Journal of Applied Microbiology 2003, 95, 913920

doi:10.1046/j.1365-2672.2003.02050.x

Antibacterial activity of honey and propolis from Apis mellifera and Tetragonisca angustula against Staphylococcus aureus
P.L. Miorin1, N.C. Levy Junior2, A.R. Custodio3 W.A. Bretz4 and M.C. Marcucci2
1

ncias Biome dicas da Universidade de Sa s-graduac cleo de Po o Paulo, Sa o Paulo, SP, Brazil, 2Nu o, Pesquisa e Instituto de Cie a lises e Pesquisas, Campinas, SP, o da Universidade Bandeirante de Sa o Paulo, Sa o Paulo, SP, Brazil, 3Natural-labor, Ana Extensa Brazil, and 4University of Pittsburgh, Pittsburgh, PA, USA

2002/439: received 8 November 2002, revised 2 May 2003 and accepted 23 May 2003

ABSTRACT
P . L . M I O R I N , N . C . L E V Y J U N I O R , A . R C U S T O D I O , W . A . B R E T Z A N D M . C . M A R C U C C I . 2003.

Aims: The antibacterial activity against Staphylococcus aureus of honey and propolis produced by Apis mellifera and Tetragonisca angustula was evaluated. Secondary aims included the study of the chemical composition of propolis and honey samples and its relationship with antibacterial activity against S. aureus. Methods and Results: The antibacterial activity of honey and propolis was determined by the method of macrodilution. The minimum inhibitory concentration (MICs) of A. mellifera honey ranged from 12623 to 18570 mg ml)1 and of T. angustula from 14287 to 21433 mg ml)1. For propolis, the MIC ranged from 036 to 365 mg ml)1 (A. mellifera) and from 044 to 201 mg ml)1 (T. angustula). Honey and propolis were evaluated by high-performance liquid chromatography. Some typical compounds of Brazilian propolis were also identied in honey samples. Principal component analysis revealed that the chemical composition of honey and propolis samples was distinct based on the geographical location of the samples. Conclusions: Propolis samples had higher antibacterial activity against S. aureus when compared with honey. However, both propolis and honey samples had antibacterial against S. aureus. Signicance and Impact of the Study: These antimicrobial properties would warrant further studies on the clinical applications of propolis and honey against S. aureus. Keywords: honey, multivariate analysis, phenolic compounds, propolis, Staphylococcus aureus.

INTRODUCTION Honey and propolis are bee products that have been used for centuries in folk medicine (Ghisalberti 1979; Zumla and Lulat 1989). These products have received increased attention from the general population because of their health claims (Dobrowolski et al. 1991). The in vitro activity of Brazilian ethanolic extract of propolis (EEP) against 118 strains of Staphylococcus aureus isolated from human infections was studied by the agar dilution method. The average minimum inhibitory concentration (MIC) was 225 mg ml)1 (Fernandes et al. 1995).
lises e pesquisas Correspondence to: Maria Cristina Marcucci, Natural Labor Ana ndida, Campinas, SP, Ltda. Rua Lauro Vanucci, 1020Jardim Santa Ca 13087-548, Brazil (e-mail: cristamarcucci@naturalab.com.br).

The susceptibility of S. aureus (ATCC 25923) to several concentrations of Brazilian EEP was also tested. The results showed that concentrations of 20% (v/v) and higher inhibited the growth of this bacteria (Fernandes et al. 1997). Azevedo et al. (1963) evaluated the antibacterial activity of beeswax and propolis elaborated by Apis mellifera and by two Brazilian Meliponinae using the agar diffusion method. Propolis from A. mellifera inhibited Bacillus subtilis, S. aureus and Mycobacterium smegmatis. Wax from Trigona rucrus inhibited B. subtilis, S. aureus and Myco. smegmatis. Wax and propolis from Trigona postica inhibited B. subtilis and S. aureus. Collectively, these studies suggested that propolis has inhibitory activities against bacterial pathogens. Phenolic substances that include cinnamic acid derivatives (mainly prenylated compounds) and some avonoids are

2003 The Society for Applied Microbiology

914 P . L . M I O R I N ET AL.

detected in honey and propolis (Martos et al. 1997; Marcucci et al. 2001). The chemical composition and antibacterial activity of propolis from A. mellifera and Tetragonisca angustula have been previously described (Bankova et al. 1996; Velikova et al. 2000a,b; Marcucci et al. 2001). However, there are few studies on the phenolic composition of Brazilian honey from A. mellifera and T. angustula, and its antibacterial activity. The aims of this investigation were: (i) to study the antibacterial activity of honey and propolis samples against S. aureus from A. mellifera Linnaeus 1758 and from T. angustula Moure 1946 using the method of macrodilution; (ii) to determine the chemical composition of propolis and honey samples by reverse phase high-performance liquid chromatography (HPLC); and (iii) to determine by employing multivariate statistical analysis the relationship between the compounds identied by HPLC with the antibacterial activity of propolis and honey samples.

EEP The EEP was prepared by using a modied technique initially described by Szewczak and Godoy (1984). Thirty grams of crude propolis were dissolved in 100 ml of ethanol by shaking the solution at room temperature for 7 days. The solution was then ltered with Wattman paper no. 1 and placed in amber asks. Propolis solutions were dried and weighed to obtain the correct concentration of each extract. These concentrations were used for determining MICs. Microbiological procedures The MIC of honey and propolis samples (A. mellifera and T. angustula) against S. aureus (ATCC 6538) was determined by the method of macrodilution according to the National Committee of Clinical Laboratory Standards (NCCLS 2000). The strains of S. aureus were inoculated in nutrient agar (Merck, Darmstadt, Germany) for 24 h at 37C in aerobiose and the number of colony-forming units (CFU ml)1) was adjusted according to the Mac Farland scale (10 108 CFU ml)1) and then further diluted to obtain a nal concentration of 50 105 CFU ml)1. The test tubes contained 50 ml of antibiotic broth medium (Merck) and were inoculated with 250 l)1 of S. aureus suspension (in 085% saline solution) with different concentrations of honey or propolis. As a control, a tube was prepared with the ingredients described above except for propolis and honey. All tubes were incubated at 37C for 24 h, and bacterial growth was measured using an UV-Vis

M A T E R I A LS A N D M E T H O D S Honey and propolis sources Honey samples of A. mellifera Linnaeus 1758 honey (mAm) and Minas Gerais were collected in the States of Parana (Brazil). The T. angustula Moure 1946 honey samples (mTa) were collected in Minas Gerais. Propolis samples from A. and Minas mellifera (pAm) were from the States of Parana Gerais. Propolis samples from T. angustula (pTa) were from Minas Gerais. Area and plant source of honey and propolis samples are shown in Table 1.

Table 1 Area and plant source of honey and propolis samples from Apis mellifera and Tetragonisca angustula Bee Africanized A. mellifera Honey mAm1 mAm2 mAm3 mAm4 mAm5 mAm6 mAm7 mAm8 mTa1 mTa2 mTa3 mTa4 mTa5 mTa6 mTa7 mTa8 mTa9 mTa10 Propolis pAm1* pAm2* pAm3* pAm4* pAm5* pAm6* pAm7* pAm8* pTa1* pTa2* pTa3* pTa4* pTa5* pTa6* pTa7* pTa8* pTa9* City/state o da Vito ria/PR Unia o da Vito ria/PR Unia o da Vito ria/PR Unia o da Vito ria/PR Unia o da Vito ria/PR Unia Brumadinho/MG Mateus Leme/MG Esmeralda/MG Betim/MG Betim/MG Betim/MG Betim/MG Betim/MG Betim/MG Betim/MG Betim/MG Betim/MG Betim/MG Collection date 6 August 1999 6 August 1999 6 August 1999 6 August 1999 6 August 1999 15 August 1999 Plant source Typical Typical Typical Typical Typical Typical Typical Typical Typical Typical Typical Typical Typical Typical Typical Typical Typical Typical native native native native native native native native native native native native native native native native native native forest forest forest forest forest forest forest forest forest forest forest forest forest forest forest forest forest forest

T. angustula

16 16 16 16 16 16 16 16 16 16

August August August August August August August August August August

1999 1999 1999 1999 1999 1999 1999 1999 1999 1999

with with with with with with with with with with with with with

Baccharis dracunculifolia B. dracunculifolia B. dracunculifolia B. dracunculifolia B. dracunculifolia B. dracunculifolia B. dracunculifolia B. dracunculifolia B. dracunculifolia B. dracunculifolia B. dracunculifolia B. dracunculifolia B. dracunculifolia

and MG-Minas Gerais). *Propolis samples and honey samples were obtained from the same hives (PR-Parana
2003 The Society for Applied Microbiology, Journal of Applied Microbiology, 95, 913920, doi:10.1046/j.1365-2672.2003.02050.x

BEE PRODUCTS EFFECTS ON S. AUREUS

915

spectrometer (Cary 50-Varian Inc., Walnut Creek, CA, USA). All the assays were performed in triplicate (Fava Netto 1975). Minimal inhibitory concentration of propolis an honey samples Bacterial growth for the honey samples was measured by turbidity with a UV-Vis spectrometer at 600 nm. The MIC was calculated graphically (Graph Pad, Prism 30, San Diego, CA, USA) and expressed in mg ml)1 as the percentage (v/v) of anhydrous honey contained in the antibiotic broth able to totally inhibit the bacterial growth. The antibacterial activities of different EEP were expressed as the percentage of inhibition in relation to the control (without inhibition) incubated without EEP. The results were based on the bacterial glucose uptake from the antibiotic broth. For each propolis dilution and for the control tube, aliquots of 250 ll were transferred to a second tube containing 10 ml of a glucose reagent (PAP-Liquiform o Paulo, Brazil) and shaken for 30 min. The Labtest, Sa optical density (O.D.) was determined at 505 nm. The concentration of propolis, where there was no glucose consumption, was considered as 100% of bacterial inhibition (Fava Netto 1975). The MIC was determined graphically by the method of minimum squares (Graph Pad, Prism 30). HPLC The phenolic compounds from honey samples (A. mellifera and T. angustula) were extracted according to methodology previously described in the literature (Martos et al. 2000a,b). The honey samples were analysed by HPLC in a D-7000 Merck-Hitachi equipment (Merck-Hitachi, Darmstadt, Germany) with a L-7100 pump and L-7200 autosampler. The chromatographic conditions were: reverse phase column (Lichrochart 100 RP-18; 125 04 cm, particle diameter of 5 lm (Merck). The mobile phase was water-formic acid (95 : 5, solvent A) and methanol (solvent B), at a ow of 1 ml min)1 using a linear gradient. The time of analysis was 60 min and the detection was performed at 280 and 340 nm using a diode array detector (Merck-Hitachi, L-7450). For propolis analysis, the same chromatographic conditions were used, with a different linear gradient. The time of analysis was 50 min and the detection was monitored at 280 and 340 nm. The software used for the data analysis was the one provided to the Merck-Hitachi model D-7000 (DAD, Merck-Hitachi, Darmstadt, Germany). Multivariate statistical analysis Multivariate analysis was employed to analyse the data (Pirouette 26; Infometrix Inc., Seattle, WA, USA). Data

were denoted by a matrix where the rows represented the honey or propolis samples while the columns represented the chemical composition of honey or propolis and their respective MIC values. RESULTS Figure 1a,b show the MIC values for honey from A. mellifera and T. angustula against S. aureus. The MIC of honey samples from A. mellifera ranged from 12623 to 18570 mg ml)1. The MIC of honey samples from T. angustula ranged from 14287 to 21433 mg ml)1. Figure 2a,b depicts chromatograms showing the separation of the honey samples from A. mellifera and T. angustula. Table 2 describes the chemical composition of honey samples by HPLC. Principal component analysis of A. mellifera honey samples (Fig. 3a) shows differences in chemical composition based on the geographical origin of the sample. Figure 3b shows the compounds that distinguished the samples in the statistical analysis (Fig. 3a). The MIC for propolis samples from A. mellifera are shown in Fig. 4a and for T. angustula in Fig. 4b. MIC of propolis samples from A. mellifera ranged from 036 to 365 mg ml)1. MIC of propolis from T. angustula ranged from 044 to 201 mg ml)1. The chemical composition of propolis samples from A. mellifera analysed by HPLC is shown in Table 3 and for T. angustula in Table 4. Figure 5 shows multivariate statistical analysis of the chemical
(a) 300

MIC (mg ml1)

200 100 0 1 2 3 4 5 6 7 8

(b)

300 200 100 0 1 2 3 4 5 6 7 8 9 10

Fig. 1 (a) MIC (mg ml)1) of Apis mellifera honey, numbers 18 are samples mAm1, mAm2, mAm3, mAm4, mAm5, mAm6, mAm7 and mAm8, respectively. (b) Tetragonisca angustula honey MIC, numbers 110 are samples mTa1, mTa2, mTa3, mTa4, mTa5, mTa6, mTa7, mTa8, mTa9 and mTa10, respectively

2003 The Society for Applied Microbiology, Journal of Applied Microbiology, 95, 913920, doi:10.1046/j.1365-2672.2003.02050.x

MIC (mg ml1)

916 P . L . M I O R I N ET AL.

1
(a)
014 012
165 235 397

Absorbance (AU)

010 008 006 004 002 000 0 5 10 15 20 25 30 35 40 45 50 55 60 65 70

117

4
1045

2288 2776

(b)

006

Absorbance (AU)

005 004 003 002 001 000 0 5 10 15 20 25 30 35 40 45 50

4304 163

117

007

5584

Retention time (min)

Fig. 2 Chromatograms of honey samples. (a) Apis mellifera honey samples. The identied peaks have the following retention times: (1) HBEN, 2.32 min; (2) p-CUM, 3.97 min; (3) BEN, 6.48 min; (4) I, 10.8 min and (5) H, 13.57 min. (For abbreviations see Table 2). (b) Tetragonisca angustula honey samples. No compounds were detected

Table 2 Quantitative analysis of the compounds identied by HPLC in Apis mellifera honey samples in mg (100 g))1 of honey Compound* [mg (100 g))1 of A. mellifera honey] HBEN BENZ p-CUM Compound H (unknown) (H) 2-[1-Hydroxymethyl]vinyl-6acetyl-5-hydroxy-cumarane (I) Total in mg (100 g))1 MIC in mg ml)1

mAm1 288

mAm2 591 4315

mAm3 1304 4559

mAm5 12775

mAm6 2668 730

mAm7 1337 537

mAm8

288 4906 5863 12775 3398 1874 000 12623 2318 16410 1299 16770 867 17331 22,29 18471 13,71 18570 2587 16753 1843

BENZ, benzoic acid; HBEN, 4-hydroxybenzoic acid; HPLC, high-performance liquid chromatography; p-CUM, p-Coumaric acid; MIC, minimum inhibitory concentration. *The abbreviations in parenthesis after the name of each compound where used in multivariate analysis.

composition and MIC for propolis samples from A. mellifera and T. angustula. Figure 5a shows the chemical composition of propolis from different geographical regions and for

both bees. Figure 5b illustrates which compounds are responsible for the separation of the samples in the statistical analysis.

2003 The Society for Applied Microbiology, Journal of Applied Microbiology, 95, 913920, doi:10.1046/j.1365-2672.2003.02050.x

BEE PRODUCTS EFFECTS ON S. AUREUS

917

(a) 2 1 PC2 0 1 2 3 2 1 0 PC1 1 2 3 m5 m3 m2 m1 m7

(a)
MIC (mg ml1)
m6

4 3 2 1 0 1 2 3 4 5 6 7 8

(b)
MIC (mg ml1)

3 2 1 0 1 2 3 4 5 6

(b) 06 04 02 PC2 00 02 04 06 H 04 02 00 PC1 02 04 06 Hbenz pCum I benz

MIC

Fig. 4 MIC (mg ml)1) of propolis samples. (a) Apis mellifera propolis samples. Samples 18 are pAm1, pAm2, pAm3, pAm4, pAm5, pAm6, pAm7 and pAm8, respectively. (b) Tetragonisca angustula propolis samples. Samples 16 are 1: pTa1, 2: pTa2, 3: pTa5, 4: pTa6, 5: pTa7 and 6: pTa8

Fig. 3 Principal components of honey samples from Apis mellifera. m1m8 are respectively mAm1, mAm2, mAm3, mAm5, mAm6, mAm7 and mAm8. (a) PC2 PC1 scores for samples, mAm8 is not shown because the concentration of chemical compounds is zero. (b) PC2 PC1 loadings of compounds identied by HPLC and MIC. For abbreviations see Table 2

DISCUSSION All honey and propolis samples evaluated in this study showed antibacterial activity against S. aureus. Although honey samples exhibited weaker activity in comparison with propolis, sample mAm1 from A. mellifera (Fig. 1a) and mTa1 from T. angustula (Fig. 1b) exhibited the highest activity. Antibacterial activity results have shown similar MIC values against S. aureus of honey samples for both bee species. There was no signicant statistical difference between the MIC of A. mellifera and T. angustula honey samples (data not shown). Many compounds identied in A. mellifera honey samples are phenolics (Fig. 2a) and some are still unknown which could be avonoids and phenolic acids (Table 2). p-Coumaric acid (p-CUM) was identied in honey samples mAm6 and mAm7 (Table 2). Most of the compounds of

T. angustula honey samples could not be identied by HPLC because of their low concentration (Fig. 2b). Compound H isolated from propolis was identied in samples mAm1 and mAm5 at high concentrations. 4-Hydroxybenzoic acid (HBEN) was identied in samples mAm2, mAm6 and mAm7, and benzoic acid was detected in honey sample mAm3. 2-[1-Hydroxymethyl] vinyl-6-acetyl-5-hydroxycumarane which is commonly found in Brazilian propolis was also identied in honey samples mAm2 and mAm3 (Table 2). This may be due to the fact that for asepsis bees cover the comb with a thin layer of propolis before the honey is deposited. Principal component analysis revealed different proles for and from Minas Gerais (Fig. 3a). honey samples from Parana Figure 3b demonstrates that the two groups of samples have different chemical composition with similar MIC values indicating that all compounds are active against S. aureus. For instance, honey samples from Minas Gerais (mAm6 and mAm7) have a high content of the HBEN and p-CUM have high concentrawhereas honey samples from Parana tions of 2-[1-hydroxymethyl] vinyl-6-acetyl-5-hydroxycumarane (I) (mAm2 and mAm3) and benzoic acid (BENZ) ) has a high concentra(mAm3). The sample mAm5 (Parana ) has a very low tion of compound H and mAm1 (Parana concentration of chemical compounds identied by HPLC. Propolis sample pAm8 was the most active against S. aureus, followed by pAm5, pAm6, pAm7, pAm4, pAm2, pAm3 and pAm1 (Fig. 4a). MIC values for propolis from T. angustula showed that sample pTa6 was the most active followed by pTa2, pTa1, pTa5, pTa7 and pTa8

2003 The Society for Applied Microbiology, Journal of Applied Microbiology, 95, 913920, doi:10.1046/j.1365-2672.2003.02050.x

918 P . L . M I O R I N ET AL.

Table 3 Quantitative analysis of the compounds identied by HPLC in Apis mellifera propolis samples (in mg g)1 of propolis) Compound* (mg g)1 of propolis from A. mellifera) 2,2-Dimethyl-6-carboxyethenyl-2H1-benzopyran (C) 2-2-Dimethyl-8-prenyl-2H-1-benzopyran6-propenoic acid (F) 3,5-Diprenyl-4-hydroxycinnamic acid (D) 3-Methoxy-4-hydroxycinnamaldehyde (G2) 3-Prenyl-4-hydroxycinnamic acid (B) Benzoic acid (Hn) Caffeic acid (Cf) Caffeic acid derivative 1 (Caf1) Cinnamic acid derivative 1 (E) Compound H (unknown) (H) Crysin (Crys) Ferulic acid (Fr) Galangin (Gal) Kaempferol derivative 1 (K1) p-Coumaric acid (p-C) p-Coumaric acid derivative 1 (D1) p-Coumaric acid derivative 2 (D2) p-Coumaric acid derivative 3 (D3) p-Coumaric acid derivative 4 (D4) p-Coumaric acid derivative 5 (D5) p-Coumaric acid derivative 6 (D6) p-Coumaric acid derivative 7 (D7) Pinobanksin (Pk) Pinobanksin derivative 1 (Pk1) Vanillin (V) Total amount in mg g)1 MIC in mg ml)1

pAm1 320

pAm2 401 1011 790 134 200 023 115 2192 032 160 054

pAm3 105 218 380 590 848

pam4 188 974 240 223 4857 1506 24249 1871

pAm5 766 088 726 1078 022 155 2321 2584 383

pAm6 1504 1959 603 079 923 5950 520 1702 1602 116 225 132 137 151 1357

pAm7 1077 1692 400 046 734 5585 488 1463 1074 029 079 155 242 1692 025 186 1419

pAm8 1338 1594 550 069 840 5610 550 1525 1457 066 131 252 1481 148 1167

320 5112 2141 32236 7740 16960 16577 16778 365 008 081 013 228 087 074 011 038 022 058 017 066 019 036 010

*The abbreviations in parenthesis after the name of each compound were used in the multivariate analysis. Table 4 Quantitative analysis of the compounds identied by HPLC in Tetragonisca angustula propolis samples(in mg g)1 of propolis) Compound* (mg g)1 of propolis from T. angustula) 3,5-Diprenyl-4-hydroxycinnamic acid (D) 3-Prenyl-4-hydroxycinnamic acid (B) 2,2-Dimethyl-8-prenyl-2H-1-benzopiran6-propenoic acid (F) Cinnamic acid derivative 1 (E) Total in mg g MIC in mg ml)1
)1

pTa1 0011 0010 0030

pTa2 0006 0019

pTa3 pTa4 pTa5 0007 0056 0053

pTa6 0075

pTa7 0008 0143 0955

pta8

pTa9 0025

0051 0025 0000 0063 0053 0075 1110 0000 0025 077 016 047 012 107 026 044 016 117 029 201 029

*The abbreviations in parenthesis after the name of each compound were used in the multivariate analysis.

(Fig. 4b). We veried that there was no statistically signicant difference between MIC of propolis samples from et al. (1994) invesA. mellifera and T. angustula. Bonveh tigated propolis from A. mellifera against S. aureus where MIC values ranged from 0080 to 0100 mg ml)1. These

MIC results are different than the MICs found to be active in this study because the authors used different methodologies to determine antibacterial activity. The compounds analysed by HPLC in propolis from A. mellifera were identied and quantied (Table 3). The

2003 The Society for Applied Microbiology, Journal of Applied Microbiology, 95, 913920, doi:10.1046/j.1365-2672.2003.02050.x

BEE PRODUCTS EFFECTS ON S. AUREUS

919

(a) 35 30 25 20 15 10 05 00 05 10 15 A4 A2 PRApis MGApis MGTa

A3

A1

T7 T1T8 A7 0 1 PC1 2 3 4 A8 A6 5

(b) 06 04 PC2 02 00 02 MIC 04 04 03 02 01 00 01 PC1 02 Cf Caf1 Fer PC Pk K1 03 04 G2 (B) C F E B D

Fig. 5 Principal components of propolis samples. (a) PC2 PC1 of scores. Samples A1A8, are samples pAm1, pAm2, pAm3, pAm4, pAm5, pAm6, pAm7 and pAm8, respectively. T1: pTa1, T2: pTa2, T5: pTa5, T6: pTa6, T7: pTa7, T8: pTa8. PR-Apis is propolis from (Apis mellifera); MG-Apis is propolis from Minas Gerais Parana (A. mellifera) and MG-Ta is propolis from Minas Gerais (Tetragonisca angustula). (b) PC2 PC1 of loadings (compounds identied by HPLC and MICs). For abbreviations see Tables 2 and 3

compounds found in these propolis samples are in accordance with those reported by Marcucci et al. (2000, 2001) in Brazilian propolis samples from A. mellifera. The main compounds identied in this study were acids such as caffeic, p-CUM, 3-prenyl-4-hydroxycinnamic, 3,5-diprenyl4-hydroxycinnamic (artepillin C), 2,2-dimethyl-8-prenyl2H-1-benzopyran-6-propenoic, a kaempferol derivative (avonoid) and 2,2-dimethyl-6-carboxyethenyl-2H-1-benzopyran (Marcucci and Bankova 1999; Bankova et al. 2000). The propolis samples were further classied as Brazilian propolis with prenylated compounds and 3methoxy 4hydroxycinnamaldehyde as main bioactive markers (BRPG) (pAm2, pAm3, pAm4 and pAm5) and Brazilian propolis with prenylated compounds (BRP) (pAm6, pAm7 and pAm8) as established by Marcucci (2000).

Bosio et al. (2000) reported that the antibacterial activity of propolis may be related to the presence of avonoids. Brumtt et al. (1990) observed no bacterial inhibition from propolis extracts. However, other investigators (TakaisiKikuni and Schilcher 1994; Nieva Moreno et al. 1999) have shown that propolis extracts have antibacterial properties against some micro-organisms. Bankova et al. (1996) reported antibacterial activity against S. aureus of triterpenic acids from Brazilian propolis where communic and imbricatoloic acids and isocupressic methyl ester were the most active compounds against S. aureus. Table 4 shows the values in mg g)1 of the compounds identied in propolis from T. angustula. The following acids were identied: 3-prenyl-4-hydroxycinnamic, 3,5-diprenyl4-hydroxycinnamic, 2,2-dimethyl-8-prenyl-2H-1-benzopiran6-propenoic and a cinnamic acid derivative. These compounds were found in lower concentrations than those identied in propolis from Apis mellifera. Other compounds not detected by HPLC could be responsible for the antibacterial activity of propolis from T. angustula. Velikova et al. (2000a) described the antibacterial activity of ent-kaurenes isolated from native Brazilian bee propolis (Melipona quadrifasciata anthidioides, mandac aia bee) against S. aureus and Escherichia coli (E. coli). Kaurenoic acid which is a newly identied compound showed higher inhibitory activity than streptomycin against S. aureus. In another report, Velikova et al. (2000b) investigated the chemical composition and antibacterial activity of propolis from native Brazilian bees including T. angustula. The propolis extracts were active against S. aureus, E. coli and Candida albicans (C. albicans). Principal component analysis for the propolis samples revealed similar patterns of those found for honey samples (Fig. 5a,b) with differences in chemical composition of propolis depending on the bee species. Samples with a high concentration of the compounds identied by HPLC were shown to be more active against S. aureus with lower MIC values indicating that the identied compounds were responsible for the antibacterial activity. Collectively, our ndings have shown that propolis samples had higher antibacterial activity against S. aureus when compared with honey samples. The chemical composition of A. melifera and T. angustula honey and propolis samples exhibited different proles based on geographical origin. Accordingly, different MICs where found for different samples albeit inhibitory against S. aureus. These antimicrobial properties would warrant further studies on the clinical applications of propolis and honey against S. aureus. ACKNOWLEDGEMENTS o de Amparo a ` Pesquisa The authors acknowledge Fundac a o Paulo, Fapesp (nos 00/06975-3 and 95/ do Estado de Sa

2003 The Society for Applied Microbiology, Journal of Applied Microbiology, 95, 913920, doi:10.1046/j.1365-2672.2003.02050.x

PC2

920 P . L . M I O R I N ET AL.

09306-5) for nancial support. Mr Henrique Breyer and Mr Getulio Ferreira provided the honey and propolis samples. Dr Olga Panhocas provided support with statistical analysis. REFERENCES
Azevedo, J.L., Flechtmann, C.H.W. and Prado Filho, I.G. (1963) ncias de substa ncias antimicrobianas em produtos de alguns Ocorre insetos sociais. Revista de agricultura 38, 12936. Bankova, V.S., Marcucci, M.C., Simova, S., Nikolova, N. and Kujumgiev, A. (1996) Antibacterial diterpenic acids from Brazilian propolis. Zeitschrift fu r Naturforschung 51C, 277280. Bankova, V.S., De Castro, S.L. and Marcucci, M.C. (2000) Propolis: recent advances and plant origin. Apidologie 31, 315. , J.S., Coll, F.V. and Jorda , R.E. (1994) The composition, Bonveh active components and bacteriostatic activity of propolis in dietetics. Journal of the American Oil Chemists Society 71, 529532. Bosio, K., Avanzini, C., DAvolio, A., Ozino, O. and Savoia, D. (2000) In vitro activity of propolis against Streptococcus pyogenes. Letters in Applied Microbiology 31, 174177. Brumtt, W., Hamilton-Miller, J.M.T. and Franklin, I. (1990) Antibiotic activity of natural products: 1. Propolis. Microbios 63, 1922. Dobrowolski, J.W., Vohora, S.B., Sharma, K., Shah, S.A., Naqvi, S.A.H. and Dandlya, P.C. (1991) Antibacterial, antifungal, antiamoebic, antiinammatory and antipyretic studies on propolis bee products. Journal of Ethnopharmacology 35, 7782. ticos, 3rd Edn. ed. Fava Netto, C. (1975) Antibiograma, in Antibio o Paulo, Edgar Blu Lacaz, C.S. pp. 221233. Sa cher Ltda. Fernandes, A.Jr, Sugizaki, M.F., Fogo, M.L., Funari, S.R.C. and Lopes, C.A.M. (1995) In vitro activity of propolis against bacterial and yeast pathogens isolated from human infections. Journal of Venomous Animals and Toxins 1, 6369. Fernandes, A.Jr, Lopes, C.A.M., Sforcin, J.M. and Funari, S.R.C. (1997) Population analysis of susceptibility to propolis in reference strains of Staphylococcus aureus and Escherichia coli. Journal of Venomous Animals and Toxins 3, 287294. Ghisalberti, E.L. (1979) Propolis: a review. Bee Word 60, 5984. Marcucci, M.C. (2000) Process to typing natural products. Requested patent. Brazilian National Institute for Intellectual Property, INPI no. PI 0105471-6, 12/22/2000. Marcucci, M.C. and Bankova, V.S. (1999). Chemical composition, plant origin and biological activity of Brazilian propolis. Current Topics in Phytochemistry 2, 115123.

dio, A.R., Ferreira, M.M.C., Marcucci, M.C., Ferreres, F., Custo a-Viguera, C. and Bretz, W.A. (2000) Bankova, V.S., Garc Evaluation of phenolic compounds in Brazilian propolis from different geographic regions. Zeitschrift fu r Naturforschung 55C, 7681. a-Viguera, C., Bankova, V.S., De Marcucci, M.C., Ferreres, F., Garc Castro, S.L., Dantas, A.P., Valente, P.H.M. and Paulino, N. (2001) Phenolic compounds from brazilian propolis with pharmacological activities. Journal of Ethnopharmacology 74, 105112. s-Barbera n, F. Martos, I., Conssentini, M., Ferreres, F. and Toma (1997) Flavonoid composition of Tunisian honey and propolis. Journal of Agricultural and Food Chemistry 54, 28242829. s-Brabera n, F.A. (2000a) IdentiMartos, I., Ferreres, F. and Toma cation of avonoid markers for the botanical origin of eucalyptus honey. Journal of Agricultural and Food Chemistry 48, 14981502. sMartos, I., Ferreres, F., Yao, L.H., Darcy, B., Cafn, N. and Toma n, F.A. (2000b). Flavonoids in monospecic eucalyptus Barbera honeys from Australia. Journal of Agricultural and Food Chemistry 48, 47444748. National Committee for Clinical Laboratory Standards (NCCLS) (2000) Approved Standard M7-A5: Methods for Dilution Antimicrobial Susceptibility Test for Bacteria that Grow Aerobically, 5th Edn. Wayne, PA: NCCLS. Nieva Moreno, M.I., Isla, M.I., Cudmani, N.G., Vattuone, M.A. and Sampietro, A.R. (1999) Sreeening of bacterial activity of Amaicha Del Valle (Tucuman, Argentina) propolis. Journal of Ethnopharmacology 68, 97102. co sobre a Szewczak, E.H. and Godoy, G.F. (1984) Um estudo cient polis. Apicultura no Brasil 1, 2829. pro Takaisi-Kikuni, N.B. and Schilcher, H. (1994) Electron microscopy and microcalorimetric investigations of the possible mechanism of the antibacterial action of a dened propolis provenance. Planta Medica 60, 222227. Velikova, M., Bankova, V., Tsvetkova, I., Kujumgiev, A. and Marcucci, M.C. (2000a) Antibacterial ent-kaurene from Brazilian propolis of native stingless bees. Fitoterapia 71, 693696. Velikova, M., Bankova, V., Marcucci, M.C., Tsvetkova, I. and Kujumgiev, A. (2000b) Chemical composition and biological activity of propolis from Brazilian meliponinae. Zeistschrift fu r Naturforschung 55C, 785789. Zumla, A. and Lulat, A. (1989) Honey a remedy rediscovered. Journal of Royal Society of Medicine 82, 384385.

2003 The Society for Applied Microbiology, Journal of Applied Microbiology, 95, 913920, doi:10.1046/j.1365-2672.2003.02050.x