Abstract
Listeria monocytogenes is regularly isolated from food processing environments and is endemic in some facilities. Bacteriophage have potential as biocontrol strategies for L. monocytogenes. In this study, the lytic capacity of a commercial Listeria phage cocktail was evaluated against a library of 475 Listeria spp. isolates (426 L. monocytogenes and 49 other Listeria spp.) with varied genotypic and phenotypic characteristics. The lytic capacity of the Listeria phages was measured by spot assays where lysis was scored on a scale of 0–3 (0 = no lysis; 1 = slight lysis; 2 = moderate lysis; 3 = confluent lysis). Only 5% of all tested Listeria spp. isolates, including L. monocytogenes, were either moderately or highly susceptible (score 2 or 3) to lysis by Listeria phage when scores were averaged across temperature and phage concentration; 155 of 5700 treatment (multiplicity of infection [MOI] and temperature) and characteristic (genotype, sanitizer tolerance, and attachment capacity) combinations resulted in confluent lysis (score = 3). Odds ratios for susceptibility to lysis were calculated using multinomial logistic regression. The odds of susceptibility to lysis by phage decreased (p < 0.05) if the L. monocytogenes isolate was previously found to persist or if the phage-bacteria culture was incubated at 30°C; neither isolate persistence or temperature was significant (p ≥ 0.05) when all factors were considered. In addition, lytic efficacy varied (p < 0.05) among pulse field gel electrophoresis (PFGE) pulsotypes and may be affected by host MOI (p < 0.05). There was no effect (p > 0.05) of attachment capacity or sanitizer tolerance on phage susceptibility. This study underscores the complexity of using Listeria phage as a biocontrol for Listeria spp. in food processing facilities and highlights that phage susceptibility is most greatly impacted by genotype. Further studies are needed to evaluate these findings within a processing environment.
Introduction
L
Listeria spp. may persist owing to varying biofilm-forming capacity conferring increased resistance to environmental stressors and commonly utilized cleaning and sanitation protocols compared with planktonic isolates (Lou and Yousef, 1996; Lundén et al., 2003; Soni and Nannapaneni, 2010; Fagerlund et al., 2017; Sadekuzzaman et al., 2017; Rodríguez-Campos et al., 2019). Bacteriophage are a promising biocontrol strategy against biofilms and have potential to overcome some of the challenges associated with Listeria control (Moye et al., 2018).
ListShield™ was the first bacteriophage cocktail approved for use in food processing environments by the Food and Drug Administration (FDA; Silver Spring, MD) and Environmental Protection Agency (EPA; Washington, DC; Gutiérrez et al., 2017). The FDA has since approved additional bacteriophage products for application to foods to control pathogen contamination, including SalmoFresh™, EcoShield™, and Listex P100™ (Sadekuzzaman et al., 2017). Recent studies have evaluated the efficacy of single and multi-bacteriophage cocktails against L. monocytogenes in ready-to-eat foods (Gutiérrez et al., 2017; Henderson et al., 2019), on food contact surfaces (FCS) (Gutiérrez et al., 2017; Vongkamjan et al., 2017), and against biofilms on foods and the environment (Sadekuzzaman et al., 2017; Rodríguez-Melcón et al., 2018). These studies have found that under the appropriate conditions for each food product and environment tested, both single and multi-bacteriophage preparations can be effective control strategies against Listeria spp. contamination, specifically L. monocytogenes.
However, there are limited data on the efficacy of commercially available Listeria phage against a wide range of L. monocytogenes and other Listeria spp. strains, as well as the potential for Listeria phage as a NFCS control strategy. In addition, to the best of our knowledge, there are no studies evaluating the effects of using a multiphage cocktail against Listeria isolates with varying phenotypes and genotypes. Therefore, the aim of our study was to measure the lytic capacity of a cocktail of six Listeria phages against a library of 475 Listeria spp. isolated from FCS and NFCS in retail delicatessen environments in an effort to better determine whether phage application could limit Listeria environmental contamination.
Materials and Methods
Listeria spp. isolates used this study
A panel of 426 L. monocytogenes and 49 other Listeria spp. isolates from retail deli environments (Supplementary Table S1) was assembled from previous studies (Simmons et al., 2014; Wang et al., 2015; Etter et al., 2017; Assisi et al., 2020). Among the isolates were 41 L. monocytogenes pulse field gel electrophoresis (PFGE) types, isolates previously defined as persistent (n = 320; isolated at least three times from the same store), transient isolates (n = 91), isolates with increased sanitizer tolerance (n = 47), and isolates with enhanced attachment capacity (n = 99).
Listeria phages
The Listeria phage cocktail has FDA approval for use as food additive in the United States and consists of six commercially available Listeria phages (ListShield; Intralytix, Inc., Baltimore, MD). The stock phage solution concentration was 109 plaque-forming unit per milliliter (PFU/mL).
Listeria phage spot assay
Lytic capacity of the phage cocktail was tested against 475 Listeria spp. isolates through spot assay as described by Vongkamjan et al. (2013). In brief, overnight broth cultures of each strain were prepared by inoculating a single colony into 5 mL Luria Bertani broth buffered with morpholinepropanesulfonic acid at 5% (LB MOPS, pH 7.6; Fisher Scientific, Fair Lawn, NJ), then incubated for 18 h with shaking (30°C, 220 rpm).
Cultures were diluted 10-fold in LB MOPS supplemented with 40% glucose (w/v; LMG; Fisher Scientific) and grown (30°C) to early log phase. A 100 μL aliquot of culture was combined with 67.5 μL CaCl2 and an overlay of LMG supplemented with agar at 15% (w/v). Serially diluted Listeria phage (in phosphate-buffered saline) was spotted on to LMG agar in duplicate to result in final treatment concentrations of 107, 106, 105, 104, 103, and 102 PFU/mL. These application concentrations are reported here as 10 × , 1 × , 0.1 × , 0.01 × , 0.001 × , and 0.0001 × , with 1 × (106 PFU/mL) being the suggested application concentration for the Listeria phage cocktail.
LMG plates were incubated at room temperature (∼21.5°C) or 30°C for 24 h. Spot assays of each isolate were tested in technical duplicates for each temperature and replicated on three different days (six total assays per isolate). Lysis was scored on a scale of 0–3, where 0 = no lysis, 1 = partial lysis with no lysis in center, 2 = ring of confluent lysis around lawn in center (moderate lysis), and 3 = confluent lysis. Average lytic capacity scores for individual isolates were averaged across incubation temperature and phage application for each isolate, resulting in continuous values for average counts; therefore, half values were used as the cutoff for scores (i.e., 0–0.5 = no lysis, 0.5–1.5 = partial lysis, 1.5–2.5 = marginal lysis, 2.5–3 = confluent lysis). Listeria monocytogenes LM94 was used as a reference and control.
Statistical analyses
Multinomial logistic regression was used to determine odds ratio estimates for phage susceptibility (spot assay score) based on persistence versus transience, attachment capacity, sanitizer tolerance, and incubation temperature. The odds ratio estimate is the ratio of odds of having a specific spot assay score (0, 1, and 2) given the odds of having a complete lysis score of 3. Odds ratios >1.0 indicate that there is a greater likelihood (e.g., odds ratio 1.7 = 1.7 times more likely) of that particular score occurring given the odds of having complete lysis of the cell (i.e., 0 vs. 3) for the tested parameter; however, odds ratios <1.0 indicate a decreased likelihood of that particular score occurring in the tested parameter. In this study, three comparisons were made: (1) score 0 versus 3, (2) score 1 versus 3, and (3) score 2 versus 3.
A single-factor analysis of variance (ANOVA) model was conducted to determine the impact of MOI and PFGE type with individual isolates (n = 475) considered a random effect. Finally, a factorial ANOVA was conducted to study the combined effect of phenotypic and genotypic characteristics on lytic capacity of Listeria phage, with fixed effects of species type (L. monocytogenes and other Listeria spp.), PFGE type, persistence, sanitizer tolerance, and attachment capacity. All statistical analyses were performed using SAS 9.4 software (Cary, NC).
Results and Discussion
Spot assays revealed that 60.6% Listeria spp. isolates were at least partially susceptible to lysis by Listeria phage
Spot assay scores for 475 isolates are given in Table 1. Taken together, 60.6% (288/475) of isolates (L. monocytogenes and other Listeria spp.) were susceptible (score >0.5) to Listeria phage treatment (Table 1); however, 91.7% (264/288) of spot assays resulted in scores between 0.5 and 1.5, indicating that most Listeria isolates were only partially susceptible to the phage treatment.
Spot Assay Results for Each Listeria Phage Treatment Concentration and Temperature Tested
Incubation temperature of the plates with Listeria spp. isolates and phage treatment.
Concentration of the Listeria phage treatments and corresponding label use concentration.
Number of isolates within in each temperature and phage concentration resulting in a specific spot assay score.
Total test conditions (isolate, temperature, and concentration).
PFU, plaque-forming unit; RT, room temperature.
Listeria monocytogenes isolates were less susceptible to phage treatment than the other Listeria spp. isolates. Approximately 58% (247/426) of the tested L. monocytogenes isolates and 34.7% (17/49) of the other Listeria spp. isolates had average spot assay scores of 0.5–1.5 (partial lysis), whereas only 2.1% (9/426) of the L. monocytogenes and 30.6% (15/49) other Listeria spp. had average spot assay scores of 1.5–2.5 (moderate lysis). However, 39.9% (179/426) of L. monocytogenes and 34.7% (17/49) of other Listeria spp. were resistant to the Listeria phage with average spot assay scores of 0–0.5. The phage treatment did not produce confluent lysis (score 3) in any of the isolates tested when scores were averaged across phage concentration and incubation temperature.
Previous studies have evaluated lytic capacity of different Listeria phage against Listeria spp. Guenther et al. (2009) reported reductions up to 5-log units in animal-based ready-to-eat (RTE) products (i.e., hot dogs, sliced turkey, and smoked salmon), concluding these bacteriophages could be an effective biocontrol strategy against L. monocytogenes. In contrast, Kim et al. (2008) reported that 9 of 12 phage were effective against all tested Listeria spp., including L. monocytogenes. The other three phages were less effective against all isolates they were tested against, which is more consistent with this study.
There were no susceptibility differences to Listeria phage between persistent and transient L. monocytogenes isolates
Spot assay scores were compared among 320 persistent L. monocytogenes isolates and 91 transient isolates. The odds of L. monocytogenes isolates previously identified as persistent having a spot assay score of 0, 1, and 2 were 1.1, 1.0, and 1.2, respectively (Table 2); these odds scores were not statistically significant (p ≥ 0.05). Persistence was not a significant effect when scores were averaged across phage concentration and temperature in an ANOVA model. Persistence data were not available for the other Listeria spp. isolates and therefore was not included in this analysis.
Odds Ratio Estimates for Multinomial Regression Analysis of Spot Assay Data
Listeria spp. isolate effect evaluated for susceptibility to lysis by Listeria phage using spot assay across all isolates.
Point estimates of the ratio of odds of having a score (0, 1, 2) when the effect is compared with the odds of having a score of 3.
Control of persistent L. monocytogenes strains has been extensively studied (Holah et al., 2002; Romanova et al., 2002; Lundén et al., 2003; Leong et al., 2016; Rodríguez-Campos et al., 2019; Ochiai et al., 2020). A previous study by Vongkamjan et al. (2013) observed significant susceptibility differences in 50 persistent L. monocytogenes strains to 28 different Listeria phages isolated from food processing and dairy production facilities. These studies suggest L. monocytogenes strains with persistent phenotypes are no more susceptible to phage lysis compared with transient strains.
Temperature did not significantly impact susceptibility to Listeria phage
There was a trend between temperature (p = 0.0818) on lysis score when scores were averaged across all phage concentrations and genotypic and phenotypic characteristics; however, this trend was not statistically significant. When incubation temperature was changed from 21.5 to 30°C, the odds of lysis score being 0, 1, and 2 (vs. 3) were 1.5, 1.7, and 1.3, respectively, suggesting that phage lysis capacity may be reduced at higher temperatures (Table 2). Temperature had a significant effect when comparisons of score 0 versus 3 (p = 0.0135) and score 1 versus 3 (p = 0.0011) were considered.
Guenther et al. (2009) reported similar results for L. monocytogenes susceptibility to Listeria phages evaluated at different storage temperatures over a 13-d storage period and found lytic capacity was increased at 20 versus 6°C. In addition, Kim and Kathariou (2009) reported that plaques formed in epidemic clone II strains grown at 37°C, but these strains were resistant to phage treatment at lower temperatures, indicating that higher incubation temperature had a positive impact on the effectiveness of these Listeria phage on L. monocytogenes. However, the temperatures utilized in this study were <37°C (30 and 21.5°C), suggesting that an inability for the Listeria phage to produce confluent lysis should be expected. Therefore, the refrigerated storage and processing temperatures commonly utilized in meat processing may reduce the efficacy of Listeria phage as a biocontrol against L. monocytogenes.
Listeria monocytogenes isolates with evidence of increased attachment capacity for biofilm formation were more susceptible to lysis by Listeria phage
Isolates with increased attachment capacity (n = 99) were marginally more susceptible to lysis; however, 1 versus 3 was the only significant odds score (p = 0.0341). The odds of isolates with increased attachment capacity producing spot assay scores of 0, 1, and 2 (vs. 3) were 0.3, 0.2, and 0.2, respectively. Isolates with increased attachment capacity were much less likely to result in a score of 0, 1, or 2 (vs. 3) than isolates with weak attachment capacity (Table 2). In the ANOVA model, with all other factors included, attachment capacity was not considered a significant determinant of spot score (p = 0.2185).
To our knowledge, there are no studies comparing the efficacy of Listeria phage against Listeria spp. biofilm and planktonic cell cultures. However, in a previous study, the authors (Soni and Nannapaneni, 2010) reported a significant reduction (5.4 log/cm2) of L. monocytogenes cells in biofilms collected from stainless steel coupons, compared with the controls, irrespective of strain phenotype. Similar results have been reported using bacteriophages against biofilms of Staphylococcus aureus (Cha et al., 2019) and Salmonella spp. (Islam et al., 2019).
Increased sanitizer tolerance is associated with reduced susceptibility to lysis
The odds of sanitizer tolerant isolates (n = 47) having a lysis score being 0, 1, and 2 (vs. 3) were 3.2, 3.3, and 3.5 (Table 2), respectively, when compared with sanitizer sensitive isolates (n = 91). However, sanitizer tolerance was significant for lysis scores in all three comparisons; 0 versus 3 (p = 0.060), 1 versus 3 (p = 0.056), and 2 versus 3 (p = 0.047). Factorial ANOVA results suggested that sanitizer tolerance did not have a significant effect (p = 0.1999) on the lysis of L. monocytogenes when all other effects were included in the model. Whereas isolates with evidence of sanitizer tolerance were less susceptible to phage lysis, this also suggests that a combination of phenotypic traits (sanitizer resistance, attachment capacity, persistence, etc.) have a greater impact on lytic capacity. It has been suggested that frequent, widespread use of disinfectants and sanitizers has lead to increased resistance among many L. monocytogenes strains (Lou and Yousef, 1996; Romanova et al., 2002; Lundén et al., 2003; Heir et al., 2004).
MOI had a significant effect on lysis when other predictors were included in the model
Six concentrations of Listeria phage were evaluated for efficacy against the Listeria spp. isolates tested here. Phage concentration was a significant factor in both the single factor model (p < 0.001) and factorial model (p < 0.001). Least-squares means of the lysis scores for the main effect of Listeria phage concentration averaged across isolate and temperature are given in Table 3. These results suggest the 1 × and 10 × treatments had the greatest efficacy against the Listeria spp. isolates (p < 0.05). MOIs lower than 1 × resulted in average lysis scores <1.0, and therefore can be concluded the efficacy of these concentrations against a broad host range of Listeria spp. were minimal (score = 1). In addition, of the 155 treatment combinations resulting in confluent lysis (score = 3), only 14 were observed with MOIs <1 × .
Least-Squares Means for Average Spot Assay Score Averaged Across Temperature and Isolate for Each Phage Concentration Tested
Least-squares means without a common superscript letter differ (p < 0.05).
PFU, plaque-forming unit.
Guenther et al. (2009) reported that higher MOIs resulted in greater reductions of L. monocytogenes on experimentally contaminated processed meats and cheeses. Rodríguez-Melcón et al. (2018) reported the greatest reduction in biofilm biomass after application of bacteriophage with a concentration of 108 PFU/mL compared with other concentrations tested (100–107 PFU/mL). These studies suggest that higher concentrations of Listeria phage are more effective in controlling Listeria spp. contamination in food and food processing environments than lower concentrations.
Lytic efficacy of Listeria phage treatment varies across L. monocytogenes genotypes
A total of 41 different L. monocytogenes PFGE pulsotypes were included in the factorial ANOVA model. PFGE type had a significant effect (p < 0.05) on the lytic capacity of tested Listeria phages, indicating phage susceptibility differed across L. monocytogenes genotypes (Supplementary Table S1). However, least-squares means for the tested PFGE pulsotypes indicated that the highest lysis score for an individual strain was 1.6 when all other factors were considered in the calculation. This indicates that the tested Listeria phage were only able to achieve partial lysis across all genotypes tested, suggesting the tested bacteriophage may not have broad host range against L. monocytogenes. Finally, factorial ANOVA results suggest that PFGE pulsotype had a significant effect (p = 0.0013) when all tested factors were considered in the model, indicating that different L. monocytogenes genotypes can have varying susceptibility to Listeria phage treatment.
Previous studies have suggested that different Listeria spp. strains have varying susceptibility and resistance to bacteriophages (Sword and Pickett, 1961; Kim et al., 2008; Denes et al., 2015). Kim et al. (2008) examined the susceptibility of L. monocytogenes, L. innocua, L. ivanovii, and/or L. welshimeri to a set of phages. They found no differences (p ≥ 0.05) in lysis among L. monocytogenes isolates of serotypes previously identified as susceptible to bacteriophages. Denes et al. (2015) reported that differences in host gene expression is a likely source of phage resistance in L. monocytogenes strains; therefore, further studies evaluating the host gene expression patterns may be necessary to further understand if broad host range can be reached for the tested Listeria phage cocktail.
Conclusion
This study establishes that phage susceptibility of Listeria spp. may be affected by multiple factors, including potential persistence capabilities of each isolate, enhanced attachment capacity, tolerance to sanitizers, and genotypic characteristics. The six-phage cocktail utilized here may not have sufficient lytic capacity against a broad host range of Listeria spp. isolated from retail deli environments as tested. Further studies involving application of Listeria phage within processing environments, accounting for environmental effects, temperature variation, and phage concentration should be conducted to evaluate true efficacy of these Listeria phage as a control strategy in food processing facilities. Although many isolates in this study were previously sequenced, future work could include comprehensive whole genome sequence analyses to further understand characteristics driving phage susceptibility.
Footnotes
Authors' Contributions
All authors have participated in the conception and design, or analysis and interpretation of the data, drafting the article or revising it critically for important intellectual content, and approval of the final version.
Disclosure Statement
No competing financial interests exist.
Funding Information
The data reported in this communication were supported, in part, by the Proctor and Gamble Company. Dr. Oliver is supported by the USDA National Institute of Food and Agriculture Hatch project 2016-67017-24459. Dr. Ebner is supported by the USDA National Institute of Food and Agriculture Hatch project no. IND010930.
Supplementary Material
Supplementary Table S1
References
Supplementary Material
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
