International Journal of Antimicrobial Agents 22 (2003) 48 /53 www.ischemo.

org

Berberis aetnensis C. Presl. extracts: antimicrobial properties and interaction with ciprooxacin
Rosario Musumeci a,*, A. Speciale a, R. Costanzo a, A. Annino b, S. Ragusa c, A. Rapisarda d, M.S. Pappalardo b, L. Iauk a
a

Department of Microbiological and Gynaecological Sciences, Section of Microbiology, University of Catania,Via Androne 81, 95124 Catania, Italy b Department of Pharmaceutical Sciences, University of Catania, Catania, Italy c Department of Pharmacobiological Sciences, University of Catanzaro, Catanzaro, Italy d Pharmacobiological Department, University of Messina, Messina, Italy Received 27 August 2002; accepted 16 December 2002

Abstract Previous research showed that berberine-containing Berberis species synthesise the substances 5?-methoxyhydnocarpin-D (5?MHC-D) and pheophorbide a , which have no antimicrobial activity but inhibit the expression of multidrug resistant efflux pumps (MDRs) in Staphylococcus aureus and potentiate the action of berberine. The MDR pumps extrude synthetic and natural antimicrobials from bacterial cells. We searched for these compounds in Berberis aetnensis C. Presl. (Berberidaceae), an endemic plant of the volcano Mount Etna. This work confirms the presence of pheophorbide a and permits us to hypothesise the presence of 5?-MHC-D in leaf extracts. In fact, the activity of ciprofloxacin was improved when two chromatographic fractions isolated from leaf extracts were added. These results are indicative of the presence of MDR pump inhibitors. Moreover, crude extracts were tested on several micro-organisms and showed antimicrobial activity mainly against Gram-positive bacteria and yeasts. # 2003 Elsevier Science B.V. and the International Society of Chemotherapy. All rights reserved.
Keywords: Berberis aetnensis ; Berberine; Efux pump inhibitors; Pheophorbide a ; 5?-MHC-D

1. Introduction In traditional medicine the extracts of various Berberidaceae (Berberis aquifolium , Berberis vulgaris and Berberis aristata) are used for rheumatic complaints and other types of chronic inflammations [1]. Some authors demonstrated that these extracts have a significant activity against bacteria, viruses, fungi, protozoa, helminthes and chlamydia [2]. Studies carried out on the properties and chemical composition of the extracts show that their principal activity is due to their alkaloid constituents with an isoquinolinic nucleus such as berberine, oxyacanthine, berbamine and palmatine [1]. It has been shown that berberine has febrifugal, hypotensive, immuno-stimulating, anti-inflammatory and antimicrobial properties and there are on-going

* Corresponding author. Tel.: '/39-095-31-2386; fax: '/39-095-325032. E-mail address: docsaro@yahoo.it (R. Musumeci).

studies regarding a possible anti-tumour activity [1 /5]. The most prominent clinical uses include bacteria related diarrhoea, parasitic intestinal infections and ocular infections (conjunctivitis, trachoma) [2,6]. As Stermitz et al. [7,8] stated, the weak antibacterial activity of berberine in some bacterial strains could be due to the expression of multidrug resistant efflux pumps (MDRs) that extrude the drugs, rather than their intrinsic activity. MDR efflux pumps are important mechanisms of active and unidirectional transport. Their presence makes bacterial strains resistant to various classes of antibiotics and prevents the intracellular action of berberinic alkaloids that are thus promptly expelled [9]. Several studies have shown that plants of the genus Berberis (Berberis repens , B. aquifolium and Berberis fremontii ) producing berberine (Fig. 1a) synthesise two substances, the flavonolignan 5?-MHC-D and the porphyrin pheophorbide a (Fig. 1b and c), which have no antibacterial activity but have an inhibiting property against MDR efflux pumps found so far in Staphylo-

0924-8579/03/$30 # 2003 Elsevier Science B.V. and the International Society of Chemotherapy. All rights reserved. doi:10.1016/S0924-8579(03)00085-2

R. Musumeci et al. / International Journal of Antimicrobial Agents 22 (2003) 48 /53

49

by the vegetation that is characteristic of the volcanic slopes, Astragaletum siculi : xerophyte and thorny plants are prevalent. It is believed that this form of Astragaletum established itself in the area where once forests were dominant; it is thus secondary vegetation, in substitution [14,15]. In the past some authors considered the species aetnensis a simple adaptation of B. vulgaris due to the dry conditions of the South [12]. Currently the Kewensis Index [13] gives the species aetnensis its own characteristics and distinguishes it from the species vulgaris .

2. Materials and methods 2.1. Plant material Roots of B. aetnensis C. Presl. were collected on the slopes of the volcano Mount Etna (1800 /1900 m) in November 2001, leaves were collected in May 2001. B. aetnensis specimens were obtained thanks to the Regional Forest Corps Detachment of Catania-Nicolosi. A voucher specimen of the plant was deposited in the herbarium of the Pharmacobiological Department of the University of Messina (Italy). The fresh material was air-dried and powdered. 2.2. Preparation of extracts For the preparation of the ethanol and ether extracts of roots, and ethanol and chloroform extracts of leaves of B. aetnensis exhaustive extraction of 100 g of single drugs was carried out at room temperature by maceration for 24 h on a rotating shaker with 600 ml of ethanol 70%, with ether and with chloroform [16]. The extracts were then filtered and dried under vacuum. The residue of the ethanol and ether extracts of the roots were 4.64 and 0.215 g, respectively; the ethanol and chloroform extracts of the leaves were 10.31 and 12.58 g, respectively. The residues of the extracts were solubilised in dimethylsulphoxide (DMSO). 2.3. Chromatographic analysis of the chloroform extract One hundred grams of dry powdered leaves of B. aetnensis were first steeped in hexanes (drug/solvent ratio 1:6 w/v) for 24 h, filtered under vacuum and then steeped in CHCl3 (drug/solvent ratio 1:5 w/v) for 24 h and subjected to a final filtration. Twenty microlitres of chloroform extract of B. aetnensis leaves (test solution) were used for thin layer chromatographic (TLC) analysis. A standard solution was prepared dissolving 1 mg of commercial pheophorbide a (ICN) in 10 ml of CHCl3 (reference solution). Twenty microlitres of test solution and 10 ml of reference solution were laid on chromatographic silica gel 60 F254 pre-coated TLC plates (Merck,

Fig. 1. Chemical structure of berberine. Chemical structure of 5?MHC-D. Chemical structure of pheophorbide a .

coccus aureus (allowing berberine to carry out its activity) [7,8,10]. Because of the results obtained with the Berberidaceae so far, we assayed the antimicrobial activity and verified the presence of eventual MDR efflux pump inhibitors in Berberis aetnensis C. Presl., a berberidaceae species endemic to Mount Etna, Sicily, Italy. Plant material was collected in the A zone of the Etna Volcano Park. B. aetnensis C. Presl is a bushyspiny shrub with dense twisted branches whose grey, smooth bark has fine longitudinal channels. Leaves on long branches are transformed into spines 3-fid, the shorter branches have normal leaves. The leaves are oblong-egg shaped, rigid, coriaceous and have spinyserrated margins. The flowers, yellow in colour and are grouped in hanging racemes. The fruit is an oblong berry, dark red in colour [11 /13]. As soon as they are cut, the branches and the roots have a penetrating odour and are intensely yellow in colour. B. aetnensis grows on rocky slopes, on the rocky bed of water courses, at the upper limit of the wooded zone and beyond. In Sicily it is common on Mount Etna, in the area of vegetation between 1800 and 2450 /2500 m. This zone is dominated

50

R. Musumeci et al. / International Journal of Antimicrobial Agents 22 (2003) 48 /53

Germany). The mobile phase was made up of chloroform /methanol (9.5:0.5). The developed plate was then examined by UV lamp at 254 and 365 nm. Ratio front (Rf) values were calculated from the extract spots, corresponding to that of the reference solution. The chloroform extract was separated by flash chromatography on silica gel 60 (230 /400 mesh, supplied by Merck, Germany) eluting with chloroform /methanol (9.5:0.5). Two intensely coloured fractions were recovered, one yellow and the other green. These fractions were dried, weighed and then solubilised in DMSO, obtaining two solutions with concentrations of 7000 (Y, yellow solution) and 13 500 mg/ml (G, green solution). The two solutions were diluted in each well to obtain the following final concentrations: 233 mg/ml (Y 233), 23.3 mg/ml (Y 23.3) and 0.76 mg/ml (Y 0.76) for solution Y and 450 mg/ml (G 450), 45 mg/ml (G 45) and 4.5 mg/ml (G 4.5) for solution G. 2.4. Strains The activity of B. aetnensis extracts were assayed against seven standard Gram-positive and Gram-negative bacterial strains: S. aureus ATCC 29213, Bacillus subtilis ATCC 6603, Enterococcus faecalis ATCC 29212, Escherichia coli ATCC 30218, E. coli ATCC 25922, Pseudomonas aeruginosa ATCC 27853, Stenotrophomonas maltophilia NCTC 10257 and against 14 strains of nosocomial origin: two strains of S. aureus (1 Met-S, 1 Met-R); four strains of Staphylococcus epidermidis (2 Met-S, 2 Met-R); three strains of E. coli ; four strains of P. aeruginosa (three from patients affected by cystic fibrosis-CF) and Hafnia alvei . We also assayed C. albicans ATCC 3183, C. parapsilosis ATCC 22019 and C. krusei ATCC 6258. 2.5. Antimicrobials We used ciprofloxacin (Cip; Bayer), amphotericin B (AMB; Sigma) and berberine chloride (Sigma). The susceptibility breakpoints for ciprofloxacin were chosen according to the National Committee for Clinical Laboratory Standard (NCCLS 2001) [17]. For B. subtilis, studies are not yet adequate to develop reproducible results. As regards amphotericin B, the sensitivity breakpoints for C. albicans , C. parapsilosis and C. krusei were chosen according to the NCCLS 1997 suppl. M-27A [18]. 2.6. Determination of the minimum inhibitory concentrations (MICs) The MICs for the 21 bacterial strains and the three Candida strains were determined following the method of double-serial microdilution, conforming to the procedures recommended by the NCCLS 2001 [17]. After

aerobic incubation for 18/24 h at 37 8C, the bacterial and fungal cultures were diluted to a turbidity of 0.5 McFarland (1.5 )/108 CFU/ml) and further diluted in saline solution to obtain an inoculum of 5 )/105 CFU/ well in a final volume of 100 ml. The plates were incubated aerobically at 37 8C for about 18 h. The MIC is defined as the lowest concentration at which there was no visible growth after incubation at 37 8C for 18 h. With the aim of verifying eventual synergistic activity of ciprofloxacin with different dilutions of our chromatographic fractions, the activity of these associations was compared with that of ciprofloxacin plus commercial pheophorbide a (0.5 mg/ml) whose which antimicrobial activity was also tested.

3. Results 3.1. Activity of the extracts The root and leaf extracts generally showed a greater activity against Gram-positive bacteria and yeasts than against Gram-negative bacteria, except for P. aeruginosa CF 03 (MIC 0/312 mg/l). The lowest MIC value obtained with the ethanol extract of roots was 312 mg/l against S. epidermidis . S. aureus ATCC 29213, two strains of S. epidermidis , B. subtilis ATCC 6603 and the three species of Candida were inhibited at 625 mg/l. In most cases the ether extract was more efficacious than the ethanol extract except for one strain of S. epidermidis (MIC ]/10 000 mg/l), C. albicans ATCC 3183 and C. parapsilosis ATCC 22019 (MIC 0/1250 mg/l in both cases). With the root ethanol extract the lowest MIC value was found against E. faecalis ATCC 29212 and C. krusei ATCC 6258 (MIC 0/78 mg/l in both cases). The lowest MIC value for the leaf ethanol extract was 1250 mg/l against S. epidermidis strains. In the remaining cases there was no difference in the activity of the extracts against Gram-positive, Gram-negative bacteria and yeasts. The chloroform extract of leaves was more active (MIC range0/39/2500 mg/l) than the ethanol one (MIC range 1250/10 000 mg/l) both against bacterial strains and yeasts. The lowest MIC value obtained with this extract against Gram-positives was 78 mg/l for both B. subtilis ATCC 6603 and E. faecalis ATCC 29212. S. aureus ATCC 29213, the four strains of S. epidermidis and H. alvei were inhibited at concentrations of 156 mg/ l. Among the yeasts, the chloroform extract of leaves was particularly active against C. krusei ATCC 6258 with a MIC of 39 mg/l, the lowest value found in our study (data not shown in Table). 3.2. TLC analysis of the chloroform leaf extract TLC analysis of the chloroform leaf extract (Fig. 2) showed a green spot (Rf 0/0.95) corresponding to the

R. Musumeci et al. / International Journal of Antimicrobial Agents 22 (2003) 48 /53

51

Fig. 2. TLC analysis of the chloroform leaf extract of B. aetnensis C. Presl. (Lane A). The reference compound (Lane B) is a sample of pheophorbide a (eluting system CHCl3/MeOH 9.5:0.5).

lipid constituents; a yellow spot (Rf 0/0.42) and a green spot (Rf 0/0.24) corresponding to the reference pheophorbide a. Based on these results, we hypothesise that the yellow spot with Rf 0/0.42 (Lane A) is the 5?-MHCD and, moreover, we confirm the presence of pheophorbide a in the chloroform leaf extract of B. aetnensis C. Presl. (Lane A) compared with the commercial compound (Lane B). 3.3. Activity of the chromatographic fractions The solutions at the lowest concentration prepared from the two coloured fractions isolated by chromatography, namely Y 0.76 and G 4.5, showed no synergistic action with ciprofloxacin and pheophorbide a (data not shown). The results of the activity of ciprofloxacin and of its associations with dilutions Y 233 and G 450 and with commercial pheophorbide a (PPB 0.5) against Gram-positive and Gram-negative microrganisms are shown in Table 1. For most Gram-positive and Gramnegative strains there was an increase in the activity of ciprofloxacin as demonstrated by the values of the MICs when it was associated both with dilutions Y 233 and G

450. The most interesting results were seen against S. aureus MR Br11, which, initially resistant to ciprofloxacin (MIC 0/32 mg/l), was inhibited by 1 mg/l using Cip'/Y 233 and by 2 mg/l using Cip'/G 450 associations. Of the Gram-negatives, P. aeruginosa P 54, initially resistant (MIC 0/64 mg/l), was inhibited by B/ 0.5 mg/l of both Cip'/Y 233 and Cip'/G 450 associations. For a more accurate evaluation of the MIC values of the associations tested, we used dilutions Y 23.3 and G 45 and the results are shown in Table 2. The synergistic activity of the two associations was confirmed for all the bacterial strains except for S. epidermidis MR 26 and P. aeruginosa CF 09. Moreover, of the the Gram-positive bacteria, the most interesting results were observed for S. aureus MS DG630; in fact, both Cip'/Y 23.3 and Cip'/G 45 dilutions reduced the MIC of ciprofloxacin by 2 to 5/0.015 mg/l. As regards the Gram-negative microrganisms the above mentioned dilutions reduced the MIC of ciprofloxacin of P. aeruginosa P 54, initially resistant, both by 64 /0.25 mg/l. These results were compared with those of the association of ciprofloxacin with commercial pheophorbide a .

52

R. Musumeci et al. / International Journal of Antimicrobial Agents 22 (2003) 48 /53

Table 1 MICs (expressed in mg/l) of ciprooxacin (Cip alone), Cip'/Y 233, Cip'/G 450, Cip'/PPB 0.5 and commercial pheophorbide a (PPB alone) against Gram-positive and Gram-negative bacteria Bacterial strains MIC (mg/l) Cip alone S. aureus ATCC 29213 S. aureus MS DG630 S. aureus MR Br11 S. epidermidis MS 29 S. epidermidis MR 26 E. coli ATCC 30218 E. coli ATCC 25922 E. coli 116 E. coli 131 E. coli 147 P. aeruginosa ATCC 27853 P. aeruginosa P54 P. aeruginosa CF 03 P. aeruginosa CF 04 P. aeruginosa CF 09 0.25 2 32 2 0.25 0.03 0.015 0.015 2 8 0.5 64 2 2 0.06 Cip'/Y 233 5/0.5 5/0.5 1 1 5/0.5 5/0.5 5/0.5 5/0.5 5/0.5 1 5/0.5 5/0.5 5/0.5 5/0.5 5/0.5 Cip'/G 450 5/0.5 5/0.5 2 1 5/0.5 5/0.5 5/0.5 5/0.5 5/0.5 5/0.5 5/0.5 5/0.5 5/0.5 5/0.5 5/0.5 Cip'/PPB 0.5 5/0.015 5/0.015 5/0.015 0.25 0.12 5/0.015 5/0.015 5/0.015 0.06 0.5 0.5 0.5 5/0.015 5/0.015 0.06 PPB alone 500 250 250 250 4 64 64 125 125 125 125 125 125 125 125

4. Discussion Comparing the activity of the ethanol root extract with that of the ethanol leaf extract of B. aetnensis , it clear that the former had greater antibacterial activity against Gram-positive microrganisms and the Candida strains. Moreover, S. epidermidis and C. krusei were the microrganisms most susceptible to all our extracts. The Gram-negative bacteria, perhaps due to their external lipopolysaccharide structure, were more resistant. Ether is among the solvents with the greatest extraction activity for alkaloids and terpenoids, [16]. The good

antimicrobial activity of our ether extract could be explained by its better quantitative extraction of berberine-alkaloids and terpenoids present in B. aetnensis . The chloroform gave the lowest MIC values, particularly against Gram-positive bacteria, thus demonstrating a superior antimicrobial activity. The antimicrobial activity of ciprofloxacin was improved when it was added to the dilutions of the chromatographic fractions of the chloroform leaf extract of B. aetnensis . These results confirm that our extracts contain pheophorbide a and allow us to hypothesise the presence of 5?-MHC-D.

Table 2 MICs (expressed in mg/l) of ciprooxacin (Cip alone), Cip'/Y 23.3, Cip'/G 45, Cip'/PPB 0.5 and commercial pheophorbide a (PPB alone) against Gram-positive and Gram-negative bacteria Bacterial strains MIC (mg/l) Cip alone S. aureus ATCC 29213 S. aureus MS DG630 S. aureus MR Br11 S. epidermidis MS 29 S. epidermidis MR 26 E. coli ATCC 30218 E. coli ATCC 25922 E. coli 116 E. coli 131 E. coli 147 P. aeruginosa ATCC 27853 P. aeruginosa P54 P. aeruginosa CF 03 P. aeruginosa CF 04 P. aeruginosa CF 09 0.25 2 32 2 0.25 0.03 0.015 0.015 2 8 0.5 64 2 2 0.06 Cip'/Y 23.3 5/0.015 5/0.015 1 1 0.25 5/0.015 5/0.015 5/0.015 0.25 1 0.12 0.25 5/0.015 5/0.015 0.06 Cip'/G 45 5/0.015 5/0.015 2 1 0.25 5/0.015 5/0.015 5/0.015 0.06 0.25 0.12 0.25 5/0.015 5/0.015 0.06 Cip'/PPB 0.5 5/0.015 5/0.015 5/0.015 0.25 0.12 5/0.015 5/0.015 5/0.015 0.06 0.5 0.5 0.5 5/0.015 5/0.015 0.06 PPB alone 500 250 250 250 4 64 64 125 125 125 125 125 125 125 125

R. Musumeci et al. / International Journal of Antimicrobial Agents 22 (2003) 48 /53

53

The improvement in the antibacterial activity of ciprofloxacin, when associated with fractions Y (Y 233 and Y 23.3) and G (G 450 and G 45), is attributable to the inhibiting properties of pheophorbide a and 5?MHC-D against the NorA specific MDR pumps found in S. aureus [7 /10].

[6]

[7]

Acknowledgements The authors thank the Regional Forest Corps Detachment of Catania, Sicily, in particular the Commandant of the Detachment of Catania-Nicolosi, Warrant Ofcer Gianluca Ferlito, for his helpful and effective collaboration.

[8]

[9] [10]

References
[1] Ivanovska N, Philipov S. Study of the anti-inammatory action of Berberis vulgaris root extract, alkaloid fractions and pure alkaloids. Int J Immunopharmacol 1996;18:553 /61. [2] Birdsall TC, Kelly GS. Berberine: therapeutic potential of an alkaloid found in several medicinal plants. Altern Med Rev 1997;2:94 /103. [3] Sethi ML. Enzyme inhibition VI: inhibition of reverse transcriptase activity by protoberberine alkaloids and structure-activity relationship. J Pharm Sci 1983;72(5):538 /41. [4] Tsai TH. Analytical approaches for traditional Chinese medicines exhibiting antineoplastic activity. J Chromatogr B Biomed Sci Appl 2001;764(1 /2):27 /48. [5] Cernakova M, Kostalova D, Kettmann V, Plodova M, Toth J, Drimal J. Potential antimutagenic activity of berberine, a

[11] [12] [13] [14] [15] [16] [17]

[18]

constituent of Mahonia aquifolium . Complement Altern Med 2002;2:2. Khosla PK, Neeraj VI, Gupta SK, Satpathy G. Berberine, a potential drug for trachoma. Rev Int Trach Pathol Ocul Trop Subtrop Sante Publique 1992;69:147 /65. Stermitz FR, Lorenz P, Tawara JN, Zenewicz LA, Lewis K. Synergy in a medicinal plant: antimicrobial action of berberine potentiated by 5?-methoxyhydnocarpin, a multidrug pump inhibitor. Proc Natl Acad Sci 2000;97:1433 /7. Stermitz FR, Tawara-Matsuda J, Lorenz P, Mueller P, Zenewicz L, Lewis K. 5?-Methoxyhydnocarpin-D and pheophorbide A: Berberis species components that potentiate berberine growth inhibition of resistant Staphylococcus aureus . J Nat Prod 2000;63:1146 /9. Marshall NJ, Piddock LJV. Antibacterial efux systems. Microbiologia 1997;13(3):285 /300. Stermitz FR, Beeson TD, Mueller PJ, Hsiang JF, Lewis K. Staphylococcus aureus MDR efux pump inhibitors from a Berberis and a Mahonia (sensu strictu) species. Biochem Syst Ecol 2001;29:793 /8. Tutin TG, Heywood VH, Burges NA, et al. Flora Europaea, vol. I. Cambridge: University Press, 1964. Pignatti S. Flora dItalia, vol. I. Edagricole, 1982. Index Kewensis 2.0. Oxford University Press, 1997. Poli Marchese E. Piante e Fiori dellEtna. Palermo: Sellerio, 1991. Poli Marchese E. La vegetazione altomontana dellEtna. Boll Ist Bot Univ Catania 1965;s. 3(5):1 /6. Marjorie Murphy C. Plant products as antimicrobial agents. Clin Microbiol Rev 1999;12:564 /82. National Committee for Clinical Laboratory Standard (NCCLS); Performance Standards for Antimicrobial Susceptibility Testing. Eleventh Informational Supplement. Villanova, PA, 2001. National Committee for Clinical Laboratory Standards (NCCLS); Reference Method for Broth Dilution Antifungal Susceptibility Testing of Yeasts. Approved standard. Document M-27A. Wayne, PA, 1997.