International Journal of Food Microbiology 164 (2013) 99–107

Contents lists available at SciVerse ScienceDirect

International Journal of Food Microbiology

journal homepage: www.elsevier.com/locate/ijfoodmicro

The antimicrobial effects of wood-associated on food pathogens and spoilage organisms

Carme Plumed-Ferrer a,1, Kati Väkeväinen a,1, Heli Komulainen a, Maarit Rautiainen a, Annika Smeds b, Jan-Erik Raitanen b, Patrik Eklund c, Stefan Willför b, Hanna-Leena Alakomi d, Maria Saarela d, Atte von Wright a,⁎ a University of Eastern Finland, Institute of Public Health and Clinical Nutrition, P.O. Box 1627, FI-70211 Kuopio, Finland b Åbo Akademi University, Process Chemistry Centre, Porthansgatan 3, FI-20500 Turku, Finland c Åbo Akademi University, Laboratory of Organic Chemistry, Biskopsgatan 8, FI-20500 Turku, Finland d VTT Technical Research Centre of Finland, P.O. Box 1000, FI-02044, VTT, Finland article info abstract

Article history: The antimicrobial effects of the wood-associated polyphenolic compounds pinosylvin, pinosylvin monomethyl Received 20 December 2012 ether, , , , , cycloXMe, dHIMP, ArX, and ArXOH were assessed Received in revised form 25 March 2013 against both Gram-negative (Salmonella) and Gram-positive bacteria (Listeria monocytogenes, Staphylococcus Accepted 1 April 2013 epidermidis, Staphylococcus aureus) and yeasts (Candida tropicalis, Saccharomyces cerevisiae). Particularly the Available online 9 April 2013 stilbenes pinosylvin, its monomethyl ether and piceatannol demonstrated a clear antimicrobial activity, which in the case of pinosylvin was present also in food matrices like sauerkraut, gravlax and berry jam, but not in Keywords: wood associated phenolic compounds milk. The destabilization of the outer membrane of Gram-negative microorganisms, as well as interactions stilbenes with the cell membrane, as indicated by the NPN uptake and LIVE/DEAD viability staining experiments, can be pinosylvin one of the specific mechanisms behind the antibacterial action. L. monocytogenes was particularly sensitive Salmonella to pinosylvin, and this effect was also seen in L. monocytogenes internalized in intestinal Caco2 cells at Listeria monocytogenes non-cytotoxic pinosylvin concentrations. In general, the antimicrobial effects of pinosylvin were even more Food spoilage prominent than those of a related stilbene, , well known for its various bioactivities. According to our results, pinosylvin could have potential as a natural disinfectant or biocide in some targeted applications. © 2013 Elsevier B.V. All rights reserved.

1. Introduction and therapeutic agents (Mattson et al., 2011; Puupponen-Pimia et al., 2005; Rauha et al., 2000). One of the most studied individual substances The antimicrobial properties of several plant-associated substances, is (E)-resveratrol (3,4′,5-trihydroxystilbene) (Fig. 1), a compound such as flavonoids, tannins, and essential oils have received a lot of present in red grapes and berries (Adrian et al., 2000; Lyons et al., attention during recent years as potential preservatives, disinfectants, 2003). It has both antifungal and antibacterial properties (Man-Ying Chan, 2002), and, furthermore, its practical use as an antimicrobial

Abbreviations: ArX, 4,4′-dihydroxy-3,3′-dimethoxy-9-nor-6′,7-cycloligna-7,7′- compound has been suggested (Montero et al., 2003). In bioassays diene-9′-oic acid; ArXOH, 4,4′-dihydroxy-3,3′-dimethoxy-9-nor-6′,7-cycloligna-7,7′- with cancer cell lines and in animal studies, resveratrol appears to diene-9′-ol; As, astringin; bw, bodyweight; cycloX/cyX, (8′S,7R)-4,4′-dihydroxy-3, interfere with the major steps of the carcinogenesis (Boissy et al., 3′-dimethoxy-6′,7-cyclo-9-norlign-9′-oic acid and (8′S,7S)-4,4′-dihydroxy-3,3′-dimethoxy- 2005; Boocock et al., 2007; Jang et al., 1997; Tolomeo et al., 2005). 6′,7-cyclo-9-norlign-9′-oic acid; cycloXMe/CyXMe, (8′S,7R)-4,4′-dihydroxy-3, Other recorded activities include positive effects in processes associated 3′-dimethoxy-6′,7-cyclo-9-norlign-9′-oic acid and (8′S,7S)-4,4′-dihydroxy-3, 3′-dimethoxy-6′,7-cyclo-9-norlign-9′-oic acid; cycloXOH/cyXOH, (8′S,7R)-4,4′- with cardiovascular diseases (Paceasciak et al., 1995), hormone-like dihydroxy-3,3′-dimethoxy-6′,7-cyclo-9-norlign-9′-ol and (8′S,7S)-4,4′-dihydroxy-3, activities (Klinge et al., 2005), and the prevention of type II diabetes 3′-dimethoxy-6′,7-cyclo-9-norlign-9′-ol; dHIMP, (-)-(8′S)-4,4′-dihydroxy-3, by SIRT1 activation (Lagouge et al., 2006). ′ ′ ′ ′ ′ 3 -dimethoxy-9-norlign-9 -ol; dHX, (-)-(8 S)-4,4 -dihydroxy-3,3 -dimethoxy-9- We have previously reported antimicrobial and cytotoxic activities of norlign-9′-oic acid; HEPES, n-Heptadecanoic acid methyl ester; IMP, (-)-(7E)- (8′R)-4,4′-dihydroxy-3,3′-dimethoxy-9-norlign-7-en-9′-ol; IRh, isorhapontin; (E)-pinosylvin (3,5-dihydroxystilbene, PS) (Fig. 1), a naturally occurring NOAEL, No Observed Adverse Effect level; NPN, 1-N-phenylnaphthylamine; stilbene related to resveratrol, present particularly in the heartwood PS, pinosylvin; PSMME, pinosylvin monomethyl ether; X, (-)-(7E)-(8′R)- andknotwoodoftreesofthePinus family (Välimaa et al., 2007). PS 4,4′-dihydroxy-3,3′-dimethoxy-9-norlign-7-en-9′-oic acid; XMe, (-)-Methyl was active against both Gram-negative bacteria (Salmonella, Escherichia (7E)-(8′R)-4,4′-dihydroxy-3,3′-dimethoxy-9-norlign-7-en-9′-oat. coli, Pseudomonas fluorescens) and Gram-positives like Bacillus cereus, ⁎ Corresponding author. Tel.: +358 505376030. E-mail address: atte.vonwright@uef.fi (A. von Wright). Listeria monocytogenes and Staphylococcus aureus, but not against 1 These two authors have equally contributed to the publication. Lactobacillus plantarum. In later studies (unpublished results) with

0168-1605/$ – see front matter © 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.ijfoodmicro.2013.04.001 100 C. Plumed-Ferrer et al. / International Journal of Food Microbiology 164 (2013) 99–107

R' OH

R R OR'

OH OH

R= OH, R'= H: Pinosylvin R= O-Glu, R'= O-Glu: Astringin R= OMe, R'= H: Pinosylvin monomethylether R= OH, R'= H: Piceatannol R=OH,R'=OH:Resveratrol R= O-Glu, R'= Me: Isorhapontin

O MeO MeO MeO R OH OMe

HO HO HO

OMe Me OMe OH OH OH

dHIMP CyXMe R= COOH: ArX R= CH2OH: ArXOH

Fig. 1. Molecular structures of the studied compounds.

different genera of lactic acid bacteria, both homo- and hetero- 2.2.1. PS, PSMME, As, and IRh fermentative lactobacilli and leuconostocs were markedly resistant PS and PSMME were isolated from Scots pine (Pinus sylvestris L.) to pinosylvin, whereas both lactococci and enterococci were sensitive. heartwood, and As and IRh from Norway spruce (Picea abies) bark. The antifungal activity could be demonstrated both against yeasts The wood material was splintered, freeze-dried and ground, and (Candida albicans, Saccharomyces cerevisiae)andfilamentous fungi approximately 500 g was extracted in a Soxhlet apparatus with (Aspergillus fumigatus, Penicillium brevicompactum). hexane overnight to remove lipophilic extractives, and then with In this study, we have further compared the antimicrobial activity (for PS and PSMME) or ethanol (EtOH) (for As and IRh) over- of pinosylvin with some other naturally occurring stilbenes and with night to yield hydrophilic extractives. The acetone and EtOH extracts some semisynthetically prepared norlignans of similar structure were evaporated to dryness using a rotary evaporator, and approxi- and also assessed the effect in certain food matrices (sauerkraut, mately 40 g of the freeze-dried extracts was mixed with silica gel 60 cold smoked fish, and milk). Since L. monocytogenes was found to be (Fluka), and placed in a sample injection module of a Flash 75i appara- particularly susceptible to pinosylvin, the effect of pinosylvin against tus (Biotage UK Ltd., Hertford, England). PS and PSMME were separated the internalized L. monocytogenes was also studied in a Caco2 cell on a silica cartridge, whereas As and IRh were separated on an RP-18 model. cartridge (75 mm × 30 cm). PS and PSMME, were eluted using a step gradient of dichloromethane:EtOH (98.8:1.2 v/v–96:4 v/v); As and IRh using a step gradient of EtOH:distilled water (10:90 v/v–100:00 v/v). 2. Materials and methods Fractions were collected in 500 ml bottles, and fractions containing PS were evaporated to dryness using a vacuum evaporator to yield an oil. 2.1. Chemicals Crystallization from gave PS as a white solid. The purity was 98% (GC) with an (E)-:(Z)-isomer ratio of 100:0. n-Heptadecanoic acid methyl ester (HEPES) and 1-N- Fractions containing both PS and PSMME were evaporated to phenylnaphthylamine (NPN) were purchased from Sigma-Aldrich dryness using a vacuum evaporator. Distilled water was added to (Steinheim, Germany), and EDTA from Riedel-de-Haen (Seelze, the resulting oil, pH was raised to 13 using KOH (10 M), and the Germany). mixture was extracted three times with methyl tert-butyl ether. The organic phases were combined, acidified to pH 5 using HCl (0.5 M)

2.2. The compounds studied and dried under N2-stream. Acetone was then added, and after evaporation of the solvent, an oil was obtained. Crystallization from Resveratrol was obtained from Sigma-Aldrich (Steinheim, Germany). concentrated acetic acid (AcOH) yielded PSMME as a white solid. The The other compounds, i.e., the stilbene aglycones pinosylvin (PS), PS purity was 99% (GC) with an (E)-:(Z)-isomer ratio of 100:0. Fractions monomethyl ether (PSMME), and piceatannol, the stilbene glucosides containing As and IRh were combined into two separate fractions. astringin (As) and isorhapontin (IRh), and the norlignans cycloXMe, These were further purified using a smaller flash chromatograph dHIMP, ArX and ArXOH (see Fig. 1) were prepared in our laboratories. (Biotage UK Ltd., Flash40i). Approximately 2 g of the extract was The purity and identity of the compounds were determined using GC– separated on a Biotage RP-4 flash cartridge (40 mm × 15 cm) using a FID and GC–MS after silylation (Willför et al., 2003). For the bioactivity step gradient of EtOH:water (5:95 v/v–80:20 v/v). Fractions were tests the compounds were dissolved in 98% (v/v) ethanol. collected in 50 ml test tubes, and the fractions containing As and IRh C. Plumed-Ferrer et al. / International Journal of Food Microbiology 164 (2013) 99–107 101

Table 1 The microbial strains used in the experiments.

Strain Routine growth medium Cultivation Origin temperature (°C)

Listeria monocytogenes EELA Hki L211 Tryptone soy agar or broth (Lab M Ltd., UK) 37 The Finnish Food Safety Authority (EVIRA), Helsinki, Finland L. monocytogenes EELA Oulu ba 2392/4 Tryptone soy agar or broth (Lab M Ltd., UK) 37 The Finnish Food Safety Authority (EVIRA), Oulu, Finland L. monocytogenes EELA Hki L627 b Tryptone soy agar or broth (Lab M Ltd., UK) 37 The Finnish Food Safety Authority (EVIRA), Helsinki, Finland L. monocytogenes KTL IHD 43573 Tryptone soy agar or broth (Lab M Ltd., UK) 37 National Institute for Health and Wellfare, THL, Helsinki, Finland L. monocytogenes 01 REB 3 Tryptone soy agar or broth (Lab M Ltd., UK) 37 The University of Eastern Finland (UEF), Kuopio, Finland Salmonella enterica ssp. enterica Blood agar, BHI-broth (Lab M Ltd., UK) 37 The Finnish Food Safety Authority (EVIRA), serovar Infantis EELA 72 Helsinki, Finland S. enterica ssp. enterica Typhimurium VTT E-95582T Iso-sensitest broth (Oxoid, UK) 37 Technical Research Centre of Finland (VTT), Espoo, Finland S. enterica ssp. enterica Typhimurium E-981151 Iso-sensitest broth (Oxoid, UK) 37 Technical Research Centre of Finland (VTT), Espoo, Finland Staphylococcus epidermidis VTT E-97768 Iso-sensitest broth (Oxoid, UK) 37 Technical Research Centre of Finland (VTT), Espoo, Finland S. aureus VTT E-94531T Iso-sensitest broth (Oxoid, UK) 37 Technical Research Centre of Finland (VTT), Espoo, Finland S. aureus S770 Blood agar, BHI-broth (Lab M Ltd., UK) 37 The Finnish Food Safety Authority (EVIRA), Finland Candida tropicalis 4068 OGYE agar or broth (Lab M Ltd., UK) 30 Valio Finnish Cooperative Dairies' Association, Espoo, Finland Saccharomyces cerevisiae 4054 OGYE agar or broth (Lab M Ltd., UK) 30 Valio Finnish Cooperative Dairies' Association, Espoo, Finland

were again combined into two separate fractions. The purity of As was 2.3. The microbial strains and culture conditions 93%(GC)withan(E)-:(Z)-isomer ratio of 99.7:0.3. The purity of IRh was 83% (GC). IRh was further purified using a previously described pre- The microbial strains, their origins and the general culture conditions parative HPLC system (Smeds et al., 2012). The mixture was separated are indicated in Table 1. on a Phenomenex Luna RP-18 column (250 × 21.2 mm, particle size 10 μm) at a flow rate of 20 ml/min using gradient elution with 0.1% 2.4. Comparison of the antimicrobial activities AcOH in methanol (B) and 0.1% AcOH in Milli-Q water (A). The gradient was from 10% B to 30% B in 9 min, then to 35% B in 14 min, to 50% B in The comparison of antimicrobial activities was done using 8 min, and to 95% B in 7 min, which was held for 3 min. The UV/VIS de- L. monocytogenes L211, Salmonella infantis EELA 72 and Candida tropicalis tector was set at 254 nm and 340 nm. The fractions containing pure IRh 4068. The assay was performed using an automated incubator and a were collected, the solvent was evaporated, and the product was dried turbidity reader (Bioscreen C, ThermoLabsystems, Helsinki, Finland), in a vacuum oven. The purity was 95% (GC) with an (E)-:(Z)-isomer which enables the simultaneous testing of 200 samples and monitoring ratio of 98.6:1.4. of bacterial growth in real time during the test. The preparation of the test cultures and the practical performance of the assays have been 2.2.2. Piceatannol described by Välimaa et al. (2007). The compounds were studied Pure As was enzymatically hydrolyzed in 0.2 M acetate buffer at equimolar concentrations. The growth inhibition was calculated by using β-D-glucosidase isolated from almonds (Fluka), at 37 °C over- subtracting the percentage turbidity of the test culture from that of night, according to a method described previously (Matsuda et al., the control (taken as 100%), when the latter had reached the stationary 2001). The enzyme was then filtered off, and the reaction mixture phase. was extracted three times with ethyl acetate. The ethyl acetate phase was washed with a 30% NaCl solution, and dried with Na2SO4, 2.5. The mechanisms of the antibacterial action of PS against Gram- which was then filtered off. The extract was evaporated to dryness negative bacteria using a vacuum evaporator and dried in a vacuum oven. The purity of piceatannol was 90% (GC) with an (E)-:(Z)-isomer ratio of 98.3:1.7. 2.5.1. Permeability assay A hydrophobic probe, 1-N-phenylnaphthylamine (NPN), was used to 2.2.3. Norlignans measure changes in the permeability of Gram-negative cells. NPN uptake The 9-norlignans were semisynthetically prepared from the by bacterial suspensions was measured using black fluorotiter plates readily available lignan 7-hydroxymatairesinol (HMR). First, HMR (ThermoLabSystems, Helsinki, Finland) and the automated fluorometer was transformed to the carboxylic acid derivative of imperanene, Fluoroskan Ascent FL (LabSystems) as described earlier (Alakomi et al., as previously described (Eklund et al., 2002). This compound was 2000, 2003). Briefly, cells grown into A630 = 0.5 ± 0.02 were deposited further transformed to cycloXMe by cyclization and conversion to by centrifugation at room temperature for 10 min at 1000 g and the methyl ester. ArX was prepared from cycloXMe by aromatization suspended into a half volume of 5 mM HEPES buffer (pH 7.2). Aliquots of the aliphatic ring with DDQ, followed by hydrolysis. Reduction of (100 μl) of this cell suspension were pipetted into fluoroplate wells, the aromatized methyl ester (ArXMe) by LiAlH4 gave ArXOH. dHIMP which contained NPN (10 μM), and as test substances either EDTA (1.0 was obtained by the hydrogenation of the carboxylic acid derivative and 0.1 mM), PS or its derivatives, resveratrol (0.5 mM), or HEPES buffer of imperanene, followed by esterification and reduction. The purity (control) to make up a total volume of 200 μl. If desired, MgCl2 was (>95%) and the structure of all compounds were analyzed by GC– added to the cell suspension before the addition of NPN. Fluorescence MS and NMR-spectroscopy (the details of the synthesis will be pub- was monitored within 3 min from four parallel wells per sample. Each lished elsewhere). assay was performed at least three times. Results from the permeability 102 C. Plumed-Ferrer et al. / International Journal of Food Microbiology 164 (2013) 99–107 assays were analyzed statistically using two-tailed unpaired Student's GK 113D apparatus making as many packages as there were anticipated t-test to determine differences; levels of significance are denoted sampling times. The packages were incubated in dark at three different as*p b 0.02, **p b 0.01, and ***p b 0.001. temperatures (3, 8, and 20 °C). Packages were removed and examined for L. monocytogenes (triplicate samples per package) either on every 2.5.2. Growth inhibition tests and antibiotic susceptibility third (the 20 °C samples) or seventh day (the 3 and 8 °C samples). Antimicrobial activity of the compounds and susceptibility of The enrichment of L. monocytogenes was done according to the ISO Salmonella typhimurium VTT E-95582T and S. typhimurium E-981151 11290 standard. At the later stages, when the CFU counts started cells to hydrophobic antibiotic, novobiocin, alone and in combination to rise, the enrichment step was not necessary, and direct plating with PS, PSMME, and resveratrol were examined in liquid cultures as from a dilution series made in 0.1% peptone water could be done. described by Alakomi et al. (2006) with an automated turbidimeter Both PALCAM (Polymyxin–Acriflavine–Lithium chloride–Ceftazidime, Bioscreen C. The test procedures and the interpretation of the results Esculine–Mannitol, Lab M Ltd., UK) and Harlequin™ Aloa (Lab M Ltd., were done essentially as in Section 2.4 above. UK) agars were used.

2.5.3. Membrane damage 2.6.3. Berry jam The effect of PS, PMME and resveratrol against Staphylococcus Strawberry jam (Brix value 30) was obtained from the Suonenjoki epidermidis VTT E-97768, S. aureus E-94531T and Salmonella enterica Jam Factory (Valio Ltd.) before any preservatives had been added. The sv. Typhimurium E-981151 cells were examined with LIVE/DEAD® jam was inoculated with S. cerevisiae 4054, which had proven to be BacLight™ Bacterial Viability Kit (Molecular Probes). Briefly, overnight more resistant to PS than the other yeasts screened (see Results). in Iso Sensitest Broth (ISB) grown cells were harvested by centrifuga- The inoculation levels were 100 and 1000 CFU per ml. The PS levels tion, washed with buffered peptone saline (PBS pH 7.2), resuspended applied were 150, 300, and 600 μg/g. Potassium sorbate (0.2% w/w) and diluted 1:10 in PBS. The tested compounds were added at a was used as a control antimicrobial. Jam samples were cultured on 2.5-mM concentration, and viability and membrane damage of the OGYE agar every 7 days. In conjunction with CFU determination, cells were observed after 2 h of incubation at 37 °C with Live/Dead enrichment cultures were also prepared by withdrawing an aliquot of staining. To reveal changes in membrane functions caused by PS and 1mlfromthefirst 10-fold dilution and mixing with 10 ml of OGYE resveratrol, cells were stained with bis-(1, 3-dibutylbarbituric acid) broth and incubating for 120 h at 30 °C. Streak plates were subsequently (3) trimethine oxonol dye, DiBAC4 (Molecular Probes, Inc.) (Saarela prepared from the enrichment cultures to verify whether the sample et al., 2009). Increased depolarization results in more influx of the contained cultivable yeast cells. anionic dye and thus an increase in fluorescence, whereas hyper- polarization is indicated by a decrease in fluorescence. Briefly, 2.6.4. Milk treated cells were pipetted in microtiter plates, and after addition Aliquots of 9.8 ml of UHT-treated low-fat milk (fat percentage 1.5) of the dye, the microplate was incubated at 37 °C in Fluoroskan were inoculated either with 100 CFU/ml of L. monocytogenes EELA 63 Ascent FL (LabSystems), and changes in fluorescence were mea- or S. aureus S770. The PS concentrations were 75 and 200 μg/ml. The sured kinetically for 16 min (measurement interval 2 min). milk samples (final volume 10.0 ml) were incubated at 30 °C in dark, and samples were withdrawn at 10, 22, 34, and 46 h after the addi- 2.6. The antibacterial effects of PS in food matrices tion of bacteria and PS. L. monocytogenes was determined as described above, and S. aureus on Baird–Parker agar (Lab M Ltd., UK). The antimicrobial potency of pinosylvin was studied in four different food matrices: sauerkraut, berry jam, gravlax and milk. In the case of 2.7. The effects of PS on L. monocytogenes internalized in Caco2 cells sauerkraut, the effects on the microbial populations during the natural fermentation were followed, while the other foods were intentionally For each experiment, fresh brain heart infusion broth (BHI, Lab M) inoculated with contaminants or pathogens typical for the product. cultures of L. monocytogenes were prepared. After 18 h incubation at 37 °C, the bacterial cells were collected by centrifugation, washed 2.6.1. Sauerkraut twice with 3 ml PBS and resuspended in non-supplemented cell Sauerkraut was prepared in a traditional way by slicing cabbage, culture medium (DMEM) to an appropriate dilution (2 × 107 CFU/ml). mixing the slices with shredded carrots and salt (the proportions of The human colon carcinoma cell-line Caco2 (ATCC HTB-37) these ingredients being 10:1:0.2) and allowing the spontaneous was used between passages 30 and 55. In brief, cells were seeded fermentation to proceed in a plastic container at anaerobic conditions to 24-well tissue culture plates at a concentration of 2.5 × 105 cells (first 3–4 days at ambient temperature, then at 5 °C for 2 weeks). PS per well. The culture medium (DMEM supplemented with 10% (v/v)) was added in an ethanol solution, giving final concentrations of 25, heat inactivated fetal bovine serum, 2 mM L-glutamine, 1% (v/v) 50, 100 and 200 μg/g sauerkraut. The concentration of ethanol was non-essential amino acids and 100 IU penicillin/ml and 100 μg kept constant (1 ml per dl sauerkraut), and an ethanol control was streptomycin/ml was replaced every two days until the cells included. During the first week, samples were cultured daily, and had reached the differentiated stage (21 days). All the reagents after that once in a week. Lactic acid bacteria were assessed using for cell culture were purchased from EuroClone (Siziano, Italy). both MRS agar (de Man, Rogosa and Sharpe Agar, Lab M Ltd., UK) and Cells were incubated with bacteria for 2 h at 37 °C. After incubation, M17 agar (Lab M Ltd., UK). Enterobacteriaceae were analyzed using monolayers were washed three times with 1 ml of PBS warmed to the VRBG (Violet Red Bile Glucose Agar, Lab M Ltd., UK). 37 °C. To eliminate non-internalized bacteria, the cells were covered for 1.5 h with 500 μl DMEM supplemented with 100 μg/ml gentamicin 2.6.2. Gravlax (Sigma-Aldrich, St Louis, MO). After incubation, the cell monolayers Commercial fresh gravlax (Apetit Kala, Kuopio, Finland) was were washed once and the internalized bacteria were enumerated prepared from rainbow trout (Oncorhynchus mykiss) by surface treat- from a portion of monolayers by plating on TSA-agar after lysing the ment of the fish filets with a salt–spice mixture. The fish was inoculated cells with 0.5 ml Triton X-100 (0.1% in ice-cold PBS) for 10 min. To with L. monocytogenes 01 REB 3 by streaking the surface of a 50 g piece investigate the ability of the internalized bacteria to multiply inside of filet with bacterial suspension (in 0.9% saline) giving a CFU number of cells, the balance of monolayers was covered with 1 ml DMEM ca. 1000/g fish. PS was subsequently applied onto the surface in an supplemented with 5 μg/ml gentamicin. After 1.5 h, the medium ethanol solution (1 ml per fish sample) to give a final concentration of was replaced using DMEM containing both gentamicin and variable 140 μg/g. The filet pieces were then vacuum-packed using a Supervac® amounts of PS and incubation was extended up to 21.5 h, after which C. Plumed-Ferrer et al. / International Journal of Food Microbiology 164 (2013) 99–107 103 the enumeration of bacteria was performed using cell lysis and plating Table 3 as described above. Multiplication rates were calculated as a ratio Antimicrobial effects of pinosylvin and related compounds alone and in combination with novobiocin (50 μg/ml) against S. enterica serovar enterica Typhimurium of bacterial counts observed after multiplication and after invasion. expressed as growth inhibition percentages (mean ± SD of two experiments)1. Experiments were carried out in triplicate for each strain tested, including positive and negative controls. One uninfected well was included as a Test concentration cross-contamination control in every 24-well plate used. 0.5 mM 0.5 mM + novobiocin 50 μg2

S. Typhimurium VTT E-95582T 3. Results PS 69.5 ± 2.5 94.8 ± 1.7 Resveratrol 58.2 ± 3.2 56.1 ± 4.5 PSMME 72.8 ± 6.5 89.5 ± 2.5 3.1. Antimicrobial activity of the investigated compounds S. Typhimurium E-981151 Antimicrobial effects of PS, resveratrol, PSMME, piceatannol, IRh, PS 57.6 ± 3.1 89.6 ± 2.5 Resveratrol 45.7 ± 2.7 52.4 ± 4.5 cycloXMe, dHIMP, ArX, and ArXOH are given in Table 2, where PS has PSMME 62.4 ± 2.5 91.6 ± 1.5 the widest antimicrobial activity range. Also resveratrol, piceatannol 1 Growth inhibition was calculated by subtracting the percentage turbidity of and PSMME displayed a considerable activity, while the antimicrobial each treatment from the control (assumed as 100%), when the latter had reached the activities of the other compounds were marginal. Results obtained stationary phase. with PSMME were somewhat variable regarding effects against 2 Novobiocin alone did not cause growth inhibition, (growth inhibition percentages Salmonella. In experiments with S. Infantis, it had a relatively low in the two experiments 2.0 ± 1.5 and 0.4 ± 2.1, respectively). antibacterial activity (Table 2), whereas with S. Typhimurium it turned out to be active (Table 3). PSMME was also exceptionally antifungal 3.2.2. Growth inhibition assay against C. tropicalis. PS and PSMME efficiently inhibited the growth of S. typhimurium VTT E-95582T and E-981151strains (inhibition 70%) whereas resveratrol had 3.2. Effects of the investigated compounds on membrane permeability a lower antimicrobial activity (approx. 50%, Table 3). PS and PSMME and integrity destabilized the OM of examined Salmonella strains and increased susceptibility of the cells to the hydrophobic antibiotic novobiocin 3.2.1. Effect of the tested compounds on the uptake of NPN (growth inhibition 95 and 92%, respectively). To reveal changes in OM permeability, three S. enterica strains were selected for NPN uptake studies. Detailed results of the NPN 3.2.3. Membrane damage uptake experiments with the treatments, including the effect of the LIVE/DEAD viability staining of treated cells revealed that PS addition of MgCl2 in the assay buffer, are shown in Fig. 2a–c. PS, efficiently caused membrane damage in Staphylococcus epidermis PSMME, and resveratrol caused significant (p b 0.001) NPN uptake VTT E-97768, S. aureus E-94531T and S. typhimurium E-981151 cells in S. typhimurium E-981151, S. typhimurium E-95582T and S. infantis (Fig. 3a–c). PS inactivated the cells more efficiently than resveratrol E-97738 cells. In S. typhimurium E-981151 cells the NPN uptake (Fig. 3b–c). With resveratrol, only part of the treated cells died or were caused by these compounds was higher, compared to NPN uptake damaged (stained red or orange), whereas a certain subpopulation of caused by the classical chelator EDTA at 1 mM/l concentration. cells was able to retain their membranes undamaged (green cells). For other examined strains the NPN destabilizing activity was at the PS and PSMME depolarized the cytoplasmic membrane of the tested same level as with EDTA. Staphylococcus strains, whereas other tested PS derivatives and resvera- Divalent cations are known to inhibit the action of many outer trol had a minor effect on their membrane potential (data not shown). membrane permeabilizers, which act by chelation or replacement of divalent cations in the OM. We, therefore, tested whether Mg2+ 3.3. The antimicrobial effect of PS in food matrices could affect the activity of the tested agents. Fig. 2a–c shows that for the smooth strains, 1 mM MgCl2 addition almost totally abolished 3.3.1. Sauerkraut fermentation the activity of EDTA. In the case of pinosylvin and resveratrol treat- At the concentration range of 25–200 μg/g fermenting sauerkraut, ments 1 mM MgCl2 addition decreased the NPN uptake approx. 50% PS consistently caused a reduction of 2–4 logs of Enterobacteriaceae, indicating that their membrane destabilizing activity is not merely which typically dominate the early stages of fermentation. The effect related to chelation of divalent cations. seemed even more pronounced at lower PS concentrations (Fig. 4a).

Table 2 Antimicrobial effects of stilbenes and norlignans against S. infantis, L. monocytogenes and C. tropicalis expressed as growth inhibition percentages (mean ± SD of three experiments)1.

Concentration PS Resveratrol PSMME Piceatannol As IRh cycloXMe arX arXOH dHIMP mM

S. Infantis 0.25 12. 5 ± 6.2 12.0 ± 0.6 22. 3 ± 1.9 12. 4 ± 0.5 6. 0 ± 1.3 5. 1 ± 1.1 3.9 ± 0.2 7.4 ± 1.2 7.7 ± 3.4 4.4 ± 0.9 0.5 95.9 ± 0.7 20.1 ± 1.1 29.0 ± 1. 7 20.7 ± 3. 2 10. 7 ± 1. 3 5.9 ± 0.7 4.6 ± 0.5 7.5 ± 1.2 8.9 ± 2.2 4.8 ± 0.8 1.0 ND2 ND 38.2 ± 0.5 92.1 ± 0.6 13.5 ± 2.6 6.8 ± 0.6 10.7 ± 0.8 8.0 ± 1.3 13.7 ± 3.3 4.6 ± 0.6

L. monocytogenes 0.125 76.4 ± 9.9 42.7 ± 10.3 93.4 ± 0.0 38.3 ± 4.1 22.1 ± 1.5 15.8 ± 0.7 17.4 ± 0.2 16.8 ± 1.0 22.9 ± 3.0 9.4 ± 2.2 0.25 94.6 ± 0.6 84.2 ± 3.4 94.9 ± 0.2 84.8 ± 0.9 27.6 ± 2.6 17.0 ± 3.2 32.2 ± 1.1 28.6 ± 4.5 39.6 ± 1.2 15.9 ± 0.9 0.5 93.2 ± 0.4 92.7 ± 1.5 ND 89.1 ± 0.8 37.9 ± 6.3 30.5 ± 1.0 55.0 ± 4.1 67.8 ± 6.0 76.7 ± 2.3 28.3 ± 2.6

C. tropicalis 0.0625 2.6 ± 9.1 47.4 ± 5.0 91.7 ± 1.5 22.1 ± 3.2 18.4 ± 6.2 23.3 ± 19.1 −4.6 ± 16.0 29.4 ± 2.5 −9.3 ± 6.4 26.8 ± 7.9 0.125 90.7 ± 2.8 48.9 ± 2.2 91.1 ± 1.4 23.0 ± 8.5 24.9 ± 13.1 19.5 ± 19.0 5.6 ± 18.2 37.7 ± 4.1 1.8 ± 4.8 32.8 ± 8.0 0.25 ND ND 93.8 ± 1.4 30.6 ± 3.3 31.8 ± 14.1 22.3 ± 16.8 11.8 ± 9.9 48.9 ± 2.8 20.9 ± 2.2 37.9 ± 6.9

1 Growth inhibition was calculated by subtracting the percentage turbidity of each treatment from the control (assumed as 100%), when the latter had reached the stationary phase. 2 ND = Not determined. 104 C. Plumed-Ferrer et al. / International Journal of Food Microbiology 164 (2013) 99–107 a 300 *** *** Treatment 250 + 1 mM 200 MgCl2 ** *** 150

100 NPN uptake

50

0

b

c

Fig. 2. a. NPN uptake of S. typhimurium E-981151. Error bars present standard deviations, ***p b 0.001 compared with the controls. b. NPN uptake of S. typhimurium VTT E-95582T. Error bars present standard deviations, ***p b 0.001 compared with the controls. c. NPN uptake of S. infantis VTT E-97738. Error bars present standard deviations, ***p b 0.001 compared with the controls.

The counts of lactic acid bacteria and the development of pH were not the Listeria-counts were reduced, but the effects on growth could not be affected, and the appearance and flavor of the product were normal observed, since even the growth of controls was almost nonexistent (data not shown). during the observation period. At 20 °C, the antibacterial effect could be observed during the first 72 h, but after that, the growth of 3.3.2. Gravlax L. monocytogenes became logarithmic and profuse. The antimicrobial activity of PS (140 μg/g) against inoculated L. monocytogenes was most apparent at the storage temperature 3.3.3. Antifungal activity in berry jam of 8 °C, at which the multiplication of the bacterium was completely At the concentration of 150 μg/g, PS had no effect on the growth of inhibited in comparison to the control (Fig. 4b). At lower temperatures, inoculated S. cerevisiae. However, at the concentration of 300 μg/g, C. Plumed-Ferrer et al. / International Journal of Food Microbiology 164 (2013) 99–107 105 a

b

c

Fig. 3. Effect of PS and resveratrol on S. epidermis E-97768 (a), S. aureus E-94531T (b) and S. enterica sv. Typhimurium E-981151 (c) cells observed with LIVE/DEAD® BacLight™ Bacterial Viability Kit (Molecular Probes). Order of the samples from left to right: control cells, PS and resveratrol.

the yeast was completely eliminated by PS already on the day of astringin and isorhapontin, which are the monoglucosides of piceatannol inoculation and stayed below the limit of detection even during extended and isorhapontigenin, respectively, showed no clear antimicrobial storage of seven days (data not shown). effects. They are more hydrophilic than their mother compounds, why this may be a reason for their inactivity. Furthermore, resveratrol is 3.3.4. Milk more hydrophilic (has one hydroxyl more) than the more antimicrobial Milk completely eliminated the antibacterial activity of PS against PS, and this small chemical difference was associated with a clear S. aureus. Even the highest concentration applied (200 μg/ml) was decrease of the antimicrobial activity. PS, again, is more hydrophilic ineffective (data not shown), although in laboratory media, S. aureus than PSMME, and the higher hydrophobicity of PSMME may facilitate is completely inhibited already at PS concentrations of 75 μg/ml its passage through biological membranes. Indeed, PSMME shows a (unpublished results). higher antimicrobial activity than PS at least in C. tropicalis and in L. monocytogenes. Whether the higher lipophilicity could be an explana- 3.4. The effects of PS on internalized L. monocytogenes in Caco2 cells tion for the marked antifungal properties observed with PSMME cannot be concluded on the basis of available data. PS clearly affected the intracellular multiplication of all four The norlignans were chosen for this study since their L. monocytogenes strains tested in the Caco2 model, and the diphenylpentane skeleton structure resembles the stilbene structure effect was generally detectable even at the lowest concentration with phenolic hydroxyls and double bonds (Fig. 1). Some norlignans (50 μg/ml) tested (Fig. 5). are also found in nature, but the now-tested compounds can be semisynthetically prepared from HMR, which is an abundant bioactive 4. Discussion lignan that can be produced in a large scale from knotwood (Holmbom et al., 2003; Willför et al., 2006); however, none of the norlignans now The wide antimicrobial spectrum of PS and PSMME, reported tested showed any effect. Quite obviously, the norlignan structure is earlier by Välimaa et al. (2007), spanning both Gram-positive and not favorable for this kind of activity, which may be due to the longer Gram-negative bacteria, as well as yeasts, was further confirmed in carbon chain between the phenolic structures or steric hindrance due this study. The activity of PS was consistently and markedly higher to the ring structure in e.g. ArX (Fig. 1). both in the growth inhibition assays and in the tests measuring Plant-derived phenolic compounds and their metabolites have the membrane integrity than that of the chemically closely related earlier been shown to destabilize and weaken outer membrane of resveratrol. Gram-negative bacteria (Alakomi et al., 2007; Nohynek et al., 2006), One of the original intentions of the study was to find structure– which could also be one of the antibacterial action mechanisms of activity relationships among the different stilbenes and norlignans. PS. The results obtained in the NPN uptake assay demonstrate that The only compounds having clear antimicrobial properties were PS and PSMME destabilized the outer membrane of Salmonella cells the stilbenes PS, PSMME, piceatannol, and resveratrol. The stilbenes and that the MgCl2 addition partially abolishes the permeabilization 106 C. Plumed-Ferrer et al. / International Journal of Food Microbiology 164 (2013) 99–107 a 400 1.E+10 350 1.E+08 300 1.E+06 250

cfu/g 1.E+04 200 1.E+02 150 1.E+00 12345671417 100 Multiplication factor Multiplication Day 50 PS 25 µg/g PS 200 µg/g control 0 ControlDMSO Pinosylvin Pinosylvin Pinosylvin Pinosylvin b -50 1.E+06 Control 50 µM 75 µM 100 µM 200 µM

Fig. 5. The effect of pinosylvin on the intracellular replication of L. monocytogenes in Caco2 cells. Average results of four strains (EELA Oulu ba 2392/4, EELA Hki L627 b, KTL IHD 43573 and 01 REB 3) and four experiments. 1.E+04

cfu/g of action against eukaryotic micro-organisms cannot even be speculated 1.E+02 at the moment. The antimicrobial efficacy of PS was not affected in sauerkraut and gravlax, indicating that neither vegetable nor proteinaceous and fatty food matrix abolishes the activity. The complex microbial interactions 1.E+00 in a fermentation process typical for sauerkraut may account that 0 7 14 21 the antimicrobial activity against Enterobacteriaceae appears to be Day reduced at higher PS concentrations. The results obtained with gravlax control PS 140 µg/g have been also described by Gözü et al. (2010).However,milktotally neutralized the antimicrobial action. A relatively high binding of resver- c atrol to milk proteins has been demonstrated by Xiao et al. (2011),anda 1.E+08 decrease of antioxidative properties of plant polyphenolics was also observed as a result of interaction with milk proteins. It is, therefore, 1.E+06 likely that similar types of interactions occur with PS and milk proteins, and this could be also the reason for the decreased antimicrobial action. 1.E+04 In the cell culture conditions, PS was able to inhibit the intracellular

cfu/g multiplication of L. monocytogenes strains in Caco2 cells at doses roughly similar to the effective doses in growth inhibition experiments 1.E+02 performed in liquid cultures. Although the multiplication factor varies widely between the strains and experiments, which makes it difficult 1.E+00 to statistically evaluate the EC50 dose, there is a clear tendency for de- μ 012345 creased intracellular multiplication from the PS concentration of 50 M μ Week onwards, leading to an almost complete inhibition at 200 M. The cytotoxicity (24 h) of PS becomes apparent at doses between 125 and potassium sorbate control PS 150 µg/g 250 μM (unpublished results), with practically no effects at lower

Fig. 4. a. The effect of PS on the Enterobacteriaceae in sauerkraut fermentation (a repre- concentrations. Although reduction of the multiplication factors can sentative experiment). b. The effect of PS on the levels of inoculated L. monocytogenes thus be seen already at concentrations well below the clearly cytotoxic 01 REB 3 in gravlax stored at 8 °C. Means of three experiments. PS doses, the possibility of Caco2 cytotoxicity to contribute to the bacteriocidal effect of PS on the intracellular L. monocytogenes cannot be excluded. In any case, the results indicate the ability of PS to penetrate activity of both PS and resveratrol indicating that part of their activity the cellular membranes of the intestinal epithelium, and this could have is linked to the chelation of outer membrane stabilizing divalent consequences both regarding the eventual uses and the safety aspects cations. PS and PSMME caused severe damage on the outer membrane of PS. of Salmonella cells as indicated by increased susceptibility to the hydro- In any eventual food uses of PS or its derivatives, the safety aspects phobic antibiotic novobiocin. In contrast resveratrol was not able to have to be taken into account. While the actual toxicological studies destabilize the outer membrane sufficiently to cause increased sensitivity on pinosylvin are apparently still lacking, it has been administered to novobiocin. Also, viability staining with Live/Dead kit demonstrated either intravenously or intraperitoneally to animals (rats or mice) in that PS and PSMME caused more severe damage on bacterial cell pharmacokinetic trials or in studies on antimetastatic activity using membranes than resveratrol. doses of up to 10 mg/kg bw without immediate adverse effects (Park The specific mechanism of the action of PS against Gram-positive et al., 2012; Roupe et al., 2005, 2006). A subchronic 90 day oral toxicity bacteria – and why certain bacteria, like L. monocytogenes, B. cereus, study on resveratrol administered in doses up to 1000 mg/kg bw/day and S. aureus, are very sensitive, while lactobacilli are not affected – (rats) or 1200 mg/kg bw/day (dogs) indicated no biologically significant is not known. The present results cannot give any answer to this, adverse effects in either species the NOAEL for rats being 200 mg/kg bw/ either, except for the observed effect on the depolarization of the day and 600 mg/kg bw/day for dogs (Johnson et al., 2011). Because cytoplasmic membrane of staphylococci. Likewise, the mechanisms of the chemical similarity of resveratrol and PS, the latter actually being C. Plumed-Ferrer et al. / International Journal of Food Microbiology 164 (2013) 99–107 107 metabolized to certain extent to resveratrol by liver microsomes Johnson,W.D.,Morrissey,R.L.,Usborne,A.L.,Kapetanovic,I.,Crowell,J.A.,Muzzio,M., (Roupe et al., 2005) one can assume that these compounds share similar McCormick, D.L., 2011. Subchronic oral toxicity and cardiovascular safety pharmacology studies of resveratrol, a naturally occurring with cancer preventive activity. toxicological properties. However, confirmatory studies are certainly Food and Chemical Toxicology 49, 3319–3327. needed to verify this assumption. Klinge, C., Blankenship, K., Risinger, K., Bhatnagar, S., Noisin, E., Sumanasekera, W., Zhao, L., Brey, M., Keynton, R., 2005. Resveratrol and estradiol rapidly activate MAPK signaling through estrogen receptors alpha and beta in endothelial cells. 5. Conclusions Journal of Biological Chemistry 280, 7460–7468. Lagouge,M.,Argmann,C.,Gerhart-Hines,Z.,Meziane,H.,Lerin,C.,Daussin,F.,Messadeq,N., Pinosylvin was shown to have more prominent antimicrobial Milne, J., Lambert, P., Elliott, P., Geny, B., Laakso, M., Puigserver, P., Auwerx, J., 2006. Resveratrol improves mitochondrial function and protects against metabolic disease properties than the related compound resveratrol that has received by activating SIRT1 and PGC-1 alpha. Cell 127, 1109–1122. much attention as a multifunctional natural antimicrobial compound Lyons, M., Yu, C., Toma, R., Cho, S., Reiboldt, W., Lee, J., Van Breemen, R., 2003. with promising health effects. The antibacterial and antifungal action of Resveratrol in raw and baked blueberries and bilberries. Journal of Agricultural and Food Chemistry 51, 5867–5870. pinosylvin is not hampered by food components of either vegetable or Matsuda, H., Morikawa, T., Toguchida, I., Park, J., Harima, S., Yoshikawa, M., 2001. animal origin, with the notable exception of milk, which effectively Antioxidant constituents from rhubarb: structural requirements of stilbenes for the neutralizes the activity. Consequently, pinosylvin could also be considered activity and structures of two new anthraquinone glucosides. Bioorganic & Medicinal Chemistry 9, 41–50. as a natural disinfectant or a biocide, in particular as a targeted agent Mattson, T.E., Johny, A.K., Amalaradjou, M.A.R., More, K., Schreiber, D.T., Patel, J., against Listeria, Bacillus,andStaphylococcus. These types of applications, Venkitanarayanan, K., 2011. Inactivation of Salmonella spp. on tomatoes by plant of course, would require the elucidation of the mechanism(s) of action molecules. International Journal of Food Microbiology 144, 464–468. Montero, C., Cristescu, S., Jimenez, J., Orea, J., Hekkert, S., Harren, F., Urena, A., 2003. Trans- and the safety aspects. resveratrol and grape disease resistance. A dynamical study by high-resolution laser- based techniques. Plant Physiology 131, 129–138. Acknowledgments Nohynek, L.J., Alakomi, H., Kahkonen, M.P., Heinonen, M., Helander, K.M., Oksman- Caldentey, K., Puupponen-Pimia, R.H., 2006. Berry phenolics: antimicrobial properties and mechanisms of action against severe human pathogens. Nutrition and Cancer — We thank Niina Torttila, Mirja Rekola and Riitta Venäläinen An International Journal 54, 18–32. for their excellent technical assistance. Financial support from the Paceasciak, C., Hahn, S., Diamandis, E., Soleas, G., Goldberg, D., 1995. The red wine phenolics trans-resveratrol and quercetin block human platelet-aggregation and Finnish Funding Agency for Technology and Innovation (Tekes) is eicosanoid synthesis — implications for protection against coronary heart-disease. gratefully acknowledged. This work was also part of the activities of Clinica Chimica Acta 235, 207–219. the Åbo Akademi Process Chemistry Centre. Park, E.-J., Park, H.J., Chung, H.-J., Shin, Y., Min, H.-Y., Hong, J.-Y., Kang, Y.-J., Ahn, Y.-H., Pyee, J.-E., Lee, S.K., 2012. Antimetastatic activity of pinosylvin, a natural , is associated with the suppression of matrix metalloproteinases. Journa of Nutritional References Biochemistry 23, 946–952. Puupponen-Pimia, R., Nohynek, L., Hartmann-Schmidlin, S., Kahkonen, M., Heinonen, M., Adrian, M., Jeandet, P., Douillet-Breuil, A., Tesson, L., Bessis, R., 2000. Stilbene content of Maatta-Riihinen, K., Oksman-Caldentey, K., 2005. Berry phenolics selectively mature Vitis vinifera berries in response to UV-C elicitation. Journal of Agricultural inhibit the growth of intestinal pathogens. Journal of Applied Microbiology and Food Chemistry 48, 6103–6105. 98, 991–1000. Alakomi, H., Skytta, E., Saarela, M., Mattila-Sandholm, T., Latva-Kala, K., Helander, I., Rauha, J., Remes, S., Heinonen, M., Hopia, A., Kahkonen, M., Kujala, T., Pihlaja, K., 2000. Lactic acid permeabilizes gram-negative bacteria by disrupting the outer Vuorela, H., Vuorela, P., 2000. Antimicrobial effects of Finnish plant extracts membrane. Applied and Environmental Microbiology 66, 2001–2005. containing flavonoids and other phenolic compounds. International Journal of Alakomi, H., Saarela, M., Helander, I., 2003. Effect of EDTA on Salmonella enterica Food Microbiology 56, 3–12. serovar Typhimurium involves a component not assignable to lipopolysaccharide Roupe, K., Halls, S., Davies, N.M., 2005. Determination and assay validation of release. Microbiology-Sgm 149, 2015–2021. pinosylvin in rat serum: application to drug metabolism and pharmacokinetics. Alakomi, H., Paananen, A., Suihko, M.-, Helander, I.M., Saarela, M., 2006. Weakening Journal of Pharmaceutical and Biomedical Analysis 38, 148–154. effect of cell permeabilizers on gram-negative bacteria causing biodeterioration. Roupe, K.A., Yáñez, J.A., Teng, X.W., Davies, N.M., 2006. Pharmacokinetics of selected Applied and Environmental Microbiology 72, 4695–4703. stilbenes: , piceatannol and pinosylvin in rats. Pharmacy and Alakomi, H., Puupponen-Pimia, R., Aura, A., Helander, I.M., Nohynek, L., Oksman- Pharmacology 58, 1443–1450. Caldentey, K., Saarela, M., 2007. Weakening of Salmonella with selected microbial Saarela, M.H., Alakomi, H.-, Puhakka, A., Matto, J., 2009. Effect of the fermentation pH metabolites of berry-derived phenolic compounds and organic acids. Journal of on the storage stability of Lactobacillus rhamnosus preparations and suitability Agricultural and Food Chemistry 55, 3905–3912. of in vitro analyses of cell physiological functions to predict it. Journal of Applied Boissy, P., Andersen, T., Abdallah, B., Kassem, M., Plesner, T., Delaisse, J., 2005. Resveratrol Microbiology 106, 1204–1212. inhibits myeloma cell growth, prevents osteoclast formation, and promotes osteoclast Smeds, A., Eklund, P., Willför, S., 2012. Content, composition, and stereochemical char- differentiation. Cancer Research 65, 9943–9952. acterization of lignans in berries and seeds. Food Chemistry 134, 1991–1998. Boocock, D.J., Faust, G.E.S., Patel, K.R., Schinas, A.M., Brown, V.A., Ducharme, M.P., Booth, T.D., Tolomeo,M.,Grimaudo,S.,DiCristina,A.,Roberti,M.,Pizzirani,D.,Meli,M.,Dusonchet,L., Crowell, J.A., Perloff, M., Gescher, A.J., Steward, W.P., Brenner, D.E., 2007. Phase I dose Gebbia,N.,Abbadessa,V.,Crosta,L.,Barucchello,R.,Grisolia,G.,Invidiata,F.,Simoni,D., escalation pharmacokinetic study in healthy volunteers of resveratrol, a poten- 2005. and 3′-hydroxypterostilbene are effective apoptosis-inducing tial cancer chemopreventive agent. Cancer Epidemiology, Biomarkers & Preven- agents in MDR and BCR-ABL-expressing leukemia cells. The International Journal of tion 16, 1246–1252. Biochemistry & Cell Biology 37, 1709–1726. Chan, M., 2002. Antimicrobial effect of resveratrol on dermatophytes and bacterial Välimaa, A., Honkalampi-Hämäläinen, U., Pietarinen, S., Willför, S., Holmbom, B., von pathogens of the skin. Biochemical Pharmacology 63, 99–104. Wright, A., 2007. Antimicrobial and cytotoxic knotwood extracts and related pure Eklund, P., Riska, A., Sjoholm, R., 2002. Synthesis of R-(-)-imperanene from the natural compounds and their effects on food-associated microorganisms. International Journal lignan hydroxymatairesinol. Journal of Organic Chemistry 67, 7544–7546. of Food Microbiology 115, 235–243. Gözü, B.B., Komulainen, H., Hyvönen, P., von Wright, A., 2010. The impact of pinosylvin Willför, S., Ahotupa, M., Hemming, J., Reunanen, M., Eklund, P., Sjöholm, R., Eckerman, C., on the development of Listeria monocytogenes in the salted rainbow trout Pohjamo, S., Holmbom, M., 2003. Antioxidant activity of knotwood extractives (Oncorhynchus mykiss, Walbaum, 1792) stored at different temparatures. Journal of and phenolic compounds of selected tree species. Journal of Agricultural and Food Fisheries Sciences.com 4, 419–426 (article in Turkish). Chemistry 51, 7600–7606. Holmbom, B., Eckerman, C., Eklund, P., Hemming, J., Nisula, L., Reunanen, M., Sjöholm, R., Willför, S., Smeds, A., Holmbom, B., 2006. Chromatographic analysis of lignans. Sundberg, A., Sundberg, K., Willför, S., 2003. Knots in trees — a new rich source of Journal of Chromatography 1112, 64–77. lignans. Phytochemistry Reviews 2, 331–340. Xiao, J., Mao, F., Yang, F., Zhao, Y., Zhang, C., Yamamoto, K., 2011. Interaction of dietary Jang, M., Cai, E., Udeani, G., Slowing, K., Thomas, C., Beecher, C., Fong, H., Farnsworth, N., polyphenols with bovine milk proteins: molecular structure–affinity relationship Kinghorn, A., Mehta, R., Moon, R., Pezzuto, J., 1997. Cancer chemopreventive activity and infl uencing bioactivity aspects. Molecular Nutrition & Food Research 55, of resveratrol, a natural product derived from grapes. Science 275, 218–220. 1637–1645.