model TE, USA) equipped with a temperature-controlled system which stabilised the temperature during measurement .

Colour analysis
The honey surface colour was measured using a colourimeter (HunterLab, model D25 L optical sensor, Hunter Associates Reston, VA, USA) calibrated with a white and a black standard plate. Honey solution (10 g of honey in 75 mL free CO 2 distilled water) was placed into a cylindrical (base diameter 11.3 cm and height 2 cm) optical cell. Reflectance values were obtained using a 45 mm viewing aperture. The results reported were the mean of five determinations (Shafiee et al., 2014).

Statistical analysis
Experiments were performed in duplicate, and the results were expressed as mean values with standard deviations (SD). The significant differences were obtained by one-way analysis of variance (ANOVA, Tukey's test).

Disc diffusion assay
For the Disc Diffusion Assay, only S. aureus showed inhibition with a clear zone in all types of honey. For TH 1, the inhibition zone at 25% was 10.08 mm, and at 50% the inhibition zone was 15.01 mm. For TH 2, the inhibition zone at 25% was 6.25 mm, and at 50% it was 12.92 mm. For AH, the inhibition zone at 25% was 8.82 mm, and at 50% it was 13.06 mm. For YSH, the inhibition zone at 25% was 12.49 mm, and at 50% it was 20.35 mm, as shown in Table 1. The diameter of the inhibition zone of the positive control was between 22.00 -24.24 mm, which was higher than the 50% concentration of treatments. The rest of the bacteria did not show any inhibition zone. There was a significant (p < 0.05) difference between the samples.
The use of disc could lead to the exclusion of large molecules which are not properly absorbed by the paper disks and may contribute to inaccurate results as honey is FULL PAPER a complex solution consisting of different sizes of chemicals and compounds (Zainol et al., 2013). Kolayli et al. (2016) examined the antibacterial activities of chestnut honey using a disc, found that the highest inhibition was achieved against pathogenic bacteria such as S. aureus, E. coli, Enterococcus faecalis and Yersinia pseudotuberculosis.

Well diffusion assay
In the preliminary screening, all the honey showed various range diameter of inhibition zone (DIZ) against most of the tested bacteria. More than half of the isolates showed antibacterial activity in a concentration between 12.5, 25 and 50% (w/v). The inhibitory activity was significantly (p < 0.05) affected by the type of bacteria used as well as the honey samples. TH 1 and YSH had the highest DIZ. For TH 1, as shown in Table 2, the range was between 15.82 mm and 31.9 mm, lowest DIZ was seen on S. aureus, which was 15.82 mm at 25%. At 12.5 and 25%, both E. and S. enterica ser. Typhimurium ATCC 13311 almost showed the same DIZ, which was 18.13 mm and 18.37 mm, respectively. The DIZ was seen on S. enterica ser. Typhimurium ATCC 13311 at 50%, which was 31.95 mm. S. aureus ATCC 2973, E. coli, and C.
jejuni ATCC 29428 were almost in the same range 25.88, 25.23 and 25.44 mm, respectively. Only the positive control of S. aureus (food isolate) and P. aeruginosa ATCC 27853 had bigger DIZ. The rest of the controls showed lower DIZ in which 50% of honey was effective inhibiting than the antibiotic.
For TH 2 , DIZ was in the range of 13.91 mm and 27.16 mm as seen in Table 3 The highest DIZ was seen on S. aureus ATCC 29737 which was 27.26 mm at 50% and the lowest DIZ was seen on both S. enterica ser. Typhimurium ATCC 13311 and E. coli which were almost the same 13.91 mm and 13.93 mm at 12.5% and 25%, respectively. In the positive control, S. aureus (food isolate) had bigger DIZ of 2.29 mm. The rest of the controls showed lower DIZ. While for AH, the range was between 11.17 mm and 33.87 mm as shown in Table  4, S. enterica ser. Typhimurium ATCC 13311 showed bigger DIZ than TH 1, which was 33.87 mm followed by both S. aureus and S. aureus ATCC 29737, which had 26.06 mm and 29.73 mm DIZ respectively, at 50%. The lowest DIZ was seen on S. enterica ser. Typhimurium ATCC 13311 same as TH 2 which was 11.1 mm at 12.5%. All positive control had smaller DIZ as compared to 50% P. aeruginosa ATCC 27853 6.00±0.00 a 6.00±0.00 a 6.00±0.00 a CN 12.56 C. jejuni ATCC 29428 6.00±0.00 b 6.00±0.00 b 25.44±0.16 a CIP 0.00 L. monocytogenes ATCC 19112 6.00±0.00 a 6.00±0.00 a 6.00±0.00 a CIP 0.00 C. jejuni ATCC 29428 6.00±0.00 b 6.00±0.00 b 22.26±1.02 a CIP 0.00 L. monocytogenes ATCC 19112 6.00±0.00 a 6.00±0.00 a 6.00±0.00 a CIP 0.00 Meanwhile, at 12.5%, only S. aureus was inhibited with DIZ of 15.13 mm, and it was the lowest DIZ. All positive control had smaller DIZ as compared to 50% honey concentration. There was no inhibition of B. cereus and L. monocytogenes ATCC 19112 in all four types of honey.
Agar Diffusion Assay to assess the antibacterial activity is commonly performed in three different ways; well/cup diffusion, disc diffusion or agar dilution. The method usually depends on the nature of the antibacterial or the antibacterial to be tested, and on the kinetic properties of the molecules present in the honey. Agar well allows direct contact of honey components and bacteria immediately after application. The diffusion mechanism may also represent in-vivo conditions when honey is applied on infected wounds, and therefore, may provide information about the kinetic system of honey application (Zainol et al., 2013).
Both well diffusion and disc diffusion were performed to compare between them, and there was a huge difference between the results which was obtained by both methods. Agar well exercise allows direct contact of honey components and bacteria immediately after application. In one study, it was shown that Tualang Honey could inhibit the growth of food spoilage bacteria which were S. aureus, S. enterica ser. Typhimurium, E. coli and P. aeruginosa as the DIZ was 24.50 mm, 18 mm, 16.50 mm and 16 mm, respectively (Aween et al., 2014). In the present work, for TH 1 , S. aureus had 15.82 mm DIZ at 25% which was less than they reported, but for TH 2 it was similar to the previous study for both S 6.00±0.00 a 6.00±0.00 a 6.00±0.00 a CIP 0.00 Acacia honey showed the lowest inhibitory activity as compared to the other honey samples (Zainol et al., 2013). In this study, YSH showed bigger zones than Seder honey (Aween et al., 2014 (Aween et al., 2014). Furthermore, in this study, AH showed bigger DIZ when compared to another study which also used Acacia honey (Zainol et al., 2013) The antibacterial activity of Tetragonisca angustula honey has been well known and was found to possess good antibacterial activity against S. aureus. Another study revealed that T. angustula honey had significant antibacterial activity against several different bacterial strains, including B. cereus and P. aeruginosa (Miorin et al., 2003). A recent study has confirmed the antibacterial activity of T. angustula honey on Gram-positive bacteria such as S. aureus, and Gram-negative bacteria such as P. aeruginosa and E. coli (Rao et al., 2016). Wilkinson and Cavanagh (2005) studied the antibacterial activity of different types of honey against E. coli and P. aeruginosa. They found that all types of honey showed a zone of inhibition. A study was done using Nigerian honey against S. aureus, E. coli, and P. aeruginosa. It was found that 50% concentration could inhibit P. aeruginosa with DIZ of 8 mm and 100% could inhibit all the other bacteria with DIZ of 11 mm, 13.3 mm and 8 mm for S. aureus, E. coli and P. aeruginosa, respectively (Agbaje et al., 2006).
The inconsistency in the observed antibacterial activity could be due to a number of reasons. One of them could be related to the differences in susceptibility of each species of bacterium to the antibacterial activity of honey tested. Similar results were reported by Taormina et al. (2001). Furthermore, the different results observed between honey samples might also be due to the different floral sources used by the bees as well as the geographical factors like humidity, temperature, and where the honey was produced (Tumin et al., 2005). Previously, Ainul Hafiza et al. (2005) conducted a study of five types of local honey (Belimbing, Gelam, Durian, Kelapa and Tualang); using agar diffusion method for antibacterial activity against S. aureus. All samples had low colony counts ranging from 37 CFU/mL in Kelapa honey to 161 CFU/mL in Durian honey. All samples showed a clear zone of inhibition, and there was a significant difference (p < 0.05) among the samples. Tumin et al. (2005) studied the antibacterial properties of several local Malaysian brands of honey which were Tualang, Hutan, Gelang, Pucuk Daun and Ee Feng Gu against six bacterial species in 20 µL/100 µL, 40 µL/100 µL, 60 µL/100 µL and 80 µL/100 µL concentrations. They found that Tualang honey exhibited antibacterial effect against S. enterica ser. Typhimurium, Streptococcus pyogenes and E. coli, with the highest activity seen against S. enterica ser. Typhimurium. In this finding, Tualang honey exhibited antibacterial properties against most of the bacteria with S. aureus, E. coli, and S. enterica ser. Typhimurium showed higher effects.
The antibacterial activity of chestnut honey at 100% concentration against S. aureus and E. coli were higher when compared to the flower honey at the same concentration. At 50% concentration, only chestnut honey had average inhibition against S. aureus. At 25% concentration, both of the honey had a slight inhibition against S. aureus. The diversity of flowers affect the composition of the honey produced. According to the literature, S. aureus is considered to be the most sensitive bacterium towards chestnut honey (Güneş et al., 2017). In the present work, all types of honey inhibited S. aureus and E. coli especially Tualang honey as it is well known for its antibacterial activity against S. aureus. Estevinho et al. (2008) studied the antibacterial activity of 20 honey samples from the nectar of Lavandula, Echium and Erica plants, and concluded that S. aureus was the most affected bacterium, followed by B. subtilis, S. lentus, K. pneumoniae and E. coli which were affected gradually. Silici et al. (2010) examined the antibacterial activity of Rhododendron honey at concentrations of 10, 25, 50 and 75% against 13 different microorganisms. P. aeruginosa and P. mirabilis were the most sensitive microorganisms followed by S. aureus, Aeromonas hydrophila, L. monocytogenes, B. subtilis, Mycobacterium smegmatis, and S. enterica ser. Typhimurium while B. cereus, E. coli O157: H7 and Y. enterocolitica were resistant, and they did not show any DIZ. In this study, S. aureus, E. coli, S. enterica ser. Typhimurium and P. aeruginosa were inhibited by all types of honey, but Tualang and Yemeni Sumur honey showed higher effects. In another study, Nzeako and Hamdi (2000) investigated the antibacterial activity of six different honey samples in which Turkish honey showed higher antibacterial activity against S. aureus.

Minimum Inhibitory Concentration (MIC) and Minimum Bactericidal Concentration (MBC) of honey samples
The disc and well diffusion methods are primarily a qualitative test for detecting the susceptibility of bacteria to antibacterial substances. However, both the MIC (the lowest concentration of the sample that will prevent at least 99% of the bacterial growth) and the MBC (the lowest concentration of honey required to kill at least 99% of the bacteria) reflects the quantity needed for bacterial inhibition (Balouiri et al., 2016). The MIC and MBC of the four types of honey are shown in Tables 6 and 7. TH 1 and YSH had the most bacteriostatic effect as compared to other honey. They showed almost the same inhibitory concentration except for P. aeruginosa ATCC 27853, which was higher in TH 1 12.5% MIC instead of 25% in YSH. In TH 2 , only S. aureus ATCC 29737, S. enterica ser. Typhimurium ATCC 13311, S. aureus (food isolate) and L. monocytogenes ATCC 1112 had 25% MIC. The rest of the bacteria did not show any inhibitory in the range which was developed. Also in AH only S. aureus ATCC 29737, S. enterica ser. Typhimurium ATCC 13311, S. aureus (food isolate) and P. aeruginosa ATCC 27853 showed a MIC of 25%, the rest of other bacteria exceeded the range.
A previous study which used Malaysian Tualang honey showed antibacterial activity against foodborne pathogen and food spoilage bacteria. The results were as follows; S. aureus, E. coli, P. aeruginosa and S. enterica ser. Typhimurium with a MIC of 12.5, 25, 12.5 and 3.125%, respectively (Shehu et al., 2015). The result was in line with this study except for S. enterica ser. Typhimurium, which had a lower MIC. It was also supported by Tan et al. (2009) who reported that Tualang honey exhibited lower MIC against different microorganisms.
Tualang honey closely resembled New Zealand Manuka honey (Comvita UMF 18+) as found by a study in which Malaysian honey was used as a comparison (Zainol et al., 2013) (Zainol et al., 2013).
TH 1 and YSH had the most bactericidal effect compared to other honey. These two honey showed almost the same inhibitory concentration except for P. aeruginosa ATCC 27853, which was higher in TH 1 25% MBC instead of 50% in YSH. In TH 2 , only S. aureus ATCC 29737, S. enterica ser. Typhimurium ATCC 13311, S. aureus and L. monocytogenes ATCC 19112 showed MBC of 50%. In contrast, the rest of the bacteria did not show any bactericidal effect in the range which was developed. Also in AH only S. aureus ATCC 29737, S. enterica ser. Typhimurium ATCC 13311, S. aureus and P. aeruginosa ATCC 27853 showed an MBC of 50%, while the rest of the bacteria exceeded the range. Malaysian Tualang honey showed antibacterial activity against S. aureus, E. coli, P. aeruginosa and S. enterica ser. Typhimurium with MBC of 25, 50, 25 and 12.5, respectively (Shehu et al., 2015). The results were in line with this study except for S. enterica ser. Typhimurium, which had lower MBC. The reaction of each bacterium towards honey is different depending on the bacterial growth and also the active compounds which are present in the honey.
In another study, Tualang honey had the same effect as Manuka honey (Comvita UMF 18+) against S. aureus, P. aeruginosa and E. coli with 15, 20 and 25% MBC. In B. cereus, Manuka honey showed lower MBC of 12.5%. In Acacia honey, S. aureus and B. cereus had the same MBC which was 25%, and E. coli and P. aeruginosa also had the same MBC of 50% (Zainol et al., 2013). Earlier in 2009, the effect of Tualang and Manuka honey was also studied against enteric microorganisms such as S. aureus and E. coli (>25% MBC), P. aeruginosa (25% MBC) and S. enterica ser. Typhimurium (20% MBC) (Tan et al., 2009).

Time-Kill
Time-Kill Kinetics test, also known as the "suspension tests or suspension time-kill analysis", determines the time required by the antibacterial agent to kill the challenged test microorganism. This test is utilised in microbiological studies to assess a test article's in vitro antibacterial activity in relation to time. In TH 1 , S. aureus (food isolate) (Figure 1) and S. aureus ATCC 29737 (Figure 2) showed 4.01 and 4.83 log CFU/g at 4 h in 1× MIC. After 6 h, S. aureus (food isolate) showed complete inhibition and 3.84 log CFU/g for S. aureus ATCC 29737 at 2× MIC. In one study, the growth of methicillin-resistant S. aureus (MRSA49) was entirely inhibited by 50% honey concentration after 8 hrs, whereas the growth of S. aureus was inhibited after 20 hrs. The examination of the killing kinetic of honey collected from the longan (Dimocarpus longan) flower at a concentration of 50% (v/v) showed a reduction in the survival of MRSA49 and  S. aureus, although, S. aureus showed more resistance to honey. Therefore, the potent activity of Thai honey against antibiotic-resistant pathogenic bacteria, MRSA, was the significant finding of that study (Jantakee and Tragoolpua, 2015). The antibacterial activity of honey is a property that is of significant interest in its function as an agent in treating bacterial infections that are resistant to the action of antibiotics.
By using the time-kill test, treatment with all stingless bee honey showed a significant reduction in viability of S. aureus (p < 0.001) and P. aeruginosa (p < 0.001), with a reduction in viability after 60 mins in the order of 1 to 3 log and >3 log for each bacteria, respectively (Boorn et al., 2010). Moreover, results showed that stingless bee honey had a relatively rapid bactericidal action while the Apis mellifera honey did not. A previous study reported a similar trend where a little effect was seen after treatment with 25% (w/v) A. mellifera honey, whereas notable cell death was seen after treatment with 25% Trigona biroi honey. Although the majority of results showed clear overall trends, there were a few instances where outcomes were untypical.
The majority of constituents in a ripen honey are sugars accounting for 80% of the components which include glucose and fructose, sucrose and maltose are in a lesser amount, and less than 18% of water. The large percentages of sugars and low moisture content results in osmotic stress, hence, prevent the spoilage of honey by microorganisms. It has been observed that a slight increase in moisture content can result in yeast growth. However, the high percentage of sugar still prevents bacterial growth even after the honey is diluted to approximately 30 to 40% (Kwakman and Zaat, 2012). It is noteworthy that the antibacterial activity of honey after dilutions is due to other compounds found in honey, among which is the hydrogen peroxide (H 2 O 2 ). The production of H 2 O 2 is catalysed by a glucose oxidase enzyme, which is incorporated into collected nectar by the honey bees during honey production. Upon adequate dilution of honey, this enzyme is activated and in turn, catalyses the conversion of glucose to H 2 O 2 and gluconic acid in the presence of H 2 O (Kwakman and Zaat 2012).
Although H 2 O 2 is the major antibacterial compound in honey, there are many honey varieties that have significant antibacterial activity resulting from nonperoxide compounds. Methylglyoxal and bee defensin-1 has been recently identified in Manuka honey and RS respectively as non-peroxide antibacterial compounds (Kwakman and Zaat 2012). The role of low pH (3.2 to 4.5) for the antibacterial activity of honey has been debatable in the world of science with a number of studies reporting opposing results with regards to the antibacterial contribution of low pH (Adams et al., 2008). Furthermore, studies have suggested the presence of additional unknown antibacterial compounds in honey.

Determination of pH
Overall, the pH values of the honey samples in this study were acidic ranging between 3.69 to 3.94 and within the recommended limit pH (3.4 to 6.1) (Moniruzzaman et al., 2013) as shown in (Figure 3). AH was the most acidic honey (pH 3.69), and the least acidic honey was YSH with pH 3.94. TH 2 and TH 1 had pH of 3.72 and 3.91, respectively. There was a significant difference (p < 0.05) between the samples. Although honey is considered to be acidic, however, the high sugar content masks the acidity in the honey taste. The average pH of honey is pH 3.9. Formic acid and citric acid were originally believed to be the dominant acidic compounds in honey (Ball, 2007). However, it was recently discovered that gluconic acid is the predominant acid compound in honey, which is produced from bee secretions under the action of oxidase enzyme on glucose.
A variation in the pH values of Brazilian, Spanish, Turkish and Algerian honey was found to be in the range of 3.10 to 4.05, 3.63 to 5.01, 3.67 to 4.57 and 3.49 to 4.53 respectively (Amri and Ladjama, 2013). A pH value between 3.5 to 5.0 for honey from Northwest Spain was also reported (Olga et al., 2012). Similarly, pH values between 3.8 to 4.5 for Moroccan honey were also reported (Karabagias et al., 2014). The acidity is likely to contribute to the antibacterial potency of the honey (Almasaudi et al., 2016). Properties and the composition of honey depend on its geographical floral origin, season, environmental factors as well as treatment of beekeepers (El Sohaimy et al., 2015).

Determination of water activity
The water activity (a w ) of honey samples were in  (Figure 4). The a w of honey should be below 0.60 . TH 2 exceeded the level with 0.62 a w ; the rest of the honey was within the limit. TH 1 and YSH had the same a w of 0.53, and for AH it was 0.60. There was a significant difference (p < 0.05) in a w between the samples. These results are quite similar to those of Greek honey for which the a w ranged from 0.53 to 0.67 (Lazaridou et al., 2004). The a w is an important factor which governs the food stability by preventing or limiting microbial growth. The increase in a w influences the shelf life of honey and supports the growth of undesirable microflora, especially osmotolerant yeasts. The osmotolerant yeasts are able to grow at a minimal a w of 0.60 (Saxena et al., 2010). Although osmolality plays a significant role in the antibacterial activity of honey, there are other factors that influence the antibacterial effect of honey. The presence of such factors in honey depends to a greater extent on the source of nectar, location of the flowers, the storage time, preservation method and related water conditions (Almasaudi et al., 2016). Some of the quality problems found in honey such as stability, viscosity and crystallisation are due to a w . During glucose crystallisation, glucose monohydrate is formed, and other water molecules are dehydrated, resulting in a decrease in solute concentration in the liquid phase and increase in a w .

Colour analysis
Colour is the first attractive attribute of honey, and as such, it is very important for commercialisation. It is a significant parameter in the quality, acceptance and preference of consumers (Missio et al., 2016). As shown in Table 8, the L* parameter ranged from 2.27 to 13.63. The a* parameter ranged between 1.44 to 1.62. The b* parameter ranged from 12.09 to 2.27. There was a significant (p < 0.05) difference between the samples. The maximum and minimum lightness was seen in YSH and AH, respectively. The maximum and minimum redness was seen in TH 1 and AH, respectively. The maximum and minimum brightness was seen in YSH and AH, respectively. Honey's colour normally differs over a wide range of tones, ranging from light yellow to amber, dark amber and, in rare cases, black. Occasionally, even green or red hues may occur.
The colour of unprocessed honey depends on its botanical or floral origins. For this reason, colour is significant in the classification of monofloral honey for commercial activities (Moniruzzaman et al., 2013). It also depends on its ash content, temperature and time of storage (Gámbaro et al., 2006). The Codex Alimentarius Committee on Sugars (2001) stipulates that the colour of honey should be nearly colourless to dark brown. In one study, seven reputed commercial Indian honey brands showed L*, a* and b* values ranging from 26.3 to 36.8, 0.12 to 4.9 and 0.7 to 14.4, respectively (Saxena et al., 2010). Boussaid et al. (2014) evaluated six samples of honey which were collected from different regions of Tunisia and found that L* values ranged from 36.64 to 51.37. The lightness of honey plays an important role as it is what the consumers prefer. The analysed honey also contains other colours orange, green and yellow. Rosemary honey showed green colour, which is the negative value of a* while mint honey had the highest redness followed by eucalyptus honey. The orange, thyme eucalyptus horehound and mint honey showed values between 10 and 20 for b* (Boussaid et al., 2014).

Conclusion
In disc diffusion, TH 1 and YSH showed higher DIZ of 10.08 mm and 12.49 mm at 25%, and 15.01 mm and 20.35 mm at 50% against S. aureus (food isolate), respectively. While in well diffusion, TH 1 and AH showed the highest DIZ of 31.95 mm and 33.87 mm by S. enterica ser. Typhimurium ATCC 13311 at 50%. Meanwhile, in TH 2 and YSH, S. aureus ATCC 29737 showed the highest DIZ of 27.16 mm and 35.75 mm at 50%. In TH 1 , positive control of S. aureus (food isolate) and P. aeruginosa ATCC 27853 showed bigger DIZ than 50% honey concentration while TH 2 on S. aureus (food isolate) positive control had bigger DIZ. In both AH and  and 2.73 log CFU/g at 2× MIC, respectively. The highly acidic honey was AH with pH of 3.69, and the lowest water activity was measured in TH 1 and YSH at 0.53. Maximum L* and b* was by YSH, and maximum a* was by TH 1 .