- Email: [email protected]
Damage to the cytoplasmic membrane of Escherichia Coli by catechin-copper (II) complexes
Free Radical Biology & Medicine, Vol. 27, Nos. 11/12, pp. 1245–1250, 1999 Copyright © 1999 Elsevier Science Inc. Printed in the USA. All rights reserved 0891-5849/99/$–see front matter
PII S0891-5849(99)00157-4
Original Contribution DAMAGE TO THE CYTOPLASMIC MEMBRANE OF ESCHERICHIA COLI BY CATECHIN-COPPER (II) COMPLEXES NOBUO HOSHINO,* TAKAHIDE KIMURA,† AKIRA YAMAJI,*
and
TAKASHI ANDO†
*Department of Pharmacy and †Department of Chemistry, Shiga University of Medical Science, Seta, Otsu, Shiga 520-2192 Japan (Received 24 May 1999; Revised 6 July 1999; Accepted 16 July 1999)
Abstract—In the presence of a nonlethal concentration of Cu(II), washed Escherichia coli ATCC11775 cells were killed by (-)-epigallocatechin (EGC) and (-)-epicatechin (EC). Cell killing was accompanied by a depletion in both the ATP and potassium pools of the cells, but the DNA double strand was not broken, indicating that the bactericidal activity of catechins in the presence of Cu(II) results from damage to the cytoplasmic membrane. Induction of endogenous catalase in E. coli cells increased their resistance to being killed by the combination of catechins and Cu(II). In all cases studied, EGC and EC with Cu(II) were found to generate hydrogen peroxide, but its concentration was too low to account for the bactericidal activity. The bactericidal activity of EGC in the presence of Cu(II) was completely suppressed by ethylenediaminetetraacetate, bathocuproine, catalase, superoxide disumutase (SOD), heated catalase, and heated SOD, but not by dimethyl sulfoxide. When catalase, either heated or unheated, was added to the cells incubated with EGC in the presence of Cu(II), it completely inhibited further killing of the cells. These findings suggest that recycling redox reactions between Cu(II) and Cu(I), involving catechins and hydrogen peroxide on the cell surface, must be important in the mechanism of the killing. © 1999 Elsevier Science Inc. Keywords—Catechins, Copper (II), Escherichia coli, Bactericidal activity, ATP, Potassium, Catalase, SOD, Free radicals
INTRODUCTION
was much less effective [8]. However, the mechanism of the bactericidal activity remains unknown. In this article, we demonstrate that hydrogen peroxide produced on the cell surface is involved in the bactericidal activity of catechins in the presence of Cu(II) and that the cytoplasmic membrane is a target organelle of the damage.
Several recent reports have indicated that tea catechins show antibacterial or bactericidal activities [1–3]. Ikigai et al. [3] have reported that the minimal growth inhibitory concentration of (-)-epigallocatechin gallate against Escherichia coli K-12 is 1.25 mM, the highest reported level of activity among tea catechins. However, tea catechins are also known as powerful antioxidants [4]. Some antioxidants with trace amounts of metals have been reported to show a pro-oxidative activity to cause DNA damage and lipid peroxidation under certain conditions [5,6]. We have already reported that, in the presence of Cu(II) and molecular oxygen, catechins promote extensive DNA cleavage and fatty acid peroxidation [7]. In a recent study, we found that (-)-epigallocatechin (EGC) showed bactericidal activity in the presence of a nonlethal concentration of Cu(II), whereas EGC alone
MATERIALS AND METHODS
Chemicals Tea catechins [(-)-epigallocatechin (EGC) and (-)-epicatechin (EC)] were purchased from Funakoshi Co. (Tokyo, Japan), superoxide dismutase (SOD) from Attest (Kyoto, Japan), horseradish peroxidase and catalase from Sigma Chemical Co. (St. Louis, MO, USA), vitamin-free casamino acid, proteose peptone, and yeast extract from Difco Lab. Inc. (Detroit, MI, USA). All other chemicals of reagent grade were purchased from Attest (Japan) and used without additional purification.
Address correspondence to: Takahide Kimura, Department of Chemistry, Shiga University of Medical Science, Seta, Otsu, Shiga 520-2192 Japan; Tel/Fax: ⫹81-77-548-2102; E-Mail: [email protected]. 1245
1246
N. HOSHINO et al.
Cultures Escherichia coli ATCC 11775 was grown in Davis minimal medium [9] supplemented with 1% casamino acids, 0.5% glucose and 1 g/ml thiamin. Cultures were grown for at least three generations to late logarithmic phase (optical density of 0.8 at 600 nm) [10]. Preparation of cell suspension The culture was sedimented by centrifugation, washed twice in a buffer [1-mM phosphate buffer (pH 7.4) and 1-mM MgSO4], and resuspended in a reaction buffer (10-mM 2-(N-morpholino)ethanesulfonic acid (MES) (pH 6.5) and 1-mM MgSO4) to 4 – 6 ⫻ 107 cells/ml. Determination of viable cell counts (survival) Cell suspensions were incubated for 2 min at 37°C. EGC, EC, or hydrogen peroxide (1–500 M), then CuSO4 (1 M) were added to the cell suspensions. The concentration of Cu(II) when added, was 1 M in all the present studies. At this concentration of Cu(II), E. coli was not killed by Cu(II) itself [8]. For inhibition experiments, ethylenediaminetetraacetate (EDTA) (100 M), bathocuproine (100 M), dimethyl sulfoxide (DMSO) (100 M), catalase (50 g/ml), SOD (50 g/ml), heated catalase (50 g/ml) or heated SOD (50 g/ml), which were prepared by a 5-min treatment at 95°C [11], was added to the cell suspensions. Samples were withdrawn at various time intervals and diluted 1:10 in a stop mixture (wash buffer supplemented with 0.1 mM diethylenetriaminepentaacetic acid). When hydrogen peroxide was added to the cell suspensions, the stop mixture was supplemented with catalase (0.5 mg/ml). After additional serial dilution in wash buffer the viable cell number was determined by a standard plate assay on heart infusion agar. Determination of ATP levels in E. coli cells Cellular ATP levels were determined after incubation of E. coli with EGC or EC in the presence of Cu(II) for 60 min at 37°C. ATP was measured by using an ATP Bioluminescence Assay Kit HSII (Boehringer Mannheim, Mannheim, Germany) based on the method of Stanley [12]. Determination of potassium levels in E. coli cells Cellular potassium levels were measured according to the method of Kohen et al. [13]. Briefly, E. coli cells were filtered on to a Millipore membrane filter (Bedford,
MA, USA) (pore size 0.22 m) after incubation with EGC or EC in the presence of Cu(II) for 60 min at 37°C. The filter was washed six times with 10 ml of distilled water, and the cells were resuspended in water. The cell suspension was heated to 70°C for 5 min, cooled, centrifuged (10,000 rpm ⫻ 10 min), and kept at 4°C overnight. The supernatant was separated and potassium concentration was determined by atomic absorption with a Shimadzu spectrophotometer (AA-660) (Kyoto, Japan).
Analysis of DNA of E. coli cells DNA was isolated from E. coli cells, incubated with EGC or EC in the presence of Cu(II) for 60 min at 37°C, by using Easy-DNA Kit (Invitrogen Co., Carlsbad, CA, USA). Electrophoretic studies were performed on 0.8% agarose gel (50 V, 40 mA) as previously described [7].
Induction of catalase The cells were grown at 37°C on a shaking incubator in a medium containing proteose peptone (1%), yeast extract (0.5%), and sodium chloride (0.5%). Catalase was induced by the addition of 0.44-mM hydrogen peroxide [13]. The E. coli were washed twice in a wash buffer and resuspended in a reaction buffer to 4 – 6 ⫻ 107 cells/ml. Catalase was assayed by spectrophotometric analysis of the rate of decomposition of hydrogen peroxide at 240 nm [14].
Determination of hydrogen peroxide The amount of hydrogen peroxide was determined according to the method of Guilbault et al. [15]. Briefly, EGC or EC and Cu(II) was kept for 60 min at 37°C. To 0.1 ml of the sample, 2.7 ml of 0.1 M Tris-HCl buffer (pH 8.5), 0.1 ml of 2.5-mg/ml homovanilic acid solution and 0.1 ml of 1-mg/ml peroxidase solution were added. The fluorescence of the solution was measured at ex. 315 nm and em. 425 nm. The standard curve, prepared with reaction buffer (pH 6.5) containing catechin and hydrogen peroxide, was used for calibration.
Binding of copper ions to E. coli cells Binding of copper ions to E. coli cells were determined after incubation of E. coli with EGC in the presence of Cu(II) at 37°C. After 60 min, the supernatant was separated by centrifugation. Copper concentration of the supernatant was determined by atomic absorption with a Shimadzu spectrophotometer (AA-660).
Bactericidal catechin-Cu(II)
1247
Table 2. Effect of EGC and EC on the Cellular ATP Level and Cellular Potassium Level in E. Coli in the Presence of Cu(II)
Catechins
Concentrations (M)
Residual ATP (%)
Residual potassium (%)
— 1 10 100 1 10 100
100 ⫾ 12 24 ⫾ 8† 18 ⫾ 9† 18 ⫾ 14† 45 ⫾ 7† 39 ⫾ 13† 43 ⫾ 7†
108 ⫾ 17 43 ⫾ 11† 32 ⫾ 16† 43 ⫾ 16† 64 ⫾ 14† 48 ⫾ 14† 42 ⫾ 19†
Control ECG EC
Washed E. coli cells were incubated with various concentrations of catechins in the presence of Cu(II) (1 M) in reaction buffer (pH 6.5) at 37°C. After 60 min, cellular ATP and cellular potassium were determined. Residual ATP and residual potassium were calculated from those at time zero. (mean ⫾ SD, n ⫽ 3– 6). The control experiments were carried out in a similar manner without treatment of any catechins.Comparison with unpaired t-test reveals significant difference (*p ⬍ .05, †p ⬍ .01) versus control group.
Fig. 1. Inhibitory effect of catalase and heated catalase on the bactericidal activity of EGC in the presence of Cu(II). EGC (10 M), then Cu(II) (1 M) were added to the washed E. coli cells. After incubation of the cells in reaction buffer (pH 6.5) at 37°C, samples were withdrawn at various time intervals and viable cell numbers were determined (䊐). Catalase (50 g/ml) (●) or heated catalase (50 g/ml) (Œ) was added to the cell suspension at 2, 30, and 60 min.
RESULTS
In the presence of 1-M Cu(II), 10-M EGC caused exponential killing of E. coli cells after an induction period, as reported previously [8]. This bactericidal activity of catechins in the presence of Cu(II) was completely inhibited by EDTA and bathocuproine, a specific chelator of Cu(I). Catalase and SOD also completely suppressed the bactericidal activity of EGC-Cu(II), but even heated catalase and heated SOD showed the same ability. The addition of catalase, or heated catalase, to the cells after 2, 30, and 60 min incubation completely
inhibited additional killing (Fig. 1). Addition of DMSO, a hydroxyl radical scavenger, did not influence the bactericidal activity of EGC-Cu(II) (Table 1). Table 2 shows that, in the presence of Cu(II), a wide range of concentration of EGC and EC caused considerable depletion in both ATP and potassium pools of the cells. DNA degradation induced by catechins and Cu(II) was not detected by electrophoretic analysis. E. coli cells which contained a 2.4-fold concentration of endogenous catalase due to having been grown in a medium containing hydrogen peroxide (0.44 mM), showed an increased resistance to the killing caused by the combination of catechins and Cu(II), compared with uninduced cells. The percent survival of the catalase-rich cells after 60-min treatment with EGC and Cu(II) was above 72%, while that of uninduced cells was less than 8% (Table 3). In all the cases studied, EGC and EC with Cu(II)
Table 1. Effect of Chelators and Scavengers on the Survival of E. coli Incubated with EGC or EC in the Presence of Cu(II) % Survival After 60 min
Catechins Control EGC EC
Concentrations Without chelator EDTA Bathocuproine Catalase SOD Heated catalase Heated SOD DMSO (M) and scavenger (100 M) (100 M) (50 g/ml) (50 g/ml) (50 g/ml) (50 g/ml) (100 M) — 1 10 100 1 10 100
105 ⫾ 4 8 ⫾ 4† 6 ⫾ 4† 4 ⫾ 2† 93 ⫾ 14 82 ⫾ 18 68 ⫾ 16*
— 116 ⫾ 7 94 ⫾ 9 103 ⫾ 4 — — 100 ⫾ 6
— 116 ⫾ 8 108 ⫾ 10 103 ⫾ 2 — — 113 ⫾ 3
— — 128 ⫾ 2 — — — —
— — 126 ⫾ 10 — — — —
— — 133 ⫾ 9 — — — —
— — 124 ⫾ 11 — — — —
— — 10 ⫾ 4† — — — —
Washed E. coli cells were incubated with various concentrations of catechins in the presence of Cu(II) (1 M) in reaction buffer (pH 6.5) at 37°C. After 60 min, viable cell numbers were determined and % survival was calculated from that at time zero. (mean ⫾ SD, n ⫽ 3– 6). The control experiments were carried out in a similar manner without treatment of any catechins. Comparison with unpaired t-test reveals significant difference (*p ⬍ .05, †p ⬍ .01) versus control group.
1248
N. HOSHINO et al.
Table 3. Effect of Endogeneous Catalase on the Survival of E. coli Incubated with ECG or EC in the Presence of Cu(II) % Survival after 60 min Catechins ECG EC
Concentrations (M)
Uninduced cells
Induced cells
1 10 100 1 10 100
8⫾4 6⫾4 4⫾2 93 ⫾ 14 82 ⫾ 18 68 ⫾ 16
72 ⫾ 7* 78 ⫾ 15* 122 ⫾ 11* 89 ⫾ 11 88 ⫾ 5 116 ⫾ 14*
E. coli cells were grown in a medium in the presence of 0.44 mM hydrogen peroxide as an inducer. The cells were harvested at midlog phase and incubated with catechins (1–100 M) in the presence of Cu(II) (1 M) in reaction buffer (pH 6.5) at 37°C. After 60 min, % survival was determined. Comparison with unpaired t-test reveals significant differences (*p ⬍ .01) versus uninduced cells.
produced hydrogen peroxide (Table 4), the concentration of which increased with the concentration of catechins. EGC was more effective than EC in the formation of hydrogen peroxide. In the absence of Cu(II), hydrogen peroxide showed a bactericidal effect at 500 M, but no effect was detected at 200 M or less. In the presence of Cu(II), the bactericidal effect of hydrogen peroxide was detected even at 200 M and no effect was shown below 100 M (Fig. 2). As shown in Table 5, the percent of copper ions bound to the cells was approximately 67%, independent of the concentration of EGC. DISCUSSION
In 1991 Aronovitch et al. [10] reported the bactericidal activity of epinephrine-Cu(II) complexes. They noticed that, although epinephrine-Cu(II) complex caused little killing of E. coli, rapid killing was induced by the addition of 0.5 mM hydrogen peroxide. The presence of
Table 4. Concentration of Hydrogen Peroxide Formed in Reaction Buffer Containing EGC or EC in the Presence of Cu(II)
Catechins EGC EC
Concentrations (M)
Concentration of hydrogen peroxide formed after 60 min (M)
1 10 100 1 10 100
0.6 ⫾ 0.2 9.0 ⫾ 0.7 41.9 ⫾ 2.6 0.1 ⫾ 0.1 0.8 ⫾ 0.2 3.0 ⫾ 2.3
Various concentrations of catechins were incubated in the presence of Cu(II) (1 M) in reaction buffer (pH 6.5) at 37°C. After 60 min, concentration of hydrogen peroxide was determined. (mean ⫾ SD, n ⫽ 4 –7).
Fig. 2. Effect of hydrogen peroxide on the survival of E. coli in the presence (●) or absence (E) of Cu(II). Washed E. coli cells were incubated with various concentrations of hydrogen peroxide in the presence or absence of 1 M Cu(II) in reaction buffer (pH 6.5) at 37°C. After 60 min, viable cell numbers were determined (mean ⫾ S.D., n ⫽ 4). Comparison using the unpaired t-test reveals significant difference (*p ⬍ .01) versus the group treated without hydrogen peroxide.
hydrogen peroxide is considered to form a recycling redox system to damage the cytoplasmic membrane of E. coli, which is apparently the main reason for its bactericidal activity. The extent of damage to the cytoplasmic membrane was determined by the cellular ATP level [10,13] and cellular potassium level [13]. We have recently reported that EGC killed E. coli cells in the presence of a nonlethal concentration of Cu(II) (1 M) without the addition of hydrogen peroxide [8]. Although Cu(II) (1 M), EGC and EC (100 M each) showed no effect separately on the viability of E. coli, the combination of Cu(II) (1 M) with EGC (1, 10, 100 M) or EC (100 M) killed E. coli cells (Table 1) [8]. Although the killing was accompanied by a depletion in both the ATP and potassium pools of the cells (Table Table 5. Effect of EGC on the Binding of Copper to E. coli Cells at 37°C EGC concentrations (M)
Bound copper (%)
1 10 100
63 ⫾ 2 70 ⫾ 4 67 ⫾ 11
Washed E. coli cells (8 ⫻ 107 cells) were incubated with varous concentrations of EGC in the presence of Cu(II) (1 M) in reaction buffer (pH 6.5) at 37°C. After incubation of the cell suspension for 60 min, supernatant was separated by centrifugation and copper concentration was determined (mean ⫾ SD, n ⫽ 3).
Bactericidal catechin-Cu(II)
2), it was found that the DNA double strand did not break in the killing, indicating that the bactericidal activity of catechins in the presence of Cu(II) is derived from damage to the cytoplasmic membrane. The bactericidal activity of EGC-Cu(II) was completely inhibited by bathocuproine, a specific chelator of Cu(I) (Table 1). This fact indicates that the bactericidal activity of EGC-Cu(II) is mediated through reduction of Cu(II) to Cu(I) by EGC. Reoxidation of Cu(I) to Cu(II) by molecular oxygen generates active oxygen species, superoxide anion at first and then hydrogen peroxide and hydroxyl radicals [7,16]. In order to investigate the possibility of the involvement of active oxygen species in the bactericidal activity of EGC-Cu(II), we examined the generation of superoxide anion, hydrogen peroxide, and hydroxyl radicals after the addition of SOD, catalase, and DMSO. As shown in Table 1, catalase and SOD completely suppressed the bactericidal activity of EGCCu(II). However, similar suppression against the bactericidal activity was observed by the addition of heated enzymes (Table 1). As shown in Fig. 1, addition of either catalase or heated catalase to the cells incubated with EGC and Cu(II) for 2, 30, and 60 min, completely inhibited additional killing. These observations suggest that the protective effects of these enzymes are not derived from the elimination of active oxygen species, but rather from their binding of Cu(I) as proposed by Løvstad [17]. Because these enzymes can not pass through the cell membrane, they must suppress the bactericidal activity by sequestering Cu(I) outside the cells. That is, Cu(I) is not present in the cells but on the surface of the cells where catalase can easily approach. The bactericidal activity of EGC-Cu(II) was not influenced by the addition of DMSO, a hydroxyl radical scavenger (Table 1). Induction of endogenous catalase in the cells increased their ability to resist being killed by the combination of catechins and Cu(II) (Table 3). Therefore it is evident that hydrogen peroxide inside the cells must be critical for the bactericidal activity of catechin-Cu(II). EGC and EC with Cu(II) can generate hydrogen peroxide under aerobic conditions (Table 4). The highest concentration of hydrogen peroxide, 41.9 M, was detected in the case of 100 M EGC and 1 M Cu(II) (Table 4). However, no killing was observed after the addition of 41.9 M hydrogen peroxide and 1 M Cu(II) to the cell suspension (Fig. 2). A hydrogen peroxide concentration of more than 500 M was necessary to manifest the same bactericidal activity as that observed in the presence of 100-M EGC and 1-M Cu(II) (Table 1, Fig. 2). These results suggest that the concentration of hydrogen peroxide in the vicinity of the cells must be higher than that detected in the bulk reaction buffer. It is known that Cu(II) is complexed by catechins [18,19] and
1249
that catechins are apt to be present on bacterial membranes [3]; atomic absorption showed that 63% or more of the copper ions were bound to the cells in the presence of EGC (Table 5). Aronovitch et al. [10] suggested that with hydrogen peroxide added, epinephrine-Cu(II) complex binds to the cell surface to induce oxidative membrane damage. We believe that catechin-Cu(II) on the cell surface reacts locally with molecular oxygen to produce hydrogen peroxide. The hydrogen peroxide generated on the surface can then enter easily into cells and can cause damage to the cytoplasmic membrane. The apparently low concentration of hydrogen peroxide observed in the bulk suspension will not reflect the local concentration of hydrogen peroxide generated on the cell surface. Therefore, we propose that Cu(I), and its redox reactions involving catechins and hydrogen peroxide on the cell surface, must be involved in the killing of the cells. Whether hydrogen peroxide is the only critical substance or not, should be clarified in due course.
REFERENCES [1] Toda, M.; Okubo, S.; Hiyoshi, R.; Shimamura, T. The bactericidal activity of tea and coffee. Lett. Appl. Microbiol. 8:123–125; 1989. [2] Toda, M.; Okubo, S.; Hara, Y.; Shimamura, T. Antibacterial and bactericidal activities of tea extracts and catechins against methicilline resistant Staphylococcus aureus (in Japanese). Jpn. J. Bacteriol. 46:839 – 845; 1991. [3] Ikigai, H.; Nakae, T.; Hara, Y.; Shimamura, T. Bactericidal catechins damage the lipid bilayer. Biochim. Biophys. Acta 1147: 132–136; 1993. [4] Huang, S.-W.; Frankel, E. N. Antioxidant activity of tea catechins in different lipid systems. J. Agric. Food Chem. 45:3033–3038; 1997. [5] Sotomatu, A.; Nakao, M.; Hirai, S. Phospholipid peroxidation induced by the catechol-Fe3⫹ (Cu2⫹) complex: a possible mechanism of nigrostriatal cell damage. Arch. Biochem. Biophys. 283: 334 –342; 1990. [6] Sahu, S. C.; Gray, G. C. Interactions of flavonoids, trace metals, and oxygen: nuclear DNA damage and lipid peroxidation induced by myricetin. Cancer Lett. 70:73–79; 1993. [7] Hayakawa, F.; Kimura, T.; Maeda, T.; Fujita, M.; Sohmiya, H.; Ando, T. DNA cleavage reaction and linoleic acid peroxidation induced by tea catechins in the presence of cupric ion. Biochim. Biophys. Acta 1336:123–131; 1997. [8] Kimura, T.; Hoshino, N.; Yamaji, A.; Hayakawa, F.; Ando, T. Bactericidal activity of catechin-copper(II) complexes on Escherichia coli ATCC11775 in the absence of hydrogen peroxide Lett. Appl. Microbiol. 27:328 –330; 1998. [9] Davis, B. D.; Mingioli, E. S. Mutants of E. coli requiring methionine or vitamin B12. J. Bacteriol. 60:17–28; 1950. [10] Aronovitch, J.; Godinger, D.; Czapski, G. Bactericidal activity of catecholamine copper complexes. Free Radic. Res. Commun. 12-13:479 – 488; 1991. [11] Nishida, K.; Takagi, M.; Yano, K. Escherichia coli cell inactivation by superoxide anion and hydrogen peroxide. J. Gen. Appl. Microbiol. 27:477– 485; 1981. [12] Stanley, P. E. Extraction of adenosine triphosphate from microbial and somatic cells. Methods Enzymol. 60:17–28; 1986. [13] Kohen, R.; Chevion, M. Cytoplasmic membrane is the target organella for transition metal mediated damage induced by paraquat in Escherichia coli. Biochemistry 27:2597–2603; 1988. [14] Aebi, H. Catalase in vitro. Methods Enzymol. 105:121–126; 1984. [15] Guilbault, G.; Brignac, P. J.; Juneau, M. New substrates for the
1250
N. HOSHINO et al.
fluorometric determination of oxidative emzymes. Anal. Chem. 40:1256 –1263; 1968. [16] Rahman, A.; Shahabuddin; Hadi, S. M.; Parish, J. H.; Ainley, K. Strand scission in DNA induced by quercetin and Cu(II): role of Cu(I) and oxygen free radicals. Carcinogenesis 10:1833–1839; 1989. [17] Løvstad, R. A. Copper catalyzed oxidation of ascorbate (vitamin C). Inhibitory effect of catalase, superoxide dismutase, serum
proteins (ceruloplasmin, albumin, apotransferin) and amino acids. Int. J. Biochem. 19:309 –313; 1987. [18] Cetta, G.; Pallavicini, G.; Tenni, R.; Bisi, C. Influence of flavonoid-copper complexes on cross linking in elastin. Itali. J. Biochem. 26:317–327; 1977. [19] Weber, G. HPLC with electrochemical detection of metal-flavonoid-complexes isolated from food. Chromatographia 26:133– 138; 1988.
PII S0891-5849(99)00157-4
Original Contribution DAMAGE TO THE CYTOPLASMIC MEMBRANE OF ESCHERICHIA COLI BY CATECHIN-COPPER (II) COMPLEXES NOBUO HOSHINO,* TAKAHIDE KIMURA,† AKIRA YAMAJI,*
and
TAKASHI ANDO†
*Department of Pharmacy and †Department of Chemistry, Shiga University of Medical Science, Seta, Otsu, Shiga 520-2192 Japan (Received 24 May 1999; Revised 6 July 1999; Accepted 16 July 1999)
Abstract—In the presence of a nonlethal concentration of Cu(II), washed Escherichia coli ATCC11775 cells were killed by (-)-epigallocatechin (EGC) and (-)-epicatechin (EC). Cell killing was accompanied by a depletion in both the ATP and potassium pools of the cells, but the DNA double strand was not broken, indicating that the bactericidal activity of catechins in the presence of Cu(II) results from damage to the cytoplasmic membrane. Induction of endogenous catalase in E. coli cells increased their resistance to being killed by the combination of catechins and Cu(II). In all cases studied, EGC and EC with Cu(II) were found to generate hydrogen peroxide, but its concentration was too low to account for the bactericidal activity. The bactericidal activity of EGC in the presence of Cu(II) was completely suppressed by ethylenediaminetetraacetate, bathocuproine, catalase, superoxide disumutase (SOD), heated catalase, and heated SOD, but not by dimethyl sulfoxide. When catalase, either heated or unheated, was added to the cells incubated with EGC in the presence of Cu(II), it completely inhibited further killing of the cells. These findings suggest that recycling redox reactions between Cu(II) and Cu(I), involving catechins and hydrogen peroxide on the cell surface, must be important in the mechanism of the killing. © 1999 Elsevier Science Inc. Keywords—Catechins, Copper (II), Escherichia coli, Bactericidal activity, ATP, Potassium, Catalase, SOD, Free radicals
INTRODUCTION
was much less effective [8]. However, the mechanism of the bactericidal activity remains unknown. In this article, we demonstrate that hydrogen peroxide produced on the cell surface is involved in the bactericidal activity of catechins in the presence of Cu(II) and that the cytoplasmic membrane is a target organelle of the damage.
Several recent reports have indicated that tea catechins show antibacterial or bactericidal activities [1–3]. Ikigai et al. [3] have reported that the minimal growth inhibitory concentration of (-)-epigallocatechin gallate against Escherichia coli K-12 is 1.25 mM, the highest reported level of activity among tea catechins. However, tea catechins are also known as powerful antioxidants [4]. Some antioxidants with trace amounts of metals have been reported to show a pro-oxidative activity to cause DNA damage and lipid peroxidation under certain conditions [5,6]. We have already reported that, in the presence of Cu(II) and molecular oxygen, catechins promote extensive DNA cleavage and fatty acid peroxidation [7]. In a recent study, we found that (-)-epigallocatechin (EGC) showed bactericidal activity in the presence of a nonlethal concentration of Cu(II), whereas EGC alone
MATERIALS AND METHODS
Chemicals Tea catechins [(-)-epigallocatechin (EGC) and (-)-epicatechin (EC)] were purchased from Funakoshi Co. (Tokyo, Japan), superoxide dismutase (SOD) from Attest (Kyoto, Japan), horseradish peroxidase and catalase from Sigma Chemical Co. (St. Louis, MO, USA), vitamin-free casamino acid, proteose peptone, and yeast extract from Difco Lab. Inc. (Detroit, MI, USA). All other chemicals of reagent grade were purchased from Attest (Japan) and used without additional purification.
Address correspondence to: Takahide Kimura, Department of Chemistry, Shiga University of Medical Science, Seta, Otsu, Shiga 520-2192 Japan; Tel/Fax: ⫹81-77-548-2102; E-Mail: [email protected]. 1245
1246
N. HOSHINO et al.
Cultures Escherichia coli ATCC 11775 was grown in Davis minimal medium [9] supplemented with 1% casamino acids, 0.5% glucose and 1 g/ml thiamin. Cultures were grown for at least three generations to late logarithmic phase (optical density of 0.8 at 600 nm) [10]. Preparation of cell suspension The culture was sedimented by centrifugation, washed twice in a buffer [1-mM phosphate buffer (pH 7.4) and 1-mM MgSO4], and resuspended in a reaction buffer (10-mM 2-(N-morpholino)ethanesulfonic acid (MES) (pH 6.5) and 1-mM MgSO4) to 4 – 6 ⫻ 107 cells/ml. Determination of viable cell counts (survival) Cell suspensions were incubated for 2 min at 37°C. EGC, EC, or hydrogen peroxide (1–500 M), then CuSO4 (1 M) were added to the cell suspensions. The concentration of Cu(II) when added, was 1 M in all the present studies. At this concentration of Cu(II), E. coli was not killed by Cu(II) itself [8]. For inhibition experiments, ethylenediaminetetraacetate (EDTA) (100 M), bathocuproine (100 M), dimethyl sulfoxide (DMSO) (100 M), catalase (50 g/ml), SOD (50 g/ml), heated catalase (50 g/ml) or heated SOD (50 g/ml), which were prepared by a 5-min treatment at 95°C [11], was added to the cell suspensions. Samples were withdrawn at various time intervals and diluted 1:10 in a stop mixture (wash buffer supplemented with 0.1 mM diethylenetriaminepentaacetic acid). When hydrogen peroxide was added to the cell suspensions, the stop mixture was supplemented with catalase (0.5 mg/ml). After additional serial dilution in wash buffer the viable cell number was determined by a standard plate assay on heart infusion agar. Determination of ATP levels in E. coli cells Cellular ATP levels were determined after incubation of E. coli with EGC or EC in the presence of Cu(II) for 60 min at 37°C. ATP was measured by using an ATP Bioluminescence Assay Kit HSII (Boehringer Mannheim, Mannheim, Germany) based on the method of Stanley [12]. Determination of potassium levels in E. coli cells Cellular potassium levels were measured according to the method of Kohen et al. [13]. Briefly, E. coli cells were filtered on to a Millipore membrane filter (Bedford,
MA, USA) (pore size 0.22 m) after incubation with EGC or EC in the presence of Cu(II) for 60 min at 37°C. The filter was washed six times with 10 ml of distilled water, and the cells were resuspended in water. The cell suspension was heated to 70°C for 5 min, cooled, centrifuged (10,000 rpm ⫻ 10 min), and kept at 4°C overnight. The supernatant was separated and potassium concentration was determined by atomic absorption with a Shimadzu spectrophotometer (AA-660) (Kyoto, Japan).
Analysis of DNA of E. coli cells DNA was isolated from E. coli cells, incubated with EGC or EC in the presence of Cu(II) for 60 min at 37°C, by using Easy-DNA Kit (Invitrogen Co., Carlsbad, CA, USA). Electrophoretic studies were performed on 0.8% agarose gel (50 V, 40 mA) as previously described [7].
Induction of catalase The cells were grown at 37°C on a shaking incubator in a medium containing proteose peptone (1%), yeast extract (0.5%), and sodium chloride (0.5%). Catalase was induced by the addition of 0.44-mM hydrogen peroxide [13]. The E. coli were washed twice in a wash buffer and resuspended in a reaction buffer to 4 – 6 ⫻ 107 cells/ml. Catalase was assayed by spectrophotometric analysis of the rate of decomposition of hydrogen peroxide at 240 nm [14].
Determination of hydrogen peroxide The amount of hydrogen peroxide was determined according to the method of Guilbault et al. [15]. Briefly, EGC or EC and Cu(II) was kept for 60 min at 37°C. To 0.1 ml of the sample, 2.7 ml of 0.1 M Tris-HCl buffer (pH 8.5), 0.1 ml of 2.5-mg/ml homovanilic acid solution and 0.1 ml of 1-mg/ml peroxidase solution were added. The fluorescence of the solution was measured at ex. 315 nm and em. 425 nm. The standard curve, prepared with reaction buffer (pH 6.5) containing catechin and hydrogen peroxide, was used for calibration.
Binding of copper ions to E. coli cells Binding of copper ions to E. coli cells were determined after incubation of E. coli with EGC in the presence of Cu(II) at 37°C. After 60 min, the supernatant was separated by centrifugation. Copper concentration of the supernatant was determined by atomic absorption with a Shimadzu spectrophotometer (AA-660).
Bactericidal catechin-Cu(II)
1247
Table 2. Effect of EGC and EC on the Cellular ATP Level and Cellular Potassium Level in E. Coli in the Presence of Cu(II)
Catechins
Concentrations (M)
Residual ATP (%)
Residual potassium (%)
— 1 10 100 1 10 100
100 ⫾ 12 24 ⫾ 8† 18 ⫾ 9† 18 ⫾ 14† 45 ⫾ 7† 39 ⫾ 13† 43 ⫾ 7†
108 ⫾ 17 43 ⫾ 11† 32 ⫾ 16† 43 ⫾ 16† 64 ⫾ 14† 48 ⫾ 14† 42 ⫾ 19†
Control ECG EC
Washed E. coli cells were incubated with various concentrations of catechins in the presence of Cu(II) (1 M) in reaction buffer (pH 6.5) at 37°C. After 60 min, cellular ATP and cellular potassium were determined. Residual ATP and residual potassium were calculated from those at time zero. (mean ⫾ SD, n ⫽ 3– 6). The control experiments were carried out in a similar manner without treatment of any catechins.Comparison with unpaired t-test reveals significant difference (*p ⬍ .05, †p ⬍ .01) versus control group.
Fig. 1. Inhibitory effect of catalase and heated catalase on the bactericidal activity of EGC in the presence of Cu(II). EGC (10 M), then Cu(II) (1 M) were added to the washed E. coli cells. After incubation of the cells in reaction buffer (pH 6.5) at 37°C, samples were withdrawn at various time intervals and viable cell numbers were determined (䊐). Catalase (50 g/ml) (●) or heated catalase (50 g/ml) (Œ) was added to the cell suspension at 2, 30, and 60 min.
RESULTS
In the presence of 1-M Cu(II), 10-M EGC caused exponential killing of E. coli cells after an induction period, as reported previously [8]. This bactericidal activity of catechins in the presence of Cu(II) was completely inhibited by EDTA and bathocuproine, a specific chelator of Cu(I). Catalase and SOD also completely suppressed the bactericidal activity of EGC-Cu(II), but even heated catalase and heated SOD showed the same ability. The addition of catalase, or heated catalase, to the cells after 2, 30, and 60 min incubation completely
inhibited additional killing (Fig. 1). Addition of DMSO, a hydroxyl radical scavenger, did not influence the bactericidal activity of EGC-Cu(II) (Table 1). Table 2 shows that, in the presence of Cu(II), a wide range of concentration of EGC and EC caused considerable depletion in both ATP and potassium pools of the cells. DNA degradation induced by catechins and Cu(II) was not detected by electrophoretic analysis. E. coli cells which contained a 2.4-fold concentration of endogenous catalase due to having been grown in a medium containing hydrogen peroxide (0.44 mM), showed an increased resistance to the killing caused by the combination of catechins and Cu(II), compared with uninduced cells. The percent survival of the catalase-rich cells after 60-min treatment with EGC and Cu(II) was above 72%, while that of uninduced cells was less than 8% (Table 3). In all the cases studied, EGC and EC with Cu(II)
Table 1. Effect of Chelators and Scavengers on the Survival of E. coli Incubated with EGC or EC in the Presence of Cu(II) % Survival After 60 min
Catechins Control EGC EC
Concentrations Without chelator EDTA Bathocuproine Catalase SOD Heated catalase Heated SOD DMSO (M) and scavenger (100 M) (100 M) (50 g/ml) (50 g/ml) (50 g/ml) (50 g/ml) (100 M) — 1 10 100 1 10 100
105 ⫾ 4 8 ⫾ 4† 6 ⫾ 4† 4 ⫾ 2† 93 ⫾ 14 82 ⫾ 18 68 ⫾ 16*
— 116 ⫾ 7 94 ⫾ 9 103 ⫾ 4 — — 100 ⫾ 6
— 116 ⫾ 8 108 ⫾ 10 103 ⫾ 2 — — 113 ⫾ 3
— — 128 ⫾ 2 — — — —
— — 126 ⫾ 10 — — — —
— — 133 ⫾ 9 — — — —
— — 124 ⫾ 11 — — — —
— — 10 ⫾ 4† — — — —
Washed E. coli cells were incubated with various concentrations of catechins in the presence of Cu(II) (1 M) in reaction buffer (pH 6.5) at 37°C. After 60 min, viable cell numbers were determined and % survival was calculated from that at time zero. (mean ⫾ SD, n ⫽ 3– 6). The control experiments were carried out in a similar manner without treatment of any catechins. Comparison with unpaired t-test reveals significant difference (*p ⬍ .05, †p ⬍ .01) versus control group.
1248
N. HOSHINO et al.
Table 3. Effect of Endogeneous Catalase on the Survival of E. coli Incubated with ECG or EC in the Presence of Cu(II) % Survival after 60 min Catechins ECG EC
Concentrations (M)
Uninduced cells
Induced cells
1 10 100 1 10 100
8⫾4 6⫾4 4⫾2 93 ⫾ 14 82 ⫾ 18 68 ⫾ 16
72 ⫾ 7* 78 ⫾ 15* 122 ⫾ 11* 89 ⫾ 11 88 ⫾ 5 116 ⫾ 14*
E. coli cells were grown in a medium in the presence of 0.44 mM hydrogen peroxide as an inducer. The cells were harvested at midlog phase and incubated with catechins (1–100 M) in the presence of Cu(II) (1 M) in reaction buffer (pH 6.5) at 37°C. After 60 min, % survival was determined. Comparison with unpaired t-test reveals significant differences (*p ⬍ .01) versus uninduced cells.
produced hydrogen peroxide (Table 4), the concentration of which increased with the concentration of catechins. EGC was more effective than EC in the formation of hydrogen peroxide. In the absence of Cu(II), hydrogen peroxide showed a bactericidal effect at 500 M, but no effect was detected at 200 M or less. In the presence of Cu(II), the bactericidal effect of hydrogen peroxide was detected even at 200 M and no effect was shown below 100 M (Fig. 2). As shown in Table 5, the percent of copper ions bound to the cells was approximately 67%, independent of the concentration of EGC. DISCUSSION
In 1991 Aronovitch et al. [10] reported the bactericidal activity of epinephrine-Cu(II) complexes. They noticed that, although epinephrine-Cu(II) complex caused little killing of E. coli, rapid killing was induced by the addition of 0.5 mM hydrogen peroxide. The presence of
Table 4. Concentration of Hydrogen Peroxide Formed in Reaction Buffer Containing EGC or EC in the Presence of Cu(II)
Catechins EGC EC
Concentrations (M)
Concentration of hydrogen peroxide formed after 60 min (M)
1 10 100 1 10 100
0.6 ⫾ 0.2 9.0 ⫾ 0.7 41.9 ⫾ 2.6 0.1 ⫾ 0.1 0.8 ⫾ 0.2 3.0 ⫾ 2.3
Various concentrations of catechins were incubated in the presence of Cu(II) (1 M) in reaction buffer (pH 6.5) at 37°C. After 60 min, concentration of hydrogen peroxide was determined. (mean ⫾ SD, n ⫽ 4 –7).
Fig. 2. Effect of hydrogen peroxide on the survival of E. coli in the presence (●) or absence (E) of Cu(II). Washed E. coli cells were incubated with various concentrations of hydrogen peroxide in the presence or absence of 1 M Cu(II) in reaction buffer (pH 6.5) at 37°C. After 60 min, viable cell numbers were determined (mean ⫾ S.D., n ⫽ 4). Comparison using the unpaired t-test reveals significant difference (*p ⬍ .01) versus the group treated without hydrogen peroxide.
hydrogen peroxide is considered to form a recycling redox system to damage the cytoplasmic membrane of E. coli, which is apparently the main reason for its bactericidal activity. The extent of damage to the cytoplasmic membrane was determined by the cellular ATP level [10,13] and cellular potassium level [13]. We have recently reported that EGC killed E. coli cells in the presence of a nonlethal concentration of Cu(II) (1 M) without the addition of hydrogen peroxide [8]. Although Cu(II) (1 M), EGC and EC (100 M each) showed no effect separately on the viability of E. coli, the combination of Cu(II) (1 M) with EGC (1, 10, 100 M) or EC (100 M) killed E. coli cells (Table 1) [8]. Although the killing was accompanied by a depletion in both the ATP and potassium pools of the cells (Table Table 5. Effect of EGC on the Binding of Copper to E. coli Cells at 37°C EGC concentrations (M)
Bound copper (%)
1 10 100
63 ⫾ 2 70 ⫾ 4 67 ⫾ 11
Washed E. coli cells (8 ⫻ 107 cells) were incubated with varous concentrations of EGC in the presence of Cu(II) (1 M) in reaction buffer (pH 6.5) at 37°C. After incubation of the cell suspension for 60 min, supernatant was separated by centrifugation and copper concentration was determined (mean ⫾ SD, n ⫽ 3).
Bactericidal catechin-Cu(II)
2), it was found that the DNA double strand did not break in the killing, indicating that the bactericidal activity of catechins in the presence of Cu(II) is derived from damage to the cytoplasmic membrane. The bactericidal activity of EGC-Cu(II) was completely inhibited by bathocuproine, a specific chelator of Cu(I) (Table 1). This fact indicates that the bactericidal activity of EGC-Cu(II) is mediated through reduction of Cu(II) to Cu(I) by EGC. Reoxidation of Cu(I) to Cu(II) by molecular oxygen generates active oxygen species, superoxide anion at first and then hydrogen peroxide and hydroxyl radicals [7,16]. In order to investigate the possibility of the involvement of active oxygen species in the bactericidal activity of EGC-Cu(II), we examined the generation of superoxide anion, hydrogen peroxide, and hydroxyl radicals after the addition of SOD, catalase, and DMSO. As shown in Table 1, catalase and SOD completely suppressed the bactericidal activity of EGCCu(II). However, similar suppression against the bactericidal activity was observed by the addition of heated enzymes (Table 1). As shown in Fig. 1, addition of either catalase or heated catalase to the cells incubated with EGC and Cu(II) for 2, 30, and 60 min, completely inhibited additional killing. These observations suggest that the protective effects of these enzymes are not derived from the elimination of active oxygen species, but rather from their binding of Cu(I) as proposed by Løvstad [17]. Because these enzymes can not pass through the cell membrane, they must suppress the bactericidal activity by sequestering Cu(I) outside the cells. That is, Cu(I) is not present in the cells but on the surface of the cells where catalase can easily approach. The bactericidal activity of EGC-Cu(II) was not influenced by the addition of DMSO, a hydroxyl radical scavenger (Table 1). Induction of endogenous catalase in the cells increased their ability to resist being killed by the combination of catechins and Cu(II) (Table 3). Therefore it is evident that hydrogen peroxide inside the cells must be critical for the bactericidal activity of catechin-Cu(II). EGC and EC with Cu(II) can generate hydrogen peroxide under aerobic conditions (Table 4). The highest concentration of hydrogen peroxide, 41.9 M, was detected in the case of 100 M EGC and 1 M Cu(II) (Table 4). However, no killing was observed after the addition of 41.9 M hydrogen peroxide and 1 M Cu(II) to the cell suspension (Fig. 2). A hydrogen peroxide concentration of more than 500 M was necessary to manifest the same bactericidal activity as that observed in the presence of 100-M EGC and 1-M Cu(II) (Table 1, Fig. 2). These results suggest that the concentration of hydrogen peroxide in the vicinity of the cells must be higher than that detected in the bulk reaction buffer. It is known that Cu(II) is complexed by catechins [18,19] and
1249
that catechins are apt to be present on bacterial membranes [3]; atomic absorption showed that 63% or more of the copper ions were bound to the cells in the presence of EGC (Table 5). Aronovitch et al. [10] suggested that with hydrogen peroxide added, epinephrine-Cu(II) complex binds to the cell surface to induce oxidative membrane damage. We believe that catechin-Cu(II) on the cell surface reacts locally with molecular oxygen to produce hydrogen peroxide. The hydrogen peroxide generated on the surface can then enter easily into cells and can cause damage to the cytoplasmic membrane. The apparently low concentration of hydrogen peroxide observed in the bulk suspension will not reflect the local concentration of hydrogen peroxide generated on the cell surface. Therefore, we propose that Cu(I), and its redox reactions involving catechins and hydrogen peroxide on the cell surface, must be involved in the killing of the cells. Whether hydrogen peroxide is the only critical substance or not, should be clarified in due course.
REFERENCES [1] Toda, M.; Okubo, S.; Hiyoshi, R.; Shimamura, T. The bactericidal activity of tea and coffee. Lett. Appl. Microbiol. 8:123–125; 1989. [2] Toda, M.; Okubo, S.; Hara, Y.; Shimamura, T. Antibacterial and bactericidal activities of tea extracts and catechins against methicilline resistant Staphylococcus aureus (in Japanese). Jpn. J. Bacteriol. 46:839 – 845; 1991. [3] Ikigai, H.; Nakae, T.; Hara, Y.; Shimamura, T. Bactericidal catechins damage the lipid bilayer. Biochim. Biophys. Acta 1147: 132–136; 1993. [4] Huang, S.-W.; Frankel, E. N. Antioxidant activity of tea catechins in different lipid systems. J. Agric. Food Chem. 45:3033–3038; 1997. [5] Sotomatu, A.; Nakao, M.; Hirai, S. Phospholipid peroxidation induced by the catechol-Fe3⫹ (Cu2⫹) complex: a possible mechanism of nigrostriatal cell damage. Arch. Biochem. Biophys. 283: 334 –342; 1990. [6] Sahu, S. C.; Gray, G. C. Interactions of flavonoids, trace metals, and oxygen: nuclear DNA damage and lipid peroxidation induced by myricetin. Cancer Lett. 70:73–79; 1993. [7] Hayakawa, F.; Kimura, T.; Maeda, T.; Fujita, M.; Sohmiya, H.; Ando, T. DNA cleavage reaction and linoleic acid peroxidation induced by tea catechins in the presence of cupric ion. Biochim. Biophys. Acta 1336:123–131; 1997. [8] Kimura, T.; Hoshino, N.; Yamaji, A.; Hayakawa, F.; Ando, T. Bactericidal activity of catechin-copper(II) complexes on Escherichia coli ATCC11775 in the absence of hydrogen peroxide Lett. Appl. Microbiol. 27:328 –330; 1998. [9] Davis, B. D.; Mingioli, E. S. Mutants of E. coli requiring methionine or vitamin B12. J. Bacteriol. 60:17–28; 1950. [10] Aronovitch, J.; Godinger, D.; Czapski, G. Bactericidal activity of catecholamine copper complexes. Free Radic. Res. Commun. 12-13:479 – 488; 1991. [11] Nishida, K.; Takagi, M.; Yano, K. Escherichia coli cell inactivation by superoxide anion and hydrogen peroxide. J. Gen. Appl. Microbiol. 27:477– 485; 1981. [12] Stanley, P. E. Extraction of adenosine triphosphate from microbial and somatic cells. Methods Enzymol. 60:17–28; 1986. [13] Kohen, R.; Chevion, M. Cytoplasmic membrane is the target organella for transition metal mediated damage induced by paraquat in Escherichia coli. Biochemistry 27:2597–2603; 1988. [14] Aebi, H. Catalase in vitro. Methods Enzymol. 105:121–126; 1984. [15] Guilbault, G.; Brignac, P. J.; Juneau, M. New substrates for the
1250
N. HOSHINO et al.
fluorometric determination of oxidative emzymes. Anal. Chem. 40:1256 –1263; 1968. [16] Rahman, A.; Shahabuddin; Hadi, S. M.; Parish, J. H.; Ainley, K. Strand scission in DNA induced by quercetin and Cu(II): role of Cu(I) and oxygen free radicals. Carcinogenesis 10:1833–1839; 1989. [17] Løvstad, R. A. Copper catalyzed oxidation of ascorbate (vitamin C). Inhibitory effect of catalase, superoxide dismutase, serum
proteins (ceruloplasmin, albumin, apotransferin) and amino acids. Int. J. Biochem. 19:309 –313; 1987. [18] Cetta, G.; Pallavicini, G.; Tenni, R.; Bisi, C. Influence of flavonoid-copper complexes on cross linking in elastin. Itali. J. Biochem. 26:317–327; 1977. [19] Weber, G. HPLC with electrochemical detection of metal-flavonoid-complexes isolated from food. Chromatographia 26:133– 138; 1988.