Nisin application delays growth of Listeria monocytogenes on fresh-cut iceberg lettuce in modified atmosphere packaging, while the bacterial community structure changes within one week of storage

Postharvest Biology and Technology 147 (2019) 185–195

Contents lists available at ScienceDirect

Postharvest Biology and Technology journal homepage: www.elsevier.com/locate/postharvbio

Nisin application delays growth of Listeria monocytogenes on fresh-cut iceberg lettuce in modified atmosphere packaging, while the bacterial community structure changes within one week of storage Oisin McManamon, Thomas Kaupper1, Johann Scollard2, Achim Schmalenberger

T

⁎

University of Limerick, School of Natural Sciences, Department of Biological Sciences, Plassey Park Road, Schrodinger Building, Limerick, V94T9PX, Ireland

A R T I C LE I N FO

A B S T R A C T

Keywords: Nisin A Anti-listerial agent Modified atmosphere Next generation sequencing Lactuca sativa

Listeria monocytogenes poses a risk to minimally processed ready-to-eat foods such as lettuce due to its ability to grow under refrigeration conditions. Since many natural anti-listerial products render Iceberg lettuce unsuitable for consumption within 2 d of storage, this study investigated the efficacy of Nisin A as anti-listerial agent and its sensory impact on lettuce. In addition, the evolution of the bacterial community on fresh-cut lettuce was monitored for the duration of storage. In-vitro assays confirmed the efficacy of Nisin A to inhibit growth of a three strain mix of L. monocytogenes in model atmospheres and air. The L. monocytogenes strain mix was added to lettuce that was subsequently treated either with Nisin, L. lactis DSM20729 (a Nisin A producer) or was kept without inoculation. Incubation took place at 4 and 8 °C under various atmospheres. On days 0, 2, 5 and 7, L. monocytogenes was enumerated on selective agar and a sensory panel graded the lettuce on visual appearance. At 4 and 8 °C a 10 to 100-fold reduction of L. monocytogenes growth was achieved with 5 mg kg−1 Nisin over a seven-day period, while lettuce kept an acceptable sensory appearance over the first 5 d. Direct application of L. lactis had no detectable effect on L. monocytogenes growth in situ. The bacterial community structure changed substantially from each sampling day to the next over the seven days of incubation. However, Pseudomonadaceae with the genus Pseudomonas were most abundant at all times and increased in relative abundance to over 90% by day 7. In conclusion, the application of Nisin A to minimally processed vegetables like lettuce seems to be a viable alternative to reduce and delay growth of pathogen L. monocytogenes, while not impacting the sensory appearance for 2–5 d.

1. Introduction Production in the global fresh fruit and vegetable industry has increased in the first half of this decade by almost 40% (FAOSTAT - Food and Agriculture Organization of the United Nations (FAO, 2017). A central part of this increase is the rise in consumption of foods that are minimally processed and alongside with it food related illnesses (Omac et al., 2015). Most fresh fruit and vegetables harbour considerable amounts of natural microbes, some for which have the potential to be pathogenic such as Listeria monocytogenes (Francis et al., 1999). The food industry that includes fresh cut fruit and vegetables minimizes contamination through its use of chlorine washes, low storage temperatures, modified atmospheres (Francis and O’Beirne, 1997) and the adherence to the Hazard Analysis Critical Control Point (HACCP) procedure. Despite these efforts, L. monocytogenes is still a common

occurrence (Leong et al., 2017) and is estimated to be responsible for about 260 food-related deaths each year in the USA alone (Centers for Disease Control and Prevention - CDC, 2016). L. monocytogenes is able to grow in adverse environmental conditions, all while being regularly detected in soil, water, vegetation, livestock, food processing and storage facilities (Harris et al., 2003). Its ubiquitous presence results in easy entry into the food processing chain at various points, where it is able to survive for extended periods despite sanitation efforts (Leong et al., 2014). The contamination of freshcut produce that support L. monocytogenes growth are of particular concern as control methods including refrigeration and modified atmospheres (MAP) may be inefficient to inhibit growth over the course of storage (Scollard et al., 2016). Even lowest levels of contamination may result in multiplication beyond the limits that are considered to be safe for consumption (McManamon et al., 2017). The use of chlorine

⁎

Corresponding author. E-mail address: [email protected] (A. Schmalenberger). 1 Current address: Leibniz University Hannover, Institute of Microbiology, Herrenhäuser Straße, 30419 Hannover, Germany. 2 Current address: University College Cork, School of Food & Nutritional Sciences, Cork, Ireland. https://doi.org/10.1016/j.postharvbio.2018.10.002 Received 7 June 2018; Received in revised form 28 September 2018; Accepted 1 October 2018 0925-5214/ © 2018 Elsevier B.V. All rights reserved.

Postharvest Biology and Technology 147 (2019) 185–195

O. McManamon et al.

and to CO2 of 19 nmol m−2 s-1 kPa-1 (Amcor Flexibles, Gloucester, UK).

based products as anti-microbial treatment on fresh cut produce has raised safety concerns due to the formation of halogenated carcinogenic by-products (Ölmez and Kretzschmar, 2009; Zhang and Farber, 1996). The application of essential oil based natural anti-listerial agents on fresh-cut fruit appeared to be a viable alternative to chlorine dipping, but their application on fresh-cut lettuce resulted in product browning beyond acceptance levels (Scollard et al., 2016). The antimicrobial peptide Nisin has been used in dairy products, dressings, sauces and post fermented meats (Adams and Smid, 2003). Nisin is regarded as a safe food additive (Cotter et al., 2005; Delves-Broughton et al., 1996) and is naturally produced by Lactococcus lactis. The European Union has already approved the use of Nisin (E 234) to various foodstuffs such as semolina pudding and dairy products including cheese at concentration limits of 3–12.5 mg kg−1 (European Council, 2008). A paucity of studies looked into the use of Nisin on ready to eat vegetable products. Xu et al. (2007) showed that a dipping solution with Nisin (50 mg kg−1), 0.5% citric acid and 200 mg kg−1 grapefruit seed extract could reduce L. monocytogenes on lettuce leaves by over 2 logs (CFU g-1) without a negative effect on its organoleptic properties. The aim of this study was to test the effect of Nisin onto growth of L. monocytogenes on fresh cut lettuce as a single anti-listerial additive at concentrations that would be in accord with current EU guidelines of Nisin as a food additive. In addition, this study aimed to characterise the bacterial community structure of fresh-cut lettuce with and without the use of Nisin and L. monocytogenes. The hypotheses of this study were that i) Nisin application can sustainably inhibit growth of L. monocytogenes on fresh-cut lettuce at concentrations used in the food industry, ii) that the application of a Nisin producing L. lactis strain onto fresh-cut lettuce may be an economical substitute for application of Nisin, and that iii) Nisin application and storage conditions have a pronounced effect onto the lettuce bacterial community structure.

2.3. Preparation of L. monocytogenes cultures and inoculation of fresh-cut lettuce Three L. monocytogenes strains (cultured separately at 4 °C for 14 days or 8 °C for 7 days in accordance to the temperatures being used in the test conditions in 10 ml of tryptic soy broth; TSB, Oxoid CM129, Fannin Healthcare, Cork, Ireland) from the Listeria strain collection at Teagasc Food Research Centre (Moorepark, Ireland; strains 959 - vegetable isolate, 1382 – EUR Lm reference strain, 6179 – food processing plant isolate) were used for all tests in order to follow the European Guidelines for challenge tests on ready-to-eat foods (Beaufort et al., 2014). Equal quantities of each strain were mixed after cultivation for inoculation of fresh-cut produce and diluted to 102−3 CFU g-1 in phosphate buffered saline (PBS, pH 7.3, Oxoid BR014, Fannin Healthcare). Aliquots of 0.1 mL of L. monocytogenes suspension were distributed uniformly over the lettuce contained within each of the packages immediately before antimicrobial treatments were applied (see section 2.4 below). 2.4. Antimicrobial treatments The effectiveness of the anti-listerial treatments were tested as follows and compared to controls that received 1 ml of PBS instead: 2.4.1. Lactococcus lactis application L. lactis subsp. lactis DSM 20729 (ATCC 11454) was grown in TSB as described for L. monocytogenes in section 2.3 at either 4 or 8 °C. The culture was diluted in 20 ml PBS (pH 7.3) to allow inoculation of freshcut produce at 103 CFU g−1. Aliquots of 1 mL of L. lactis suspension were distributed uniformly over the lettuce contained within each of the packages.

2. Materials and methods 2.1. In vitro Nisin assay

2.4.2. Nisin A application Commercially available Nisin A (2.5% w/w in salt) (Handary SA, Brussels, Belgium) was suspended in 10 ml of distilled water (100 mg kg-1) in Float-A-Lyzer membranes of 1–5 kDa (Spectrum Laboratories, Inc. California, USA) in 2 L of distilled water for 24 h. The dialyzed solution in the tubes were adjusted with distilled water to 100 mg L−1 nisin and autoclaved at 121 °C for 15 min. One part of the stock solution was diluted with sterile water to 50 mg L−1 nisin. Both, the 50 and 100 mg L−1 nisin was diluted in PBS to 25 and 50 mg L−1 nisin for application at a final concentration of 2.5 and 5 mg kg-1. Aliquots of 1 mL of the Nisin A suspension were distributed uniformly over the lettuce contained within each of the packages. Control packages received 1 mL of 0.5x PBS instead.

A mini titre plate assay was conducted (24-well plate) in order to establish the effectiveness of the purified Nisin A at various quantities (0, 5, 10 and 25 mg kg−1) in vitro under 3 different atmospheric conditions: (a) air, (b) 8 kPa CO2, 4 kPa O2, 88 kPa N2 (modified atmosphere packaging, MAP) or (c) 15 kPa CO2, 1 kPa O2, 84 kPa N2 (MAP). The L. monocytogenes mix from Section 2.3 was used for inoculation at 106 CFU mL-1. Into each well 1000 μL of TSA broth, 1000 μL of PBS containing the appropriate amount of purified Nisin (0, 5, 25 and 50 mg kg−1 final concentration) and 10 μL of L. monocytogenes was added aseptically in triplicate. The 24-well plate was then placed into a 35 μm think orientated polypropylene (OPP) packaging bags sealed using heat sealed at the selected atmosphere as described in Section 2.4.3. The packs were then stored (incubated) at 37 °C for 48 h. At the selected time point the plates were removed from the 37 °C incubator and analysed for cell density using a spectrophotometer (VWR, USA) at 600 nm. Loops were used to streak the well contents on LSA to confirm the presence of viable L. monocytogenes. Individual 24-well plates were used for each of the sampling points and then discarded after reading.

2.4.3. Atmospheric treatments, package sealing and storage conditions After completion of the inoculation and the antimicrobial treatment, packs were flushed with gas atmospheres: (a) air, (b) 8 kPa CO2, 4 kPa O2, 88 kPa N2 or (c) 15 kPa CO2, 1 kPa O2, 84 kPa N2 modified atmosphere packaging (MAP) and heat sealed using a vacuum packer (Multivac Mobil 3000, Wolfertschwenden, Germany) as described previously (Scollard et al., 2016) into 18 x 10 cm packs (see also Section 2.2). The packs were then incubated for up to 7 d at either 4 or 8 °C. Oxygen concentrations in the packs were determined immediately before destructive sampling with a Systech 6600 (Systech Instruments, Thame, UK) headspace oxygen analyser.

2.2. Preparation of fresh-cut iceberg lettuce Iceberg lettuce (Class I Spain) was obtained from a local vegetable supplier (Limerick, Ireland) at the day of testing and kept under refrigeration (4 °C) until processed. Outer layers and stem were manually removed using a disinfected sharp stainless steel knife. Any damaged leaves and the core of the heads were excluded. The remaining inner leaves were sliced with a disinfected sharp stainless steel knife and cut into pieces of approx. 20 mm portions of 10 g cut lettuce were transferred into 35 μm thick orientated polypropylene (OPP) packaging bags (18 x 10 cm) which had a permeability to O2 of 5.7 nmol m−2 s-1 kPa-1

2.5. Enumeration of L. monocytogenes Bacterial cell counts were carried out throughout storage from three replicate packs on days 0 (day of inoculation) 2, 5 and 7. The lettuce samples from each package were homogenised for 120 s at 260 pedal strokes per minute in PBS in a 2-fold dilution (20 ml for 10 g of lettuce) 186

Postharvest Biology and Technology 147 (2019) 185–195

O. McManamon et al.

lettuce incubation in air with 5 mg kg−1 Nisin). Briefly, DNA was amplified directly without blocking of eukaryotic DNA using the 341 F/ 806R primer set (341 F: 5′- CCTAYGGGRBGCASCAG -3′, 806R: 5′-GGACTACNNGGGTATCTAAT -3′), targeting the V3-V4 region of the bacterial 16S rDNA. Phusion high-fidelity PCR Mastermix (New England Biolabs, Ipswich, MA) was used for amplification and PCR products were purified using a Gel Extraction Kit (Qiagen, Dusseldorf, Germany). Sequencing libraries were generated with a TruSeq DNA PCR-Free Sample Preparation Kit (Illumina, San Diego, CA) following manufacturer's instructions and index codes were added. The library quality was assessed using the Agilent Bioanalyzer 2100 system and quantified via Qubit Fluorometer (Thermo Scientific). The library was sequenced on an IlluminaHiSeq2500 platform to generate 250 bp paired-end reads.

using a laboratory stomacher (Seward 400, AGB Scientific, Ireland). For expected low cell counts, samples were concentrated 10 fold via centrifugation at 4000 g for 240 s and re-suspended in PBS (Scollard et al., 2016). Aliquots of 0.1 mL were plated in duplicate on Listeria selective agar (LSA, Oxoid CM856) containing a modified L. monocytogenes selective supplement (Oxoid SR0206) at a detection limit of 2 CFU g−1. Serial dilutions were performed in PBS when required and plated on LSA during the testing period. Incubation took place at 37 °C for 48 h. 2.6. Enumeration of total cultivable heterotrophic bacteria Cell counts of cultivable heterotrophic bacteria were carried out as in Section 2.5 from three replicate packs, but Tryptone Soya Agar (TSA, Oxoid CM0131) was used instead of LSA. As higher CFU were expected when compared to L. monocytogenes, additional serial dilutions were performed in PBS and plated on TSA during the testing period.

2.9. DGGE fingerprinting and analysis DGGE was carried out in a TV400-DGGE (Scie-Plas, Cambridge, UK). Gels (200 × 200 x 1 mm) with a 10% (w/v) acrylamide/bisacrylamide gel using a linear 35–65% gradient in 1 x TAE buffer (60 °C) for 16.5 h at 63 V as described previously (Fox et al., 2014). Gels were stained in 10,000 times diluted SYBR Gold (Invitrogen, Carlsbad, CA) for 30 min for visualization on a trans-illuminator (G:Box, Syngene, Cambridge, UK). Selected fingerprints were re-run as above and bands were excised on a visiblue trans-illuminator (UVP, Upland, CA). Bands were treated as described previously (Schmalenberger et al., 2008) to extract DNA for direct amplification using the same primers as before but without a GC clamp and 32 cycles instead of 20. Nucleic acid sequences were deposited in the GenBank Nucleotide Archive (MH229463- MH229467).

2.7. Evaluation of sensory quality Evaluation of appearance was performed on the fresh-cut produce packages during storage (days 0, 2, 5 and 7) by an untrained sensory panel as described recently (Scollard et al., 2013). In brief, the panel consisted of 5 evaluators and before analysis, panellists were familiarised with the product and scoring procedure. The panellists were asked to score the appearance of samples, on an 11-point scale ranging appearance from 10 (excellent) to 0 (poor). A combined score of 6 was considered the lowest acceptable commercial score. Panellists were asked to evaluate colour and appearance against that of a fresh-cut control sample. The evaluations were carried out under typical indoor daylight conditions and at a temperature of 18–20 °C (Scollard et al., 2016).

2.10. Next generation sequencing and analysis 2.8. Extraction of DNA, 16S rRNA gene amplification for DGGE and next generation sequencing

Based on their unique barcode, paired-end reads were assigned to samples and truncated by removing barcode and primer sequence. Paired-end reads were merged using FLASH (Magoc and Salzberg, 2011). The QIIME (Caporaso et al, 2010) quality controlled process was followed for filtering the raw tags at specific filtering conditions to obtain high-quality clean tags (Bokulich et al, 2013). Tags were compared with the reference database (Gold database, http://drive5.com/ uchime/uchime_download.html) via the UCHIME algorithm (Edgar et al, 2011). Sequence analyses were performed with the Uparse software (Edgar, 2013). Sequences with ≥ 97% similarity were assigned to the same OTUs. The GreenGene Database (DeSantis et al, 2006) was used for OTU screening based on the RDP classifier algorithm (Wang et al, 2007). Multiple sequence alignments were carried out with the PyNAST software (Caporaso et al 2010b) to study the phylogenetic relationships between OTUs. Alpha and beta diversities were calculated using QIIME (Version 1.7.0), principal coordinates analyses (PCoA) were performed and displayed with the WGCNA package, stat packages and ggplot2 package in R (Version 2.15.3). R was also used to create heat maps in order to display the abundance distribution of dominant taxa among the differently treated lettuce samples. Nucleic acid sequences were deposited in the Nucleotide Archive (Project: PRJEB26864, accession numbers ERS2492060-71).

DNA was extracted from replicate model packages by homogenization for 120 s at 200 pedal strokes per minute in PBS (2-fold dilution) using a laboratory stomacher (Seward 400). The obtained suspension was transferred into conical tubes and pelleted in a centrifuge at 4500 g (15 min at 4 °C). DNA was extracted from 0.5 g of the lettuce derived pellets using the PowerFood DNA Isolation kit (MO BIO Laboratories, Carlsbad, CA) according to manufacturer’s instruction. Quantification of the extracted DNA was conducted with the Qubit dsDNA HS Assay kit in a Qubit Fluorometer (Life Technologies, Carlsbad, CA). Amplifications of the bacterial 16S rRNA gene fragments for denaturing gradient gel electrophoresis (DGGE) were conducted in a nested PCR approach in order to exclude the majority of the lettuce DNA from being co-amplified. In the first PCR reaction primers 101 F (5′ACTGG CGGACGGGTGAGTAA’3) and 784R (5′TACCMGGGTATCTAATCCKG’3) were selected to amplify preferentially 16S rRNA gene fragments from bacteria. A third oligonucleotide was added to the reaction (784RinvT; 5′TACTGGGGTATCTAATCCCA’3T’5) that was optimised to bind lettuce DNA from chloroplasts. An inverted T at the 3 strand end of the oligo prevented the polymerase from extending the oligonucleotide when bound to plant DNA. Each 25 μl reaction contained 1 x buffer (2 mM MgCl2), 0.2 mM dNTP mix, 0.4 mmol of each primer (and 784RinvT), 0.5 U of DreamTaq polymerase (Fisher Scientific, Waltham, MA) and around 10 ng template DNA. Amplification was conducted in a G-Storm cycler with 28 cycles of 94 °C denaturation (45 s), 58 °C annealing (45 s), 72 °C extension (60 s). The obtained PCR product was 10 times diluted and used as template for the second PCR with GC-341 F/518R primer pair (Muyzer et al., 1993) and 20 cycles of 94 °C denaturation (45 s) 55 °C annealing (45 s) and 72 °C extension (45 s). Sequencing was performed by Novogene Bioinformatics Technology Co., Ltd from 12 DNA samples (Day 0, 2, 5, 7 in triplicates from 4 °C

2.11. Statistical analyses Populations were reported as the means of three independent values and ( ± ) standard deviations. Experimental results were tested for homoscedasticity (Leven’s test) and normality (Shapiro-Wilk test). Where conditions of normality and homoscedasticity were met, an ANOVA with Tukey posthoc test was carried out. Where only homoscedasticity was met, an ANOVA with Games-Howell posthoc was conducted. Where both requirements were not met, even after data transformation, a Kruskal-Wallis test and manual posthoc was applied 187

Postharvest Biology and Technology 147 (2019) 185–195

O. McManamon et al.

to identify significant differences (P ≤ 0.05 for all tests). 3. Results 3.1. Effect of Nisin and modified atmosphere on the growth of L. monocytogenes in vitro The effect of different concentrations of Nisin A onto growth of a three strain L. monocytogenes mix was tested in vitro in 24-well mini titre plates. The plates were sealed in an atmosphere of: (a) air, (b) 4 kPa O2 and (c) 1 kPa O2 MAP packs at 37 °C. Each of the assays demonstrated that all Nisin treatments starting at 5 mg kg−1 were sufficient to inhibit growth over a 48 h period (Supplementary Fig. S1). In contrast, L. monocytogenes showed strong growth in wells without Nisin reaching a maximum cell density (OD of 0.4-0.6) after 16–24 h of incubation. Graphs in Supplementary Fig. S1 suggest that growth may have been slightly accelerated in air and 4 kPA oxygen over the 1 kPA oxygen atmosphere. Subsequent streaking of samples from the respective wells after 48 h of incubation on LSA plates established that viable cells of L. monocytogenes were still present in the wells. 3.2. Effect of Nisin and L. lactis on the growth of L. monocytogenes in lettuce model atmosphere packages at 8 °C Viable numbers of L. monocytogenes (LSA plate counts) differed significantly between the Nisin treatments and the control as well as the L. lactis inoculated packs (Fig. 1). The application of 5 mg kg−1 Nisin resulted in an initial 1.2 log reduction in L. monocytogenes viable numbers compared to the control in air model packs (after a starting concentration of 103 CFU g-1). While growth of L. monocytogenes appeared to be restored at subsequent sampling days (2–7), a significant difference between 5 mg kg−1 Nisin and the control was maintained until day 7 where a 1.04 log difference was recorded. The application of 2.5 mg kg−1 Nisin contributed to a 0.71 log reduction in viable numbers of L. monocytogenes compared to the control lettuce bags in air which was maintained until day 7 (significant; Fig. 1A). The application of L. lactis did not contribute to any reduction in L. monocytogenes numbers compared to the control in air. The numbers of cultivable total heterotrophs (TSA plate counts) in air dropped on day 2 of 0.9 and 1.6 logs for the 5 and 2.5 mg kg−1 Nisin treated lettuce, respectively (Suppl. Fig. S2A). However, this reduction was not detected on days 0, 5 and 7. Application of L. lactis displayed no substantial change in the aerobic counts over the 7 d. Attempts to enumerate L. lactis on lettuce using Raka-Ray agar (Oxoid CM0777, Fannin Healthcare) failed as no CFU’s were identified on the agar plates (data not shown). Further attempts to quantify L. lactis on lettuce using selective media such as MRS (Oxoid CM0361) was subsequently abandoned due to CFUs of variable appearance. Overall, counts of viable heterotrophic bacteria increased by about 1 log from 105 to 106 CFU g-1 over the 7 d course. Incubation of lettuce model packs was repeated in MAP with 4 kPa oxygen. Similar to the packs with air atmosphere, the application of 5 mg kg−1 Nisin contributed to an initial 1.17 log reduction in L. monocytogenes growth compared to the control which was maintained until day 7 where a 0.95 log reduction was still present (both significantly different). The application of 2.5 mg kg−1 Nisin contributed to a 0.92 log reduction in L. monocytogenes growth compared to the control lettuce bags (significant difference) of which was only to some extent maintained until day 7 where a 0.42 log difference was recorded (no significant difference; Fig. 1B). As in air, the application of L. lactis contributed to no reduction in L. monocytogenes growth compared to the control. The heterotrophic plate counts in 4 kPA oxygen at day 2 showed a reduction of 0.89, 1.57 and 0.79 logs for the 5 and 2.5 mg kg−1 Nisin and L. lactis treated lettuce (Suppl. Fig. S2B). However, no differences were detected at day 0 and 5. Some differences in viable counts were recorded again on day 7 with the L. lactis inoculated packs being nearly 0.5 logs lower than the control, while 2.5 mg kg−1 Nisin

Fig. 1. Cultivation of L. monocytogenes at 8 °C on fresh-cut lettuce with 0, 2.5, 5 mg kg−1 Nisin A and L. lactis (square, triangle, inverted triangle, circle, respectively) in air (A), modified atmosphere with 4 kPa oxygen (B), modified atmosphere with 1 kPA oxygen (C). Different letters indicate significant differences, error bars indicate standard deviation.

packs had nearly 0.5 logs higher counts of viable heterotrophic bacteria than the control. On average, heterotrophic growth appeared to be slightly lower in 4 kPa model packs than air model packs. Viable counts of L. monocytogenes in 1 kPa oxygen were similar to the two other tested atmospheres. The application of 2.5 and 5 mg kg−1 Nisin contributed to an initial 0.78 and 0.84 log reduction in L. monocytogenes compared to the control which was maintained until day 7 where a 1.66 and 1.22 log reduction was recorded (Fig. 1C; significant on days 0 and 7). As before, the application of L. lactis had no effect on L. monocytogenes abundance. As above, heterotrophic bacterial counts were only reduced at day 2 with a reduction of 1.4, 0.53 and 0.66 logs for the 2.5, 5 mg kg−1 Nisin and L. lactis treated lettuce (Suppl. Fig. S2C). As for 4 kPa packs, heterotrophic growth appeared to be slightly lower in 1 kPa model packs than air model packs over the 7 d course.

3.3. Effect of Nisin and L. lactis on the growth of L. monocytogenes in lettuce model atmosphere packages at 4 °C Viable numbers of L. monocytogenes among the treatments stored at 4 °C are similar to the ones stored at 8 °C (Figs. 1 and 2) with the difference that the full effect of Nisin in inhibiting L. monocytogenes growth appeared to be more pronounced at day 2. CFU of L. monocytogenes was significantly lower for the 5 mg kg−1 Nisin treatment at all days and in all atmospheres (Fig. 2). In air, CFU counts at 5 mg kg−1 Nisin were 188

Postharvest Biology and Technology 147 (2019) 185–195

O. McManamon et al.

3.4. Oxygen concentrations in lettuce model atmosphere packages (4 and 8 °C) In air, the oxygen content for both Nisin treatments, L. lactis treatment and the control decreased from around 21% to 12–17% over the course of the trial for both storage temperatures. At both temperatures and in all atmospheres, the 5 mg kg−1 Nisin treatment appeared to have higher oxygen concentrations in the model packs after 7 d of storage compared to the other treatments and control (Suppl. Figs. S4 and S5). At MAP with 4 kPa O2 all incubations started at 4% oxygen at day 0 and increased to 10–14% oxygen over the course of the trial at 8 °C. In contrast, at 4 °C storage, oxygen concentration only increased to about 8% for all treatments. An exception was 5 mg kg−1 Nisin, which showed oxygen concentrations of 20% by day 7. In MAP at a value of 1 kPa oxygen at day 0, all packs increased in their oxygen content to between 9 and 10% by day 7 at 8 °C. At 4 °C, the oxygen concentrations in the control and L. lactis treatment increased to 6–8% over the sevenday course, while both Nisin treatments had higher oxygen concentrations of 16–20% by day 7.

3.5. Sensory effect of Nisin and L. lactis on lettuce in model atmosphere packages at 8 °C An untrained sensory panel scored the lettuce appearance in the three atmospheres over the 7 d of incubation. For the 5 mg kg−1 Nisin treatment, scores in all atmospheres at days 0 and 2 remained above the lowest acceptance level (sensory appearance, level 6 out of 10, Fig. 3) of which the 1 kPa O2 atmosphere did slightly better than the other two atmospheres on day 2. On day 5, only the air sample was above the satisfactory level while at day 7 all lettuce samples treated with 5 mg kg−1 Nisin were below the acceptance level of 6. Likewise, the Nisin 2.5 mg kg−1 treatment displayed acceptable scoring from day 0 to day 2, from day 5 onward it fell below acceptance levels for all atmospheres. The L. lactis treatment displayed better sensory levels at day 2 when compared to the Nisin treatments (not significant) but also dropped below acceptance levels on day 5 for the 1 and 4 kPA oxygen atmospheres. On day 7 only the 1 kPa oxygen atmosphere treatment appeared to be above level 6 (Fig. 3). However, statistically, none of the three lettuce treatments appeared to be significantly better than any other. Similarly, the appearance of lettuce controls without addition of nisin or L. lactis deteriorated at similar rates when compared to the treatments. Nonetheless, overall appearance seemed to be better at day 5 in the controls. However, only in 4 kPA oxygen at day 5 the appearance of untreated lettuce was actually significantly better. On all other occasions, the controls were not significantly different to the treatments.

Fig. 2. Cultivation of L. monocytogenes at 4 °C on fresh-cut lettuce with 0, 2.5, 5 mg kg−1 Nisin A and L. lactis (square, triangle, inverted triangle, circle, respectively) in air (A), modified atmosphere with 4 kPa oxygen (B), modified atmosphere with 1 kPA oxygen (C). Different letters indicate significant differences, error bars indicate standard deviation.

lower by 0.46, 1.52, 1.48 and 1.04 logs over the 7 d period when compared to the control (Fig. 2A). In 4 kPa oxygen, a significant reduction of around 1.8 logs was detected with 5 mg kg−1 Nisin over the control between days 0 and 5 (Fig. 2B). The reduction in L. monocytogenes counts at 0.4 logs was found by day 7 (significantly lower than control). In 1 kPa oxygen MAP, differences between the control and 5 mg kg-1 Nisin were significant at all days and varied between 1.2 and 1.0 logs over the 7 d course (Fig. 2C). In air, the 2.5 mg kg−1 treatment closely mirrored the effect of the 5 mg kg−1 treatment, while in 4 kPa oxygen the reduction in L. monocytogenes cell counts was significant only at day 2 (1.4 logs). In 1 kPa oxygen, significant reductions were achieved with 2.5 mg kg−1 Nisin at days 0, 2 and 5 (up to 1.1 logs difference). The application of L. lactis contributed to no substantial reduction in L. monocytogenes growth compared to the control. Only in air at day 0, L. lactis addition resulted in increased numbers of L. monocytogenes over the control (not significant). Otherwise, the L. lactis treatment followed the control very closely. The counts of viable total heterotrophs was only marginally affected by any treatments over the 7 d period (Suppl. Fig. S3) as growth curves largely overlapped. Reductions in numbers peaked between 0.5 and 0.7 logs for the 2.5 mg kg−1 Nisin treatments (day 7, all atmospheres) while for 5 mg kg−1 Nisin no clear reductions over the controls were detected. For the L. lactis treatment, counts tended to be slightly higher than in the controls.

3.6. Sensory effect of Nisin and L. Lactis on lettuce in model atmosphere packages at 4 °C In contrast to the results obtained at 8 °C storage, incubation at 4 °C resulted in higher acceptance levels throughout the 7 d period (Fig. 4). All samples were above the lowest acceptance level for days 0 and 2. The 5 mg kg−1 Nisin treatments dropped below the lowest acceptance level of 6 for days 5 and 7 for the air and 4 kPa oxygen MAPs but remained above level 6 for the duration of the experiment in 1 kPa of oxygen (images of lettuce from day 0 and day 7 in air in Suppl. Fig. S6). At 2.5 mg kg−1 Nisin, only day 7 in air resulted in an acceptance level of below 6, all other levels remained above level 6 for the duration of the experiments. In case of air and 4 kPa oxygen this higher scoring when compared to 5 mg kg−1 Nisin was also significant. Treatment with L. lactis had the lowest sensory impact; all samples remained above the lowest acceptance level (Fig. 4).

189

Postharvest Biology and Technology 147 (2019) 185–195

O. McManamon et al.

Fig. 4. Sensory analysis of fresh-cut lettuce stored at 4 °C with 2.5, 5 mg kg−1 Nisin A and L. lactis (square, circle, triangle, respectively) in air (A), modified atmosphere with 4 kPa oxygen (B), modified atmosphere with 1 kPA oxygen (C). Dotted line indicates the lowest acceptance level for consumers. Different letters indicate significant differences, error bars indicate standard deviation.

Fig. 3. Sensory analysis of fresh-cut lettuce stored at 8 °C with 2.5, 5 mg kg−1 Nisin A and L. lactis (square, circle, triangle, respectively) in air (A), modified atmosphere with 4 kPa oxygen (B), modified atmosphere with 1 kPA oxygen (C). Dotted line indicates the lowest acceptance level for consumers. Different letters indicate significant differences, error bars indicate standard deviation.

species of Pseudomonas, another band appeared to be closely related to Erwinia (Fig. 5). Another band that was predominant at days 0 and 2 of the experiments turned out to be 100% identical with lettuce chloroplast DNA. Direct sequencing of the bacterial community on lettuce for the duration of the 7 days of incubation revealed that up to 90% of the identified sequences belonged to the chloroplasts of lettuce (Suppl. Fig. S8). Highest relative abundance of the chloroplast DNA was found at day 0. Successively, the relative dominance reduced slightly to around 62% at day 7. Furthermore, between 5 and 14% of the remaining sequences were identified as lettuce mitochondrial DNA. After removal of the plant sequences, the bacterial sequences identified via Illumina next generation sequencing were dominated by the family of the Pseudomonaceae (Table 1). After an initial dominance of the lettuce bacterial community at 69% of relative abundance, a drop to 38% at day 2 was recorded that increased again for day 5 (72%) and peaked at day 7 with a relative abundance of almost 95%. The Enterobacteriaceae were the second most abundant bacterial family recorded and ranged from nearly 10% at day 0 to 3–4% at day 5 and 7. A change in the relative abundances of the bacterial families over the 7 d period was visualised as a heatmap of the 35 most abundant families (including lettuce chloroplasts and mitochondria; Fig. 6). The Aeromonadaceae family showed a steady decline in abundance over the incubation period, while families of the Moraxellaceae, Leuconostocaceae,

3.7. Lettuce bacterial community structure Initial fingerprint analysis of the lettuce bacterial community structure revealed low diversity fingerprints dominated by less than four DNA bands (data not shown). Further analysis of the genomic DNA extracted from the fresh-cut lettuce revealed that the extracted nucleic acid contained predominantly DNA from lettuce (data not shown). Subsequent PCR based analysis was therefore carried out with a modified primer of Rastogi and colleagues (Rastogi et al., 2010), preferentially discriminating against plant DNA (primer 784R). Further optimization was achieved through selective bacterial amplification by adding a blocking primer that matched the lettuce SSU DNA sequence. The blocking was achieved by adding an inverted base at the 3-strand end in order to prevent extension by the polymerase (784RinvT). The obtained PCR product then served as a template for a nested PCR for DGGE analysis. Fingerprint analysis for the bacterial community structure on the lettuce at 4 and 8 °C with and without Nisin application revealed diverse banding patterns of over 30 unique bands (Fig. 5, Suppl. Fig. S7). Visual differences were detectable over the course of the experiment, especially when incubated at 8 °C. However, other bands were present throughout the duration of the experiments. In total, eight bands were removed from fingerprints obtained from the 8 °C experiment, of which six contained a dominant sequence ready for direct sequencing. Four bands revealed to be closely associated to different 190

Postharvest Biology and Technology 147 (2019) 185–195

O. McManamon et al.

Fig. 5. Denaturing gradient gel electrophoresis of 16S rRNA gene fragments from lettuce model packages incubated at 8 °C at 4 kPA oxygen atmosphere (left) and at 4 °C in air (centre) and at 4 °C in air with 5 mg kg-1 Nisin (right) over 7 d in triplicates (1–3) with L. monocytogenes. DGGE fingerprints are flanked by a 16S species standard (S) and a 16S rRNA gene fragment from L. monocytogenes (Lm). Arrows indicate positions of bands that were cut out and sequenced.

diversity on the lettuce was low. The Shannon index stayed below 2 during the experiment when the plant amplicons were included. When the lettuce amplicons were removed the Shannon index increased from 2.4 at day 0 to its highest value of 4.3 at day 2 and then slumped to below 1 by day 7. Likewise, the beta diversity changed considerably over the course of the storage at 4 °C. A Principal Coordinate Analysis (PCoA; Weighed Unifrac) was able to separate day 0 from the remainder of the samples on the first axis, while day 2 and 5 were partially separated on the first axis (Fig. 7). Sampling from day 7 was best separated on the first axis from day 0 to 5. However, this separation was in part linked to the variation of the relative abundance of lettuce derived amplicons, the unweighed unifrac PCoA separated the sampling times less clear (data not shown). When lettuce DNA was included in the analysis, the nonparametric multivariate variance test ADONIS revealed that the community at day 0 was significantly different (P < 0.01) to the ones from day 2, 5 and 7. This significant difference remained for day 0 in comparison to day 2 and 7 but not for day 5 when the lettuce amplicons were removed from the analysis.

Table 1 Relative abundance of dominating bacteria (> 0.2%; 16S rRNA gene fragments from next generation sequencing without mitochondria and chloroplasts) from lettuce model packages incubated at 4 °C in air with 5 ppm Nisin A. Family

Day 0

Day 2

Day 5

Day 7

Pseudomonadaceae Enterobacteriaceae Lactobacillaceae Streptococcaceae Moraxellaceae Leuconostocaceae Xanthobacteraceae Acidobacteriaceae Subgroup 1 Syntrophobacteraceae Staphylococcaceae Bradyrhizobiaceae Corynebacteriaceae Rhodobacteraceae Dermabacteraceae Anaplasmataceae Microbacteriaceae Desulfovibrionaceae Mycoplasmataceae Micrococcaceae Synergistaceae Enterococcaceae Aerococcaceae Sphingomonadaceae Aeromonadaceae Erysipelotrichaceae Peptostreptococcaceae Gemmatimonadaceae Comamonadaceae Other

69.30 9.71 0.88 1.32 3.02 1.04 0.28 0.79 0.00 0.22 0.47 0.00 0.00 0.25 0.00 0.00 0.00 0.00 0.09 0.00 0.00 0.00 0.38 1.41 0.03 0.72 0.69 0.35 9.05

37.78 6.98 1.78 5.56 6.88 3.97 0.15 0.03 0.39 0.75 0.15 1.65 1.80 0.59 1.80 1.65 1.18 1.37 0.90 1.01 0.93 0.49 0.34 0.67 0.62 0.70 0.00 0.49 19.39

71.59 3.41 4.34 1.56 0.66 0.20 0.75 0.71 0.59 0.54 0.48 0.53 0.10 0.44 0.15 0.11 0.37 0.03 0.29 0.18 0.25 0.23 0.21 0.14 0.18 0.14 0.17 0.10 11.56

94.72 3.82 0.19 0.08 0.10 0.05 0.02 0.02 0.00 0.03 0.02 0.02 0.01 0.01 0.00 0.00 0.00 0.00 0.02 0.00 0.07 0.00 0.02 0.02 0.01 0.02 0.01 0.00 0.75

4. Discussion The use of natural antimicrobial compounds to keep food products save has gained considerable attention over a number of years now. Examples of such natural compounds include plant derived essential oils, animal derived immunoglobulins and microbial derived bacteriocins (Cotter et al., 2013). While some of these compounds have a long history of application in traditional preservation of certain foods, recent studies have widened the horizon of applications to establish if these constituents could be used as potential additional hurdles in the implementation of the multiple hurdle strategy. One among many benefits of using natural antimicrobial agents is that they may have already GRAS status (generally recognised as safe) as this is the case for Nisin A (Jung et al., 1992). When combined with an increasing demand by consumers to move away from chemical sanitizers, this has resulted in new studies of these antimicrobials. Here, Nisin has been of particular interest due to its established application in cheese products (Burt, 2004; Cotter et al., 2005; Rico et al., 2007). The goal of this study was to establish if two purified commercial Nisin treatments (2.5 and 5 mg kg−1 Nisin A) and a Nisin producing strain of L. lactis can inhibit growth or reduce abundance of L. monocytogenes on iceberg lettuce under active and passive modified atmospheres at different storage temperatures (4 °C and 8 °C). Previous studies have shown that

Mycoplasmataceae, Rhodobacteraceae, Anaplasmataceae and Micrococcaceae were most abundant on day 2. On day 5 the Syntrophobacteraceae, Dermabacteraceae, Staphylococcaceae, Acidobactriaceae, Bradyrhizobiaceae, Lactobacillaceae and Xanthobacteraceae families were highest in their relative abundance. At day 7 Enterobacteriaceae and Pseudomonaceae had their highest relative abundance. Further bacterial families indicated changes in relative abundance over the 7 d period but their relative abundance was below 0.02% of the total bacterial community. Due to the fact that the vast majority of amplicons belonged to chloroplasts, mitochondria and Pseudomonadaceae, the bacterial alpha 191

Postharvest Biology and Technology 147 (2019) 185–195

O. McManamon et al.

Fig. 6. Heatmap of bacterial families (16S rRNA gene fragments from next generation sequencing) from lettuce model packages incubated at 4 °C in air with 5 mg kg−1 Nisin over 7 d. (For interpretation of the heatmap in colour, the reader is referred to the web version of this article).

peptide antimicrobial Nisin appeared to be reduced at the later stages of incubation where L. monocytogenes seemed to have returned to growth. The conclusion was that Nisin A would inactivate around or above 90% of L. monocytogenes cells on fresh-cut lettuce while delaying growth for a number of days for the remainder of the L. monocytogenes cells that have reverted back to growth by day 5. The development of encapsulated Nisin for a delayed release into the food environment may inactivate and delay cell division of L. monocytogenes further. This could lead to a Nisin application with an antibacterial activity for the whole duration of the shelf life of fresh-cut vegetables like lettuce (Boelter and Brandelli, 2016; da Silva-Malheiros et al., 2012; Wu et al., 2016). Randazzo et al. (2009) demonstrated the utilization of commercial nisin as a spray in hydrochloric acid onto lettuce. There, reductions in Listeria sp. counts (below 0.8 logs) was only confirmed after 5 and 7 days of storage (4 °C), which is in direct contrast to the findings from the present study. When compared with the non-treated lettuce, the application of L.

antimicrobial essential oils and their pure compounds can exert antilisterial properties on lettuce leaves but not without severely compromising the product appearance (Scollard et al., 2016). Randazzo et al., (2009) successfully reduced L. monocytogenes counts on lettuce via spraying extracts from a bacteriocin producing lactic acid bacterium onto the lettuce leaves but the sensory implications were not clearly followed up. In the present study, the application of purified Nisin A at 2.5 and 5 mg kg−1 showed a significant reduction of up to 1.3 and 1.6 logs CFU g-1 respectively at 4 °C when compared to the control. At 8 °C the significant reduction was up to around 1.0 log CFU g-1 for both Nisin applications. The effect of the L. lactis treatment in reducing the count of L. monocytogenes was negligible at both storage temperatures. Different atmospheres appeared to have only limited impact on the growth of L. monocytogenes. The effect of the Nisin treatments seemed to be directly tied to the application with the majority of the reductions occurring on the initial 2 d of the trial. However, the efficacy of the 192

Postharvest Biology and Technology 147 (2019) 185–195

O. McManamon et al.

Fig. 7. Principal Coordinate Analysis (PCoA, Weighed Unifrac) of 16S rRNA gene fragments (next generation sequencing) from lettuce model packages incubated at 4 °C in air with 5 mg kg−1 Nisin from days 0 (red squares), 2 (green circles), 5 (dark blue triangles) and 7 (light blue diamonds). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article).

significant role in the quality and microbial safely of the fresh cut produce (Allende et al., 2007; Oliveira et al., 2015). Thus, the continuation of the maintenance of this microbiota is important in order to prevent the rise of pathogenic organisms which may be unaffected by the bacteriocin application. While a paucity of studies have investigated the microbiome of lettuce (Erlacher et al., 2014), a detailed insight into the progression of the lettuce bacterial community was missing until now. In this study, the DGGE fingerprint analysis suggested a progression of the bacterial community on lettuce over the 7 d storage period that was similar for the different storage temperatures and treatments (atmosphere, Nisin, L. monocytogenes). Several of the sequenced DGGE bands turned out to be closely related to Pseudomonas species and this apparent dominance of the Pseudomonadaceae was confirmed via next generation sequencing of one of the experiments (4 °C, 5 mg kg−1 Nisin, air atmosphere) in this study. Species of Pseudomonas have been identified to release pectinolytic enzymes that can affect the surface structure of lettuce and release further nutrients from the lettuce leaves (Ragaert et al., 2007). Previous studies already highlighted the high abundance of Pseudomonas on lettuce leaves via cultivation (Garg et al., 1990) or cultivation independent methods (Rudi et al., 2002). Very recently, the bacterial community structure of ready-to-eat lettuce under various MAP packages has been studied via next generation sequencing (Ioannidis et al., 2018). There, the dominance of Pseudomonas was also reported after 10 d of storage in atmospheres of reduced oxygen, oxygen-free, or in air. DNA was extracted from stomacher bag suspensions before the extraction took place, hence only easily detached bacteria would have been included in the study. In the present investigation, DNA was extracted after the extraction with a stomacher

lactis showed little to no effect on the inhibition of L. monocytogenes growth. This observation has been reported before in a similar study (Allende et al., 2007) where cold storage reduced the Nisin production in situ. In the study conducted by Allende et al. (2007), temperatures under 10 °C only allowed a Nisin Z producer to actually form detectable levels of Nisin while other L. lactis strains grew without Nisin formation. Another possible reason for the lack of Nisin production may be due to inability to produce Nisin in the stationary growth phase (de Arauz et al., 2009; Pongtharangkul and Demirci, 2004). The lack of production of the antimicrobial Nisin A by L. lactis 20729 may have been caused by numerous other conditions such as the natural microbiota of the lettuce, the physical and chemical environment of the lettuce. A recent study by Siroli et al. (2016) showed that the use of a Nisin Z producing strain of L. lactis on fresh cut apples combined with biocontrol agent hexanal/2-(E)-hexenal significantly reduced the counts of L. monocytogenes at 6 °C over a period of 28 d. This suggests that the increase of time in combination with an effective Nisin producing strain over a longer period may produce adequate results in other fresh-cut produce. However, due to the physical characteristics, these techniques may be of little use on fresh-cut lettuce and other short shelf-life readyto-eat products. The effect of the Nisin treatments on the cultivable heterotrophic bacteria on lettuce was limited and only showed a temporary reduction at day 2 for the 8 °C storage condition. At 4 °C, Nisin appeared to have no effect on the general cultivable bacterial populations. However, the use of nutrient rich media and the selected incubation temperatures may have resulted in an underestimation of the actual numbers of viable heterotrophic bacteria. These natural bacterial populations can form part of a protective competitive “shield” which may play a 193

Postharvest Biology and Technology 147 (2019) 185–195

O. McManamon et al.

Appendix A. Supplementary data

which would potentially increase the number of bacterial cells initially attached as biofilm. As for Ioannidis and colleagues (Ioannidis et al., 2018), the present study amplified mostly lettuce DNA, especially at day 0. Unfortunately, the use of the nested PCR approach with the blocking primer as used for the DGGE analysis was not compatible with the primer selection for the chosen Illumina sequencing (V4:515F806R). Future studies may select the V2 or V3 region for Illumina sequencing which would allow a nested PCR approach with a blocking primer. In the present study the following families (and genera) beside the Pseudomonadaceae exceeded an average abundance of above 1%: Enterobacteriaceae (e.g. Pantoea & Enterobacter), Lactobacillaceae (e.g. Lactobacillus), Streptococcaceae (e.g. Lactococcus), Moraxellaceae (e.g. Acinetobacter) and Leuconostocaceae (e.g. Weissella) (all at 4 °C in air with 5 mg kg−1 Nisin). While Pseudomonas and Acinetobacter are regarded as aerobic, the other genera from above are classified as facultative anaerobes (Whitman et al., 2015). The genus Enterobacter has been previously isolated from lettuce and used to reduce growth of L. monocytogenes (Francis and O’Beirne, 1998a, 1998b). Future management of the bacterial community structure may therefore still be an interesting option to control growth of pathogens like L. monocytogenes. However, L. lactis as Nisin A producer seems to be not one of it. In contrast to the findings from this study, Ioannidis and colleagues (Ioannidis et al., 2018) found other bacteria to be abundant on lettuce leaves in air MAPs after 10 d of incubation beside Pseudomonas. These included Janthinobacterium (Oxalobacteriaceae), Rhanella (Yersiniaceae) and Mycoplasma (Mycoplasmataceae). Of these three genera, the present study found only Mycoplasma in smallest quantities at day 2 and 5. It appears that apart from Pseudomonas, bacterial community structures on lettuce leaves may vary substantially not only over time as highlighted in the present study but may also be dependent on where the lettuce has been cultivated and harvested. In addition, the present study suggested that Nisin application may have an effect on the oxygen concentration in the MAPs at 4 °C. This could be explained by a reduced respiration rate of the lettuce microbiome over the control, which in turn could further affect the progression of the community structures during the 7 d storage period.

Supplementary material related to this article can be found, in the online version, at doi:https://doi.org/10.1016/j.postharvbio.2018.10. 002. References Adams, M., Smid, E., 2003. Nisin in multifactorial food preservation. In: Roller, S. (Ed.), Natural Antimicrobials for the Minimal Processing of Foods. CRC Press, Boca Raton, pp. 11–26. Allende, A., Martinez, B., Selma, V., Gil, M.I., Suarez, J.E., Rodriguez, A., 2007. Growth and bacteriocin production by lactic acid bacteria in vegetable broth and their effectiveness at reducing Listeria monocytogenes in vitro and in fresh-cut lettuce. Food Microbiol. 24, 759–766. Beaufort, A., Bergis, H., Lardeux, A.-L., Lombard, B., 2014. EURL Lm Technical Guidance Document for Conducting Shelf-Life Studies on Listeria Monocytogenes in Ready-ToEat Foods, Version 3. Boelter, J.F., Brandelli, A., 2016. Innovative bionanocomposite films of edible proteins containing liposome-encapsulated nisin and halloysite nanoclay. Colloids Surf. B Biointerfaces 145, 740–747. Burt, S., 2004. Essential oils: their antibacterial properties and potential applications in foods–a review. Int. J. Food Microbiol. 94, 223–253. Centers for Disease Control and Prevention - CDC, 2016. List of Selected Multistate Foodborne Outbreak Investigations. https://www.cdc.gov/foodsafety/outbreaks/ multistate-outbreaks/outbreaks-list.html. Cotter, P.D., Hill, C., Ross, R.P., 2005. Food microbiology: bacteriocins: developing innate immunity for food. Nat. Rev. Microbiol. 3, 777–788. Cotter, P.D., Ross, R.P., Hill, C., 2013. Bacteriocins – a viable alternative to antibiotics? Nat. Rev. Microbiol. 11, 95–105. da Silva-Malheiros, P., Daroit, D.J., Brandelli, A., 2012. Inhibition of Listeria monocytogenes in minas frescal cheese by free and nanovesicleencapsulated Nisin. Braz. J. Microbiol. 43, 1414–1418. de Arauz, L.J., Jozala, A.F., Mazzola, P.G., Vessoni Penna, T.C., 2009. Nisin biotechnological production and application: a review. Trends Food Sci. Technol. 20, 146–154. Delves-Broughton, J., Blackburn, P., Evans, R.J., Hugenholtz, J., 1996. Applications of the bacteriocin, nisin. Antonie Leeuwenhoek 69, 193–202. Erlacher, A., Cardinale, M., Grosch, R., Grube, M., Berg, G., 2014. The impact of the pathogen Rhizoctonia solani and its beneficial counterpart Bacillus amyloliquefaciens on the indigenous lettuce microbiome. Front. Microbiol. 5, 175. European Parliament and of the Council, 2008. Regulation (EC) No 1333/2008 of 16 December 2008 on food additives. Annex II. FAOSTAT - Food and Agriculture Organization of the United Nations (FAO), 2017. http:// www.fao.org/faostat/en/#data (selection criteria: fruit, vegetables, all countries, production quantity, 2011-2016). Fox, A., Kwapinski, W., Griffiths, B.S., Schmalenberger, A., 2014. The role of sulfur and phosphorus mobilizing bacteria in biochar induced growth promotion of Lolium perenne. FEMS Microbiol. Ecol. 90, 78–91. Francis, G.A., O’Beirne, D., 1997. Effects of gas atmosphere, antimicrobial dip and temperature on the fate of Listeria innocua and Listeria monocytogenes on minimally processed lettuce. Int. J. Food Sci. Technol. 32, 141–151. Francis, G.A., O’Beirne, D., 1998a. Effects of storage atmosphere on Listeria monocytogenes and competing microflora using a surface model system. Int. J. Food Sci. Technol. 33, 465–476. Francis, G.A., O’Beirne, D., 1998b. Effects of the indigenous microflora of minimally processed lettuce on the survival and growth of Listeria innocua. Int. J. Food Sci. Technol. 33, 477–488. Francis, G.A., Thomas, C., O’Beirne, D., 1999. The microbiological safety of minimally processed vegetables. Int. J. Food Sci. Technol. 34, 1–22. Garg, N., Churey, J.J., Splittstoesser, D.F., 1990. Effect of processing conditions on the microflora of fresh-cut vegetables. J. Food Prot. 53, 701–703. Harris, L.J., Farber, J.N., Beuchat, L.R., Parish, M.E., Suslow, T.V., Garrett, E.H., Busta, F.F., 2003. Outbreaks associated with fresh produce: incidence, growth, and survival of pathogens in fresh and fresh-cut produce. Comprehen. Rev. Food Sci. Food Safety 2, 78–141. Ioannidis, A.G., Kerckhof, F.M., Riahi Drif, Y., Vanderroost, M., Boon, N., Ragaert, P., De Meulenaer, B., Devlieghere, F., 2018. Characterization of spoilage markers in modified atmosphere packaged iceberg lettuce. Int. J. Food Sci. Technol. 279, 1–13. Jung, D.-S., Bodyfelt, F.W., Daeschel, M.A., 1992. Influence of fat and emulsifiers on the efficacy of nisin in inhibiting Listeria monocytogenes in fluid milk. J. Dairy Sci. 75, 387–393. Leong, D., Alvarez-Ordonez, A., Jordan, K., 2014. Monitoring occurrence and persistence of Listeria monocytogenes in foods and food processing environments in the Republic of Ireland. Front. Microbiol. 5, 436. Leong, D., NicAogáin, K., Luque-Sastre, L., McManamon, O., Hunt, K., Alvarez-Ordóñez, A., Scollard, J., Schmalenberger, A., Fanning, S., O’Byrne, C., Jordan, K., 2017. A 3year multi-food study of the presence and persistence of Listeria monocytogenes in 54 small food businesses in Ireland. Int. J. Food Sci. Technol. 249, 18–26. McManamon, O., Scollard, J., Schmalenberger, A., 2017. Inoculation density is affecting growth conditions of Listeria monocytogenes on fresh cut lettuce. World J. Microbiol. Biotechnol. 33, 217. Oliveira, M., Abadias, M., Colas-Meda, P., Usall, J., Vinas, I., 2015. Biopreservative methods to control the growth of foodborne pathogens on fresh-cut lettuce. Int. J.

5. Conclusion Application of Nisin A to lettuce at concentrations widely used in the food sector was effective in reducing the abundance of L. monocytogenes by at least 90% and to delay growth by about 2 d. Since Nisin A had only limited effects on the sensory appearance of lettuce, its application could have commercial potential where chlorine dipping is not allowed legally or not desired by the consumers. Further optimizations of storage conditions as well as the application of encapsulated Nisin may have the potential to extend the shelf-life of iceberg lettuce further. During storage, the bacterial community on the lettuce leaves is constantly changing with Enterobacteriaceae abundant at day 0, Streptococcaceae and Lactobacillaceae most abundant at days 2 and 5, respectively, and Pseudomonadaceae being most abundant overall. This constantly changing environment makes it challenging for anti-pathogenic biological agents to be effective. This situation is exacerbated by the high cross-experimental variability of the bacterial lettuce community. Therefore, further studies on the lettuce microbiomes and how to manage them more effectively are required.

Acknowledgements We would like to thank the Irish Department of Agriculture, Food and the Marine and Safe Food Ireland for financing this study (FIRM 11F008).

194

Postharvest Biology and Technology 147 (2019) 185–195

O. McManamon et al.

Schmalenberger, A., Hodge, S., Bryant, A., Hawkesford, M.J., Singh, B.K., Kertesz, M.A., 2008. The role of Variovorax and other Comamonadaceae in sulfur transformations by microbial wheat rhizosphere communities exposed to different sulfur fertilization regimes. Environ. Microbiol. 10, 1486–1500. Scollard, J., Francis, G.A., O’Beirne, D., 2013. Some conventional and latent anti-listerial effects of essential oils, herbs, carrot and cabbage in fresh-cut vegetable systems. Postharvest Biol. Technol. 77, 87–93. Scollard, J., McManamon, O., Schmalenberger, A., 2016. Inhibition of Listeria monocytogenes growth on fresh-cut produce with thyme essential oil and essential oil compound verbenone. Postharvest Biol. Technol. 120, 61–68. Siroli, L., Patrignani, F., Serrazanetti, D.I., Vannini, L., Salvetti, E., Torriani, S., Gardini, F., Lanciotti, R., 2016. Use of a nisin-producing Lactococcus lactis strain, combined with natural antimicrobials, to improve the safety and shelf-life of minimally processed sliced apples. Food Microbiol. 54, 11–19. Whitman, W.B., Rainey, F., Kämpfer, P., Trujillo, M., Chun, J., DeVos, P., Hedlund, B., Dedysh, S., 2015. Bergey’s Manual of Systematics of Archaea and Bacteria. John Wiley & Sons. Wu, C., Wu, T., Fang, Z., Zheng, J., Xu, S., Chen, S., Hu, Y., Ye, X., 2016. Formation, characterization and release kinetics of chitosan/γ-PGA encapsulated nisin nanoparticles. RSC Adv. 6, 46686–46695. Xu, W., Qu, W., Huang, K., Guo, F., Yang, J., Zhao, H., Luo, Y., 2007. Antibacterial effect of grapefruit seed extract on food-borne pathogens and its application in the preservation of minimally processed vegetables. Postharvest Biol. Technol. 45, 126–133. Zhang, S., Farber, J.M., 1996. The effects of various disinfectants against Listeria monocytogenes on fresh-cut vegetables. Food Microbiol. 13, 311–321.

Food Microbiol. 214, 4–11. Ölmez, H., Kretzschmar, U., 2009. Potential alternative disinfection methods for organic fresh-cut industry for minimizing water consumption and environmental impact. Food Sci. Technol. 42, 686–693. Omac, B., Moreira, R.G., Castillo, A., Castell-Perez, E., 2015. Growth of Listeria monocytogenes and Listeria innocua on fresh baby spinach leaves: effect of storage temperature and natural microflora. Postharvest Biol. Technol. 100, 41–51. Pongtharangkul, T., Demirci, A., 2004. Evaluation of agar diffusion bioassay for nisin quantification. Appl. Microbiol. Biotechnol. 65, 268–272. Ragaert, P., Devlieghere, F., Debevere, J., 2007. Role of microbiological and physiological spoilage mechanisms during storage of minimally processed vegetables. Postharvest Biol. Technol. 44, 185–194. Randazzo, C.L., Pitino, I., Scifò, G.O., Caggia, C., 2009. Biopreservation of minimally processed iceberg lettuces using a bacteriocin produced by Lactococcus lactis wild strain. Food Control 20, 756–763. Rastogi, G., Tech, J.J., Coaker, G.L., Leveau, J.H.J., 2010. A PCR-based toolbox for the culture-independent quantification of total bacterial abundances in plant environments. J. Microbiol. Methods 83, 127–132. Rico, D., Martín-Diana, A.B., Barat, J.M., Barry-Ryan, C., 2007. Extending and measuring the quality of fresh-cut fruit and vegetables: a review. Trends Food Sci. Technol. 18, 373–386. Rudi, K., Flateland, S.L., Hanssen, J.F., Bengtsson, G., Nissen, H., 2002. Development and evaluation of a 16S ribosomal DNA array-based approach for describing complex microbial communities in ready-to-eat vegetable salads packed in a modified atmosphere. Appl. Environ. Microbiol. 68, 1146–1156.

195