Synergistic activity and mode of action of flavonoids isolated from smaller galangal and amoxicillin combinations against amoxicillin-resistant Escherichia coli

Authors

  • G. Eumkeb,

    1.  School of Pharmacology, Institute of Science, Suranaree University of Technology, 111 University Avenue, Suranaree Subdistrict, Muang District, Nakhonratchasima 30000, Thailand
    Search for more papers by this author
  • S. Siriwong,

    1.  School of Pharmacology, Institute of Science, Suranaree University of Technology, 111 University Avenue, Suranaree Subdistrict, Muang District, Nakhonratchasima 30000, Thailand
    Search for more papers by this author
  • S. Phitaktim,

    1.  School of Pharmacology, Institute of Science, Suranaree University of Technology, 111 University Avenue, Suranaree Subdistrict, Muang District, Nakhonratchasima 30000, Thailand
    Search for more papers by this author
  • N. Rojtinnakorn,

    1.  School of Pharmacology, Institute of Science, Suranaree University of Technology, 111 University Avenue, Suranaree Subdistrict, Muang District, Nakhonratchasima 30000, Thailand
    Search for more papers by this author
  • S. Sakdarat

    1.  School of Chemistry, Institute of Science, Suranaree University of Technology, 111 University Avenue, Suranaree Subdistrict, Muang District, Nakhonratchasima 30000, Thailands
    Search for more papers by this author

Griangsak Eumkeb, School of Pharmacology, Institute of Science, Suranaree University of Technology, 111 University Avenue, Suranaree Subdistrict, Muang District, Nakhonratchasima 30000, Thailand. E-mail: griang@sut.ac.th

Abstract

Aim:  The smaller galangal is extracted, purified and identified the bioactive compounds. The purpose of this research was to investigate whether these isolated compounds have antibacterial and synergistic activity against amoxicillin-resistant Escherichia coli (AREC) when used singly and in combination with amoxicillin. The primarily mode of action is also studied.

Method and Results:  The galangin, kaempferide and kaempferide-3-O-β-d-glucoside were isolated. The minimum inhibitory concentrations(MIC) of amoxicillin and these flavonoids against AREC were between 500 and >1000 μg ml−1. Synergistic activity was observed on combining amoxicillin with these flavonoids. The combinations of amoxicillin and these flavonoids exhibited a synergistic effect, reducing AREC cell numbers. Electron microscopy showed that these combinations damaged the ultrastructure of AREC cells. The results indicated that these combinations altered outer membrane permeability but not affecting cytoplasmic membrane. Enzyme assays showed that these flavonoids had an inhibitory activity against penicillinase.

Conclusion:  These results indicated that these flavonoids have the potential to reverse bacterial resistance to amoxicillin in AREC and may operate via three mechanisms: inhibition of peptidoglycan and ribosome synthesis, alteration of outer membrane permeability, and interaction with β-lactamases.

Significance and Impact of the Study:  These findings offer the potential to develop a new generation of phytopharmaceuticals to treat AREC.

Introduction

Smaller galangal (Alpinia officinarum Hance) is a pungent and aromatic rhizome, which is a member of the ginger family (Zingiberaceae). The rhizome is cultivated in India, Vietnam, Southern China and Thailand and is used as a spice, a condiment for flavouring, a carminative and a traditional medicine for the treatment for pyogenic diseases, abdominal discomfort and diarrhoea (Athamaprasangsa et al. 1994). The rhizomes of galangal, which have pungent and fragrant properties, are regularly used as a flavouring agent for food in Thailand. Chemical and pharmacological studies of the rhizomes of smaller galangal revealed three major groups of chemical constituents: flavonoids, glycosides and diarylheptanoids. It has been reported that smaller galangal has biological activities, including antitumor, antiulcer, antibacterial and antifungal properties (Itokawa et al. 1985; Ly et al. 2003). The two diarylheptanoids and galangin isolated from this plant were found to inhibit LPS-induced nitric oxide (NO) production with IC50 values of 33–62 μ mol l−1 (Matsuda et al. 2006). Moreover, four diarylheptanoids, kaemferide and galangin extract from the rhizomes of this galangal have been reported that these compounds inhibited meanogenesis with IC50 values of 10–48 μ mol l−1 (Matsuda et al. 2009). Isolated galangin from propolis showed a minimum inhibitory concentration (MIC) against multidrug-resistant Staphylococcus aureus (S. aureus), Enterococcus spp. and Pseudomonas aeruginosa of 0·16–0·24 mg ml−1 (Pepeljnjak and Kosalec 2004). In addition, flavonol galangin showed that it caused aggregation of S. aureus cells (Cushnie et al. 2007).

Bacterial resistance to β-lactam antibiotics is a global problem. A recent study reported the persistence of ampicillin- and nalidixic acid-resistant strains of Escherichia coli in healthy, untreated chicken flocks (Bortolaia et al. 2010). In recent years, some E. coli strains carrying plasmid-mediated ampC genes have been detected that have the ability to spread β-lactam resistance among Gram-negative bacilli, presenting a serious medical challenge (Lalitagauri et al. 2006). The emergence of faecal E. coli isolates exhibiting reduced susceptibility or resistance to extended-spectrum cephalosporins has been reported among pigs in Spain (Escudero et al. 2010). About 3–8 per cent of E. coli strains have been shown to vary in their resistance to antipseudomonal β-lactams. Moreover, more than 3500 isolates of Enterobacteriaceae, 48–69% of those tested, were found to be resistant to aminopenicillins, 11–45% were resistant to amoxicillin-clavulanic acid and 11–17% were resistant to third-generation cephalosporins (Gál et al. 2000). Antibiotics available for the treatment for E. coli infection exhibit toxicity, and their use is frequently associated with unwanted side-effects. The development of novel antibiotics and/or a new generation of phytopharmaceuticals that can reverse the resistance to well-established therapeutic agents that have lost their original effectiveness is of great importance (Wagner and Ulrich-Merzenich 2009).

In this study, we isolated and identified bioactive compounds from the rhizome of smaller galangal. Furthermore, we investigated the in vitro activity and mode of action of galangin, kaempferide and kaempferide-3-O-β-d-glucoside, the bioactive constituents isolated from smaller galangal, against amoxicillin-resistant Escherichia coli (AREC) strain DMST 20662, alone and in combination with amoxicillin.

Materials and methods

Plant materials, β-lactam antibiotics and bacterial strains

Fresh rhizomes of smaller galangal were isolated from the soil located in Nakhonratchasima Province, Thailand. The plant specimens were deposited at the Thailand National Herbarium and identified by Dr Paul J. Grote, Institute of Science, Suranaree University of Technology (Nakhonratchasima Province, Thailand). The rhizomes of smaller galangal were washed thoroughly and dried in an oven at 50°C for 3 days. The dried samples were then ground into a powder. Galangin, kaemferide and kaempferide-3-O-β-d-glucoside were purchased from Indofine Chemical (USA). Amoxicillin, clavulanic acid, TTX-100 and o-nitrophenol-β-d-galactoside were obtained from Sigma (Sigma-Aldrich, UK). Mueller–Hinton broth was obtained from Oxoid (Basingstoke, UK). AREC strain DMST 20662 was obtained from the Department of Medical Sciences, Ministry of Public Health, Thailand. E. coli American Type Culture Collection (ATCC) strain 25922, used as a positive control, was purchased from the (ATCC).

Extraction and isolation of bioactive compounds

General experiment procedures, extraction, purification and identification of bioactive compounds and the structure of three compounds were employed as previously described by Eumkeb et al. (2010). Briefly, 2 kg of dried powder of the rhizomes of smaller galangal were extracted consecutively with hexane, chloroform and methanol for 12 h in each solvent using soxhlet extraction apparatus. The extraction processes were performed until getting three compounds. The structural elucidation of the isolated compounds was carried out on the basis of spectral analyses, including UV, infrared, mass spectrometry, 1H-NMR and 13C-NMR, as well as comparisons with reported values in the literature.

Bacterial suspension standard curve

To obtain bacterial suspensions of a known viable count, the method of Richards and Xing (1993) was employed. Mueller–Hinton agar and Cation-adjusted Mueller–Hinton broth (CAMHB) were used as medium.

Minimum inhibitory concentration (MIC) and checkerboard determinations

MIC and checkerboard determinations of amoxicillin, galangin, kaempferide and kaempferide-3-O-β-d-glucoside against AREC were performed following the method of Liu et al. (2000) and the Clinical and Laboratory Standards Institute (2006). Mueller–Hinton agar and CAMHB were used as medium. E. coli ATCC 25922 and clavulanic acid were used as positive controls.

Killing curve determinations

Viable counts for the determination of killing curves were performed as previously described by Richards and Xing (1993) and the Clinical and Laboratory Standards Institute (2006). Mueller–Hinton agar and CAMHB were used as medium.

Transmission electron microscopy

The doses of amoxicillin, galangin, kaempferide and kaempferide-3-O-β-d-glucoside when used singly and in combination that dramatically decreased the MICs against AREC were chosen for analysis by transmission electron microscopy. Bacterial subcultures were prepared for transmission electron microscopy following the method of Richards and Xing (1993). CAMHB was used as medium.

Outer membrane (OM) and cytoplasmic membrane (CM) permeability

To examine the impact of amoxicillin, galangin, kaempferide and kaempferide-3-O-β-d-glucoside when used singly and in combination on the function of the OM and CM as permeability barriers, the methods of Jackson and Demoss (1965) and Marri et al. (1996) were employed. Toluene at 50 μg ml−1 was used as positive control for CM permeability.

Enzyme assays

The β-lactamases (Penicillinase) of Enterobacter cloacae (E. cloacae) were obtained from Sigma (Poole, UK). Enzymes activities were adjusted to concentrations sufficient to hydrolyse 50–60% substrate in 5 min. Flavonoids were preincubated with enzyme in 50 m mol l−1 sodium phosphate buffer (pH 7·0) at 37°C for 5 min prior to substrate addition. Time-course assays were carried out using methanol/acetic acid (100 : 1) as stop reagent and analyses of the remaining substrate determined by reverse-phase HPLC with acetronitrile/acetate in the mobile phase (Richards et al. 1995).

Statistical analyses

Values are expressed as mean ± standard error of the mean (SEM). The comparison between pre- and post-treatment with TTX-100 in outer membrane permeability data or control and each isolated flavonoid in enzyme assays data was submitted to Student’s t-test. All values were considered statistically significant at P < 0·05 (*) and P < 0·01 (**) using IBM SPSS 19·0 software (IBM Corp., Armonk, NY, USA).

Results

Isolation and structure of compounds 1–3

Bioactive compounds 1, 2 and 3 were identified as 3,5,7-trihydroxyflavone (galangin), 3,5,7-trihydroxy-4′-methoxyflavone (kaempferide) and kaempferide-3-O-β-d-glucoside, respectively (Table 1 and Fig. 1) (Agrawal 1992; Juha-Pekka et al. 2004; Rubens and Wangner 2005; Eumkeb et al. 2010).

Table 1.   The 1H-NMR and 13C-NMR chemical shifts of compound 1 (galangin), 2 (kaempferide) and 3 (kaempferide-3-O-β-d-glucoside)
Proton positionCompoundCarbon positionCompound
 1 23123
12-glu (dd, = 7·90 Hz) 5·411130·2123·0122·0
218·2(dd, = 8·72 Hz)8·18·012  100·3
22 (m) 3·22159·8146·0155·7
3-OH9·6(s)9·4 21126·1129·2130·2
317·6(dd, = 8·72 Hz) 7·022  73·7
32 (m) 3·33136·4135·9133·0
417·5(m)  31128·5114·0113·3
41-OCH3 (s)3·83·832  76·0
42 (m) 3·14178·1176·0176·8
5-OH12·4(s)12·312·441127·9160·3160·6
517·6(dd, = 8·72 Hz)7·17·041-OCH3 55·155·0
52 (m) 3·442  69·4
66·2(d, s)6·36·05169·1160·5160·6
618·2(dd, = 8·72 Hz)8·18·051128·5114·0113·3
62 (m) 3·452  76·4
7-OH10·7(s)10·810·7698·198·198·2
86·4(s)6·46·161126·1129·2130·2
     62  60·5
     7166·2163·7163·6
     897·893·393·3
     9159·7156·1155·3
     10103·2103·4103·7
Figure 1.

 Structure of compound 1(galangin), 2(kaempferide) and 3(kaempferide-3-O-β-D-glucoside).

Minimum inhibitory concentration (MIC) and checkerboard assay

The MICs for amoxicillin, galangin, kaempferide and kaempferide-3-O-β-d-glucoside against AREC are shown in Table 2. The MICs of amoxicillin, galangin, kaempferide and kaempferide-3-O-β-d-glucoside alone against AREC were >1000, 500, 500 and 600 μg ml−1, respectively. When amoxicillin was combined with each of these flavonoids, the fractional inhibitory concentrations (FICs) of amoxicillin plus galangin or kaempferide or kaempferide-3-O-β-d-glucoside were 10 + 40 (μg ml−1, FIC indices <0·09), 10 + 40 (μg ml−1, FIC indices <0·09) and 10 + 50 (μg ml−1, FIC indices <0·09), respectively. Synergy is considered at an FIC index ≤0·5 (Johnson et al. 2004). Thus, the amoxicillin plus flavonoid combinations were synergistic against the AREC strain. E. coli ATCC 25922 and clavulanic acid were used as positive controls.

Table 2.   Minimum inhibitory concentrations (MICs ± SEM), fractional inhibitory concentrations (FICs) and FIC indexes determined by checkerboard assays of amoxicillin, galangin, kaempferide and kaempferide-3-O-β-d-glucoside alone, and amoxicillin in combination with isolated flavonoids, against clinical isolates of Escherichia coli DMST 20662 [amoxicillin-resistant E. coli, (AREC)]
StrainMIC (μg ml−1)FIC (μg ml−1)FIC index
amoxgalkaekagclaamox + galamox + kaeamox + kagamox + claamox + galamox + kaeamox + kagamox + cla
  1. Each compound was measured three times.

  2. amox, amoxicillin; gal, galangin; kae, kaempferide; kag, kaempferide-3-O-β-D-glucoside; cla, clavulanic acid, N/D, not determined; ATCC, American Type Culture Collection.

  3. *E. coli ATCC 25922 was used as a positive control.

DMST 20662>1000 ± 8500 ± 4500 ± 4600 ± 4>128 ± 210 + 4010 + 4010 + 50>1000 + >128<0·09<0·09<0·092·0
ATCC 25922*4·0 ± 0·01100 ± 2100 ± 2200 ± 216 ± 0·2N/DN/DN/D0·6 + 0·2N/DN/DN/D0·15

Killing curve determinations

Figure 2 shows that the viable counts for AREC were slightly reduced in the presence of amoxicillin, galangin, kaempferide and kaempferide-3-O-β-d-glucoside alone at 200, 100, 100 and 100 μg ml−1, respectively, compared with the untreated control culture between 6 and 24 h incubation. The combination of amoxicillin at 10 μg ml−1 plus kaemferide at 40 μg ml−1 or kaempferide-3-O-β-d-glucoside at 50 μg ml−1 decreased the cells to 4 × 103 and 1 × 104 colony-forming units per ml after 6 h, respectively. In addition, amoxicillin at 10 μg ml−1 in combination with 40 μg ml−1 of galangin reduced the number of colony-forming units per ml by 1 × 103 over 6 h. The reduced cell counts did not recover to normal levels within 24 h.

Figure 2.

 Viability of Escherichia coli DMST 20662 after treatment with amoxicillin and/or isolated flavonoids; symbol represents: (inline image) control (without amoxicillin or flavonoid); (inline image) 200 μg ml−1 amoxicillin alone; (inline image)100 μg ml−1 galangin; (inline image)100 μg ml−1 kaempferide; (inline image) 100 μg ml−1 kaempferide-3-O-β-d-glucoside; (inline image) amoxicillin 10 μg ml−1 plus kaempferide-3-O-β-d-glucoside 50 μg ml−1; (inline image) amoxicillin 10 μg ml−1 plus kaempferide 40 μg ml−1; (inline image) amoxicillin 10 μg ml−1 plus galangin 40 μg ml−1. The values plotted are the means of four observations, and the vertical bars indicate the standard errors of the means.

Transmission electron microscopy

Electron micrographs of thin sections of log-phase E. coli DMST 20662 cells grown for 4 h in the presence of CAMHB, amoxicillin, flavonoids alone and the combination of amoxicillin plus flavonoids are shown in Fig. 3a–h. Fig. 3a–e shows the appearance of normal log-phase cells of E. coli DMST 20662 in the presence of CAMHB or after treatment with amoxicillin at 200 μg ml−1 or each flavonoid at 100 μg ml−1 using alone. The cell wall and the cytoplasmic membrane can be distinguished. The electron-dense ribosomes can be seen in great number in cytoplasm. Whereas amoxicillin 8 μg ml−1 plus galangin 30 μg ml−1 revealed marked morphological damage to the cells, this damage included loosening or detachment of the OM, possibly resulting from damage to the internal peptidoglycan layer. Some of the bacteria exhibited electron-transparent areas devoid of ribosomes in the cytoplasm. Most of the treated bacteria also appeared considerably larger than the control bacteria (Fig. 3f). The micrographs of this strain after exposure to amoxicillin 8 μg ml−1 plus kaempferide 30 μg ml−1 or kaempferide-3-O-β-d-glucoside 40 μg ml−1 showed that some of these bacterial cells exhibited larger gap between outer membrane and cytoplasmic membrane, and many of these bacterial cells exhibited morphological damage of cell wall and cell shape and electron-transparent area in cytoplasm owing to loosening most of organelles. Several bacterial cells showed broken cell and distortion of cell wall (Fig. 3g and h).

Figure 3.

 Ultrathin sections of log phase Escherichia coli DMST 20662 cells grown in cation-adjusted Mueller-Hinton broth (CAMHB) containing: (a) no drugs (control); (b) amoxicillin 200 μg ml−1; (c) galangin 100 μg ml−1; (d) kaempferide 100 μg ml−1; (e) kaempferide-3-O-β-D-glucoside 100 μg ml−1; (f) amoxicillin 8 μg ml−1 plus galangin 30 μg ml−1; (g) amoxicillin 8 μg ml−1 plus kaempferide 30 μg ml−1; (h) amoxicillin 8 μg ml−1 plus kaempferide-3-O-β-D-glucoside 40 μg ml−1. Bar = 0.5 μm (b,c,f,g,h); 1 μm (a,d,e).

Outer membrane (OM) and cytoplasmic membrane (CM) permeability

Figure 4 shows that amoxicillin or flavonoids did not alter the OM permeability when used alone. In contrast, the combination of 8 μg ml−1 amoxicillin plus 30, 30 or 40 μg ml−1 of galangin, kaemferide or kaempferide-3-O-β-d-glucoside, respectively, altered the OM permeability of AREC. The differences in absorbance between pre- and post-treatment of these combinations with TTX-100 were significant at P < 0·05 (at time = 30 min) and P < 0·01 (at time after 60 min), respectively. The effects of amoxicillin plus kaemferide or kaempferide-3-O-β-d-glucoside were less than those of amoxicillin plus galangin. Triton X-100 was used as a permeabilizing probe.

Figure 4.

 Permeabilization of Escherichia coli DMST 20662 by amoxicillin and/or isolated flavonoids; symbol represents: (inline image) control (without amoxicillin or flavonoid); (inline image) 300 μg ml−1 TTX-100; (inline image) 200 μg ml−1 amox; (inline image) 200 μg ml−1 amox plus 300 μg ml−1 TTX-100; (inline image) 100 μg ml−1 gal; (inline image) 100 μg ml−1 gal plus 300 μg ml−1 TTX-100; (inline image) 100 μg ml−1 kae; (inline image) 100 μg ml−1 kae plus 300 μg ml−1 TTX-100; (inline image) 100 μg ml−1 kag; (inline image) 100 μg ml−1 kag plus 300 μg ml−1 TTX-100; (inline image) amox 8 μg ml−1 plus gal 30 μg ml−1; (inline image) amox 8 μg ml−1 plus gal 30 μg ml−1 plus 300 μg ml−1 TTX-100; (inline image) amox 8 μg ml−1 plus kae 30 μg ml−1; (inline image) amox 8 μg ml−1 plus kae 30 μg ml−1 plus 300 μg ml−1 TTX-100; (inline image) amox 8 μg ml−1 plus kag 40 μg ml−1; (inline image) amox 8 μg ml−1 plus kag 40 μg ml−1 plus 300 μg ml−1 TTX-100. The lysis caused by subsequent treatment with 300 μg ml−1 Triton X-100 (TTX-100). The significance of differences between pre- and post-treatment with TTX-100 are represented at, * = P < 0.05; ** = P < 0.01. The bars represent the standard deviations of three replicates; amox, amoxicillin; gal, galangin; kae, kaempferide; kag, kaempferide-3-O-β-D-glucoside.

The effect of amoxicillin at 200 μg ml−1 or each flavonoid at 100 μg ml−1 alone and the combinations of 8 μg ml−1 amoxicillin plus each of 30, 30 or 40 μg ml−1 galangin, kaemferide or kaempferide-3-O-β-d-glucoside, respectively, on CM permeability was investigated using the cytoplasmic enzyme β-galactosidase. Our results exhibited that there was no increase in β-galactosidase activity with increasing time in the presence of amoxicillin, galangin, kaemferide or kaempferide-3-O-β-d-glucoside both alone and in combination (Table 3), indicating that these compounds, either singularly or in combination, do not alter the permeability of the CM of AREC.

Table 3. β-galactosidase activity results Escherichia coli DMST 20662 after treatment with 200 μl ml−1 amoxicillin, 100 μl ml−1 galangin, 100 μl ml−1 kaemferide, 100 μl ml−1 kaempferide-3-O-β-d-glucoside alone or in combination of amoxicillin 8 μl ml−1 plus galangin 30 μl ml−1, amoxicillin 8 μl ml−1 plus kaempferide 30 μl ml−1 and amoxicillin 8 μl ml−1 plus kaempferide-3-O-β-d-glucoside 40 μl ml−1 in Mueller-Hinton broth at 37°C. Toluene was used as positive control. The determination was carried out in triplicate
TimeControl (no drug)amoxgalkaekagamox + galamox + kaeamox + kagToluene 50 μl ml−1 (Positive control)
  1. Neg, no evidence of activity; Pos, have evidence of activity; amox, amoxicillin; gal, galangin; kae, kaempferide; kag, kaempferide-3-O-β-d-glucoside.

0 hNegNegNegNegNegNegNegNegPos
1 hNegNegNegNegNegNegNegNegPos
2 hNegNegNegNegNegNegNegNegPos
3 hNegNegNegNegNegNegNegNegPos
4 hNegNegNegNegNegNegNegNegPos
5 hNegNegNegNegNegNegNegNegPos

Enzyme assays

The ability of flavonoids to inhibit the in vitro activity of β-lactamases varied considerably. Figure 5 indicates that galangin, kaemferide and kaempferide-3-O-β-d-glucoside have an inhibitory activity against penicillinase type IV from E. cloacae. The differences in benzylpenicillin level between each isolated flavonoid and control (without flavonoid) group were significant at P < 0·01 (at time 10 and 30 min) except that kaempferide-3-O-β-d-glucoside group testing showed significant difference at P < 0·05 (at time 30 min). Galangin showed marked inhibitory activity. These results indicate that in addition to the direct effect on cell structure and cell division, the resistance-reversing activity of flavonoids against bacteria might also include inhibition of β-lactamase activity.

Figure 5.

 The inhibitory activity of flavonoids against penicillinase in hydrolyzing benzylpenicillin. β-lactamase used from E. cloacea; symbol represents flavonoids (200 μg ml−1); (inline image) galangin; (inline image) kaempferide; (inline image) kaempferide-3-O-β-D-glucoside; (inline image) control (without flavonoid). The significance of differences between control and isolated flavonoids are represented at, * = P < 0.05; ** P < 0.01. The bars represent the standard deviations of three replicates.

Discussion

The results from MIC determinations are practically agreement with those of Drago et al. (2007) that MIC of galangin in Actichelated® propolis against E. coli was 168–334 mg l−1. The results of checkerboard and viable count analyses of AREC revealed the synergistic effects of these isolated flavonoids and amoxicillin against this strain. These results were in agreement with those of Ozkan et al. (2010) who showed that the flavonol extracted from Salvia pisidica exhibited antibacterial activity against E. coli ATCC 25922 at a concentration of 13 g 100 ml−1. These findings are also in agreement with those of Hemaiswarya et al. (2008) who reported that flavonoids and synthetic drugs exhibited synergistic activity against bacteria. Previous research reported that galangin from propolis showed in vitro antimicrobial activity against E. coli and other Gram-negative bacteria (Castaldo and Capasso 2002). Similarly, these results are in substantial agreement with those of Lee et al. (2008) that galangin plus gentamicin showed synergistic effect against methicillin-resistant Staphylococcus aureus KCMM 40501 and DPS-1 strains. Likewise, previous research reported that galangin plus ceftazidime exhibited synergistic effect against penicillin-resistant S. aureus (Eumkeb et al. 2010). Furthermore, the results of Denny et al. (2002) showed that galangin inhibited metallo-β-lactamase from Stenotrophomonas maltophilia by orientation at the active site of the enzyme. In contrast, Oonmetta-aree et al. (2006) found that D,L-1-acetoxychavicol acetate, the major compound extracted from galangal (Alpinia galanga Linn.), had no effect on E. coli.

The result of OM permeability is quite similar to previous report that CTX-A altered both the outer and inner membranes of E. coli K-12 cells (Marri et al. 1996), whereas the result of CM permeability can be explained by assuming that the CM is a highly selective barrier, enabling a cell to concentrate specific metabolites and excrete waste materials. The general structure of most biological membranes is a phospholipid bilayer. The major proteins of the cell membrane generally present a highly hydrophobic external surface and form intimate associations with highly nonpolar fatty acid chains (Brock et al. 1994; Tropp 1997). Previous findings suggested that membrane proteins solubilized by sodium lauryl sarcosinate (sarkosyl) were identical to those of the CM of E. coli and that the presence of Mg2+during treatment with sarkosyl afforded partial protection of the CM from dissolution (Filip et al. 1973). The OM has a much simpler protein composition and is characterized by the presence of a single major structure with a molecular weight of about 44 kDa (Schnaitmann 1971). These results indicated that galangin, kaemferide and kaempferide-3-O-β-d-glucoside alone have not only weak activity against AREC but also the ability to reverse the resistance of such bacterial strain. This may involve three mechanisms of action by these isolated flavonoids: (i) inhibition of peptidoglycan and ribosome synthesis, resulting in damage to the peptidoglycan layer and a cell devoid of ribosomes; (ii) inhibition of the activity of certain extended-spectrum β-lactamases; and (iii) alteration of OM permeability. These mechanisms of action are virtually in substantial agreement with those of Eumkeb et al. (2010) that mechanisms of action and synergistic activity of galangin plus ceftezidime against penicillin-resistant Staphylococcus aureus (PRSA) may involve in inhibition protein synthesis, have an effect on PBP 2a, interact with penicillinase and cause cytoplasmic membrane damage. It has previously been shown that there was no cross-resistance between galangin and the 4-quinolone drugs (Cushnie and Lamb 2006).

These results can be compared to previous finding that quercetin, a citrus flavonol that has chemical structure virtually the same as galangin, suppressed the biofilm formation and emerged as potent and possibly a nonspecific inhibitor of cell–cell signalling of Ecoli O157:H7 (Vikram et al. 2010).

Previous studies demonstrated that clavulanate, which shares the same lactam ring with β-lactam antibiotics, caused a considerable increase in elevated β-lactamase production (Stapleton et al. 1995; Tzouvelekis et al. 1997). These results suggest that the current β-lactamase inhibitors could also be broken by the same mechanism as the β-lactam antibiotics. This research provides a unique example, as we show that galangin, kaemferide and kaempferide-3-O-β-d-glucoside that lack a β-lactam structure can reverse bacterial resistance to β-lactams via multiple mechanisms. Because of different structure, these isolated flavonoids are unlikely to induce β-lactamase production. It should also be noted that currently available β-lactamase inhibitors, unlike flavonoids, cannot reverse the resistance of AREC, one of the most dangerous bacterial pathogens. Galangin, kaemferide and kaempferide-3-O-β-d-glucoside as a new generation of phytopharmaceuticals may be used in combination with amoxicillin for the treatment for AREC infection that cannot be treated with amoxicillin alone.

In conclusion, galangin, kaempferide and kaempferide-3-O-β-d-glucoside have the potential to reverse bacterial resistance to amoxicillin in AREC. This is the first report of the anti-AREC activity and mode of action of these isolated flavonols. In view of their limited toxicity, these isolated flavonols offer the potential to develop a valuable adjunct to amoxicillin for the treatment of AREC; if possible, blood and tissue levels would be achievable to work synergistically.

Acknowledgements

The authors are indebted and grateful to the following persons and institutions for their invaluable assistance in carrying out this study: Miss Kanjana Wongkumsound and Miss Nuttaporn Samart for their considerable contribution, The Thailand Research Fund for grant support and The National Research Council of Thailand for research funds.

Conflict of Interests

The authors have declared that no conflict of interest exists.

Ancillary