Journal of Food Protection 86 (2023) 100104 Contents lists available at ScienceDirect Journal of Food Protection journal homepage: www.elsevier .com/ locate / j fp Research Paper Substituting Allose as the Primary Carbon Source During Enrichment Helps Improve Detection and Isolation of Lineage II Listeria monocytogenes From Food https://doi.org/10.1016/j.jfp.2023.100104 Received 19 December 2022; Accepted 8 May 2023 Available online 11 May 2023 0362-028X/Crown Copyright © 2023 Published by Elsevier Inc. on behalf of International Association for Food Protection. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/). ⇑ Corresponding author. E-mail address: jacqueline.upham@inspection.gc.ca (J.P. Upham). Jacqueline P. Upham 1,⇑, Mikaela Eisebraun 1, Alex Fortuna 2, Gustavo V. Mallo 3 1Canadian Food Inspection Agency - Dartmouth Laboratory, Dartmouth, Nova Scotia, Canada B3B 1Y9 2Ontario Institute for Cancer Research, Toronto, Ontario, Canada M5G 0A3 3Pathogen Preparedness and Test Development Unit, Public Health Ontario Laboratories, Toronto, Ontario, Canada M5G 1M1 A R T I C L E I N F O Keywords: Allose Lineage II Listeria monocytogenes lmo0734‐lmo0739 Smoked salmon A B S T R A C T Testing of foods for low levels of the human pathogen, Listeria monocytogenes (Lm), involves a selective enrich- ment procedure. A nonpathogenic species of Listeria, L. innocua (Li), is often present in foods and food‐ manufacturing environments and is an interference organism for Lm detection due to competition during enrichment. The present study investigated whether a novel enrichment strategy incorporating the sugar allose into the secondary enrichment broth (allose method) could improve the detection of Lm from foods when Li is present. First, Canadian food isolates of Listeria spp. were tested to confirm recent reports that lineage II Lm (LII‐Lm), but not Li, could metabolize allose. All LII‐Lm isolates (n = 81), but not Li (n = 36), possessed the allose genes lmo0734‐lmo0739, and could efficiently metabolize allose. Next, smoked salmon was contam- inated with mixtures of LII‐Lm and Li and tested using different enrichment procedures to compare the ability to recover Lm. Allose broth was more effective than Fraser Broth, with Lm detected in 87% (74 of 85) com- pared to 59% (50 of 85) of the samples (P < 0.05), following a common preenrichment. When evaluated against a current Health Canada method (MFLP‐28), the allose method was more effective, with LII‐Lm detected in 88% (57 of 65) compared to 69% (45 of 65) of the samples (P < 0.05). The allose method also remarkably increased the ratio of LII‐Lm to Li postenrichment, which improved the ease of obtaining isolated Lm colonies for confirmation tests. Allose may therefore provide a tool for use when the presence of back- ground flora interferes with Lm detection. As this tool is specifically applicable to a subset of Lm, the use of this method modification may provide a working example of tailoring methodology to target the known sub- type of the pathogen of interest in an outbreak investigation, or for regular monitoring activities in conjunction with a PCR screen for allose genes on preenrichment cultures. Food safety regulations in Canada require an absence of the food- borne pathogen Listeria monocytogenes (Lm) per analytical unit of ready‐to‐eat foods that are considered high‐risk (Health Canada, 2011), an approach broadly accepted internationally. Detection of such low levels of the pathogen requires a multiday selective enrich- ment procedure. Listeria innocua (Li) is a nonpathogenic species of Lis- teria that is frequently found in the environment and can overgrow Lm during selective enrichment (Cornu et al., 2002; Curiale & Lewus, 1994; MacDonald & Sutherland, 1994; Petran & Swanson, 1993). Keys et al. (2013) demonstrated that although the extent to which Li overgrew Lm was strain‐dependent, Li caused significant interference with Lm not being detected in 18 of 30 enrichment cocultures. In another example, Zittermann et al. (2016) found that when food was cocontaminated with Li and Lm, culture‐based methods failed to detect Lm postenrichment unless the initial ratio of Lm to Li was high (6.0 or higher), despite Lm being detected in all but one sample postenrichment using a sensitive real‐time PCR assay. The overgrowth of Lm by Li during enrichment has the potential to mask the presence of Lm in a sample, leading to false negative results and possible food safety risk. Recently, it was discovered that a rare but naturally occurring sugar, allose, can be used as a sole carbon source by Lm but not by Li (Liu et al., 2017). Subsequently, the cause of this metabolic differ- ence was identified; lineage II Lm (LII‐Lm) strains (serotypes 1/2a, http://crossmark.crossref.org/dialog/?doi=10.1016/j.jfp.2023.100104&domain=pdf https://doi.org/10.1016/j.jfp.2023.100104 http://creativecommons.org/licenses/by/4.0/ mailto:jacqueline.upham@inspection.gc.ca https://doi.org/10.1016/j.jfp.2023.100104 http://www.sciencedirect.com/science/journal/0362028X http://www.elsevier.com/locate/jfp J.P. Upham et al. Journal of Food Protection 86 (2023) 100104 3a, 1/2c, and 3c) possess a cassette of six genes, lmo0734‐0739, that are absent in Li and are responsible for allose metabolism (Zhang et al., 2018). Although those genes had been previously recognized as a genomic region of difference among Listeria species and lineages (Doumith, Buchrieser, et al., 2004; Doumith, Cazalet, et al., 2004; Glaser et al., 2001; Milillo et al., 2009), a specific metabolic role for them had not been identified. Zhang et al. (2018) showed that genes lmo0734; lmo0735, and lmo0736 were all required for Lm metabolism of D‐allose, while lmo0737 and lmo0738 improved the growth of Lm in the presence of allose, but were not required, and lmo0739 was dis- pensable. Based on predicted gene function and by comparison with allose operons in E. coli, it is likely that the gene cassette encodes a transcriptional regulator (lmo0734), an allulose‐6‐phosphate epimer- ase (lmo0735), an allose‐6‐phosphate isomerase (lmo0736), a kinase for phosphorylation of the allose (lmo0737), and a transporter to move allose across the membrane (lmo0738) (Zhang et al., 2018). Interest- ingly, among the lineage I Lm, multiple clades of the highly clinically relevant serotype 4b have been identified as having acquired the allose gene cassette (Burall et al., 2017; Leclercq et al., 2011). The strains are termed 4b variants and some have led to recent outbreaks of listeriosis associated with produce in the USA (Burall et al., 2017). The real‐life significance of possession of genes that confer the ability to metabolize a rare carbohydrate is unclear, but this phenotype may aid in the sur- vival of those strains in particular niches or in harsh environments where nutrient availability is limited. Differential possession of the metabolic pathway for allose may be taken advantage of to change the selectivity of a culture‐based method, and this was reported by Liu et al. (2017), prior to the discovery that allose metabolism by Lm is lineage‐specific. An enrichment method that only selects a subpopulation of the target species would obviously not be a suitable general replacement for current regulatory methods. However, such a method may be of great benefit as a tailored approach when the target subtype is known, for example during an outbreak or trace‐back investigations, or by the inclusion of a subtype‐specific screening step during enrichment. This method modification may also be useful to provide cultural confirmation in method validation studies using artificially contaminated samples when the molecular tests used are more sensitive than the reference method due to interference organisms. The present study aimed to use a collection of Canadian Listeria spp. food isolates to confirm previous findings that LII‐Lm, but not typical lineage I Lm or Li, could grow in the presence of allose as the primary carbon source, and that the phenotype is associated with possession of the lmo0734‐0739 gene cassette. Using smoked salmon as the food model, the objective was then to test whether a novel enrichment strat- egy incorporating an allose‐containing broth could serve as a tool to enable the detection and isolation of LII‐Lm when current culture methods fail due to interference by Li. Materials and Methods Media. Broth used in growth curves was prepared as described by Liu et al. (2017) with 10 g/L tryptone, 10 g/L peptone (both from Difco, BD, Mississauga, Ontario, Canada), 5 g/L lab‐lemco powder (Oxoid Ltd.), 5 g/L sodium chloride, 2.5 g/L disodium hydrogen phos- phate (both Fisherbrand, Fisher Scientific), and 2.5 g/L of either D‐ allose or D‐glucose, at pH 7.2 ± 0.1. D‐allose (CAS Number 2595‐ 97‐3) was purchased from BOC Sciences and was prepared as a filtered 20% aqueous stock solution. D‐glucose (FisherBrand) was prepared as described for D‐allose. For smoked salmon experiments, a Listeria selective broth containing allose (‘Lis‐A’) was prepared as described above, except with 0.2% allose and with the addition of Listeria selec- tive agents in the form of Oxoid supplement SR0141E (50 mg/L cyclo- heximide, 40 mg/L nalidixic acid, and 15 mg/L acriflavine). UVM, Fraser Broth, and Fraser Broth supplement were all from Difco, and 2 MOPS‐BLEB was prepared using Oxoid CM0862 base, with MOPS com- ponents (Sigma‐Aldrich, St. Louis, MO), and Listeria selective agents (SR0141E, Oxoid), as described (Health Canada, 2011). Rapid L’Mono (RLM) agar was supplied by Bio‐Rad Laboratories. Bacteria. This study used Listeria strains from the CFIA Dart- mouth Laboratory that had been isolated from ready‐to‐eat food products or from food‐processing environments between 1996 and 2018. For growth assessments in allose broth, the collection of 126 CFIA isolates was supplemented with 49 food isolates from the Public Health Ontario Laboratories collection and 19 isolates from the American Type Culture Collection and the former Interna- tional Life Sciences Institute collection maintained by the Food Safety Laboratory at Cornell University (see Supplemental Tables S1.1 and S1.2). The isolates used for smoked salmon experiments are described in Table 1. Traditional serotypes were not available for all isolates, so a predicted serotype was assigned based on the presence of serotype‐specific genes, as described by Doumith, Buchrieser, et al. (2004). To determine which isolates possessed the genes associated with allose metabolism, lmo0734‐lmo0739, gen- ome sequences were searched based on NCBI reference genome NC_003210.1 positions 764465‐765469, 765683‐766354, 766351‐ 766797, 766810‐767742, 767766‐769619, and 769639‐771012 (Altschul et al., 1990). Growth curves. An individual colony of each bacterial isolate was grown in Brain Heart Infusion broth (Oxoid) at 35°C for 22 ± 2 h and then diluted 1:100 into broth containing either 0.25% allose or glucose for growth curves. Growth was assessed at 35°C for 10 h in a CLARIOs- tar microplate reader with OD600 readings taken by orbital averaging every 30 min after 30 s of shaking. Each 96‐well microplate (Costar 3370, Corning Inc.) contained five replicate wells of each Listeria iso- late per broth type, and the blank‐corrected readings of replicates were averaged. Growth curves were performed three independent times, and the average and standard deviation of the three trials were graphed. Smoked salmon experiments. Frozen, vacuum‐packed cold‐ smoked salmon was purchased from local retail outlets (Dartmouth), thawed overnight under refrigeration, and weighed into Whirl‐Pak fil- tered stomacher bags (Nasco) for inoculation. Uninoculated salmon was tested in advance using a standard reference method (Pagotto et al., 2011), and Listeria species were not detected as part of the back- ground flora. Listeria inoculum was prepared by growing individual isolates overnight, as described in the section above, followed by serial dilution and enumeration. Cultures were held for 24 h at 4°C until the enumeration results were available. Then Lm and Li were combined to achieve a ratio of approximately one log greater CFU of Li than Lm in suspension. The suspension was then further diluted for inoculation of salmon. Concurrently, the combined inoculum suspensions were reenumerated by plating on RLM and Blood Agar to count phosphatidylinositol‐specific phospholipase C positive (blue) vs nega- tive (white) CFU and hemolytic vs nonhemolytic CFU, respectively, to confirm levels of both Lm and Li. To achieve reliable (20–200 CFU) counts per agar, various dilutions of the combined inoculum were pla- ted and the results were used to estimate low inoculum levels. Inocu- lum suspensions (0.1–0.2 mL) were pipetted onto the surface of the salmon in stomacher bags, the bags sealed and then returned to frozen storage (−20°C) for two weeks for inoculum acclimatization prior to testing. Salmon samples were tested by thawing overnight under refrigeration, then adding prewarmed UVM broth at a weight ratio of one part sample material to nine parts UVM broth. Samples in UVM were stomached for 120 s in an AES Smasher and then incubated at 30°C for primary enrichment for up to 48 h. For direct comparison of secondary enrichment broths, salmon samples of 10 g per replicate were used. UVM primary enrichments were incubated for 48 h, then mixed and transferred to Lis‐A (1–9 mL), Fraser Broth, and MOPS‐BLEB (0.1–9.9 mL), for secondary enrichment at 35°C for 24 h. Table 1 L. monocytogenes (Lm) and L. innocua (Li) strains used in this study for preliminary growth curves demonstrating proof-of-concept for metabolizing allose. N/A = not applicable Strain ID MLSTa Lineage Year isolated Source of isolation Growth in allose Presence of lmo0734-lmo0739 Lm 9 4 I 1996 deli meat − − Lm 16 199 II 2003 cold-smoked salmon + + Lm 28 321 II 2008 cold-smoked salmon + + Lm 29 292 II 2008 deli meat + + Lm 32 11 II 2009 Beans + + Li 8 448 N/A 1996 deli meat − − Li 10 603 N/A 1996 Cheese − − Li 17 599 N/A 2000 Fruit − − Li 38 132 N/A 2007 deli meat − − Li 66 43 N/A 2015 raw milk − − a multilocus sequence type (MLST) was based on the exact sequence matches to alleles within the international seven gene scheme maintained at http:// www.pasteur.fr/mlst (Ragon et al., 2008). J.P. Upham et al. Journal of Food Protection 86 (2023) 100104 To evaluate the use of Lis‐A in secondary enrichment as part of a whole method, it was compared with Health Canada methods, MFHPB‐30 (Pagotto et al., 2011) and MFLP‐28 (Health Canada, 2011). Salmon samples of 25 g per replicate were subjected to a com- mon primary enrichment in UVM at 30°C, which was then processed as described in MFHPB‐30 and MFLP‐28. The novel allose method (‘Allose Method’) consisted of UVM enrichment for 48 h, followed by the transfer of 1 mL UVM culture to 9 mL of Lis‐A for secondary enrichment for 24 h at 35°C. Both for direct comparison of secondary broths and for comparison of whole methods, the secondary enrich- ments were streaked to RLM agar for Lm isolation. Once isolated, Lm confirmation consisted of tests for hemolysis, motility, and sugar utilization as described in MFHPB‐30. Statistical analysis. Fisher’s exact test (Preacher, 2001) was used to compare enrichment broths for the number of samples Lm was detected from. In addition, the McNemar test for paired samples was used to compare the novel secondary enrichment (Lis‐A) to alternative secondary enrichments. P values were considered significant if they were less than 0.05. Statistical evaluation of the probability of detec- tion of Lm for each method was performed as described by Health Canada (2011). Results LII‐Lm isolates from food, but not Li or lineage I Lm, use allose as a carbon source for growth. Ten Listeria isolates (Table 1) were evaluated for ability to grow in an enrichment broth with either 0.25% glucose or 0.25% allose as the primary carbon source. Only the LII‐Lm isolates grew in the presence of allose (Fig. 1A), while all ten isolates grew well in the presence of glucose (Fig. 1B). Genome sequences of these isolates were searched for the gene cassette lmo0734‐lmo0739, recently reported to be involved in allose metabo- lism (Zhang et al., 2018), and all isolates from Table 1 that grew in allose possessed these genes, while the strains that did not grow lacked those genes. The correlation between the allose growth phenotype and the presence of genes lmo0734‐lmo0739 was further examined in a lar- ger collection (n = 175) of Canadian food isolates of Lm and Li with diverse MLSTs (listed in Table 2, further detailed in Supplementary Data Table S1.1) and was found to be consistent, without exception. All LII‐Lm isolates, which includes serotype 1/2a, the most frequently isolated serotype from foods in CFIA laboratories, grew well using allose as a carbon source. Although there have been reports of clinically relevant lineage I serotype 4b variants that have acquired the allose gene cassette (Burall et al., 2017; Leclercq et al., 2011), none of these variants were observed within the collection of strains tested here. The ability of different Listeria species to metabolize allose is associated with presence of the three genes lmo0734, lmo0735, 3 and lmo0736. An allose growth phenotype was then examined in a broader panel of isolates representing nonpathogenic Listeria species associated with foods (L. welshimeri and L. seeligeri) as well as other Lis- teria species not typically isolated from foods. The phenotype for these additional strains is summarized in Table 2, with details on strains and gene presence provided in Supplementary Table S1.2. Results showed that possession of at least the first three genes of the six gene cassette, lmo0734, lmo0735, and lmo0736, was associated with the ability to metabolize allose. L. welshimeri isolates (n = 17) grew in allose and possessed the three genes lmo0734‐lmo0736 and L. ivanovii (n = 3), primarily a pathogen of ruminants, grew in allose and possessed the first four genes in the cassette, lmo0734‐lmo0737. Isolates of L. seeligeri (n = 6) and L. grayi (n = 3) did not grow in allose and did not possess any of the six genes in the cassette. Single representative isolates of recently identified environmental species of Listeria were found not to grow in allose (Table 2, Supplementary Table S1.2). Although only lineage I and lineage II Lm are commonly found in food, a single iso- late of lineage III Lm was additionally tested and found to grow in allose and possess only the four genes lmo0734‐lmo0737 of the six. These growth experiments demonstrate that the ability to metabolize allose is not a feature unique to LII‐Lm, but varies among Listeria spe- cies and subtypes. The results obtained for genotype association with phenotype are also consistent with the mutational studies presented by Zhang et al. (2018) that demonstrate that only genes lmo0734‐ lmo0736 are essential for Listeria utilization of allose. Among a limited panel of other Gram‐positive organisms tested, the Listeria‐like organism Carnobacterium maltaromaticum was also found to grow with allose as a primary carbon source using genes other than lmo0734‐ lmo0736, indicating this phenotype is not unique to the Listeria genus (Table 2). An allose‐based secondary enrichment broth improves the iso- lation of LII‐Lm from smoked salmon. Smoked salmon samples arti- ficially contaminated with different inoculum levels of LII‐Lm combined with Li were preenriched in UVM broth and then compared in parallel with secondary enrichment in Lis‐A and Fraser Broth. Of the 40 inoculated salmon samples, Lm was isolated and confirmed from 85% (n = 34) using Lis‐A, which was significantly more than the 62% (n = 25) of samples Lm was detected in using Fraser Broth (P < 0.05 based on the Fisher exact test, Table 3). The McNemar test to compare all confirmed results of paired samples revealed a very sig- nificant difference between secondary enrichment broths (P < 0.01). In Fraser Broth, Lm was not detected in 2 of the 10 salmon samples inoculated with a relatively high Lm level of 13–14 CFU. Secondary enrichment in Lis‐A suppressed the growth of Li and promoted the growth of Lm, leading to easy isolation on the selective and differential media, Rapid L’Mono (RLM) agar (Fig. 2A). In contrast, when Lm was detected from Fraser Broth enrichments, it was in the presence of an abundance of Li (Fig. 2A), and required a number of restreaking steps Figure 1. Growth of ten Listeria food isolates in broth containing 0.25% allose (A) or 0.25% glucose (B) as the primary carbon source. Isolates, further described in Table 1, included one lineage I L. monocytogenes (Lm 9), four lineage II L. monocytogenes (Lm 16, 26, 29, and 32) and five L. innocua (Li 8, 10, 17, 38, and 66). Listeria growth was monitored by blank-corrected changes in optical density at 600 nm (OD600) in a CLARIOstar microplate reader over 10 h at 35°C. Averages and standard deviation of three independent growth experiments are shown, with data from each experiment resulting from the average of five replicate wells. Table 2 Isolates used in this study to assess the ability to metabolize allose. ‘und’ = undetermined Species Lineage Serotypea No. isolates MLSTs represented (No. isolates) Growth in allose Allose genesb L. monocytogenes II 1/2a (3a) 71 321(13), 120(9), 155(9), 204(6), 7(4), 11(4), 199(4), 31(2), 226(2), 551(2), 8(1), 16 (1), 19(1), 26(1), 37(1), 121(1), 226(1), 292(1), 360(1), 364(1), 369(1), 372(1), 376 (1), 451(1), 643(1), und(1) + + 1/2c 10 9 (10) + + I 1/2b 23 5 (21), 59(1), 330(1) − − 4b 35 4(12), 2(6), 145(5), 1(4), 6(5), 217(1), 388(1), und(1) − − III 4c 1 2073 (1) + partial L. innocua N/A N/A 36 132(9), 448(5), 542(4), 1489(3), 492(2), 599(2), 1008(2), 43(1), 493(1), 603(1), 731(1), 1010(1), 1481(1), 1598(1), CC140 (1), und(1) − − L. welshimeri 17 + partial L. seeligeri 6 − − L. ivanovii 3 + partial L. grayi 3 − − Other Listeria spp.c 9 − und Carnobacterium maltaromaticum 1 + − Enterococcus faecalis 1 − − Staphylococcus spp.d 4 − − Rhodococci equi 1 − − a For isolates that were not serotyped traditionally, serotype was predicted based on the presence or absence of serotype-associated genes (Doumith, Buchrieser, et al., 2004). b Presence of genes lmo0734-lmo0739 in available genome sequences (see Tables S1.1 and S1.2); all six genes present (+), none of the six genes present (−), three or four of the six genes present (‘partial’). c L. aquatica, L. booriae, L. cornellensis, L. fleischmanni, L. floridensis, L. marthii, L. newyorkensis, L. riparia, and L. grandensis. d S. aureus, S. epidermidis, and S. xylosus. J.P. Upham et al. Journal of Food Protection 86 (2023) 100104 4 Table 3 Comparison of the secondary enrichment broths Fraser with Lis-A for Lm isolation from cold-smoked salmon contaminated with low and high levels of LII-Lm combined with Li. Salmon inoculum No. replicates No. replicates Lm-positive LII-Lm Li Fraser Lis-A Strain CFU Strain CFU None None 5 0 0 Lm 16 0.7 Li 17 5.4 5 3 3 Lm 28 0.6 Li 10 5.1 5 1 2 Lm 16 1.4 Li 17 11 5 5 5 Lm 28 1.3 Li 10 10 5 3 5 Lm 16 2.9 Li 17 22 5 4 5 Lm 28 2.5 Li 10 20 5 1 4 Lm 16 14 Li 17 108 5 4 5 Lm 28 13 Li 10 102 5 4 5 Total 45 25 34 Figure 2. Representative pictures of Rapid L’Mono (RLM) agar streaked with secondary enrichments from salmon samples contaminated with a combination of L. monocytogenes and L. innocua. Samples were first enriched in UVM Broth followed by secondary enrichment in Fraser Broth (left), Lis-A (right), or MOPS-BLEB (middle for B and C), prior to streaking to RLM agar. Samples included salmon inoculated with (A) Lm 16 and Li 17 (see Table 3), (B) Lm 29 and Li L166 (see Table 4), and (C) Lm 29 and Li 8 (see Table 5, level ‘low-1’). On RLM agar, Lm appears blue and Li appears white, based on differential PIPLC activity. Agars were incubated for 48 h, however, the photos in (B) were taken after 24 h incubation. Lm was confirmed from all eight of the agar plates shown, with Lm from the Fraser Broth and MOPS-BLEB requiring multiple restreaking steps to isolate the Lm from the Li. An arrow near the top left agar plate image indicates a single blue colony on RLM. J.P. Upham et al. Journal of Food Protection 86 (2023) 100104 to isolate and subsequently confirm identity. It is noteworthy that the overgrowth of Li would have prevented the detection of Lm from all of the Fraser Broth cultures if Listeria selective agars were being used that did not distinguish Lm from Li, such as Oxford agar (data not shown). 5 A similar, independent, smoked salmon experiment was performed using a low inoculation level of LII‐Lm (2–4 CFU per sample) that is expected to be detectable using enrichment methods if an interference organism such as Li is not present. The Lis‐A secondary enrichment J.P. Upham et al. Journal of Food Protection 86 (2023) 100104 again allowed for Lm detection from significantly (P < 0.001) more samples than Fraser Broth, with Lm detection from 89% (40 of 45) of inoculated salmon samples using Lis‐A, compared to 56% (25 of 45) of samples using Fraser Broth (Table 4). Another secondary enrich- ment broth, MOPS‐BLEB, was included in parallel in this experiment and led to detection from 76% (34 of 45) of samples, lower than that of Lis‐A but not quite significantly so (P= 0.08, Fisher exact test). The McNemar test on confirmed results of all paired samples indicates stat- ically significant differences between Lis‐A and Fraser Broth (P < 0.001) as well as MOPS‐BLEB (P = 0.04). In addition, the ease of isolation of Lm from Lis‐A was remarkable compared to the other broths. For example, Figure 2B shows RLM agars incubated for 24 h after being streaked with secondary enrichments of salmon samples. Although all three broth cultures were eventually confirmed to be Lm‐positive, the blue growth of Lm on RLM agar from Fraser Broth is buried under the white growth of Li, and from MOPS‐BLEB, the Lm was not visible at all until the agar had been incubated for 48 h. The Lis‐A had almost exclusive growth of Lm that was readily visible by 24 h agar incubation. The Allose Method, using allose during secondary enrichment, was compared to two Health Canada Listeria detection methods from foods, MFHPB‐30 (Pagotto et al., 2011) and MFLP‐28 (Health Canada, 2011). Although each method uses primary enrichment in UVM, the MFHPB‐ 30 method uses Fraser Broth for secondary enrichment with diversions from the primary enrichment at both 24 and 48 h, as well as direct streaking of the primary enrichment, while MFLP‐28 uses MOPS‐ BLEB for secondary enrichment, but is streaked out to selective agars only if the MOPS‐BLEB screens positive for Lm by PCR. Smoked sal- mon sample replicates were artificially contaminated with low (2–3 CFU/sample) and high (9–14 CFU/sample) levels of LII‐Lm mixed Table 4 Comparison of the secondary enrichment broths Fraser, MOPS-BLEB and Lis-A for Lm and Li. Salmon inoculum N LII-Lm Li Strain CFU Strain CFU None None 5 Lm 16 2.7 Li 8 21 5 Lm 16 3.0 Li 17 22 5 Lm 16 2.3 Li L166 26 5 Lm 29 2.2 Li 8 25 5 Lm 29 1.8 Li 17 24 5 Lm 29 1.8 Li L166 22 5 Lm 30 3.8 Li 8 30 5 Lm 30 2.8 Li 17 38 5 Lm 30 2.8 Li L166 36 5 Total 5 Table 5 Assessment of an allose-based enrichment method (Allose Method) for detection of LI and MFLP-28. All Lm-inoculated smoked salmon was coinoculated with one log high Lm could be culturally confirmed, and the Probability of Detection (POD) per metho LII-Lm inoculum No. replicates Level CFU Control-1 0 5 Control-2a 0 5 Low-1 2–3 20 Low-2a 2–3 45 High 9–14 20 Total 95 a smoked salmon samples of 10 g per replicate, instead of the standard 25 g per b,c significant difference in POD (dPOD) of the Allose Method compared to MFLP-2 dPOD (95% CI) = 0.25 (0.02, 0.45) b, 0.16 (0.02, 0.29)c. 6 with one log higher CFU of Li, and tested by all three methods in par- allel. Of the 65 samples spiked at low levels, Lm was detected in 83% (n= 54) by MFHPB‐30, 69% (n= 45) by MFLP‐28, and 88% (n=57) by the Allose Method (Table 5). The probability of detection (POD) using the Allose Method was statistically equivalent to method MFHPB‐30 at each level, while the POD using the Allose Method was statistically greater (P < 0.05) than method MFLP‐28 at low Lm inoculum levels. Figure 2C shows the relative ease of LII‐Lm isolation on selective agar for each of the three methods, as demonstrated by a single representative salmon sample at the low inoculum level which was Lm‐positive by all three methods but overgrown by Li in Fraser Broth and MOPS‐BLEB. Allose in an agar format is a potential tool for improved LII‐Lm isolation. To explore an alternative way in which D‐allose could be used to improve the detection of LII‐Lm, it was incorporated into an agar format that cultures could be streaked to postenrichment. Agar was prepared by adding 0.2% D‐allose to purple agar base (Difco). Mixed culture containing approximately 3‐log higher CFU of Li than Lm was spread‐plated onto purple agar base containing allose, along- side RLM agar for comparison, and incubated at 35°C. On RLM agar, the Lm was difficult to see and isolate due to the abundance of Li, whereas on the allose agar, only Lm grew by 24 h, appearing as yellow colonies, indicating fermentation of the allose and greatly facilitating detection and isolation (Fig. 3). Discussion The presence of interference organisms in foods can greatly impact the outcome of culture‐based detection methods for bacterial patho- gens. Many cultural methods for Lm are two‐step enrichments that isolation from cold-smoked salmon contaminated with a combination of LII-Lm o. replicates No. replicates Lm-positive Fraser MOPS-BLEB Lis-A 0 0 0 4 4 4 0 1 4 2 5 5 5 5 5 2 2 2 5 3 5 5 5 5 2 4 5 0 5 5 0 25 34 40 I-Lm from cold-smoked salmon compared to Health Canada methods, MFHPB-30 er CFU of L. innocua. Results indicate the number of salmon samples from which d per level. No. replicates Lm-positive (POD) MFHPB-30 MFLP-28 Allose Method 0 0 0 0 0 0 15 (.75) 12 (.60b) 17 (.85) 39 (.87) 33 (.73c) 40 (.89) 20 (1.0) 19 (.95) 20 (1.0) 74 64 77 replicate, were used. 8, because the 95% confidence interval (CI) of the dPOD does not include zero. Figure 3. Representative images of Rapid L’Mono (RLM) agars (left) and allose agars (right) plated with mixed cultures of Listeria with approximately 3-log higher CFU Li than Lm. The strain combinations included Lm 16 with Li 8 (top agars) and Lm 16 with Li 10 (bottom agars). On RLM agar, both species growth well and Lm appears blue while Li appears white. On allose agar Lm readily grows and produces yellow colonies, while Li growth is suppressed. J.P. Upham et al. Journal of Food Protection 86 (2023) 100104 require multiple days, which provides an opportunity for overgrowth by interference organisms and even bias among target strains, through competition for resources, overgrowth by faster‐growing strains, or by inhibition from released phages or phage‐tail structures (Curiale & Lewus, 1994; Keys et al., 2013; Lemaître et al., 2015; Petran & Swanson, 1993). When new methods for Lm are being validated, cur- rent Health Canada guidelines recommended that at least one food type per food category being tested is contaminated with ten times more Li as an interference organism than the Lm target organism, to challenge the method (Canada, 2021). These recommendations were followed in the present study with smoked salmon, and using current regulatory methods (Health Canada, 2011; Pagotto et al., 2011), the Li was found to overgrow Lm to a large extent, such that Lm could not be detected in some samples (i.e. false negative results). To address these issues, allose was shown here to provide a cultural tool to reduce false negative LII‐Lm tests and/or greatly facilitate isolation of LII‐Lm. Results of incorporation of Lis‐A in the secondary enrichment step showed superiority to both Fraser Broth and MOPS‐BLEB broth, two secondary Listeria enrichment broths in food testing methods for Liste- ria. The Listeria selective agents that were chosen for inclusion in the Lis‐A are the same as those used in MOPS‐BLEB and differ from those used in Fraser Broth. Selective agents have been shown to play a role in differential growth rates of Li and Lm strains (Curiale & Lewus, 1994; MacDonald & Sutherland, 1994; Petran & Swanson, 1993), as well as differential induction of inhibitors (Lemaître et al., 2015), and therefore a likely explanation for the poor ability to detect Lm in Fraser Broth compared to MOPS‐BLEB. When whole methods were 7 compared, however, the MFHPB‐30 method that utilizes Fraser Broth for secondary enrichment performed better than the method that uses MOPS‐BLEB, MFLP‐28. This was due to the uniqueness of MFHPB‐30 in having three different stages during the procedure in which the cul- ture is streaked out for potential Lm isolation, rather than just at the endpoint (see Supplemental Table S4). Poor performance of the MFLP‐28 method was in part due to false negative PCR screening tests, with 5% (4 of 85) of the MOPS‐BLEB cultures producing negative PCR results despite being culturally positive for Lm (see Supplemental Table S4). These false negative PCR results may have been attributed to the Lm levels postenrichment being below the 104 CFU/mL level of detection of the assay, due to the presence of Li. The Allose Method greatly reduced interference associated with Li; however, it failed to allow the detection of LII‐Lm in all contaminated salmon samples. This result is likely from Li overgrowth during the 48 h primary enrichment in UVM prior to transfer to the Lis‐A. Further study would be needed to determine if reducing the primary enrich- ment time, or eliminating it in favor of a single‐step enrichment using allose, as was done by Liu et al. (2017), could improve the detection of LII‐Lm in contaminated samples. Use of a common UVM preenrich- ment was selected for the present study for the advantage it offers in allowing Lis‐A enrichment in parallel with the secondary enrichment of the regulatory method being used. In addition, it would facilitate the incorporation of a screening test of the UVM to determine whether Lm possessing the genes for allose utilization are present and therefore a candidate for secondary enrichment in Lis‐A. In the present study, a multiplex PCR assay for the detection of lmo0737, along with an Lm‐ specific marker (hlyA), and an internal amplification control, was suc- J.P. Upham et al. Journal of Food Protection 86 (2023) 100104 cessfully tested on a subset of samples as preliminary work toward proof‐of‐concept for that approach (data shown in Supplementary Table S4). Although this study with spiked salmon samples has shown how enrichment with allose as a primary carbon source can tip the balance in favor of LII‐Lm over Li during coculture, data presented in this study also revealed that the Listeria species L. welshimeri and L. ivanovii are both able to metabolize allose. Allose would therefore not be expected to improve LII‐Lm detection in samples cocontaminated with these species. Furthermore, while some non‐Listeriae Gram‐positive bacteria are unable to metabolize allose, an isolate of the Listeria‐like organism Carnobacterium maltaromaticum from smoked salmon was found to uti- lize allose efficiently and would therefore potentially compete with LII‐Lm enrichment and isolation. Further testing of other potential Lm interference organisms for an allose growth phenotype would be advantageous for understanding the limitations of the strategies pre- sented here. Another area worthy of further study is the applicability of an allose selective agar to improve the isolation of LII‐Lm postenrichment. Although the focus of this study was to explore an allose‐based enrich- ment method as a tool for improved isolation of LII‐Lm, the potential for use of allose in an agar format to aid in isolation was also consid- ered. Allose agar showed good potential to facilitate the isolation step of LII‐Lm when excess Li is present, and with the addition of Listeria selective agents, may serve to aid in Lm isolation independent of an allose broth enrichment. This study has shown that an enrichment step using allose as the primary carbon source can improve the detection of LII‐Lm and has the potential to serve as a tool for when the presence of background flora, including Li, interferes with LII‐Lm detection. Examples of instances when this method modification may be beneficial include: a) when a food enrichment culture screens positive (e.g. using PCR) for Lm, but cultural isolation fails, b) when a food enrichment culture shows the visible abundance of Li on selective agar, c) suspected liste- riosis cases having an epidemiological link to a particular food, but cultural isolation of Lm has failed, and d) validation studies in which the new screening method is more sensitive than the reference method and sheer persistence are required to demonstrate that the screening results are truly positive. An enrichment step in Lis‐A may provide a working example of tailoring methodology to target the known sub- type of the pathogen of interest in an outbreak scenario or during reg- ular monitoring activities through the inclusion of a screening test for allose genes following preenrichment. CRediT authorship contribution statement Jacqueline P. Upham: Conceptualization, investigation, formal analysis and writing. Mikaela Eisebraun: Investigation. Alex For- tuna: Formal analysis. Gustavo V. Mallo: Writing – original draft, Conceptualization. Declaration of Competing Interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influ- ence the work reported in this paper. Appendix A. Supplementary data Supplementary data to this article can be found online at https://doi. org/10.1016/j.jfp.2023.100104. 8 References Altschul, S. F., Gish, W., Miller, W., Myers, E. W., & Lipman, D. J. (1990). Basic local alignment search tool. Journal of Molecular Biology, 215, 403–410. Burall, L. S., Grim, C. J., Mammel, M. K., & Datta, A. R. (2017). A comprehensive evaluation of the genetic relatedness of Listeria monocytogenes serotype 4b variant strains. Frontiers in Public Health, 5, 241. Cornu, M., Kalmokoff, M., & Flandrois, J. P. (2002). Modelling the competitive growth of Listeria monocytogenes and Listeria innocua in enrichment broths. International Journal of Food Microbiology, 73, 261–274. Curiale, M. S., & Lewus, C. (1994). Detection of Listeria monocytogenes in samples containing Listeria innocua. Journal of Food Protection, 57, 1048–1051. Doumith, M., Buchrieser, C., Glaser, P., Jacquet, C., & Martin, P. (2004). Differentiation of the major Listeria monocytogenes serovars by multiplex PCR. Journal of Clinical Microbiology, 42, 3819–3822. Doumith, M., Cazalet, C., Simoes, N., Frangeul, L., Jacquet, C., Kunst, F., Martin, P., Cossart, P., Glaser, P., & Buchrieser, C. (2004). New aspects regarding evolution and virulence of Listeria monocytogenes revealed by comparative genomics and DNA arrays. Infection and Immunity, 72, 1072–1083. Glaser, P., Frangeul, L., Buchrieser, C., Rusniok, C., Amend, A., Baquero, F., Berche, P., Bloecker, H., Brandt, P., Chakraborty, T., Charbit, A., Chetouani, F., Couvé, E., de Daruvar, A., Dehoux, P., Domann, E., Domínguez-Bernal, G., Duchaud, E., Durant, L., ... Cossart, P. (2001). Comparative genomics of Listeria species. Science, 294, 849–852. Health Canada (2011). MFLP-28: the Qualicon BAX® system method for the detection of Listeria monocytogenes in a variety of food. In: Health Canada Compendium of Analytical Methods. Available at: https://www.hc-sc.gc.ca/fn-an/res-rech/analy- meth/microbio/index-eng.php. Accessed 1 April 2021. Health Canada (2011). Policy on Listeria monocytogenes in ready-to-eat foods. Available at: http://www.hc-sc.gc.ca/fn-an/legislation/pol/policy_listeria_monocytogenes_ 2011-eng.php. Accessed 01 April 2021. Health Canada (2011). Vol 1 Annex 4.4: Procedure for the statistical evaluation and calculation of performance parameters of a new alternative qualitative method compared to a reference cultural method. In: Health Canada Compendium of Analytical Methods. Available at: https://www.hc-sc.gc.ca/fn-an/res-rech/analy- meth/microbio/index-eng.php. Accessed 01 April 2021. Health Canada. 20Vol 1 Part 4: Guidelines for the relative validation of indirect qualitative food microbiological methods. In: Health Canada Compendium of Analytical Methods. Available at: https://www.hc-sc.gc.ca/fn-an/res-rech/analy- meth/microbio/index-eng.php. Accessed 01 April 2021. Keys, A. L., Dailey, R. C., Hitchins, A. D., & Smiley, R. D. (2013). Postenrichment population differentials using buffered Listeria enrichment broth: Implications of the presence of Listeria innocua on Listeria monocytogenes in food test samples. Journal of Food Protection, 76, 1854–1862. Leclercq, A., Chenal-Francisque, V., Dieye, H., Cantinelli, T., Drali, R., Brisse, S., & Lecuit, M. (2011). Characterization of the novel Listeria monocytogenes PCR serogrouping profile IVb-v1. International Journal of Food Microbiology., 147, 74–77. Lemaître, J. P., Duroux, A., Pimpie, R., Duez, J. M., & Milat, M. L. (2015). Listeria phage and phage tail induction triggered by components of bacterial growth media (phosphate, LiCl, nalidixic acid, and acriflavine). Applied and Environmental Microbiology, 81, 2117–2124. Liu, D., Wang, Y., Wang, Y., Zhang, L., Luo, L., Liu, K., & Ye, C. (2017). Development of a novel Listeria enrichment broth for the isolation of pathogenic Listeria. Journal of Food Protection., 80, 1768–1776. MacDonald, F., & Sutherland, A. D. (1994). Important differences between the generation times of Listeria monocytogenes and List. innocua in two Listeria enrichment broths. The Journal of Dairy Research, 61, 433–436. Milillo, S. R., Badamo, J. M., & Wiedmann, M. (2009). Contributions to selected phenotypic characteristics of large species- and lineage-specific genomic regions in Listeria monocytogenes. Food Microbiology, 26, 212–223. Pagotto, F., Hébert, K., & Farber, J. (2011). MFHPB-30 Isolation of Listeria monocytogenes and other Listeria spp. from foods and environmental samples. In: Health Canada Compendium of Analytical Methods. Available at: https://www.hc-sc.gc.ca/fn-an/res- rech/analy-meth/microbio/index-eng.php. Accessed 01 April 2021. Petran, R. L., & Swanson, K. M. J. (1993). Simultaneous growth of Listeria monocytogenes and Listeria innocua. Journal of Food Protection, 56, 616–618. Preacher, K. J. (2001). Calculation for Fisher's exact test: an interactive calculation tool for Fisher's exact probability test for 2×2 tables. (Computer software.) Available at: http://quantpsy.org/fisher/fisher.htm. Accessed 12 June 2021. Ragon, M., Wirth, T., Hollandt, F., Lavenir, R., Lecuit, M., Le Monnier, A., & Brisse, S. (2008). A new perspective on Listeria monocytogenes evolution. PLoS Pathogens, 4, e1000146. Zhang, L., Wang, Y., Liu, D., Luo, L., Wang, Y., & Ye, C. (2018). Identification and characterization of als genes involved in D-allose metabolism in lineage II strain of Listeria monocytogenes. Frontiers in Microbiology, 9, 621. Zittermann, S. I., Stanghini, B., See, R. S., Melano, R. G., Boleszczuk, P., Murphy, A., Maki, A., & Mallo, G. V. (2016). Assessment of Listeria sp. interference using a molecular assay to detect Listeria monocytogenes in food. Journal of Food Protection, 79, 138–143. https://doi.org/10.1016/j.jfp.2023.100104 https://doi.org/10.1016/j.jfp.2023.100104 http://refhub.elsevier.com/S0362-028X(23)06776-5/h0005 http://refhub.elsevier.com/S0362-028X(23)06776-5/h0005 http://refhub.elsevier.com/S0362-028X(23)06776-5/h0005 http://refhub.elsevier.com/S0362-028X(23)06776-5/h0010 http://refhub.elsevier.com/S0362-028X(23)06776-5/h0010 http://refhub.elsevier.com/S0362-028X(23)06776-5/h0010 http://refhub.elsevier.com/S0362-028X(23)06776-5/h0015 http://refhub.elsevier.com/S0362-028X(23)06776-5/h0015 http://refhub.elsevier.com/S0362-028X(23)06776-5/h0015 http://refhub.elsevier.com/S0362-028X(23)06776-5/h0015 http://refhub.elsevier.com/S0362-028X(23)06776-5/h0020 http://refhub.elsevier.com/S0362-028X(23)06776-5/h0020 http://refhub.elsevier.com/S0362-028X(23)06776-5/h0020 http://refhub.elsevier.com/S0362-028X(23)06776-5/h0025 http://refhub.elsevier.com/S0362-028X(23)06776-5/h0025 http://refhub.elsevier.com/S0362-028X(23)06776-5/h0025 http://refhub.elsevier.com/S0362-028X(23)06776-5/h0025 http://refhub.elsevier.com/S0362-028X(23)06776-5/h0030 http://refhub.elsevier.com/S0362-028X(23)06776-5/h0030 http://refhub.elsevier.com/S0362-028X(23)06776-5/h0030 http://refhub.elsevier.com/S0362-028X(23)06776-5/h0030 http://refhub.elsevier.com/S0362-028X(23)06776-5/h0030 http://refhub.elsevier.com/S0362-028X(23)06776-5/h0035 http://refhub.elsevier.com/S0362-028X(23)06776-5/h0035 http://refhub.elsevier.com/S0362-028X(23)06776-5/h0035 http://refhub.elsevier.com/S0362-028X(23)06776-5/h0035 http://refhub.elsevier.com/S0362-028X(23)06776-5/h0035 http://refhub.elsevier.com/S0362-028X(23)06776-5/h0035 https://www.hc-sc.gc.ca/fn-an/res-rech/analy-meth/microbio/index-eng.php https://www.hc-sc.gc.ca/fn-an/res-rech/analy-meth/microbio/index-eng.php http://www.hc-sc.gc.ca/fn-an/legislation/pol/policy_listeria_monocytogenes_2011-eng.php http://www.hc-sc.gc.ca/fn-an/legislation/pol/policy_listeria_monocytogenes_2011-eng.php https://www.hc-sc.gc.ca/fn-an/res-rech/analy-meth/microbio/index-eng.php https://www.hc-sc.gc.ca/fn-an/res-rech/analy-meth/microbio/index-eng.php https://www.hc-sc.gc.ca/fn-an/res-rech/analy-meth/microbio/index-eng.php https://www.hc-sc.gc.ca/fn-an/res-rech/analy-meth/microbio/index-eng.php http://refhub.elsevier.com/S0362-028X(23)06776-5/h0060 http://refhub.elsevier.com/S0362-028X(23)06776-5/h0060 http://refhub.elsevier.com/S0362-028X(23)06776-5/h0060 http://refhub.elsevier.com/S0362-028X(23)06776-5/h0060 http://refhub.elsevier.com/S0362-028X(23)06776-5/h0060 http://refhub.elsevier.com/S0362-028X(23)06776-5/h0065 http://refhub.elsevier.com/S0362-028X(23)06776-5/h0065 http://refhub.elsevier.com/S0362-028X(23)06776-5/h0065 http://refhub.elsevier.com/S0362-028X(23)06776-5/h0065 http://refhub.elsevier.com/S0362-028X(23)06776-5/h0070 http://refhub.elsevier.com/S0362-028X(23)06776-5/h0070 http://refhub.elsevier.com/S0362-028X(23)06776-5/h0070 http://refhub.elsevier.com/S0362-028X(23)06776-5/h0070 http://refhub.elsevier.com/S0362-028X(23)06776-5/h0070 http://refhub.elsevier.com/S0362-028X(23)06776-5/h0075 http://refhub.elsevier.com/S0362-028X(23)06776-5/h0075 http://refhub.elsevier.com/S0362-028X(23)06776-5/h0075 http://refhub.elsevier.com/S0362-028X(23)06776-5/h0075 http://refhub.elsevier.com/S0362-028X(23)06776-5/h0080 http://refhub.elsevier.com/S0362-028X(23)06776-5/h0080 http://refhub.elsevier.com/S0362-028X(23)06776-5/h0080 http://refhub.elsevier.com/S0362-028X(23)06776-5/h0080 http://refhub.elsevier.com/S0362-028X(23)06776-5/h0085 http://refhub.elsevier.com/S0362-028X(23)06776-5/h0085 http://refhub.elsevier.com/S0362-028X(23)06776-5/h0085 http://refhub.elsevier.com/S0362-028X(23)06776-5/h0085 https://www.hc-sc.gc.ca/fn-an/res-rech/analy-meth/microbio/index-eng.php https://www.hc-sc.gc.ca/fn-an/res-rech/analy-meth/microbio/index-eng.php http://refhub.elsevier.com/S0362-028X(23)06776-5/h0095 http://refhub.elsevier.com/S0362-028X(23)06776-5/h0095 http://refhub.elsevier.com/S0362-028X(23)06776-5/h0095 http://quantpsy.org/fisher/fisher.htm http://refhub.elsevier.com/S0362-028X(23)06776-5/h0105 http://refhub.elsevier.com/S0362-028X(23)06776-5/h0105 http://refhub.elsevier.com/S0362-028X(23)06776-5/h0105 http://refhub.elsevier.com/S0362-028X(23)06776-5/h0110 http://refhub.elsevier.com/S0362-028X(23)06776-5/h0110 http://refhub.elsevier.com/S0362-028X(23)06776-5/h0110 http://refhub.elsevier.com/S0362-028X(23)06776-5/h0115 http://refhub.elsevier.com/S0362-028X(23)06776-5/h0115 http://refhub.elsevier.com/S0362-028X(23)06776-5/h0115 http://refhub.elsevier.com/S0362-028X(23)06776-5/h0115 http://refhub.elsevier.com/S0362-028X(23)06776-5/h0115 Substituting Allose as the Primary Carbon Source During Enrichment Helps Improve Detection and Isolation of Lineage II Listeria monocytogenes From Food Materials and Methods Results Discussion CRediT authorship contribution statement Declaration of Competing Interest Appendix A Supplementary data References