Toward the Adoption of Loop-Mediated Isothermal Amplification for Salmonella Screening at the National Antimicrobial Resistance Monitoring System's Retail Meat Sites

Abstract

The National Antimicrobial Resistance Monitoring System (NARMS) is a One Health program in the United States that collects data on antimicrobial resistance in enteric bacteria from humans, animals, and the environment. Salmonella is a major pathogen tracked by the NARMS retail meat arm but currently lacks a uniform screening method. We evaluated a loop-mediated isothermal amplification (LAMP) assay for the rapid screening of Salmonella from 69 NARMS retail meat and poultry samples. All samples were processed side by side for culture isolation using two protocols, one from NARMS and the other one described in the U.S. Food and Drug Administration's Bacteriological Analytical Manual (BAM). Overall, 10 (14.5%) samples screened positive by the Salmonella LAMP assay. Of those, six were culture-confirmed by the NARMS protocol and six by the BAM method with overlap on four samples. No Salmonella isolates were recovered from samples that screened negative with LAMP. These results suggested 100% sensitivity for LAMP in reference to culture. Antimicrobial susceptibility testing and whole-genome sequencing analysis confirmed identities of these isolates. Using the BAM protocol, all Salmonella isolates were recovered from samples undergoing Rappaport-Vassiliadis medium selective enrichment and presumptive colonies (n = 130) were dominated by Hafnia alvei (44.6%), Proteus mirabilis (22.3%), and Morganella morganii (9.9%) based on matrix-assisted laser desorption/ionization time-of-flight mass spectrometry. This method comparison study clearly demonstrated the benefit of a rapid, robust, and highly sensitive molecular screening method in streamlining the laboratory workflow. Fourteen NARMS retail meat sites further verified the performance of this assay using a portion of their routine samples, reporting an overall specificity of 98.8% and sensitivity of 90%. As of July 2022, the vast majority of NARMS retail meat sites have adopted the Salmonella LAMP assay for rapid screening of Salmonella in all samples.

Introduction

Antimicrobial resistance (AMR) is widely recognized as a leading public health threat around the world (Antimicrobial Resistance Collaborators, 2022; CDC, 2019; WHO, 2021) and requires a global coordinated action plan (WHO, 2015a). Effective surveillance plays an essential role in the combat against AMR (Federal Task Force on Combating Antibiotic-Resistant Bacteria, 2020; WHO, 2015b). Established in 1996, the National Antimicrobial Resistance Monitoring System (NARMS) is a collaborative program of the U.S. Food and Drug Administration (FDA), the Centers for Disease Control and Prevention (CDC), the U.S. Department of Agriculture (USDA), and state and local public health and agriculture departments and universities (FDA, 2022; Karp et al, 2017). Operating under the new One Health paradigm, NARMS now tracks resistance in enteric bacteria from humans (clinical samples), animals (cecal, slaughter, retail meats, and veterinary), and the environment (surface water) in the United States.

Since its inception, the NARMS retail meat arm has been monitoring AMR trends in Salmonella, Campylobacter, Escherichia coli, and Enterococcus from retail beef, chicken, pork, and turkey products for more than two decades (Nyirabahizi et al, 2020; Tadesse et al, 2018; Tyson et al, 2018; Whitehouse et al, 2018; Yin et al, 2021). Pilot studies targeting additional bacteria and/or commodities have been carried out (Ge et al, 2017; Tate et al, 2021). Throughout these testing efforts, Salmonella remains a key pathogen tracked by the NARMS retail meat arm. Understandably, Salmonella is a ubiquitous zoonotic pathogen of significant food safety concern worldwide (WHO, 2018) and AMR issues in Salmonella are constantly evolving (Kim et al, 2020; Li et al, 2021; Tate et al, 2017; Tyson et al, 2017).

Performed at >20 NARMS retail meat sites/states, the Salmonella testing protocol relies on culture isolation followed by whole-genome sequencing (WGS) and phenotypic antimicrobial susceptibility testing (AST) (FDA, 2022). Considered the gold standard, culture methods are time-consuming and labor-intensive, demanding days to weeks of intensive work for a definitive result (Andrews et al, 2022; USDA, 2021). Previously, NARMS retail meat sites used the TECRA Salmonella Visual Immunoassay (3M Food Safety, St. Paul, MN) for screening Salmonella, which was discontinued in 2016. Some NARMS retail meat sites have since adopted new screening methods, including VIDAS (bioMérieux, Hazelwood, MO), 3M Molecular Detection Assay (MDA) 2–Salmonella, or BAX System Real-time PCR Assay Salmonella (Hygiena, Camarillo, CA). However, most sites do not use any screening methods before culture isolation for Salmonella.

Loop-mediated isothermal amplification (LAMP) has emerged as a powerful alternative to polymerase chain reaction (PCR) for detecting numerous bacterial, fungal, parasitic, and viral agents (Kumar et al, 2017; Mansour et al, 2015; Mori and Notomi, 2020). We have developed an invA-based Salmonella LAMP assay that was rapid, reliable, and robust in multiple food matrices (Chen et al, 2011; Domesle et al, 2021; Yang et al, 2013, 2015, 2016). Upon the completion of single laboratory and multilaboratory validations in animal food (Domesle et al, 2018; Ge et al, 2019), this LAMP assay was incorporated into the FDA's Bacteriological Analytical Manual (BAM) Chapter 5 in February 2020, as a regulatory method to screen for Salmonella in animal food and confirm presumptive Salmonella isolates from either animal food or human food (Andrews et al, 2022).

Given the compendium status of the Salmonella LAMP assay, we proposed for all NARMS retail meat sites to adopt this rapid screening method to improve laboratory workflow, efficiency, and Salmonella isolation rates. This method comparison study aimed to demonstrate LAMP's utility by comparing it with two culture protocols, one from NARMS (FDA, 2022) and another in the FDA's BAM (Andrews et al, 2022). The assay efficacy was verified at additional NARMS retail meat sites before wider adoption as a universal Salmonella screening method.

Materials and Methods

Sample collection and processing

Figure 1 shows a schematic diagram of this study. In September 2021, two NARMS retail meat sites (Maryland and North Carolina) obtained a total of 69 retail meat and poultry samples (Table 1) from local grocery stores as part of the routine NARMS sample collection (FDA, 2022). For each sample, a 50-g test portion was added to 250 mL of buffered peptone water (BPW; BD Diagnostic Systems, Sparks, MD) in a filtered plastic bag. After homogenization and overnight enrichment, sample pre-enrichment broths were processed for Salmonella isolation at the NARMS sites and aliquots (40 mL) from the filtered side of the bag were shipped frozen to the FDA's Center for Veterinary Medicine (CVM).

FIG. 1.

A schematic diagram of the design and workflow of this method comparison study. *Maryland site used the 3M Molecular Detection Assay 2—Salmonella as a screening method. BAP, blood agar plate; BPW, buffered peptone water; BS, bismuth sulfite agar; CVM, U.S. Food and Drug Administration’s Center for Veterinary Medicine; HE, Hektoen enteric agar; LAMP, loop-mediated isothermal amplification; LIA, lysine iron agar slant; MALDI, matrix-assisted laser desorption/ionization; NARMS, National Antimicrobial Resistance Monitoring System; RV, Rappaport-Vassiliadis medium; RVR10, Rappaport-Vassiliadis R10 broth; TSI, triple sugar iron slant; TT, tetrathionate broth; XLD, xylose lysine desoxycholate agar; XLT4, xylose lysine Tergitol 4 agar.

Table 1.

Overview of Sample Testing Results from the Method Comparison Study

Sample type	Samples collected in Maryland^a				Samples collected in North Carolina^a
Sample type	Total	Site+	LAMP+	BAM+	Total	Site+	LAMP+	BAM+
Chicken	8	1	1	1	8	0	0	0
Ground turkey	8	1	2	2	8	0	1	1
Ground beef	8	0	1^b	0	8	0	1	0
Pork	8	0	0	0	8	1	1	0
Chicken gizzard	1	1	1	1	1	0	0	0
Chicken heart	1	1	1	1	1	1	1	0
Chicken liver	0	N/A	N/A	N/A	1	0	0	0
Subtotal	34	4	6	5	35	2	4	1

Site+ means that the samples were tested at the NARMS sites following the NARMS protocol and yielded positive cultures. LAMP+ means that the LAMP screening assays performed on the samples at CVM were positive. BAM+ means the BAM confirmations performed on the samples at CVM yielded positive cultures.

Positive LAMP results in the second replicate only.

BAM, Bacteriological Analytical Manual; CVM, U.S. Food and Drug Administration's Center for Veterinary Medicine; LAMP, loop-mediated isothermal amplification; N/A, not applicable; NARMS, National Antimicrobial Resistance Monitoring System.

NARMS protocol

Fresh overnight sample pre-enrichment broths at the sites were subjected to selective enrichment in Rappaport Vassiliadis R10 broth (RVR10) and selective plating on xylose lysine Tergitol 4 agar plate (XLT4), followed by biochemical confirmation on VITEK 2 Compact system (bioMérieux) or matrix-assisted laser desorption/ionization (MALDI) time-of-flight mass spectrometry such as MALDI Biotyper (Bruker, Billerica, MA).

LAMP screening

Frozen sample pre-enrichment broths received at CVM were thawed at 4°C overnight and 1-mL aliquots were subjected to DNA extraction using the PrepMan ultra sample preparation reagent (Thermo Fisher Scientific, Waltham, MA) (Ge et al, 2019). Sample DNA extracts were stored at −20°C until use. Positive and negative extraction controls, that is, BPW with and without inoculation of Salmonella enterica serovar Typhimurium ATCC 14028 supplied in BioBall SingleShot (bioMérieux) at ∼30 CFU, were included. A positive control (Salmonella Typhimurium ATCC 19585 [LT2] at ∼10³ CFU/reaction) and a no-template control (molecular grade water) were included in each LAMP run.

The LAMP assay was performed as described previously (Domesle et al, 2020). In brief, the reaction mix (25 μL total) contained 1 × GspSSD2.0 isothermal master mix (OptiGene Ltd.), 1 × primer mix (0.1 μM each outer primer Sal4-F3 [GAACGTGTCGCGGAAGTC] and Sal4-B3 [CGGCAATAGCGTCACCTT], 1.8 μM each inner primer Sal4-FIP [GCGCGGCATCCGCATCAATATCTGGATGGTATGCCCGG] and Sal4-BIP [GCGAACGGCGAAGCGTACTGTCGCACCGTCAAAGGAAC], and 1 μM each loop primer Sal4-LF [TCAAATCGGCATCAATACTCATCTG] and Sal4-LB [AAAGGGAAAGCCAGCTTTACG]; Integrated DNA Technologies, Coralville, IA), and 2 μL of DNA template.

The reaction was carried out at 65°C for 30 min followed by 98°C to 80°C with 0.05°C decrement per second in Genie II (OptiGene, Ltd.). Time-to-peak values (T _p; min) were obtained when fluorescence ratios reached peak amplification rates and annealing temperatures (T _a; °C) were determined from peaks of anneal derivative curves. Samples were considered screen-positive for Salmonella with T _p ≤ 20 min and T _a at 89°C ± 2°C. The LAMP assay was repeated once for each sample independently using separately prepared reaction master mixes.

BAM confirmation

Regardless of LAMP screening results, all samples including extraction controls were subjected to BAM culture confirmation (Andrews et al, 2022). These included selective enrichment in Rappaport-Vassiliadis (RV) medium and tetrathionate (TT) broth, selective plating on bismuth sulfite (BS) agar, xylose lysine desoxycholate (XLD) agar, and Hektoen enteric (HE) agar, and biochemical confirmation on triple sugar iron agar and lysine iron agar slants. Bacterial identities were confirmed by serology and MALDI Biotyper (Bruker) or LAMP.

WGS characterization

All confirmed Salmonella isolates were sequenced on MiSeq using the v3 reagent kit (Illumina, Inc., San Diego, CA) (Domesle et al, 2021). Salmonella serotypes were determined using SeqSero2 (Zhang et al, 2019), whereas AMR genes were identified using AMRFinderPlus (Feldgarden et al, 2019). Phylogenetic analysis was performed based on single nucleotide polymorphism (SNP) (Davis et al, 2015). All WGS data were submitted to the National Center for Biotechnology Information (NCBI) under BioProject PRJNA292661.

Antimicrobial susceptibility testing

Minimal inhibitory concentrations (MICs) for Salmonella isolates were determined by broth microdilution using Sensititre NARMS plate CMV5AGNF (Thermo Fisher Scientific) and interpreted following guidelines from the Clinical and Laboratory Standards Institute (CLSI, 2018, 2020), except for azithromycin and streptomycin, which have no CLSI breakpoints.

Data analysis

The analysis was performed as outlined in ISO 16140-2 section 5.1.3 for a paired sensitivity study (ISO, 2016). Positive agreement (PA), negative agreement (NA), positive deviation (PD), and negative deviation (ND) were calculated (Table 2) followed by sensitivity, relative trueness, and false positive (FP) ratio calculations. The acceptability limits of 4 for ND-PD and 8 for ND+PD were used for the two categories (raw meat and raw poultry) tested.

Table 2.

Interpretation of Results for This Paired Method Comparison Study per ISO Guidelines

Interpretation	Reference method (NARMS)	Alternative method (LAMP)	Confirmed alternative method (BAM)	This study (based on duplicate LAMP)	This study (based on single LAMP)
Positive agreement (PA)	+	+	Not needed	6	6
Negative agreement (NA)	−	−	Not needed	59	60
Negative deviation due to false negative alternative-method result (ND)	+	−	Not needed	0	0
Positive deviation (PD)	−	+	+	2	2
Negative agreement due to false positive alternative-method result (NA)	−	+	−	2	1

The first four columns follow definitions used by ISO 16140-2 in a method comparison study (ISO, 2016).

LAMP performance verification before adoption

Fourteen NARMS retail meat sites performed further verification of the Salmonella LAMP screening assay when incorporated into their respective laboratory workflows using a portion of routine NARMS samples. Agreement (%) between the NARMS protocol and LAMP was calculated followed by Cohen's Kappa calculation (Microsoft Excel, Redmond, WA). FP, true negative (TN), false negative (FN), and true positive (TP) numbers were used to calculate FP rate (FP/[FP+TN]) and FN rate (FN/[FN+TP]), along with specificity and sensitivity outputs (FDA, 2019).

Results

Salmonella positive rates differed between methods

Overall, 10 (14.5%) samples screened positive with LAMP (Table 1). Six (8.7%) were culture-confirmed by the NARMS protocol and six by the BAM method with overlap on four samples. No Salmonella isolates were recovered from samples screening negative with LAMP. Of the 34 Maryland samples, 6 (17.6%) screened positive with LAMP and 3M MDA 2—Salmonella performed at the site. Of those samples, 5 (14.7%) and 4 (11.8%) were confirmed by BAM and NARMS, respectively. One ground beef sample, screened positive with LAMP in one replicate and positive with the 3M assay, was negative by both culture methods. Of the 35 North Carolina samples, 4 (11.4%) screened positive with LAMP, with 1 (2.9%) and 2 (5.7%) of them also confirmed by BAM and NARMS methods, respectively. One ground beef sample, screened positive with LAMP (in both replicates), was negative by both culture methods.

Method metrics showed 100% LAMP sensitivity

Based on duplicate LAMP testing (Table 2), sensitivity ([PA+PD]/[PA+ND+PD]) was 100% (8/8) for the alternative LAMP method and 75% (6/8) for the reference NARMS method. Relative trueness (i.e., [PA+NA]/N = 67/69) was 97.1%, and FP ratio for LAMP (i.e., FP/NA = 2/61) was 3.3%. Based on single LAMP testing (differing by one ground beef sample from Maryland), sensitivity remained 100% for LAMP and 75% for NARMS, the relative trueness stayed at 97.1%, whereas the FP ratio for LAMP decreased to 1.6% (1/61). Considering the testing efficiency, we recommended single LAMP testing for implementation at NARMS retail meat sites. Both ND-PD and ND+PD were within the acceptability limits.

LAMP improved laboratory workflows

For samples screening positive with LAMP, T _p values averaged 8.8 ± 2.6 min (range 6.0–14.5 min) and T _a values averaged 90.0°C ± 0.4°C (range 88.7–90.5°C). Following the BAM protocol, repeated efforts were made to pick typical or atypical colonies on multiple occasions. This resulted in numerous presumptive Salmonella isolates (n = 121), which were not confirmed to be Salmonella by MALDI or LAMP. They were primarily Enterobacterales, including Hafnia alvei (44.6%), Proteus mirabilis (22.3%), and Morganella morganii (9.9%) (Table 3).

Table 3.

Distribution of Non-Salmonella Isolates Recovered Using the Bacteriological Analytical Manual Protocol

Genus	Species	No. (%) of isolates
Hafnia	alvei	54 (44.6)
Proteus	mirabilis	27 (22.3)
Morganella	morganii	12 (9.9)
Citrobacter	braakii	6 (5.0)
Citrobacter	freundii	4 (3.3)
Enterobacter	cloacae	3 (2.5)
Proteus	vulgaris	3 (2.5)
Providencia	alcalifaciens	3 (2.5)
Aeromonas	veronii	2 (1.7)
Other^a	Other^a	7 (5.8)
Combined	Combined	121 (100)

Other genus/species identified by MALDI Biotyper include one isolate each of Acinetobacter dijkshoorniae, Alcaligenes faecalis, Citrobacter gillenii, Escherichia coli, and Proteus hauseri, and two undetermined.

Salmonella isolate WGS and AST profiles matched within samples

A total of five serovars were identified among Salmonella isolates (n = 18), which were Anatum, Infantis, Kentucky, Meleagridis, and Senftenberg (Table 4). All isolates (n = 6) recovered using the NARMS protocol had RVR10 selective enrichment and XLT4 selective plating. Using the BAM protocol, all Salmonella isolates (n = 12) were recovered from samples undergoing RV medium selective enrichment, and most were from XLD agar (6/12) as the selective plating medium followed by BS (4/12) and HE (2/12) agars.

Table 4.

Characteristics of Salmonella Isolates Recovered from the Method Comparison Study

Sample ID^a	Sample type	Isolation method	Selective media	No. of isolates	Serovar	Antimicrobial susceptibility profile	Antimicrobial resistance gene/mutation^b
MD8	Chicken	NARMS	RVR10-XLT4	1	Infantis	CHL-GEN-NAL-TET	aac(3)-IVa, aph(4)-Ia, floR, sul1, tet(A), and gyrA_D87Y
MD8	Chicken	BAM	RV-XLD	1	Infantis	CHL-GEN-NAL-TET	aac(3)-IVa, aph(4)-Ia, floR, sul1, tet(A), and gyrA_D87Y
MD13	Ground turkey	BAM	RV-BS, RV-HE, RV-XLD	6	Anatum	AMP-GEN-FIS	aac(3)-VIa, aadA1, bla _HER-3, and sul1
MD16	Ground turkey	NARMS	RVR10-XLT4	1	Senftenberg	None	None
MD16	Ground turkey	BAM	RV-HE, RV-XLD	2	Senftenberg	None	None
MD33	Chicken gizzard	NARMS	RVR10-XLT4	1	Kentucky	None	aph(3″)-Ib and aph(6)-Id
MD33	Chicken gizzard	BAM	RV-XLD	1	Kentucky	None	aph(3″)-Ib and aph(6)-Id
MD34	Chicken heart	NARMS	RVR10-XLT4	1	Kentucky	None	aph(3″)-Ib and aph(6)-Id
MD34	Chicken heart	BAM	RV-XLD	1	Kentucky	None	aph(3″)-Ib and aph(6)-Id
NC10	Ground turkey	BAM	RV-XLD	1	Senftenberg	None	None
NC30	Pork	NARMS	RVR10-XLT4	1	Meleagridis	None	aac(2′)-IIa and fosA7.4
NC35	Chicken heart	NARMS	RVR10-XLT4	1	Kentucky	Tet	aph(3″)-Ib, aph(6)-Id, and tet(B)

No isolates were recovered from samples MD23 or NC24, which were loop-mediated isothermal amplification (LAMP)-positive in one and both replicates, respectively.

Predicted from whole-genome sequencing data.

Where multiple Salmonella isolates were recovered by NARMS and/or BAM protocols, the isolates matched in serovar, phenotypic AST, and WGS-predicted AMR genes/mutations (Table 4). Multidrug resistance to three or more antimicrobial classes were identified for Salmonella Infantis from MD8 and Salmonella Anatum from MD13. Salmonella Kentucky isolates from MD33 and MD34 carried aminoglycoside resistance genes aph(3″)-Ib and aph(6)-Id conferring resistance to kanamycin and/or streptomycin and Salmonella Meleagridis from NC30 carried aac(2′)-IIa and fosA7.4 conferring resistance to kasugamycin (an aminoglycoside) and fosfomycin, but none of these antimicrobials were in the NARMS CMV5AGNF plates. Phylogenetic analysis confirmed isolates from the same samples having less than five SNP differences (data not shown).

Multiple NARMS retail meat sites verified LAMP performance

From 14 NARMS retail meat sites, an overall agreement of 97% between LAMP screening and the NARMS culture method was achieved, ranging from 87.5% to 100% by site (Table 5). The Cohen's Kappa statistic was 0.905, suggesting excellent agreement. The overall FP rate was 1.2% (2/161) and FN rate was 10% (4/40), thus a specificity of 98.8% and sensitivity of 90%.

Table 5.

Loop-Mediated Isothermal Amplification Performance Verification at 14 National Antimicrobial Resistance Monitoring System Retail Meat Sites Using a Portion of Routine National Antimicrobial Resistance Monitoring System Samples

NARMS retail meat site	No. of samples tested	LAMP positive		LAMP negative		Agreement between LAMP and culture (%)
NARMS retail meat site	No. of samples tested	Culture positive (true positive)	Culture negative (false positive)	Culture positive (false negative)	Culture negative (true negative)	Agreement between LAMP and culture (%)
A	8	1	0	0	7	100
B	26	8	0	0	18	100
C	14	2	0	0	12	100
D	8	1	0	1	6	87.5
E	28	4	1	0	23	96.4
F	8	1	0	1	6	87.5
G	8	0	0	0	8	100
H	8	1	0	0	7	100
I	8	3	0	0	5	100
J	8	4	0	1	3	87.5
K	26	7	0	1	18	96.2
L	12	1	1	0	10	91.7
M	26	1	0	0	25	100
N	13	2	0	0	11	100
Combined	201	36	2	4	159	97

Discussion

This collaborative study highlighted the highly sensitive nature of Salmonella LAMP with 100% sensitivity demonstrated in raw meat and raw poultry. BAM and NARMS protocols were time-consuming and labor-intensive whereas trailing in sensitivity. The benefit of incorporating the rapid, robust, and highly sensitive LAMP screening method into NARMS to prioritize Salmonella isolation from presumptive positive samples was clearly demonstrated. LAMP could also serve as a reliable method to confirm all presumptive Salmonella isolates.

It is of the highest priority to select naturally contaminated samples for use in method comparison studies (ISO, 2016). However, few studies have compared the performance of Salmonella LAMP assays and culture methods using such samples (Yang et al, 2018). Several studies in Asian Countries reported >90% Salmonella LAMP sensitivity testing naturally contaminated meat and poultry. For example, one study in China (Zhang et al, 2012) tested 160 fresh chicken and pork samples by a hisJ-based LAMP, PCR, and culture, with positivity rates of 17.5%, 16.3%, and 18.8%, respectively, and an overall sensitivity of 93.6% for LAMP and 87.1% for PCR. Another study in Thailand (Srisawat and Panbangred, 2015) compared an stn-based LAMP and BAM in 60 chicken meat and minced pork samples and showed both methods having an 88.3% positivity rate.

Studies also compared the LAMP-based 3M MDA—Salmonella with culture. One study in Italy (Bonardi et al, 2013) reported low levels of Salmonella contamination (<0.3–2.1 most-probable-number [MPN]/g) in 10.5% of 200 meat samples with a relative sensitivity of 78.9% for 3M MDA compared with ISO 6579:2002 (Bonardi et al, 2013). Another study in Poland (Sarowska et al, 2016) tested 107 meat samples and reported 100% sensitivity of the 3M MDA assay compared with ISO 6579, both detecting four positive samples.

We previously evaluated the sensitivity of this Salmonella LAMP assay run on a bioluminescent-based platform in comparison with 3M MDA in spiked ground beef and ground turkey (Yang et al, 2016). Without pre-enrichment, LAMP could detect 10⁵ CFU/25 g in both matrices, whereas 3M MDA required 10⁶ CFU/25 g in ground beef and 10⁸ CFU/g in ground turkey. With 24-h pre-enrichment, both assays accurately detected 1 to 3 CFU/25 g of Salmonella within 20 min. We also verified the LAMP performance in raw pet food, that is, raw meat-based diets for pets, in comparison with BAM (Domesle et al, 2021). LAMP consistently detected low-level (<30 CFU/25 g) Salmonella spiked in five raw pet food matrices after pre-enrichment in BPW and lactose broth. In this study, LAMP agreed 100% with 3M MDA 2 in screening Salmonella from Maryland.

As a premier AMR monitoring program, NARMS publishes resistance data on a timely basis with online integrated reports/summaries, visual displays, and open access raw data (FDA, 2022). The prevalence of Salmonella in retail chickens ranged from 3% (270/8302) in 2017 to 21% (272/1320) in 2009, whereas that in ground turkey ranged from 5% (152/2907) in 2016 to 19% (246/1309) in 2008. During the same period, Salmonella prevalence in retail beef and pork remained <2% (number of samples ranged from 613 to 2204) (FDA, 2022). With such historical prevalence, implementing a rapid, reliable, and robust screening method could significantly reduce the number of samples needed for downstream culture work.

With LAMP, laboratory personnel can quickly identify presumptive positive samples and focus effort on culturing from samples likely to generate Salmonella isolates, a critical component of the NARMS program. As shown in Table 5, initial LAMP trials in 14 NARMS retail meat laboratories showed excellent agreement with NARMS culture. The method was regarded as simple and straightforward, requiring little hands-on time, and user-friendly. As laboratory personnel further develop proficiency with the LAMP method, the value of this rapid screening method would be even more appreciated.

Although this study tested a limited number of retail meat and poultry samples, it was apparent that RV medium was the superior selective enrichment broth for BAM compared with TT. An earlier study (Hughes et al, 2003) did a pairwise comparison of RVR10 and TT for the TECRA Salmonella Visual Immunoassay and did not find a significant difference (p > 0.05). A recent study (Broadway et al, 2021) reported a 1.4% Salmonella prevalence rate (out of 865 retail samples) using the USDA protocol (USDA, 2021), much more effective compared with BAM. Further evaluation of the BAM protocol in meat and poultry may be warranted.

We recently completed a method extension study (Domesle et al, 2022), expanding the Salmonella LAMP assay to 7500 Fast (Applied Biosystems), a widely used real-time quantitative PCR platform. Both Genie II and 7500 Fast generated reliable results against an extensive collection of inclusivity and exclusivity templates and in seven animal food matrices. GspSSD2.0 master mix had the fastest time-to-positive results (as early as 3.5 min). The cost of LAMP assay per sample using GspSSD2.0 was less than 2 U.S. dollars, whereas the cost of the small and portable Genie II instrument was competitive (10,000 U.S. dollars).

Adoption of this rapid and versatile screening method by the NARMS retail meat sites for screening Salmonella and confirming presumptive Salmonella isolates will enhance the program's mission to promote and protect public health by providing real-time information about emerging bacterial resistance, limit the spread of resistance, and aid the FDA in making regulatory decisions designed to preserve the effectiveness of antibiotics for humans and animals.

Conclusions

Salmonella detection from 69 NARMS retail meat and poultry samples using the NARMS protocol, LAMP, and BAM were compared. LAMP was 100% sensitive in reference to culture. RV medium was the superior selective enrichment broth for BAM compared with TT. Enterobacterales recovered by BAM was dominated by H. alvei (44.6%), P. mirabilis (22.3%), and M. morganii (9.9%). This study clearly demonstrated the benefit of a rapid, robust, highly sensitive, and specific molecular screening method in streamlining the laboratory workflow. After verification at multiple NARMS retail meat sites, the Salmonella LAMP screening assay has been successfully adopted by the NARMS program.

Footnotes

Acknowledgments

We thank Crystal Rice-Trujillo, Gordon Martin, Shawn McDermott, Sherry Ayers, and Yesha Shrestha for the help with sample/isolate receival, sample testing, WGS analysis, metadata curation, and WGS data submission. We are grateful to Jason Abbott for the support of this project. We also appreciate the entire NARMS retail meat program for providing timely verification data as we work toward the implementation of this new screening method for NARMS Salmonella testing.

Disclaimer

The views expressed in this article are those of the authors and do not necessarily reflect the official policy of the Department of Health and Human Services, the U.S. Food and Drug Administration, or the U.S. Government. Reference to any commercial materials, equipment, or process does not in any way constitute approval, endorsement, or recommendation by the Food and Drug Administration.

Authors' Contributions

Conceptualization, methodology, investigation, data curation, visualization, and writing—review and editing by S.R.Y. and K.J.D. Investigation, visualization, and writing—review and editing by R.C.M. Investigation and data curation by K.A.L. and E.H. Project administration and supervision by P.L. and S.T. Project administration and resources by C.K. Conceptualization and resources by E.A.S. Funding acquisition and resources by P.F.M. Conceptualization, methodology, project administration, validation, visualization, writing—original draft, and writing—review and editing by B.G.

Disclosure Statement

No competing financial interests exist.

Funding Information

This project was supported by internal funds from the U.S. FDA, Center for Veterinary Medicine, Office of Research.

References

Andrews

, Wang

, Jacobson

, et al. Chapter 5: Salmonella. In: Bacteriological Analytical Manual. 2022. Available from: www.fda.gov/Food/FoodScienceResearch/LaboratoryMethods/ucm070149.htm [Last accessed: July 26, 2022].

Antimicrobial Resistance Collaborators. Global burden of bacterial antimicrobial resistance in 2019: A systematic analysis. Lancet, 2022; 399(10325):629–655; doi: 10.1016/S0140-6736(21)02724-0

Bonardi

, Alpigiani

, Bacci

, et al. Comparison of an isothermal amplification and bioluminescence detection of DNA method and ISO 6579:2002 for the detection of Salmonella enterica serovars in retail meat samples. J Food Prot, 2013; 76(4):657–661; doi: 10.4315/0362-028X.JFP-12-313

Broadway

, Brooks

, Mollenkopf

, et al. Prevalence and antimicrobial susceptibility of Salmonella serovars isolated from U.S. retail ground pork. Foodborne Pathog Dis, 2021; 18(3):219–227; doi: 10.1089/fpd.2020.2853

CDC. Antibiotic resistance threats in the United States, 2019. 2019. Available from: https://www.cdc.gov/drugresistance/pdf/threats-report/2019-ar-threats-report-508.pdf [Last accessed: July 26, 2022].

Chen

, Wang

, Beaulieu

, et al. Rapid detection of viable salmonellae in produce by coupling propidium monoazide with loop-mediated isothermal amplification. Appl Environ Microbiol, 2011; 77(12):4008–4016; doi: 10.1128/AEM.00354-11

CLSI. Methods for Dilution Antimicrobial Susceptibility Tests for Bacteria That Grow Aerobically, 11th Ed. CLSI Standard M07. Clinical and Laboratory Standards Institute: Wayne, PA, USA, 2018.

CLSI. Performance Standards for Antimicrobial Susceptibility Testing, 30th Ed. CLSI Supplement M100. Clinical and Laboratory Standards Institute: Wayne, PA, USA, 2020.

Davis

, Pettengill

, Luo

, et al. CFSAN SNP pipeline: An automated method for constructing SNP matrices from next-generation sequence data. PeerJ Comput Sci, 2015; 1:e20.

10.

Domesle

, Yang

, Hammack

, et al. Validation of a Salmonella loop-mediated isothermal amplification assay in animal food. Int J Food Microbiol, 2018; 264:63–76; doi: 10.1016/j.ijfoodmicro.2017.10.020

11.

Domesle

, Young

, Ge

. Rapid screening for Salmonella in raw pet food by loop-mediated isothermal amplification. J Food Prot, 2021; 84(3):399–407; doi: 10.4315/JFP-20-365

12.

Domesle

, Young

, McDonald

, et al. Versatility of a Salmonella loop-mediated isothermal amplification assay using multiple platforms and master mixes in animal food matrices. J AOAC Int, 2022; doi: 10.1093/jaoacint/qsac059.

13.

Domesle

, Young

, Yang

, et al. Loop-mediated isothermal amplification for screening Salmonella in animal food and confirming Salmonella from culture isolation. J Vis Exp, 2020;(159):e61239; doi: 10.3791/61239

14.

FDA. Guidelines for the Validation of Analytical Methods for Nucleic Acid Sequence-Based Analysis of Food, Feed, Cosmetics and Veterinary Products, 1st Ed. 2019. Available from: https://www.fda.gov/media/121751/download [Last accessed: July 26, 2022].

15.

FDA. The National Antimicrobial Resistance Monitoring System. 2022. Available from: https://www.fda.gov/animal-veterinary/antimicrobial-resistance/national-antimicrobial-resistance-monitoring-system [Last accessed: July 26, 2022].

16.

Federal Task Force on Combating Antibiotic-Resistant

Bacteria

. National Action Plan for Combating Antibiotic-Resistant

Bacteria

, 2020–2025. 2020. Available from: https://aspe.hhs.gov/reports/national-action-plan-combating-antibiotic-resistant-bacteria-2020-2025 [Last accessed: July 26, 2022].

17.

Feldgarden

, Brover

, Haft

, et al. Validating the AMRFinder tool and resistance gene database by using antimicrobial resistance genotype-phenotype correlations in a collection of isolates. Antimicrob Agents Chemother, 2019; 63(11):e00483-00419; doi: 10.1128/AAC.00483-19

18.

, Domesle

, Yang

, et al. Multi-laboratory validation of a loop-mediated isothermal amplification method for screening Salmonella in animal food. Front Microbiol, 2019; 10:562; doi: 10.3389/fmicb.2019.00562

19.

, Mukherjee

, Hsu

, et al. MRSA and multidrug-resistant Staphylococcus aureus in U.S. retail meats, 2010–2011. Food Microbiol, 2017; 62:289–297; doi: 10.1016/j.fm.2016.10.029

20.

Hughes

, Dailianis

, Hill

, et al. Salmonella in foods: New enrichment procedure for TECRA Salmonella Visual Immunoassay using a single rv(R10) only, TT only, or dual rv(R10) and TT selective enrichment broths (AOAC official method 998.09): Collaborative study. J AOAC Int, 2003; 86(4):775–790

21.

ISO. ISO 16140-2:2016 Microbiology of the Food Chain—Method Validation—Part 2: Protocol for the Validation of Alternative (Proprietary) Methods Against a Reference Method. Geneva. 2016. Available from: https://www.iso.org/standard/54870.html [Last accessed: July 26, 2022].

22.

Karp

, Tate

, Plumblee

, et al. National Antimicrobial Resistance Monitoring System: Two decades of advancing public health through integrated surveillance of antimicrobial resistance. Foodborne Pathog Dis, 2017; 14(10):545–557; doi: 10.1089/fpd.2017.2283

23.

Kim

, Rodriguez

, Morris

, et al. Identification of a novel plasmid-borne gentamicin resistance gene in nontyphoidal Salmonella isolated from retail turkey. Antimicrob Agents Chemother, 2020; 64(11):e00867-00820; doi: 10.1128/AAC.00867-20

24.

Kumar

, Bansal

, Jaiswal

. Loop-mediated isothermal amplification (LAMP): A rapid and sensitive tool for quality assessment of meat products. Compr Rev Food Sci F, 2017; 16(6):1359–1378; doi: 10.1111/1541-4337.12309

25.

, Tyson

, Hsu

, et al. Long-read sequencing reveals evolution and acquisition of antimicrobial resistance and virulence genes in Salmonella enterica . Front Microbiol, 2021; 12:777817; doi: 10.3389/fmicb.2021.777817

26.

Mansour

, Ali

, Chase

, et al. Loop-mediated isothermal amplification for diagnosis of 18 World Organization for Animal Health (OIE) notifiable viral diseases of ruminants, swine and poultry. Anim Health Res Rev, 2015; 16(2):89–106; doi: 10.1017/S1466252315000018

27.

Mori

, Notomi

. Loop-mediated isothermal amplification (LAMP): Expansion of its practical application as a tool to achieve universal health coverage. J Infect Chemother, 2020; 26(1):13–17; doi: 10.1016/j.jiac.2019.07.020

28.

Nyirabahizi

, Tyson

, Tate

, et al. Northeastern U.S. Salmonella strains from retail meat are more prevalent and more resistant to antimicrobials. J Food Prot, 2020; 83(5):849–857; doi: 10.4315/JFP-19-549

29.

Sarowska

, Frej-Madrzak

, Jama-Kmiecik

, et al. Detection of Salmonella in foods using a reference PN-ISO method and an alternative method based on loop-mediated isothermal amplification coupled with bioluminescence. Adv Clin Exp Med, 2016; 25(5):945–950; doi: 10.1721/9/acem/63000

30.

Srisawat

, Panbangred

. Efficient and specific detection of Salmonella in food samples using a stn-based loop-mediated isothermal amplification method. Biomed Res Int, 2015; 2015:356401; doi: 10.1155/2015/356401

31.

Tadesse

, Li

, Mukherjee

, et al. Whole-genome sequence analysis of CTX-M containing Escherichia coli isolates from retail meats and cattle in the United States. Microb Drug Resist, 2018; 24(7):939–948; doi: 10.1089/mdr.2018.0206

32.

Tate

, Folster

, Hsu

, et al. Comparative analysis of extended-spectrum-beta-lactamase CTX-M-65-producing Salmonella enterica serovar Infantis isolates from humans, food animals, and retail chickens in the United States. Antimicrob Agents Chemother, 2017; 61(7):e00488-00417; doi: 10.1128/AAC.00488-17

33.

Tate

, Li

, Nyirabahizi

, et al. A national antimicrobial resistance monitoring system survey of antimicrobial-resistant foodborne bacteria isolated from retail veal in the United States. J Food Prot, 2021; 84(10):1749–1759; doi: 10.4315/JFP-21-005

34.

Tyson

, Nyirabahizi

, Crarey

, et al. Prevalence and antimicrobial resistance of enterococci isolated from retail meats in the United States, 2002 to 2014. Appl Environ Microbiol, 2018; 84(1):e01902-01917; doi: 10.1128/AEM.01902-17

35.

Tyson

, Tate

, Zhao

, et al. Identification of plasmid-mediated quinolone resistance in Salmonella isolated from swine ceca and retail pork chops in the United States. Antimicrob Agents Chemother, 2017; 61(10):e01318-17; doi: 10.1128/AAC.01318-17

36.

USDA. Microbiology Laboratory Guidebook. 4.11: Isolation and identification of Salmonella from meat, poultry, pasteurized egg, and Siluriformes (fish) products and carcass and environmental sponges. Washington, DC; 2021. Available from: https://www.fsis.usda.gov/sites/default/files/media_file/2021-08/MLG-4.11.pdf [Last accessed: July 26, 2022].

37.

Whitehouse

, Young

, Li

, et al. Use of whole-genome sequencing for Campylobacter surveillance from NARMS retail poultry in the United States in 2015. Food Microbiol, 2018; 73:122–128; doi: 10.1016/j.fm.2018.01.018

38.

WHO. Global action plan on antimicrobial resistance. 2015a. Available from: https://apps.who.int/iris/rest/bitstreams/864486/retrieve [Last accessed: July 26, 2022].

39.

WHO. Global Antimicrobial Resistance and Use Surveillance System (GLASS). 2015b. Available from: https://www.who.int/initiatives/glass [Last accessed: July 26, 2022].

40.

WHO. Salmonella (non-typhoidal) fact sheet. 2018. Available from: https://www.who.int/news-room/fact-sheets/detail/salmonella-(non-typhoidal) [Last accessed: July 26, 2022].

41.

WHO. Antimicrobial resistance. 2021. Available from: https://www.who.int/news-room/fact-sheets/detail/antimicrobial-resistance [Last accessed: July 26, 2022].

42.

Yang

, Chen

, Ge

. Detecting Salmonella serovars in shell eggs by loop-mediated isothermal amplification. J Food Prot, 2013; 76(10):1790–1796; doi: 10.4315/0362-028X.JFP-13-140

43.

Yang

, Domesle

, Ge

. Loop-mediated isothermal amplification for Salmonella detection in food and feed: Current applications and future directions. Foodborne Pathog Dis, 2018; 15(6):309–331; doi: 10.1089/fpd.2018.2445

44.

Yang

, Domesle

, Wang

, et al. Rapid detection of Salmonella in food and feed by coupling loop-mediated isothermal amplification with bioluminescent assay in real-time. BMC Microbiol, 2016; 16(1):112; doi: 10.1186/s12866-016-0730-7

45.

Yang

, Wang

, Jones

, et al. Evaluation of loop-mediated isothermal amplification for the rapid, reliable, and robust detection of Salmonella in produce. Food Microbiol, 2015; 46:485–493; doi: 10.1016/j.fm.2014.09.011

46.

Yin

, M'Ikanatha

, Nyirabahizi

, et al. Antimicrobial resistance in non-typhoidal Salmonella from retail poultry meat by antibiotic usage-related production claims - United States, 2008–2017. Int J Food Microbiol, 2021; 342:109044; doi: 10.1016/j.ijfoodmicro.2021.109044

47.

Zhang

, Pan

, Geng

, et al. A loop-mediated isothermal amplification method targets the hisJ gene for the detection of foodborne Salmonella . Eur Food Res Technol, 2012; 234(6):1055–1062; doi: 10.1007/s00217-012-1725-8

48.

Zhang

, den Bakker

, Li

, et al. SeqSero2: Rapid and improved Salmonella serotype determination using whole-genome sequencing data. Appl Environ Microbiol, 2019; 85(23):e01746-01719; doi: 10.1128/AEM.01746-19