Transient and resident pathogens: Intra-facility genetic diversity of Listeria monocytogenes and Salmonella from food production environments.

Food production facilities are often routinely tested over time for the presence of foodborne pathogens (e.g., Listeria monocytogenes or Salmonella enterica subsp. enterica). Strains detected in a single sampling event can be classified as transient; positive findings of the same strain across multiple sampling events can be classified as resident pathogens. We analyzed whole-genome sequence (WGS) data from 4,758 isolates (L. monocytogenes = 3,685; Salmonella = 1,073) from environmental samples taken by FDA from 536 U.S. facilities. Our primary objective was to determine the frequency of transient or resident pathogens within food production facilities. Strains were defined as isolates from the same facility that are less than 50 SNP (single-nucleotide polymorphisms) different from one another. Resident pathogens were defined as strains that had more than one isolate collected >59 days apart and from the same facility. We found 1,076 strains (median = 1 and maximum = 21 strains per facility); 180 were resident pathogens, 659 were transient, and 237 came from facilities that had only been sampled once. As a result, 21% of strains (180/ 839) from facilities with positive findings and that were sampled multiple times were found to be resident pathogens; nearly 1 in 4 (23%) of L. monocytogenes strains were found to be resident pathogens compared to 1 in 6 (16%) of Salmonella strains. Our results emphasize the critical importance of preventing the colonization of food production environments by foodborne pathogens, since when colonization does occur, there is an appreciable chance it will become a resident pathogen that presents an ongoing potential to contaminate product.

Introduction

Despite the myriad guidance that exists to control the presence of foodborne pathogens (e.g., Good Manufacturing Practice (GMP), Hazard Analysis and Critical Control Point (HACCP) plans, and Food Safety Management Systems (FSMS)) [1,2] such pathogens may still be found in food production environments and ultimately may result in contamination of food. The plans to ensure food safety may consist of some form of monitoring program that includes testing the environment and products for pathogens [3]. The U.S. Food and Drug Administration (FDA) performs inspections of food production facilities it regulates. Those inspections may be initiated as part of commodity-based assignments, which seek to gain insight about a particular industry; risk-based prioritization, which is the result of a focus on food-hazard pairs (e.g., frequency of outbreaks associated with a certain food type); and for-cause inspections, which are to further investigate a specific problem regarding a facility [4]. During such investigations the FDA often collects environmental samples (e.g., swabs and sponges) that are then tested in FDA laboratories to determine whether foodborne pathogens are present. The FDA may also visit the same facility over time (e.g., years) and collect samples.

There are two general patterns that can describe the presence of a pathogen within a facility’s food production environment, over time. First, there is the possibility of the strain being a transient pathogen where it is not found during subsequent sampling events [1,5,6]. A sampling event here is defined as the collection of environmental samples from the facility at a specific point in time that are then tested for the presence of a foodborne pathogen (e.g., Listeria monocytogenes or Salmonella enterica subsp. enterica). The explanations for a negative sampling event are either the failure to detect the pathogen or the pathogen is not present in the samples. The latter may be the result of natural processes or intentional actions on behalf of the firm to eradicate pathogens.

The second general pattern involves a resident pathogen, in which the same strain is found during subsequent sampling events. This suggests that the pathogen has established itself within the environment [1,5,6]. Given the importance of controlling foodborne pathogens within food production facilities, many studies have investigated the questions of whether and how pathogens may persist in such environments. Often resident pathogens are the result of colonization of an area of the facility that is a difficult to adequately clean (e.g., cracks, junctions in the structure, drains, holes) and thus represent a niche or harborage site [7]. At such sites there is also often an accumulation of food debris or moisture that fosters the growth and persistence of pathogens such as L. monocytogenes and Salmonella [1]. Low-moisture foods and environments are known to facilitate long-term persistence of Salmonella [8,9]. Areas where biofilm formation is possible may also facilitate residency as biofilms increase the pathogen’s tolerance to many tools used as control measures (e.g., heat, desiccation, chlorine, and antimicrobials) [10]. L. monocytogenes is more likely to persist in floors and drains where conditions exist that also enable biofilm formation [11]. Persistence of L. monocytogenes or Salmonella may be on the order of years [12] and has been documented in numerous environments, including those related to produce [11], cheese production environments [13], fresh-cut vegetable processing facilities [14], and crabmeat processing environments [15].

It is important to note that a potential alternative explanation to a strain being resident is reintroduction. Here, the presence of a pathogen over time within a food facility involves successful eradication from the facility, but continued contamination from external sources results in reintroduction. Although others have noted the difficulty in differentiating among the competing explanations (i.e., resident pathogen vs reintroduction) [11], the reintroduction explanation makes a number of assumptions, including elimination of the strain via intentional or natural means, persistence of the strain within the source (e.g., ingredient, material, etc.), consistency of suppliers, and subsequent successful re-colonization of the facility from which it had previously been eradicated. These numerous assumptions render the reintroduction hypothesis less parsimonious than that of residency and, while possible, we assume the results presented here of a strain being found over time within the same facility is due to residency.

In addition to the temporal requirement (e.g., present during multiple sampling events over time), differentiating between transient and resident strains requires a method for determining whether two or more isolates represent the same strain. We use the term “strain” as defined by Tenover et al. [16] to be a group of isolates from the same lower taxonomic rank (e.g., species) that can be distinguished from other isolates from the same rank based on genetic differences. Whole-genome sequencing, with its high discriminatory power and ability to differentiate at the nucleotide level, represents the best method currently available to delineate strains [17] and has been noted for its ability to characterize persistent strains [18].

Here we determine the frequencies of transient and resident pathogens across a large number of L. monocytogenes and Salmonella isolates that were collected from food production facilities within the U.S. Given there is no formal agreed upon genetic difference threshold with which to differentiate strains, we first identified such a threshold based on single nucleotide polymorphisms (SNPs) detected through comparative genome analyses. We then determined the number of different strains that were found and classified them as either transient or resident. Given the large number of facilities investigated here, our findings provide valuable insight into the magnitude of the resident pathogen problem within food production facilities.

Materials and methods

Data curation

We first identified all FDA L. monocytogenes and Salmonella isolates for which whole-genome sequencing had been performed and had an NCBI (National Center for Biotechnology Information) BioSample accession. Additional metadata regarding the collection of each isolate (e.g., firm ID, collection date and facility description code [e.g., fish or dairy facility]) were also extracted from an internal FDA database, as were all sample collection dates for a given facility, which was used to determine sampling events that did not return a positive result for the presence of a foodborne pathogen. Only FDA isolates from environmental samples (e.g., “environmental swab” and “environmental sponge”) and from facilities located in the U.S. were included. We focused on only environmental samples because we are interested in what has directly been found in the manufacturing environment of the facility in defining transient and resident strains. We recognize this excludes other samples (e.g., product samples) that could potentially also be used to make statements about resident or transient strains. We also acknowledge that these samples are predominantly the result of inspections to assess compliance and do not represent a random sample.

This data curation resulted in a dataset of whole-genome sequence data for 4,758 isolates (L. monocytogenes = 3,685; Salmonella = 1,073) from environmental swabs taken from 536 U.S. facilities between August 2000 and March 2020 (Table 1). All sequence data analyzed here are publicly available and predominantly paired-end short-read shotgun sequence data generated using Illumina’s MiSeq platform. See S1 Table for NCBI (National Center for Biotechnology Information) SRA (sequence read archive) and BioSample accessions.

Table 1

The number of isolates (collected between August 2000 and March 2020) analyzed, and the number of different strains and resident pathogens detected.

Taxon	Isolates	Facilities	Strains	Facilities with Resident Pathogen	Resident Pathogens	TransientPathogens		Strains from facilities notRevisited
L. monocytogenes	3685	387	756	103	143 (23.75%)		459 (76.25%)	154
Salmonella	1073	151	320	23	37 (15.61%)		200 (84.39%)	83
Total	4758	536*	1076	126	180 (21.2%)		659 (78.8%)	237

* Number of unique facilities.

For these 536 facilities with WGS data, we also determined the frequency of visits, time between visits, and the frequency with which a visit resulted in a positive finding (e.g., Salmonella or L. monocytogenes was detected). However, 42 of these facilities were excluded due to either insufficient information to reliably determine the complete environmental sampling history or because sampling dates fell outside the time window investigated here (i.e., only the facilities with samples collection between August 9, 2000, and March 9, 2020 were included). As a result, a total of 494 facilities were used to describe sampling (375 facilities were sampled for L. monocytogenes; 225 facilities were sampled for Salmonella); 536 facilities are included for the analysis of WGS data and identifying whether a strain was resident or not.

Strain delineation and serotype identification

To characterize the genetic diversity among isolates within each facility, we began by generating de novo assemblies using the program SKESA v2.2 [19] with default settings. To estimate genetic variation among isolates from facilities with more than one isolate, we used the de novo assemblies and kSNP3 [20], a k-mer based variant detection program, to generate SNP matrices for isolates from a given facility. The value of k was set to 19 and the minimum fraction of isolates within which a variant had to be found to be included in the SNP matrix was 0.9 (i.e., no position in the matrix could have >10% missing data).

To determine the number of different strains within a given facility, regardless of the time over which they had persisted, we performed complete-linkage clustering using the hclust function within R [21]. We performed simulations varying the cuttree setting from 0 to 500 at increments of 5. This cuttree setting is used to delineate sub-clusters or strains in our case. Complete-linkage clustering was chosen as it produced clusters more consistent with what would be expected based on pairwise SNP distances (e.g., tree height of 50 corresponded with a 50 SNP distance threshold among samples in the sub-cluster).

It was not feasible to perform traditional serotyping on all isolates investigated in this study and, therefore, we rely on serotype prediction from the de novo assemblies to provide insight into patterns of resident and transient pathogens with respect to serotype. For L. monocytogenes, we used LisSero v0.4.1 [22], which is based on the five-locus PCR method described in Doumith [23]. That widely used five-locus serotyping schema differentiates the three major L. monocytogenes lineages (I, II, III) into five serogroups (serogroup I.1 includes 1/2a-3a, serogroup I.2 includes 1/2c-3c, serogroup II.1 includes 4b4d-4e, serogroup II.2 includes 1/2b-3b-7, and serogroup III includes 4a-4c). Although the method does not differentiate all serotypes within a lineage, the five serogroups are consistent with phylogenetic relationships and differentiate the most common serovars associated with foods and clinical samples [e.g., 1/2a, 1/2b, 1/2c and 4b; 23, 24]. For more detailed information on the lineages of L. monocytogenes we also used the program mlst [25] that uses the PubMLST [26] schema and nomenclature to provide sequence types and lineages (S2 Table). For Salmonella, we used SeqSero2 [27] to predict serotype based on the de novo assemblies. If an isolate’s serogroup or serotype could not be predicted or multiple classifications were predicted for isolates from the same strain, we assigned the most abundant serogroup or serotype to all isolates from the same strain.

Strain classification and comparison

Strains were assigned to one of three categories. 1) Resident strains were identified as those strains whose isolates were collected at least 60 days apart. 2) Transient strains were identified as those not found across multiple temporally spaced sampling events (i.e., >59 days apart) for a specific facility. 3) “strains from facilities not revisited” were identified as those from a facility that had not been sampled multiple times at least 60 days apart. Strains from a facility that was sampled multiple times >59 days apart and were only found on the last sampling event are classified as transient.

We compared the differences between L. monocytogenes and Salmonella in how likely each was to become resident via a x2-test with the p-value estimated from 2000 simulations. We used the paired Wilcoxon Signed-Rank test implemented in R to determine whether predicted serotypes differed in how likely strains were to be resident or transient. Non-typeable strains were not included in the test for differences among L. monocytogenes serotypes in how likely each was to be a resident or transient pathogen.

Fisher’s exact test was used to determine whether there were differences in facility description code (e.g., a facility produces “Cheese and Cheese Products” or “Egg and Egg Products”) in the relative number of transient and resident strains found. For facility description code analyses, we only included those facilities that were assigned a single description code, which avoids the issue of assigning strains to a specific type of product made at a facility with multiple description codes.

Results and discussion

Sampling

Sampling histories were analyzed for 494 of the 536 facilities from which we had WGS data. Of those 494 facilities, 375 were sampled for L. monocytogenes and 127 of those facilities were visited only once during the time window studied. The waiting time between visits for the rest of 248 facilities visited more than once ranged from 1 to 4,892 days with an average of 502 days. The number of visits to each of the 375 facilities ranged from 1 to 15, with an average of 3 per facility; the percent of visits resulting in a positive finding for L. monocytogenes ranged from 0% to 100%, with an average of 70.0% over all facilities. The number of environmental samples collected from each facility ranged from 1 to 15 across all visits with an average of 3 per facility (FDA samples consist of multiple subsamples); The percent of samples positive for L. monocytogenes ranged from 0% to 100% with an average of 68.6% over all facilities.

Of the 225 facilities included in this study where environmental sampling for Salmonella occurred, 82 facilities were visited only once during the time window studied. The time between visits for the rest of 143 facilities visited more than once ranged from 1 to 5,263 days with an average of 418 days. The number of visits to each of the 225 facilities ranged from 1 to 13, with an average of 3 per facility; the percent of visits yielding a detection for Salmonella ranged from 0% to 100% with an average of 37.5% over all facilities. The number of environmental samples collected from each facility ranged from 1 to 37 across all visits with an average of 4 per facility; the positive rate ranged from 0% to 100% with an average of 32.4% over all facilities.

Simulations of strain definition via complete-linkage clustering

A method for assigning isolates to the same strain is important to quantifying the prevalence of resident pathogens within facilities. To accomplish this, we applied complete-linkage clustering to a genetic distance matrix among isolates from the same facility. Based on simulations, we found that the number of strains delineated was sensitive to changes in the SNP distance threshold at SNP distances less than 50 (Fig 1A). For example, a SNP distance threshold of 10 produced a mean and maximum number of strains per facility of 3.29 and 27, respectively; a SNP distance threshold of 30 produced a mean of 2.4 and maximum of 23 strains per facility. Given this behavior, we chose to delineate strains as those isolates that were within 50 SNPs of one another. Furthermore, it is important to note that the vast majority of strains have pairwise SNP distances less than that where they are actually likely to be less than 10 SNPs different (Fig 1B). This magnitude of SNP differences and threshold for strain delineation is in line with what other studies have found. For example, a study of crabmeat processing environments observed that the SNP distances within isolates attributed to the same resident pathogen were on the order of a few to tens of SNP differences [15]. Additionally, the NCBI’s Pathogen Detection (https://www.ncbi.nlm.nih.gov/pathogens/) platform uses a 50 SNP distance threshold for single linkage clustering.

Fig 1

a) The number of clusters (strains) detected within each facility as a function of the tree height parameter, which corresponds to SNP distance, in the complete-linkage clustering algorithm. The mean (black line) and 95% confidence level (shaded gray region are plotted. b) Histograms of the pairwise SNP distance among isolates from the same strain at a sub-cluster height (i.e., SNP distance threshold) of 50.

Resident and transient strains

We found 1,076 strains of Salmonella or L. monocytogenes across 536 facilities (median = 1 and maximum = 21 strains per facility) (Table 1). Accounting for whether a facility had been sampled multiple times and, thus, there was the possibility to detect the same strain, 237 strains came from facilities that had only been sampled once and 659 were transient. As a result, 21% (180 resident strains out of 839 [= 1,076–237]) were resident. Those 180 resident pathogens were found in 126 facilities and, thus, some facilities had multiple resident pathogens (median = 1, max = 6). As noted, an alternative explanation to residency of a pathogen is the successful eradication and subsequent reintroduction of a strain. Differentiating among those hypotheses remains an issue [e.g., 11] and additional research is necessary to gauge the frequency of reintroduction. Although reintroduction is less parsimonious, it is a possibility, and such instances would reduce the number of resident pathogens identified here.

Other studies have documented the presence of resident pathogens within food production facilities, but those have often examined a single or few facilities and did not consider data collected over a larger timescale [12,28-30]. Of exception is a study by Leong, et al. [31] that investigated 54 businesses in Ireland and found 86 different pulsotypes, of which 17 (20%) were determined to be resident, which is consistent with the observations presented here of 21% of strains being resident. The reasons for persistence within a facility are numerous, and studies have shown that the easy-to-clean food contact surfaces are less likely to have subsequent positive findings compared to harder reach areas such as cracks, drains, and internal pieces on equipment. Additionally, areas such as drains have been identified as particularly vexing to efforts to adequately control L. monocytogenes; prevalence in drains was nearly the same before and after control measures were implemented within smoked fish facilities [32]. Another reason is the potential route of contamination. Sauders and Wiedmann (2007) provided two reasons for why L. monocytogenes outbreaks may occur. One possibility is contaminated raw ingredients coming into a facility that, due to handling errors, produces a few contaminated lots but the contamination then goes away. Another possibility is facility maintenance that allows for residence to be established, and resident strains may contaminate food, potentially over a long time period. Here we found that resident strains can persist upwards of 12 years; most were found across two years (Fig 2).

Fig 2

a) Histograms of the number of different strain types found per facility. b) Histogram of the maximum differences in collection dates per facility for isolates that were from resident pathogens.

Taxon and serotype specific differences

Of the strains detected, 756 (70%) were L. monocytogenes and 320 (30%) were Salmonella (Table 1, Fig 2). We found that among resident and transient strains, L. monocytogenes was more likely than Salmonella to become a resident pathogen (p = 0.039, x2 = 4.44).

The most frequently found L. monocytogenes serogroup was 1/2a, 3a with 365 strains (48%) followed by 1/2b, 3b, 7 (Table 2). Although we cannot rule out the possibility that some of those strains are 3a rather than 1/2a, the dominance of 1/2a, 3a is consistent with previous studies documenting that that 1/2a is the most abundant serogroup found in food production environments [6,24]. Although, many outbreaks and clinical infections have also historically been attributed to 4b, our results showing it to be the third most abundant serotype found supports the putative underrepresentation of that serotype within foods [33]. Interestingly, no statistical differences were detected among L. monocytogenes serotypes in the likelihood that strains were resident or transient (V = 10, p-value = 0.125), which may in part be explained by our low power to detect differences given there are only four different serotypes considered. However, the results here may support findings that there are not inherent genomic or phenotypic differences among the serogroups that enables a serogroup to establish and become resident within a facility compared to other serogroups [34]. Alternatively, Orsi, et al. [24] suggest that distinct features do exist among serotypes. Additionally, the dynamics of biofilm production and likelihood of a pathogen to persist are complex and vary among serotypes depending on environmental factors [35].

Table 2

Distribution and classification (i.e., resident, transient, from facilities not revisited) of strains across L. monocytogenes serotypes.

Percentages are relative to the total number of Transient and Resident strains.

Serotype	Transient	Resident	From Facilities Not Revisited	Total
1/2a, 3a	223 (37.04%)	75 (12.45%)	67	365
1/2b, 3b, 7	128 (21.26%)	48 (7.97%)	52	228
4b, 4d, 4e	74 (12.29%)	16 (2.65%)	22	112
1/2c, 3c	13 (2.15%)	4 (0.66%)	5	22
Nontypeable	21 (3.48%)	0 (0%)	8	29
Total	459 (76.24%)	143 (23.75%)	154	756

As for Salmonella, 100 serotypes were detected among the 320 strains. Serotype Enteritidis was the most frequently found, with 37 strains, followed by Senftenberg with 27 (Table 3). In contrast to L. monocytogenes, statistical differences were detected among the top 10 most represented Salmonella serotypes in the likelihood that strains were resident or transient (V = 55, p-value = 0.002). This is likely driven by the relatively larger differences in the numbers of transient and resident strains in Enteritidis, Newport, Anatum, Mbandaka, and Muenchen. Senftenberg and Montevideo had a greater number of resident strains compared to transient strains when compared to the most abundant serotypes in terms of transient strains (e.g., Senftenberg had 14 transient strains and 7 resident strains while Enteritidis had 19 transient and 1 resident strain; Table 3).

Table 3

Distribution and classification (i.e., resident, transient, or not revisited) of strains from the 10 most abundant Salmonella serotype detected.

Percentages are relative to the total number of Transient and Resident strains.

Serotype	Transient	Resident	From Facilities Not Revisited	Total
Enteritidis	19 (16.1%)	1 (0.85%)	19	37
Senftenberg	14 (11.86%)	7 (5.93%)	14	27
Newport	12 (10.17%)	0 (0%)	12	20
Anatum	10 (8.47%)	2 (1.69%)	10	13
Mbandaka	10 (8.47%)	1 (0.85%)	10	13
Muenchen	10 (8.47%)	0 (0%)	10	11
Cubana	6 (5.08%)	3 (2.54%)	6	10
Infantis	7 (5.93%)	1 (0.85%)	7	10
Montevideo	5 (4.24%)	3 (2.54%)	5	10
Heidelberg	6 (5.08%)	1 (0.85%)	6	9
Total	99 (83.9%)	19 (16.1%)	99	160

Resident strains and facility description codes

There were differences in the number of strain types associated with the different facility description codes (e.g., dairy facility or seafood facility) for L. monocytogenes (p-value = 0.042) (Table 4). For example, “Fishery/Seafood Products” and “Ice Cream and Related Products” tended to harbor larger fractions of resident strains than “Cheese and Cheese Products” or “Vegetables and Vegetable Products”. The differences in the number of strain types associated with facility type for Salmonella was not statistically supported (p-value = 0.328), but the distribution of strain types across facilities with positive Salmonella findings illustrates the small fraction of Salmonella resident strains compared to what is seen for L. monocytogenes (Table 1).

Table 4

The number of strain types among the top 10 most abundant facility description code types.

Facilities assigned multiple description code types were excluded.

	Listeria			Salmonella
Facility Description Code	From Facilities Not Revisited	Transient	Resident	From Facilities Not Revisited	Transient	Resident	Total
Fishery/Seafood Products	22	80	28	0	1	0	131
Cheese and Cheese Products	13	42	7	1	0	0	63
Egg and Egg Products	0	0	1	15	32	2	50
Ice Cream and Related Products	8	17	12	1	0	0	38
Nuts and Edible Seeds	0	0	0	10	24	2	36
Vegetables and Vegetable Products	2	19	3	9	2	0	35
Fruit and Fruit Products	7	5	0	6	9	0	27
Milk, Butter, and Dried Milk Products	0	0	0	2	12	4	18
Bakery Products, Doughs, Bakery Mixes, and Icings	6	5	3	0	0	0	14
Multiple Food Dinners, Gravies, Sauces, and Specialties (Total Diet)	5	3	1	0	0	0	9

We also observed differences in the numbers of strains of L. monocytogenes and Salmonella among different facility description codes (p-value = 0.001) which are consistent with the published literature. For example, from a comparison of expert elicitations and risk-based studies, the most hazardous foods regarding Listeriosis were deli meats, dairy products and seafood, but of note is the elicitation also identified produce as being important [36]. Here, “Fishery/Seafood Products”, “Cheese and Cheese Products”, and produce (“Vegetables and Vegetable Products” and “Fruit and Fruit Products) facilities had a high number of different strains, which is congruent with the discussion in Todd and Notermans [36]; however, this may be a sampling artifact, since those facilities were also the most frequently sampled. Of note is the lack of L. monocytogenes strains found in facilities with description code “Milk, Butter, and Dried Milk Products”, which is typically classified as dairy. This may suggest that L. monocytogenes is more specific to cheese products than the milk aspect of the general dairy classification, but this result is also likely due in part to differences in sampling priorities (i.e., Salmonella environmental monitoring is done for dried dairy products while L. monocytogenes is not). Salmonella strains were most frequently found in “Egg and Egg Products” facilities, followed by “Nuts and Edible Seeds” facilities (Table 4). Although both commodity types are known to be associated with Salmonella contamination [37,38], among facilities assigned a single description code “Egg and Egg Products” facilities were sampled less than “Nuts and Edible Seeds,” yet there are more strains found in the former suggesting Salmonella is more likely to be found in “Egg and Egg Products” facilities. Within “Egg and Egg Products” facilities, Enteritidis was the dominant serotype found, followed by Heidelberg, which is to be expected based on previous research [e.g., 39]. For “Nuts and Edible Seeds” facilities, serotypes Montevideo, Newport, Senftenberg, and Tymphimurium each had three strains found (not shown). Salmonella strains were also prominent in fruit and vegetable facilities (Table 4).

Conclusions

There are numerous steps along the food production chain where products could become contaminated with a foodborne pathogen. Here, we have investigated one potential source of contamination–the facility in which ingredients are processed and products are manufactured. We found 1,076 strains across 536 facilities and, most importantly, that 21% of strains within facilities that were sampled over time were found to be resident pathogens. This pattern across such a large number of facilities, isolates, and strains highlights the importance of inhibiting introduction into or ensuring efficient eradication of pathogens in the production environment, since not doing so can lead to a potentially long-term problem associated with a resident pathogen and managing the constant risk a resident strain represents to producing food free of environmental pathogens such as L. monocytogenes and Salmonella.

NCBI (National Center for Biotechnology Information) SRA (sequence read archive) and BioSample accessions for sequence data analyzed.

(XLSX)

Click here for additional data file.

Lineages and sequence types of L. monocytogenes as predicted with the program mlst (25) that uses the PubMLST (26) schema.

(XLSX)

Click here for additional data file.

5 Jan 2022

A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). You should upload this letter as a separate file labeled 'Response to Reviewers'.

A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'.

An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter. Guidelines for resubmitting your figure files are available below the reviewer comments at the end of this letter.

If applicable, we recommend that you deposit your laboratory protocols in protocols.io to enhance the reproducibility of your results. Protocols.io assigns your protocol its own identifier (DOI) so that it can be cited independently in the future. For instructions see: https://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols. Additionally, PLOS ONE offers an option for publishing peer-reviewed Lab Protocol articles, which describe protocols hosted on protocols.io. Read more information on sharing protocols at https://plos.org/protocols?utm_medium=editorial-email&utm_source=authorletters&utm_campaign=protocols.

We look forward to receiving your revised manuscript.

Kind regards,

Carlo Spanu, Ph.D.

Academic Editor

PLOS ONE

Journal requirements:

When submitting your revision, we need you to address these additional requirements.

1. Please ensure that your manuscript meets PLOS ONE's style requirements, including those for file naming. The PLOS ONE style templates can be found at

https://journals.plos.org/plosone/s/file?id=wjVg/PLOSOne_formatting_sample_main_body.pdf  and

https://journals.plos.org/plosone/s/file?id=ba62/PLOSOne_formatting_sample_title_authors_affiliations.pdf

2. In your Data Availability statement, you have not specified where the minimal data set underlying the results described in your manuscript can be found. PLOS defines a study's minimal data set as the underlying data used to reach the conclusions drawn in the manuscript and any additional data required to replicate the reported study findings in their entirety. All PLOS journals require that the minimal data set be made fully available. For more information about our data policy, please see http://journals.plos.org/plosone/s/data-availability.

Upon re-submitting your revised manuscript, please upload your study’s minimal underlying data set as either Supporting Information files or to a stable, public repository and include the relevant URLs, DOIs, or accession numbers within your revised cover letter. For a list of acceptable repositories, please see http://journals.plos.org/plosone/s/data-availability#loc-recommended-repositories. Any potentially identifying patient information must be fully anonymized.

Important: If there are ethical or legal restrictions to sharing your data publicly, please explain these restrictions in detail. Please see our guidelines for more information on what we consider unacceptable restrictions to publicly sharing data: http://journals.plos.org/plosone/s/data-availability#loc-unacceptable-data-access-restrictions. Note that it is not acceptable for the authors to be the sole named individuals responsible for ensuring data access.

We will update your Data Availability statement to reflect the information you provide in your cover letter.

3. PLOS requires an ORCID iD for the corresponding author in Editorial Manager on papers submitted after December 6th, 2016. Please ensure that you have an ORCID iD and that it is validated in Editorial Manager. To do this, go to ‘Update my Information’ (in the upper left-hand corner of the main menu), and click on the Fetch/Validate link next to the ORCID field. This will take you to the ORCID site and allow you to create a new iD or authenticate a pre-existing iD in Editorial Manager. Please see the following video for instructions on linking an ORCID iD to your Editorial Manager account: https://www.youtube.com/watch?v=_xcclfuvtxQ.

4. Please include captions for your Supporting Information files at the end of your manuscript, and update any in-text citations to match accordingly. Please see our Supporting Information guidelines for more information: http://journals.plos.org/plosone/s/supporting-information.

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: Partly

Reviewer #2: Yes

**********

2. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: I Don't Know

Reviewer #2: Yes

**********

3. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: No

Reviewer #2: Yes

**********

4. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: Yes

Reviewer #2: Yes

**********

5. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: The study describes a large-scale analysis on Listeria monocytogenes and Salmonella isolates collected from food industry by the FDA, where the authors perform analyses on whether the strains are resident or transient. I welcome such a large-scale study, in comparison with many small and often single-plant studies, as this may allow uncovering larger patterns. However, I think that the total lack of information about the sampling (no. of samples, no. of samplings, interval between samples etc.) makes it difficult for the reader (and for me as a referee) to evaluate the results and the justification of grouping strains as resident/transient. Also, the author’s fail to discuss how variation in these (sampling) parameters may influence their analysis, which they in my opinion may do. The grouping of strains according to serotype and not CC/ST makes the study less relevant for readers.

Introduction: The introduction is lacking background on studies of persistence of Salmonella in the food industry. Please include references/examples from other studies to justify the relevance of your hypothesis that Salmonella may persist in the food industry.

Line 67-73. While I realize that you need to exclude Reintroduction to perform your Resident/Transient analyses, your arguments about why it is not relevant are rather weak. Raw materials are a relevant source of introduction of pathogens and if raw materials contain a “resident” strain this may lead to overestimation of resident strains colonizing plants. Of course, food safety regulations should limit such reintroduction to hygienic zones in the factories, but results from FDA inspections reveal that it is not uncommon with sanctions for improper hygiene, lack of zoning etc. I think you should include something about how Reintroduction may have influenced your analysis in the discussion.

Introduction, Methods, Results: Serotyping? In a paper based on WGS, why do you not refer to CC or ST? At least for Listeria, almost all papers based on typing by sequencing use this type of grouping, and not serotyping. Other papers suggest that some CC/ST (e.g. 121, 9) to be dominating as resident strains, and I think it would have been interesting for readers with an analysis of which CC/ST that dominated among resident strains in your analysis, and a comparison with other studies. As you have the sequences this should be straightforward to do. This can be in addition, and not instead of the serotyping part.

Materials and Methods:

I think the authors should include information about the variation among plants/facilities in number of visits, number of samples taken, and interval between samplings. The lack of such information makes it difficult for the reader to evaluate the results. I acknowledge that there may be huge variations in these parameters but information about the distribution of this variance should be presented. Based on the information presented in the paper it does not exclude the possibility that e.g. 50% of the plants was visited twice and one sample taken at each sampling. Obviously this is a stupid assumption and even this stupid referee realizes this is not the case, but as you provide absolutely no information about the sampling, the results are very difficult to evaluate for the reader, and it is obvious that number of samplings and samples as well as the time interval between samplings can influence the analysis leading to conclusions on whether a strain is resident or transient.

Results and discussion:

As commented above; I think the authors should include information about CC/ST of the strains (at least for Listeria) and compare their results on CC/ST with other studies.

As commented above there is a need to provide more information about the variations in the samplings, and to discuss how this may have influenced your results.

I suggest to include more information about and compare with other studies on persistence/resident/transient strains. How common is persistence/resident strains. Are your results in line with other results. Such an evaluation/comparison is especially lacking for Salmonella.

Conclusions: References are usually not used in conclusions I suggest to exclude the references, and shorten the conclusion to focus on your own findings.

Reviewer #2: Summary

The authors present a SNP-based analysis of Listeria and Salmonella collected by government sampling of US food processing facilities. They use this analysis to differentiate persistent from sporadic strains by a SNP threshold, and then describe how those frequencies change given various important categories.

General Comments

Overall, this is a useful paper because only the US government would have access to the metadata for Genomtrakr isolates to do this type of analysis. From that unique position, the analysis and results seem appropriate. There is one major comment, and request for including more data:

L97: This ‘denominator’ data of all sampling events that did not include a positive sample is very useful data that could only be analyzed by the FDA, academics cannot see such data. Therefore, the reviewer would really appreciate seeing this data summarized in Table 1. Though I understand there may be legal reasons it could not be included. This presentation of negative sampling events would help support the discussion around Table 4 – facility types. It might even allow statistical control for sampling frequency.

A few additional comments follow that would improve the analysis or presentation.

Line Item Comments

L27. Clarify if 50 SNPs also applies to transient strains. Maybe: strained defined as …, and strains called resident if …

Fig 1a. Useful primary data, but I would suggest plotting the median line and a quantile range. This is more consistent with the summary statistics presented in the abstract, and more informative when there is heavy skew in the data.

Fig 2a. I would like to see statistics if the distribution of counts are different between the 3 categories. Transient does appear to have a longer tail, but that could simply be because of the large number of observations.

222. No difference in serotypes likelihood of resident or transient. OK. One additional alternative explanation could be that there is already significant selection pressure for a strain to even be transient. And therefore there is not additional genetic selection around persistence (and it might be more about location and external factors like cleaning).

**********

6. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: No

Reviewer #2: Yes: Matthew J Stasiewicz

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email PLOS at figures@plos.org. Please note that Supporting Information files do not need this step.

21 Apr 2022

See attached document

Submitted filename: Plos_reviewer_comments_reply.docx

Click here for additional data file.

1 May 2022

Transient and resident pathogens: intra-facility genetic diversity of Listeria monocytogenes and Salmonella from food production environments

PONE-D-21-35348R1

Dear Dr. Pettengill,

We’re pleased to inform you that your manuscript has been judged scientifically suitable for publication and will be formally accepted for publication once it meets all outstanding technical requirements.

Within one week, you’ll receive an e-mail detailing the required amendments. When these have been addressed, you’ll receive a formal acceptance letter and your manuscript will be scheduled for publication.

An invoice for payment will follow shortly after the formal acceptance. To ensure an efficient process, please log into Editorial Manager at http://www.editorialmanager.com/pone/, click the 'Update My Information' link at the top of the page, and double check that your user information is up-to-date. If you have any billing related questions, please contact our Author Billing department directly at authorbilling@plos.org.

If your institution or institutions have a press office, please notify them about your upcoming paper to help maximize its impact. If they’ll be preparing press materials, please inform our press team as soon as possible -- no later than 48 hours after receiving the formal acceptance. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information, please contact onepress@plos.org.

Kind regards,

Carlo Spanu, Ph.D.

Academic Editor

PLOS ONE

Additional Editor Comments (optional):

Reviewers' comments:

24 Aug 2022

PONE-D-21-35348R1

Transient and resident pathogens: intra-facility genetic diversity of Listeria monocytogenes and Salmonella from food production environments

Dear Dr. Pettengill:

I'm pleased to inform you that your manuscript has been deemed suitable for publication in PLOS ONE. Congratulations! Your manuscript is now with our production department.

If your institution or institutions have a press office, please let them know about your upcoming paper now to help maximize its impact. If they'll be preparing press materials, please inform our press team within the next 48 hours. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information please contact onepress@plos.org.

If we can help with anything else, please email us at plosone@plos.org.

Thank you for submitting your work to PLOS ONE and supporting open access.

Kind regards,

PLOS ONE Editorial Office Staff

on behalf of

Dr. Carlo Spanu

Academic Editor

PLOS ONE