Aims: To investigate the antimicrobial properties of phenolic compounds present in Finnish berries against probiotic bacteria and other intestinal bacteria, including pathogenic species.
Methods and Results: Antimicrobial activity of pure phenolic compounds representing flavonoids and phenolic acids, and eight extracts from common Finnish berries, was measured against selected Gram-positive and Gram-negative bacterial species, including probiotic bacteria and the intestinal pathogen Salmonella. Antimicrobial activity was screened by an agar diffusion method and bacterial growth was measured in liquid culture as a more accurate assay. Myricetin inhibited the growth of all lactic acid bacteria derived from the human gastrointestinal tract flora but it did not affect the Salmonella strain. In general, berry extracts inhibited the growth of Gram-negative but not Gram-positive bacteria. These variations may reflect differences in cell surface structures between Gram-negative and Gram-positive bacteria. Cloudberry, raspberry and strawberry extracts were strong inhibitors of Salmonella. Sea buckthorn berry and blackcurrant showed the least activity against Gram-negative bacteria.
Conclusions: Different bacterial species exhibit different sensitivities towards phenolics.
Significance and Impact of the Study: These properties can be utilized in functional food development and in food preservative purposes.
Flavonoids are common substances in the daily diet. These polyphenolic compounds are widely found in various types of edible plants, especially in vegetables, fruits, tea and wine. Over 4000 different flavonoids have been described and they are categorized into several subgroups. Flavonols (quercetin, myricetin and kaempferol) and flavones (apigenin and luteolin) are the most common phenolics in plant-based foods. Flavanones are typically present in citrus fruit, and flavanols in green tea. Berries, which are traditionally a part of the Finnish diet, are good sources of flavonols while the predominating group of flavonoids, especially in red berries, is anthocyanins (Fig. 1) (Heinonen et al. 1998; Häkkinen et al. 1999a).
Dietary flavonoids have attracted much interest recently because in vitro and in vivo studies suggest that they have a variety of beneficial biological properties, which may play an important role in the maintenance of human health. The flavonoids are potent antioxidants, free radical scavengers and metal chelators; they inhibit lipid peroxidation and exhibit various physiological activities including anti-inflammatory, antiallergic, anticarcinogenic, antihypertensive, antiarthritic and antimicrobial activities. Epidemiological studies have indicated that high flavonoid consumption is associated with reduced risk of chronic diseases like cardiovascular diseases (Middleton and Kandaswami 1994; Hertog et al. 1995). However, little is known about the absorption and subsequent distribution, metabolism and excretion of flavonoids in humans.
The intestinal microflora consists of approximately 1014 bacteria of over 400 species. This microflora and its consistency apparently play central roles in the metabolism and bioavailability of plant phenols such as flavonoids, the importance of which is largely unknown so far. By increasing the number and activity of probiotic bacteria such as Lactobacillus and Bifidobacterium in the colon, it has been possible to ease the symptoms of lactose intolerance, to heal and prevent diarrhoeal diseases and to stimulate the immune response (Salminen and Saxelin 1996). Interactions between plant phenolics and probiotics and the gastrointestinal flora remain poorly characterized. Many plant phenols are known to possess antimicrobial properties, so they might change the composition of the colonic flora in an unexpected way. On the other hand, microbial glucosidases and glucuronidases in the colon can possibly affect the bioactivity of glycosylated compounds by deconjugating or otherwise modifying them.
The aim of this study was to investigate the antimicrobial properties of phenolic compounds present in Finnish berries against probiotic bacteria and other intestinal bacteria, including pathogenic species. Such knowledge is important for the development of health-promoting functional foods containing both probiotic bacteria and plant material, such as berries.
MATERIALS AND METHODS
Bacterial strains and culture conditions
The bacterial strains used as test organisms are listed in Table 1. Excluding Lactobacillus plantarum VTT E-78076, all the other Lactobacillus strains have potential probiotic properties; most of the strains originate from the human gastrointestinal tract. Escherichia coli strain CM 871 (trpE65, uvrA155, recA56, lexA) is a DNA repair-deficient derivative from E. coli WP2 (Tweats et al. 1981). Salmonella enterica sv. Typhimurium SH5014 is a rough mutant strain (Stocker et al. 1979; LPS chemotype Rb2). The E. coli strains and Typhimurium SH5014 strain represented Gram-negative bacteria. All other strains were Gram-positive.
Table 1. Bacterial strains used in the study
Lactobacillus strains, Enterococcus faecalis VTT E-93203T and Bifidobacterium lactis VTT E-94508 were cultured in MRS broth (de Man, Rogosa, Sharpe-Broth, Oxoid) or on MRS agar (de Man, Rogosa, Sharpe-Agar, Oxoid). Cysteine (0·05% w/v) was added to MRS media for B. lactis. Lactobacillus strains and B. lactis were grown anaerobically at 37°C, except Lact. plantarum VTT E-78076, which was grown aerobically at 30°C. Enterococcus faecalis was cultured aerobically at 37°C. Escherichia coli strains and Typhimurium SH5014 were cultured aerobically at 37°C in Nutrient Broth (Difco) with agitation, or on Nutrient Agar (Difco). Frozen stock cultures were maintained at – 70°C. Before experimental use, cultures were transferred to solid or liquid media and incubated for 1–2 days. Cultures were then subcultured in liquid media, incubated for 12–24 h, and used as the source of inoculum for each experiment.
Apigenin, caffeic acid, (+)-catechin, chlorogenic acid, 3-coumaric acid, cyanidin chloride, delphinidin chloride, ferulic acid, isoquercitrin, kaempferol, cyanidin-3-glucoside (kuromanin), luteolin, myricetin, pelargonidin chloride, quercetin dihydrate, rutin trihydrate and trans-cinnamic acid (Fig. 1) were purchased from Extrasynthese. Folin-Ciocalteu’s phenol reagent and sodium carbonate were from Merck. All organic solvents were of HPLC grade.
Preparation and analysis of phenolic berry extracts
Phenolic extracts from blueberry (Vaccinium myrtillus), raspberry (Rubus idaeus, var. Ottawa), lingonberry (Vaccinium vitis-idaea), blackcurrant (Ribes nigrum, var. Öjeby), cloudberry (Rubus chamaemorus), cranberry (Vaccinium oxycoccus), sea buckthorn berry (Hippophae rhamnoides) and strawberry (Fragaria ananassa Senga Sengana) were prepared according to Kähkönen et al. 1999. Briefly, ground lyophilized berry material (500 mg) was weighed into a centrifuge tube, 10 ml of aqueous 70% acetone were added and the sample was homogenized (Ultra-Turrax) for 1 min. Tubes were centrifuged (3000 g, 15 min) and the clear supernatant fluid was collected. The procedure was repeated with another 10 ml of solvent. The supernatant fluids were combined and taken to dryness. The solid residue was dissolved in water. Solid-phase extraction (SPE) was used to remove sugars in the berry extracts (Heinonen et al. 1998) prior to lyophilization of the extract.
The berry extracts were analysed for total phenolics spectrophotometrically (Lambda Bio-UV/VIS spectrophotometer, Perkin Elmer, Germany) by the Folin–Ciocalteau procedure (Kähkönen et al. 1999). The total phenolic content was expressed as gallic acid equivalents (GAE) in milligrams per gram of dry material. The phenolic composition of the berry extracts was analysed by HPLC as modified from Lamuela-Raventos and Waterhouse (1994). The HPLC system consisted of three Waters 501-series pumps, a Waters Pump Control Module, a Waters WISP 700 autosampler equipped with a cooling device (Waters, Milford, MA), column oven (Merck T-6300) and a Waters PDA 996 diode array detector, all controlled by a Millennium computer programme. The column was a Nova-Pak C-18 (3·9 × 150 mm, particle size 4 μm) from Waters with a precolumn (Spherisorb S5 ODS 2·2 × 20 mm, particle size 5 μm) housed at 40°C. The gradient programme consisted of 50 mmol l–1 dihydrogen ammonium phosphate adjusted to pH 2·6 with orthophosphoric acid (solvent A), 20% solvent A with 80% acetonitrile (solvent B) and 0·2 mol l–1 orthophosphoric acid adjusted with ammonia to pH 1·5 (solvent C) (Table 2).
Table 2. The HPLC gradient programme for phenolic composition analysis
The phenolic compounds were divided into four classes, identified according to their spectral properties, and quantified as follows: flavan-3-ols as catechin equivalents at 280 nm, hydroxycinnamates as caffeic acid equivalents at 320 nm, flavonols as rutin at 365 nm, and anthocyanins as cyanidin-3-glucoside equivalents at 520 nm.
Tests for antimicrobial activity
Two methods were used to investigate antimicrobial effects of phenolic compounds: the agar diffusion assay with a soft agar overlay and measurement of bacterial growth in liquid culture. For agar diffusion assays, all reagents and extracts were dissolved in methanol. Lyophilized berry extracts were used as such in liquid culture experiments.
Agar diffusion assay.
Soft agar containing 4·5 mg ml–1 Bacto-agar (Difco) in liquid growth medium was cooled to 42°C after autoclaving, inoculated with liquid overnight culture to a cell density of 5 × 105 cfu ml–1, and plates containing 20 ml of agar media were overlaid with 5 ml of this inoculated soft agar. After boring wells (7 mm in diameter) in the agar, the plates were kept at room temperature for 30 min and then at 4°C for 2·5 h to allow liquid discharge into wells. The pure test compounds and the berry extracts in different concentrations (see Tables 4 and 5) were dissolved in methanol and pipetted into the agar wells (50 μl). Apigenin, luteolin and cyanidin-3-glucoside were tested at lower concentrations owing to the poor solubility of these substances in methanol. Polymyxin B sulphate (90 μg in 50 μl) was used as a positive control for Typhimurium, and streptomycin (150 and 600 μg in 50 μl) for all other strains. The negative control was methanol (50 μl). The diameter of the inhibition zone was measured at 24 h and 48 h. All determinations were made in triplicate.
Table 4. Antimicrobial activity of pure phenolic compounds by the agar diffusion method. (░) No inhibition, () no clear inhibition, but bacteriostatic effects
Table 4. Table 4 (Contd.)
Table 5. Antimicrobial activity of berry extracts by the agar diffusion method. (░) No inhibition, () no clear inhibition, but bacteriostatic effects, (□) not tested
Bacterial growth curve measurement.
In liquid culture experiments, 10 ml of fresh growth medium were inoculated with 1 or 5% of overnight culture. Lyophilized berry extracts or pure phenolic compounds as such were added to the culture media to give final concentrations of 0·5, 1 or 5 mg ml–1. The cultures were shaken well and incubated as described above. Bacterial growth was followed by taking samples from the cultures 5–6 times during an incubation period of 9–34 h, depending on the strain’s growth rate. The samples were diluted with peptone saline (Maximal Recovery Diluent, Lab M) and the proper dilutions were plated. The plates were incubated as above and the bacterial counts were recorded. The inhibitory (or stimulatory) effects of test compounds on the bacteria were measured by comparing the control growth curves with those obtained from cultures with berry extracts or pure phenolic compounds.
Phenolic composition of the berry extracts
The amounts of total phenolics and the composition of the phenolic compounds (expressed as mg g–1 phenolic extract) in berry extracts are shown in Table 3. The spectrophotometric total phenolic measurement resulted in higher amounts of phenolics than the HPLC analysis of the four phenolic subclasses. This could be partly due to dissimilar responses of different phenolic compounds in the Folin–Ciocalteu procedure, but it also indicates the presence of unidentified phenolic compounds in the extracts, for example, benzoic acids and tannins. Especially in cloudberry, raspberry and strawberry, only 10% or less of the total phenolics were identified as anthocyanins, flavonols, hydroxy-cinnamates or flavanols. In the blueberry extract, more than 80% of total phenolics belonged to the above-mentioned groups, anthocyanins being the major group.
Table 3. Phenolic composition of berry extracts
Antimicrobial activity by the agar diffusion method
The antimicrobial activity of 17 pure phenolic compounds representing 13 flavonoids and four phenolic acids, and, in addition, eight berry extracts, was first measured by agar diffusion, which is suitable for semi-quantitative estimation. Due to the scarcity of material, the antimicrobial properties of cranberry, cloudberry and sea buckthorn berry extracts were only tested against the most interesting bacterial strains, especially Gram-negative bacteria. The results are presented in Table 4 for pure compounds and in Table 5 for berry extracts.
Sensitivity to the compounds was found to differ significantly among the test organisms. Lactic acid bacteria (LAB), in general, were more resistant than the other bacteria to pure phenolic compounds. Escherichia coli CM 871 exhibited marked sensitivity to phenolic compounds.
Myricetin was the only compound which showed strong inhibitory effects on the growth of all LAB strains of human origin. Lactobacillus plantarum, a beer isolate, was resistant to myricetin at all test concentrations, as well as to all other pure phenolic compounds. Also, growth of E. coli strains was inhibited by myricetin. Myricetin seemed to retard the growth of both Ent. faecalis and B. lactis, although no clear-cut inhibition zones in the agar diffusion assay were detected. Typhimurium was not affected by myricetin.
The phenolic acids (cinnamic acid, 3-coumaric acid, caffeic acid, ferulic acid and chlorogenic acid) showed activity only against Gram-negative bacteria at high concentrations (500 μg well–1). Chlorogenic acid was the weakest inhibitor, exhibiting activity only against E. coli CM 871.
Luteolin was slightly inhibitory to Gram-positive LAB but not to Gram-negative bacteria. No clear-cut inhibition zones were detected, but the density of bacteria (cfu) in agar plates containing luteolin was significantly less than in control cultures, indicating weak bacteriostatic effects.
The anthocyanidins pelargonidin, delphinidin and cyanidin, as well as cyanidin-3-glucoside, only inhibited growth of E. coli CM 871 and had no effect on other bacterial strains.
All berry extracts possessed strong antimicrobial activity against Gram-negative bacteria. At higher concentrations, raspberry, cloudberry and blueberry extracts also inhibited the growth of some LAB strains. Raspberry exhibited the strongest antimicrobial activity, followed by cloudberry and blueberry. Sea buckthorn berry and blackcurrant showed the least antimicrobial activity. Among the seven LAB species tested, Lact. rhamnosus VTT E-97800 and GG VTT E-96666 were the most sensitive to the berry extracts, while Lact. johnsonii and Lact. crispatus were the least sensitive. Strawberry extract was active against Ent. faecalis. All berry extracts except sea buckthorn berry and blackcurrant showed antimicrobial activity against Typhimurium.
Bacterial growth curve measurement
For more accurate determination of antimicrobial activity, liquid culture experiments were performed. The antimicrobial activity of selected pure phenolic compounds and eight berry extracts on bacterial strains was studied by measuring the differences in bacterial growth curves in liquid cultures fortified with the test compounds. The results are presented in Table 6 and selected growth curves are shown in Fig. 2.
Table 6. Antimicrobial activity of selected pure phenolic compounds and berry extracts in liquid culture. () No inhibition: plate counts differ by <5 × 101; (░) clear inhibition: plate counts differ by 5 × 101–5 × 102; () strong inhibition: plate counts differ by 5 × 102–5 × 104; (▮) very strong inhibition: plate counts differ by >5 × 104; (□) not tested
In general, the results were in agreement with those obtained in the agar diffusion experiments. The growth of Lactobacillus strains was not inhibited by any of the berry extracts at low concentrations (1 mg ml–1). However, when five times higher concentrations of raspberry and blueberry extracts were used, growth of tested Lactobacillus strains was clearly inhibited by raspberry extract (Fig. 2a) and slightly inhibited by blueberry extract (data not shown). Bifidobacterium lactis was slightly inhibited by raspberry, strawberry and cloudberry extracts (1 mg ml–1), and Ent. faecalis by blueberry, raspberry and strawberry extracts (Table 6). Other extracts tested had no effects on these strains (Fig. 2b, c).
Escherichia coli strain 50 was sensitive to all phenolic extracts except blackcurrant (Fig. 2d, Table 6). In the presence of 0·5 mg ml–1 or 1·0 mg ml–1 raspberry extract, the number of viable cells remained approximately one logarithm lower than in the control culture (Fig. 2e). Growth of the DNA repair mutant strain E. coli CM 871 was strongly inhibited by all eight berry extracts; in the presence of 0·5 mg ml–1 raspberry extract, the number of viable cells decreased by two logarithms, and in the presence of 1·0 mg ml–1 raspberry extract, the number of viable E. coli CM 871 cells decreased rapidly beneath the detection limit (105 cfu ml–1) (Fig. 2f). Growth of Typhimurium was totally inhibited by cloudberry, raspberry and strawberry extracts (1 mg ml–1). Cloudberry extract exerted the strongest inhibitory effect, as indicated by the rapid inhibition of the Typhimurium culture in the presence of cloudberry extract (Fig. 2g).
In summary, in liquid culture, cloudberry showed the strongest antimicrobial activity against Gram-negative bacteria. Raspberry exhibited almost similar activity, whereas blackcurrant was least active (Table 6). The results are in accord with those obtained in the agar diffusion assay, although the strong inhibitory effects of cloudberry were particularly evident in liquid cultures. Blackcurrant slightly stimulated the growth of Lact. rhamnosus VTT E-97800 and GG VTT E-96666, and Lact. paracasei (data not shown).
The pure phenolic compound myricetin (0·5 mg ml–1) slightly inhibited the growth of Lact. rhamnosus VTT E-97800, Lact. rhamnosus GG VTT E-96666 (Fig. 2h), Ent. faecalis and B. lactis, whereas E. coli CM 871 was strongly inhibited (Fig. 2i) (Table 6). Lactobacillus rhamnosus VTT E-97800 was also weakly inhibited by rutin, and Lact. rhamnosus GG VTT E-96666 by quercetin (Table 6). The results are in agreement with agar diffusion data, although the inhibitory effects of myricetin were generally more prominent in agar diffusion than in liquid cultures.
Antimicrobial activity of 17 pure phenolic compounds representing flavonoids and phenolic acids, and eight berry extracts, was measured against selected Gram-positive and Gram-negative bacteria, including probiotic bacteria and an intestinal pathogen. The results showed that different bacterial species exhibit different sensitivities towards phenolics. In addition, different strains of the same bacterial species showed differences in sensitivity to one flavonoid. In general, berry extracts inhibited Gram-negative but not Gram-positive bacteria. These variations may reflect differences in cell surface structures between Gram-negative and Gram-positive bacteria. In particular, the outer membrane of Gram-negative bacteria functions as a preventive barrier against hydrophobic compounds (Helander et al. 1998).
Among the Gram-negative test bacteria, the DNA repair mutant strain E. coli CM 871 was particularly strongly affected by phenolic compounds. This strain lacks the repairing mechanism of DNA and is therefore expected to be more sensitive than strain E. coli 50 to damage caused by mutagenic agents. The latter strain represented in this study a normal E. coli strain from the human gastrointestinal flora. The antimicrobial effect of phenolics on E. coli CM 871 may be an indication of the potential mutagenic activity of these compounds. Anthocyanidins as a group of compounds, and quercetin and chlorogenic acid as single compounds, inhibited the growth of E. coli CM 871, with no inhibition of E. coli 50. Therefore, it can be hypothesized that the main reason for the antibacterial activity of these compounds is their reaction with DNA. On the other hand, phenolic acids showed activity against all Gram-negative test bacteria. In these cases, other mechanisms are also apparently involved in the antimicrobial action, genotoxicity being one (Stammati et al. 1999). Helander et al. (1998) have studied the action of plant-derived essential oil components on Gram-negative bacteria. They found that small phenolic compounds such as carvacrol and thymol inhibited E. coli and Salmonella. Inhibitory effects involved the disruptive action of these compounds on the outer membrane.
On the basis of the present results, the degree of hydroxylation might affect the antimicrobial activity of pure phenolic compounds. The flavonol myricetin, as a pure compound, clearly inhibited the growth of all LAB of human gastrointestinal tract origin, as well as the Gram-positive Ent. faecalis and B. lactis. The other flavonols tested, quercetin and kaempferol, are more lipophilic in nature (one and two hydroxyl groups less in the B ring than in myricetin, respectively), and they showed no inhibition against the above bacteria. The flavone luteolin was bacteriostatic against some of the tested LAB as well as against Ent. faecalis and B. lactis. No such effects were found with the flavone apigenin, which has one hydroxyl group less in the B ring. The results showed that the number of hydroxyl groups in the B ring in flavonols and flavones is associated with the antimicrobial activity against LAB. No other structure–activity relationship was found. Padmavati et al. (1997) studied the antimicrobial effects of flavonoids on major rice pathogens. The tested flavonoids differed in their hydroxylation patterns in the B and C rings. They showed, contrary to the present results, that the non-polar flavonoids were the most effective compounds, showing appreciable inhibition of spore germination of Pyricularia oryzea (the fungal blast pathogen). The surface of the eukaryotic spore, however, differs considerably from the cell wall of active bacterial cells, and may therefore affect the response towards phenolic compounds. In the present experiments, myricetin did not affect the growth of Typhimurium and E. coli 50. Typhimurium is a rough-type mutant with a truncated lipopolysaccharide component of the outer membrane (Stocker et al. 1979). However, the outer membrane of the mutant strain is not functionally impaired.
The berry extracts mainly inhibited the growth of Gram-negative bacteria but had no effect on Gram-positive bacteria. The antimicrobial activities of the naturally-occurring phenolics from olives, tea and wine have been widely studied (Ruiz-Barba et al. 1990; Vivas et al. 1997; Chou et al. 1999). However, there is very little information about the antimicrobial capacity of phenolics present in berries, except in cranberry. In our studies, cranberry and blueberry extracts rich in anthocyanins inhibited Gram-negative bacteria. The antibacterial properties of cranberry juice have been known for a long time (Clague and Fellers 1934; Moen 1962; Kinney and Blount 1979), and the effect may be associated with the inhibition of E. coli adherence to mucosal surfaces (Sobota 1984; Schmidt and Sobota 1988); Howell et al. (1998) recently suggested that proanthocyanidins (condensed tannins) were responsible for this. Similar inhibitory activities were reported for blueberry juice by Ofek et al. (1996). The present results agree with these observations.
Similar to other berry extracts, lingonberry extract showed activity only against Gram-negative bacteria. Annuk et al. (1999) have studied the antimicrobial activity of aqueous extracts of bearberry (Arctostaphylos uva-ursi (L.) Spreng., Ericaceae) and cowberry (Vaccinium vitis-idaea L., Ericaceae) against the Gram-negative pathogen Helicobacter pylori. Tannic acid seemed to be the responsible component. According to Holopainen et al. (1988), extracts of the aerial parts of bearberry and lingonberry were active against the Gram-negatives E. coli and Proteus vulgaris. The activity is known to be due to the phenolic glycosides arbutin and metylarbutin. Lingonberries are also rich in benzoic acid, a commonly used antimicrobial agent in foods.
In the liquid culture experiments, strawberry extract strongly inhibited the growth of Typhimurium and E. coli CM 871. Pratt et al. (1960) and Powers et al. (1960) studied the antimicrobial effects of anthocyanins and anthocyanidins from grapes and strawberries against several bacterial strains. Delphinidin-3-monoglycoside, pelargonidin-3-monoglycoside and malvidin-3,5-diglycoside were the most effective compounds. They concluded that the glucosides were hydrolysed and the aglycone was the active fraction. Also, anthocyanin extracts from the leaves and pericarp of the pigmented rice cultivar, Purpleputtu, showed inhibition of the Gram-negative species Xanthomonas, and the pigments have been characterized as cyanidin and peonidin glycosides (Reddy et al. 1995). Anthocyanidins in the agar diffusion experiments inhibited the growth of mutant E. coli strain CM 871 but had no effect on the other test bacteria. These results suggest that compounds other than anthocyanidins in the strawberry extract are mainly responsible for the inhibition of Typhimurium.
In the present study, cloudberry, raspberry and strawberry extracts were the strongest inhibitors of Gram-negative bacteria, especially Typhimurium. Recently, Rauha et al. (2000) studied the antimicrobial effects of several berry extracts. They also found that the widest bactericidal activity was expressed by berries belonging to the genus Rubus (cloudberry and raspberry). Häkkinen et al. (1999a) detected selected flavonoids and phenolic acids from 19 berries by HPLC. They found that ellagic acid was the main phenolic compound in the hydrolysed berry extracts of the genera Rubus and Fragaria (strawberry). Ellagic acid is a hydrolysis product from ellagitannins, which, together with gallotannins, form the predominant group of tannins in these berries; thus, it is not present in fresh berries or unhydrolysed berry extracts (Macheix et al. 1990). Ellagic acid has been reported to exhibit a dose-dependent inhibitory effect (IC50=1 mmol l–1) on Helicobacter pylori isolated from peptic ulcer patients (Chung 1998). Also, ellagitannin extracts inhibit a range of pathogenic organisms including Vibrio cholerae, Shigella dysenteriae and Campylobacter spp. (Scalbert 1991; Silva et al. 1997). It can be hypothesized that ellagitannins could be one of the components in cloudberries, raspberries and strawberries causing the inhibition against Salmonella.
Häkkinen et al. (1999b) recently determined the contents of the flavonols quercetin, myricetin and kaempferol in 25 edible berries. Cranberry and blueberry contained the highest concentration of myricetin, 108 and 71 mg kg–1, respectively. Interestingly, the present results showed that although myricetin itself was a strong inhibitor, berries rich in myricetin were not. In addition, compared with the high antimicrobial activity of berry extracts against Gram-negative strains, these bacteria, excluding the mutant strain E. coli CM 871, were not inhibited by pure phenolic compounds. These results suggest that the inhibitory effects of berry extracts may not be due to simple phenolics but to more complex phenolic polymers such as ellagitannins, tannins and proanthocyanidins. The antimicrobial activity of berry extracts is evidently a synergistic effect of various phenolic compounds, many of which are still unidentified. Also, other bioactive compounds in plant extracts, alone or in combination with phenols, might be responsible for the antimicrobial effects.
In conclusion, phenolic extracts of eight berries commonly consumed in Finland inhibited the growth of selected Gram-negative bacteria and were not active against Gram-positive LAB. Cloudberry, raspberry and strawberry extracts were strong inhibitors of the intestinal pathogen Salm. enterica. The antimicrobial effects of berry extracts against Gram-negative bacteria decreased in the following order: cloudberry > raspberry > strawberry > lingonberry > blueberry > cranberry > sea buckthorn berry> blackcurrant. In further investigations, fractionation of phenolic berry extracts will be carried out in order to identify the components responsible for antimicrobial activity. In addition, the possible role of berry seeds in antimicrobial action will be studied. Antimicrobial effects of berry extracts against a selection of Salmonella strains, and also against other intestinal pathogens, are under investigation, and the synergistic effects of plant phenols are being studied. The results of these studies will be used in functional food development and for food preservative purposes.
The authors thank Professor Veli Kauppinen for his ideas and valuable advice. Professor Atte von Wright, and Drs Maria Saarela and Ilkka Helander, are also thanked for useful discussions. The skilful technical assistance of Tuuli Teikari and Niina Torttila is gratefully acknowledged. This study was financially supported by Tekes, the National Technology Agency.