International Journal of Hygiene and Environmental Health
Int. J. Hyg. Environ. Health 206, 217 ± 221 (2003) ¹ Urban & Fischer Verlag http: // www.urbanfischer.de/journals/intjhyg
Bactericidal activity of herbal extracts Nikolaus Thuille, Manfred Fille, Markus Nagl Institute of Hygiene and Social Medicine, Leopold-Franzens-University of Innsbruck, Innsbruck, Austria Received January 20, 2003 ¥ Accepted March 16, 2003
Abstract The antimicrobial activity of total herbal extracts has been investigated. The MIC of extracts of Evodia rutaecarpa and grape kernel ranged between 0.25 and 1 mg/ml against grampositive cocci and P. aeruginosa. Cocci were killed after 30 ± 90 min of incubation in grape kernel extract (0.5 ± 1.5 mg/ml), and after 8 h in evodia extract (0.5 ± 1 mg/ml), respectively. C. albicans was only susceptible to evodia (MIC 0.5 mg/ml). The organic solvents of the preparations contributed to the antibacterial effect of herbal extracts with MICs of ethanol between 4 and 10 vol% and those of methanol between 6 and 10 vol%. Taking this into consideration, mastic and thyme extracts exerted hardly any microbicidal activity, while grape kernel extract and evodia were still effective at 2- to 6-fold dilutions below the MIC of the solvent. Because of the weak antimicrobial activity of total herbal extracts we recommend to give preference to preparations of single or more purified compounds. Key words: Evodia rutaecarpa ± grape kernel extract ± mastic ± thyme ± bactericidal activity
Introduction There has been a rising interest in antimicrobial herbal drugs in the last years which may be explained by the occurrence of multiresistant pathogens and by increasing popularity of alternative medicine. Indeed, plants use a huge, mainly unknown reservoir of substances for their defence against microorganisms, insects, and herbivors. Although some single substances like phenols, phenolic derivatives (quinones, flavones, flavonoids, flavonoles, tannins, and cumarins), terpenoids, essential oils, alkaloids, lectins, and polypeptides have been identified (Cowan, 1999), whole extracts of plants are still in use. Four of them have been chosen for investigation in our study, mastic, thyme, evodia and grape kernel extract.
Mastic, the resin of Pistacia lentiscus, has been used to treat abdominal pain, dyspepsia, and gastric and duodenal ulceration (Al Said et al., 1986). Signs of efficacy in abdominal disorders have been reported to be partially due to destruction of Helicobacter pylori by mastic (Huwez et al., 1998). The high content of a-pinen, an unsaturated pinan compound, has been postulated to be responsible for the antimicrobial activity of the resin (MIC 1.8 ± 4.5 mg/ml) (Magiatis et al., 1999). Thymol, a compound of the organic fraction of Thymus vulgaris, is known for its antimicrobial properties (Ettayebi et al., 2000) against grampositive and gram-negative bacteria (Ettayebi et al., 2000; Inouye et al., 2001). Also the organic whole extract of thyme was found to be active
Corresponding author: Dr. Nikolaus Thuille, Institute of Hygiene and Social Medicine, University of Innsbruck, FritzPregl-Str. 3, A-6010, Innsbruck, Austria. Phone: 43 512 507 3430, Fax 43 512 507 2870, E-mail: Nikolaus.
[email protected]
1438-4639/03/206/03-217 $ 15.00/0
218
N. Thuille
against different test strains (Essawi and Srour, 2000), while the aqueous one revealed the best activity against H. pylori with an MIC of 3.5 mg/ml (Tabak et al., 1996). The extract or the dried fruit of Evodia rutaecarpa is used in the traditional chinese medicine preferentially in combination with other herbs as a medication called goshuyu for the treatment of gastric and duodenal ulcer and diarrhea (Yu et al., 2000). Several chinolone alkaloids could be isolated which demonstrated activity against H. pylori with MICs from 0.05 to 20 mg/ml (Hamasaki et al., 2000; Rho et al., 1999). Testing the extract of red wine against H. pylori, a MIC50 of 25 mg/ml and a MIC90 of 50 mg/ml were detected (Mahady and Pendland, 2000). Since the MIC50 and the MIC90 of resveratrol, a phytoalexin present in grapes and red wine known for antioxidative and preserving properties, were 12.5 and 25 mg/ml, respectively, this compound has been assumed to contribute significantly to the antimicrobial activity of red wine (Mahady and Pendland, 2000). The aim of this study was to compare the activity of the mentioned herbal extracts against standard pathogens under the same conditions and to address the influence of the organic solvent.
Materials and methods Reagents Mastic resin (preparation 1999) was a gift from John Perikos, Kalimassia, Chios, Greece. The resin was grinded with a IKA-Universal-mill M20 and stored at 70 8C. For usage, it was dissolved in 96% ethanol to 50 mg/ml. All other extracts were kindly provided by Dr. Volker Christoffel (Bionorica Arzneimittel, Neumarkt, Germany). The extract of evodia (preparations PKE00-020 and PKE00-021) was dissolved in 60% ethanol to 10 mg/ml. Thyme was provided as aqueous and alcoholic extract. Stock solutions of 20 mg/ml were prepared in water and 70% ethanol, respectively, which was the maximum soluble concentration. Grape kernel extract (preparation 13.99001.PC/SL1,SL2) was dissolved in pure methanol to 25 mg/ml. All stock solutions were sterile filtered and stored at 4 8C.
Microorganisms and media Bacterial strains (Staphylococcus aureus ATCC 25923, Staphylococcus aureus ATCC 6538, Streptococcus pyogenes D 68 (a slime-producing and highly encapsulated strain kindly provided by Dr. J. Hildebrandt, Sandoz Scientific Centre, Vienna), Escherichia coli ATCC 11229, Proteus mirabilis ATCC 14159, Pseudomonas aeruginosa
ATCC 27853) and the yeast Candida albicans CBS 5982 deep frozen for storage were grown on M¸ller-Hinton agar (Oxoid, Hampshire, UK). Colonies from this agar were grown in M¸ller-Hinton broth (Oxoid) at 37 8C overnight to > 109 colony forming units (cfu)/ml (107 for C. albicans). M¸ller-Hinton broth or 0.1 M phosphate buffer (pH 7.0) served as a test medium for the antimicrobial activity of the herbal extracts.
Minimal inhibitory and minimal bactericidal concentrations MICs of all extracts were determined by the standard broth dilution method (NCCLS, 1997) after 24 h incubation of 2 105 cfu/ml. MBCs were determined by subsequent evaluation of 99.9% kill. MICs and MBCs of the solvents ethanol and methanol were tested by the same methods.
Bactericidal activity (time kill curves) Stock solutions of the herbal extracts were diluted in M¸ller-Hinton broth or 0.1 M phosphate buffer to different final concentrations. The microorganisms were washed twice, suspended in saline to 1 ± 2 107 cfu/ml, and diluted 100-fold in the test solutions. At time zero and after incubation times ranging from 15 min to 8 h, aliquots were removed. Portions (50 ml) of undiluted aliquots as well as of 10-fold dilutions in saline were spread in duplicate onto M¸ller-Hinton agar plates with an automatic spiral plater (Don Whitley Scientific Limited, Shipley, UK), allowing a detection limit of 10 cfu/ml. The plates were incubated at 37 8C, and the cfu were counted after 24 and 48 h. Controls without the herbal extracts were treated the same way.
Statistics One way ANOVA and Dunnett's Multiple Comparison test (Graphpad Software Inc., CA, USA) were applied for statistical evaluation of the bactericidal activity. P values < 0.05 were considered significant.
Results The MICs and MBCs of the extracts are shown in Table 1. Mastic and thyme exerted the lowest antimicrobial activity with an MIC partially exceeding their solubility threshold of 5 mg/ml. The alcoholic thyme extract was slightly more active than the aqueous one against gram-positive strains. In contrast, the inhibitory concentrations of evodia and grape kernel extract were mostly tenfold lower (Table 1), and both extracts demonstrated killing properties dependent on the concentration, bacterial strain and incubation time (Fig. 1). Par-
Bactericidal activity of herbal extracts
219
Table 1. MIC and MBC after 24 h incubation in M¸ller-Hinton broth. Strain
Agent
MIC (mg/ml)
MBC (mg/ml)
S. aureus ATCC 6538
Thyme Thyme * Evodia Mastic Grape kernel Thyme Thyme * Evodia Mastic Grape kernel Thyem Thyme * Evodia Mastic Grape kernel Thyme Thyme * Evodia Mastic Grape kernel Thyme Thyme * Evodia Mastic Grape krenel Thyme Thyme * Evodia Mastic Thyme Thyme * Evodia Mastic Grape kernel
2.5 > 5.0 1.0 > 5.0 0.5 ± 1.0 < 2.5 > 5.0 0.5 ± 1.0 5.0 1.0 1.0 5.0 0.25 1.0 0.25 > 5.0 > 5.0 > 1.0 > 5.0 > 2.5 > 5.0 > 5.0 1.0 5.0 < 1.5 > 5.0 > 5.0 > 1.0 > 5.0 > 5.0 > 5.0 0.5 5.0 2.5
5.0 > 5.0 > 1.0 > 5.0 1.0 2.5 > 5.0 1.0 > 5.0 1.0 ± 1.5 2.5 > 7.0 0.5 2.0 0.5 > 5.0 > 5.0 > 1.0 > 5.0 > 2.5 > 5.0 > 5.0 1.0 > 5.0 1.5 > 5.0 > 5.0 > 1.0 > 5.0 > 5.0 > 5.0 > 1.0 > 5.0 > 2.5
S. aureus ATCC 25923
S. pyogenes D 68
E. coli ATCC 11129
P. aeruginosa ATCC 27853
P. mirabilis ATCC 14153
C. albicans CBS 5982
* aqueous extract
ticularly, the rapid killing of S. pyogenes and S. aureus within 30 min by grape kernel extract was remarkable (Fig. 1). In general, the selected gram-positive bacteria proved to be more susceptible than the gramnegative ones. The MIC of evodia (0.5 mg/ml) against C. albicans was similar to gram-positive bacteria, whereas the grape kernel extract was less active against the yeast (Table 1). The results of the experiments testing the bactericidal activity of the solvents are summarized in Table 2. MICs of the alcohols ranged from 4 to 10 vol%. Methanol was slightly less effective than ethanol. When phosphate buffer was used instead of M¸ller-Hinton broth, the activity increased. Cfu of both S. aureus strains were reduced for more than 4 log ranges after 24 h in the presence of 10% ethanol, whereas this concentration was only bacteriostatic in broth.
Table 2. MIC and MBC of the solvent after 24 h incubation in M¸ller-Hinton broth. Strain
Agent
MIC (vol%)
MBC (vol%)
S. aureus 6538
Ethanol Methanol Ethanol Methanol Ethanol Methanol Ethanol Methanol Ethanol Methanol Ethanol Methanol Ethanol Methanol
9 ± 10 8 ± 10 8 9 ± 10 4 6 10 > 10 6 7±8 10 n.d.a 6 8
> 10 > 10 > 10 > 10 6 8 ± 10 > 10 > 10 8 ± 10 10 > 10 not determined 8 ± 10 10
S. aureus 25923 S. pyogenes D 68 E. coli 11129 P. aeruginosa 27853 P. mirabilis 14153 C. albicans 5982
220
N. Thuille
Fig. 1. Antimicrobial activity of 0.5 mg/ml (S. pyogenes) and 1.0 mg/ml (S. aureus 25923, P. aeruginosa) evodia extract (closed symbols), and of 0.5 mg/ml (S. pyogenes) and 1.5 mg/ml (S. aureus 25923) grape kernel extract (open symbols) in M¸ller-Hinton broth at pH 7 and 37 8C. The dotted lines show the controls without extract. The results are presented as mean values and standard errors of the mean of two to three independent experiments. All differences between control and test samples from 0.25 to 2 hours for grape kernel extract, and from 4 to 8 hours for evodia extract were significant (P < 0.01).
Discussion This study revealed significant differences in the herbal extracts tested. Mastic and the aqueous extract of thyme demonstrated antimicrobial properties at concentrations hardly below the threshold of solubility which was 5 mg/ml for mastic and 7 mg/ml for thyme. Investigating the essential oils of the leaves, branches and the resin of Pistacia lentiscus by macro dilution assay, the highest activity has been reported for the oil of the resin with MICs of 1.8 to 4.5 mg/ml against bacteria and fungi. The activity of the whole resin was lower and comparable to our results (Magiatis et al., 1999). Discrepancies in activity may be caused by a variation in the content of antimicrobial components, e.g. apinen, myrcen and linanol, of different mastic preparations (Lauk et al., 1996). The improved activity of the ethanolic thyme extract compared to the aqueous one may be explained by the presence of thymol, an alkyl phenol that causes perforation of bacterial membranes and rapid efflux of cytosolic components (Shapiro and Guggenheim, 1995). Accordingly, MICs of essential oils of thyme with high concentration of phenol were lower (0.12 and 4 mg/ml) compared to those of oils containing only traces of phenol (2 ± 16 mg/ml) (Kulevanova et al., 2000). Similar to our results, the activity of thyme oils has been reported to be higher against gram-positive strains (S. aureus, S.
pyogenes, S. pneumoniae) than against gram-negative ones (E. coli and others) (Cosentino et al., 1999; Inouye et al., 2001). On the other hand, we were not able to unequivocally prove activity of whole thyme extract at a concentration clearly below the MIC of the solvent. This was probably due to a content of thymol too low to exert a cidal effect. Particularly at low concentrations neutralisation phenomena in broth must be considered as shown with bovine serum albumin which binds phenolic compounds and inhibits their reaction with bacterial membrane proteins (Juven et al., 1994). Evodia and grape kernel extract demonstrated significant activity against bacteria. The MICs of whole evodia extract in our study were 20 times higher than those of the purified chinolone alkaloids against H. pylori (Hamasaki et al., 2000; Rho et al., 1999). The apolar fraction of grape kernels containing aldehydes, fatty acids and sterols as well as the polar fraction containing phenolic compounds proved to be active against gram-positive and gram-negative bacteria in disc diffusion tests (Palma et al., 1999). S. pyogenes d68 showed elevated susceptibility against all test extracts which is in accordance to previous findings using N-chlorotaurine, another ™natural∫ antimicrobial agent (Nagl et al., 2000). The strain is extraordinarily virulent in the mouse peritonitis model (Nagl et al., 2000), but obviously very susceptible to a number of antimicrobials. If this is generally true for clinical isolates of S. pyogenes and if adjuvant topical application of plant extracts as throat disinfectants in tonsillitis is of advantage, remains to be elucidated. Judging the antimicrobial activity of alcoholic extracts, participation of the solvents should not be underestimated. For solubility reasons the concentrations of alcohols in herbal preparations are usually high and may pretend microbicidal effects of the plant compounds. Taking this into consideration, in our study the only extracts still active at dilutions below the MIC of the solvents were grape kernel extract against gram-positives and P. aeruginosa, and evodia against gram-positives and C. albicans. Moreover, grape kernel extract demonstrated a rapid killing of cocci (Fig. 1), which might indicate a synergistic effect with methanol. The stronger bactericidal activity of ethanol compared to methanol can be explained by its higher lipophilicity which enhances penetration into the target organism. The lower activity of alcohols in broth than in buffer solution is caused by reaction with broth components which leads to consumption of the active agent. To summarize, only whole extracts of grape kernels and evodia had significant antimicrobial
Bactericidal activity of herbal extracts
properties. We were not able to prove a sufficient bacteriostatic effect of thyme and mastic at dilutions to rule out activity of the solvent. Since components and their concentrations vary between different batch dates and sources, we recommend to give preference to preparations of single components of herbal extracts. Acknowledgements. This study was supported by the Jubil‰umsfonds of the Austrian National Bank (grants no. 6801/1 and 8366) and by the Austrian Science Fund (grant no. P15240). We are grateful to Prof. Hartmut Glossmann, Department of Biochemical Pharmacology, University of Innsbruck, for productive discussions and help in preparation of the manuscript.
References Al Said, M. S., Ageel, A. M., Parmar, N. S., Tariq, M.: Evaluation of mastic, a crude drug obtained from Pistacia lentiscus for gastric and duodenal anti-ulcer activity. J. Ethnopharmacol. 15, 271 ± 278 (1986). Cosentino, S., Tuberoso, C. I., Pisano, B., Satta, M., Mascia, V., Arzedi, E., Palmas, F.: In-vitro antimicrobial activity and chemical composition of Sardinian Thymus essential oils. Lett. Appl. Microbiol. 29, 130 ± 135 (1999). Cowan, M. M.: Plant products as antimicrobial agents. Clin. Microbiol. Rev. 12, 564 ± 582 (1999). Essawi, T., Srour, M.: Screening of some Palestinian medicinal plants for antibacterial activity. J. Ethnopharmacol. 70, 343 ± 349 (2000). Ettayebi, K., El Yamani, J., Rossi-Hassani, B.: Synergistic effects of nisin and thymol on antimicrobial activities in Listeria monocytogenes and Bacillus subtilis. FEMS Microbiol. Lett. 183, 191 ± 195 (2000). Hamasaki, N., Ishii, E., Tominaga, K., Tezuka, Y., Nagaoka, T., Kadota, S., Kuroki, T., Yano, I.: Highly selective antibacterial activity of novel alkyl quinolone alkaloids from a Chinese herbal medicine, Gosyuyu (Wu-Chu-Yu), against Helicobacter pylori in vitro. Microbiol. Immunol. 44, 9 ± 15 (2000). Huwez, F. U., Thirlwell, D., Cockayne, A., Ala'Aldeen, D. A.: Mastic gum kills Helicobacter pylori [letter]. N. Engl. J. Med. 339, 1946 (1998). Inouye, S., Takizawa, T., Yamaguchi, H.: Antibacterial activity of essential oils and their major constituents
221
against respiratory tract pathogens by gaseous contact. J. Antimicrob. Chemother. 47, 565 ± 573 (2001). Juven, B. J., Kanner, J., Schved, F., Weisslowicz, H.: Factors that interact with the antibacterial action of thyme essential oil and its active constituents. J. Appl. Bacteriol. 76, 626 ± 631 (1994). Kulevanova, S., Kaftandzieva, A., Dimitrovska, A., Stefkov, G., Grdanoska, T., Panovski, N.: Investigation of antimicrobial activity of essential oils of several Macedonian Thymus species. Boll. Chim. Farm. 139, 276 ± 280 (2000). Lauk, L., Ragusa, S., Rapisarda, A., Franco, S., Nicolosi, V. M.: In vitro antimicrobial activity of Pistacia lentiscus L. extracts: preliminary report. J. Chemother. 8, 207 ± 209 (1996). Magiatis, P., Melliou, E., Skaltsounis, A. L., Chinou, I. B., Mitaku, S.: Chemical composition and antimicrobial activity of the essential oils of Pistacia lentiscus var. chia [letter]. Planta Med. 65, 749 ± 752 (1999). Mahady, G. B., Pendland, S. L.: Resveratrol inhibits the growth of Helicobacter pylori in vitro [letter]. Am. J. Gastroenterol. 95, 1849 (2000). Nagl, M., Hess, M., Pfaller, K., Hengster, P., Gottardi, W.: Bactericidal activity of micromolar N-chlorotaurine ± evidence for its antimicrobial function in the human defence system. Antimicrob. Agents Chemother. 44, 2507 ± 2513 (2000). NCCLS: Methods for dilution antimicrobial susceptibility tests for bacteria that grow aerobically. 4th edn; approved standard 17, M7 ± A4 (1997). Palma, M., Taylor, L. T., Varela, R. M., Cutler, S. J., Cutler, H. G.: Fractional extraction of compounds from grape seeds by supercritical fluid extraction and analysis for antimicrobial and agrochemical activities. J. Agric. Food Chem. 47, 5044 ± 5048 (1999). Rho, T. C., Bae, E. A., Kim, D. H., Oh, W. K., Kim, B. Y., Ahn, J. S., Lee, H. S.: Anti-Helicobacter pylori activity of quinolone alkaloids from Evodiae fructus. Biol. Pharm. Bull. 22, 1141 ± 1143 (1999). Shapiro, S., Guggenheim, B.: The action of thymol on oral bacteria. Oral Microbiol. Immunol. 10, 241 ± 246 (1995). Tabak, M., Armon, R., Potasman, I., Neeman, I.: In vitro inhibition of Helicobacter pylori by extracts of thyme. J. Appl. Bacteriol. 80, 667 ± 672 (1996). Yu, L. L., Liao, J. F., Chen, C. F.: Anti-diarrheal effect of water extract of Evodiae fructus in mice. J. Ethnopharmacol. 73, 39 ± 45 (2000).
