Listeria monocytogenes is a Gram-positive bacterium that remains one of the leading causes of mortality among foodborne pathogens in the United States (Scallan et al., 2011; de Noordhout et al., 2014). Listeriosis, the disease caused by L. monocytogenes, largely effects vulnerable populations, including pregnant women, children, elderly, and people who are immunocompromised (Hamon et al., 2006; Buchanan et al., 2017). Due to the immune status of this vulnerable population and the invasive nature of the disease, antibiotics are critical to the successful treatment of listeriosis. The first choice antibiotic for the treatment is typically ampicillin (ß-lactam), alone or in combination with gentamicin (aminoglycoside) (Hof, 2004). While L. monocytogenes remains largely susceptible to a wide range of antibiotics, there have been reports of multidrug resistance in isolates recovered from food production environments (Prazak et al., 2002; Jorgensen et al., 2021).

Listeria monocytogenes is well adapted to agricultural environments, where it proliferates in decaying plant matter in the soil (Linke et al., 2014; Liao et al., 2021). It is also frequently recovered from diverse food processing environments (Gray et al., 2006; Freitag et al., 2009; Cherifi et al., 2018; Kim et al., 2018; Hurley et al., 2019; Sullivan and Wiedmann, 2020). Listeria monocytogenes prevalence in these environments presents a particular concern to processors of minimally processed ready-to-eat (RTE) products, such as fresh produce, that are prone to contamination during handling and processing (Gallagher et al., 2003; Farber et al., 2020). Often, control of L. monocytogenes in these environments relies on cleaning and sanitation programs. In the United States, sanitizers are regulated by the Environmental Protection Agency (EPA) and are required to achieve a 5-log reduction of a test organism on a food contact surface and a 3-log reduction on a non-food contact surface (Office of Chemical Safety and Pollution Prevention [OCSPP], 2012). Sanitizers are formulated at concentrations that will deliver a bactericidal effect, and in general many-fold higher than the minimum bactericidal concentration (MBC) for foodborne pathogens of concern (Cruz and Fletcher, 2012).

Quaternary ammonium compounds (QACs) are one group of sanitizers commonly used in the food industry. QACs are cationic antimicrobials with medium-to-long alkyl side chains (Gilbert and Moore, 2005; Wessels and Ingmer, 2013). The mechanism of bacterial cell inhibition is generally thought to be by the hydrophobic chain interpolating into the lipid bilayer of the cellular membrane leading to issues in osmoregularity and leaking of cell contents. The manufacturer recommended concentrations (MRCs) for commercial QAC (cQAC) application range from 200 to 800 ppm (Cruz and Fletcher, 2012; Boucher et al., 2021). The reported minimum inhibitory concentrations (MICs) for L. monocytogenes to QACs range from ≤2 ppm (Romanova et al., 2006; Martínez-Suárez et al., 2016; Møretrø et al., 2017; Yu et al., 2018; Roedel et al., 2019) to 40 ppm (Elhanafi et al., 2010; Dutta et al., 2013). Ideally, appropriate cleaning and sanitation procedures are paired with optimum hygienic design of the facility and equipment; however, this ideal combination is rare, particularly in fresh produce facilities. These less-than-ideal environments can create scenarios where L. monocytogenes could be exposed to sublethal sanitizer concentrations. Links between reduced susceptibility to QACs and antibiotic resistance have been reported for various foodborne bacteria, including L. monocytogenes (Heir et al., 1999; Braoudaki and Hilton, 2004; Langsrud et al., 2004; Rakic-Martinez et al., 2011; Gnanadhas et al., 2013; Kovacevic et al., 2013; Bansal et al., 2018; Kode et al., 2021). This phenomenon of cross-resistance can occur when microorganisms develop survival methods that are effective against different antimicrobial agents with similar mechanisms of action (SCENIHR, 2009). For example, a mutation or upregulation of a gene initiated by adaptation or exposure to one of the antimicrobials may subsequently affect the efficacy of another antimicrobial agent (Jiang et al., 2018; Amsalu et al., 2020).

We previously evaluated the prevalence and distribution of antimicrobial resistance (AMR) in L. monocytogenes from produce handling and processing facilities in the Pacific Northwest (Jorgensen et al., 2021). These isolates have diverse antibiogram profiles and represent clonal complexes associated with hypervirulent phenotypes. The present study investigated the potential for antibiotic cross-resistance to develop in these strains following adaptation to a commercial sanitizer (cQAC). Culture-based assays and whole genome sequence (WGS) comparisons were used to evaluate differences between the wild-type (WT) and cQAC-adapted (qAD) strains.

Produce processing facilities, which have frequent turnover of raw agricultural products, present a unique risk for L. monocytogenes contamination. In these environments, L. monocytogenes is largely controlled through effective cleaning and sanitation practices. Previous reports of sanitizer and antibiotic cross-resistance have proposed broader consequences for lapses in sanitation efficacy in the food industry. In the present study, we demonstrated the ability of L. monocytogenes strains, including those representing hypervirulent clonal complexes, to adapt and tolerate slightly elevated concentrations of a commercial QAC (MIC 3 ppm; MBC 4 ppm) compared to the WT strains.

The six L. monocytogenes isolates evaluated were adaptable to a commercial QAC at 1 ppm above their WT MIC (2 ppm) and MBC (3 ppm). Previous studies using benzalkonium chloride (BC) compound noted a greater increase in MIC following adaptations. Similar to WT strains in the present study, Aase et al. (2000) reported five L. monocytogenes isolates with an initial MIC of 2 ppm; however, these isolates were then adapted to MICs of 6–7 ppm. It is of interest to note that other isolates tested within their study, with an initial MIC of 6 and 7 ppm, did not exceed 7 ppm MIC following the adaptation process (Aase et al., 2000). This suggests a potential plateau at which isolates are no longer readily adaptable, as we have observed in our study. Similar to Aase et al. (2000), a study by To et al. (2002) reported BC adaptation of L. monocytogenes isolates from 1 or 4 ppm MIC to 6 or 8 ppm MIC, respectively. Others have reported adaptation to BC from 2 ppm in WT strains to 10 ppm MIC in adapted strains (Yu et al., 2018). While these adaptations achieved a higher MIC than was observed in the present study, additional components of the formulated commercial QAC product may have inhibited the adaptation from progressing to BC-adaptation levels. In fact, the majority of published studies on QAC and Listeria spp. have been performed using BC as opposed to commercial sanitizer preparations, such as the cQAC used in this study. BC contains an alkyl chain length distribution from C8 to C18. Previous research using narrower alkyl chain lengths has determined that Gram-positive bacteria, such as L. monocytogenes, are most affected at chain lengths of C12–C14 whereas longer chain lengths of C14–C16 are more effective against Gram-negative bacteria, such as Escherichia coli (Gilbert and Moore, 2005). Commercial QAC products have been formulated to contain QACs with a specific distribution of alkyl chain lengths (e.g., C14: 50%, C12: 40%, and C16: 10%) to optimize antimicrobial activity. These may also be formulated with additional ingredients, including divalent chelators like EDTA, that assist in destabilizing the bacterial cell membrane structures (Gilbert and Moore, 2005). The distribution of alkyl chain lengths and other additives in the commercial product formulation likely influence adaptation and complicate comparing MIC and MBC values from independent studies.

The primary mechanism of QAC adaptation and tolerance is largely attributed to the presence and upregulation of specific efflux pumps (Aase et al., 2000; Romanova et al., 2006; Kovacevic et al., 2013; Yu et al., 2018). Previous studies have reported the assistance of QAC specific efflux systems playing a role in resistance or adaptation, specifically the bcrABC cassette (Elhanafi et al., 2010; Dutta et al., 2013), emrE (Kovacevic et al., 2016), emrC (Kremer et al., 2017), qacC and qacH (Müller et al., 2013). cQAC tolerance of the adapted strains in the present study was lost with the addition of reserpine, indicating that efflux systems are playing a significant role in the experimentally achieved cQAC tolerance. While the previously reported genomic analyses of six WT strains examined here (Bland et al., 2021) revealed the lack of efflux pumps commonly linked to increased tolerance to QAC (e.g., bcrABC, qacC, qacH, emrE, or emrC), the restoration of the WT phenotype in the presence of reserpine, a known efflux pump inhibitor, suggests the role of other efflux pumps in the cQAC adaptation in these isolates.

Research studies have reported that two efflux pumps, mdrL and lde, ubiquitous in L. monocytogenes (Romanova et al., 2006; Yu et al., 2018), are upregulated following the exposure or adaptation to QAC, suggesting their role in tolerance toward the compound. Yu et al. (2018) found that the relative expression of mdrL was significantly higher in six L. monocytogenes strains following adaptation to a QAC. In contrast, Jiang et al. (2018) reported that MICs were not affected by the absence of lde in a deletion mutant. While it can be speculated that mdrL and lde may have played a role in adaptation, as they were both present in our studied isolates, there are additional, more broadly found multidrug resistant (MDR) efflux systems in Gram-positive bacteria that may be assisting in tolerance and cross-resistance (Tamburro et al., 2015; Schindler and Kaatz, 2016; Meier et al., 2017; Du et al., 2018). Some of these major families of MDR efflux systems have been described by Schindler and Kaatz (2016) including: (1) ATP-binding cassette, (2) major facilitator superfamily (lde), (3) multidrug and toxic extrusion systems (mdrL), (4) small multidrug resistance, and (5) resistance nodulation cell division (RND) family.

Other mechanisms of QAC tolerance have been explored in isolates without known efflux pumps associated with QACs. In a study by Meier et al. (2017) 45 L. monocytogenes isolates were found to possess high tolerance to BC (MIC > 20 ppm). When exposed to reserpine, only four isolates exhibited reduced tolerance to BC. They suggested that the majority of their BC-tolerant strains did not rely on efflux systems. Studies have suggested that the decrease in cell membrane permeability plays an important role in the reduction of tolerance to QACs (McDonnell and Russell, 1999; To et al., 2002; Wessels and Ingmer, 2013) and chlorine and selective antibiotics (Bansal et al., 2018). An increase in the size (cells were elongated and filamentous) and a shift in the fatty acid composition of the cell (shift from shorter to longer fatty acid following adaptation-resulting in decrease in fluidity of the cell), was reported following BC adaptation, suggesting changes in the cell structure/membrane as a possible tolerance mechanism (To et al., 2002). While outside the scope of the data presented here, due to the lack of major efflux systems that have previously been linked to assisting in QAC tolerance development, it would be of interest to consider changes in cellular permeability and fatty acid composition in our isolates as a possible mechanism aiding the adaptation and cross resistance process.

Following cQAC adaptation, we observed a shift in susceptibility to CIP, KAN, NOV, PEN, and to a lesser degree CLI and CHL amongst the six L. monocytogenes strains tested. In previous studies, L. monocytogenes adapted to QAC resulted in phenotypic changes comparable to those seen in the present study with some minor differences. Similar to our results, the shift in susceptibility toward CIP following QAC adaptation has been described by Rakic-Martinez et al. (2011) and Yu et al. (2018), who both reported an increase in MIC of CIP following adaptation to BC. The decreased susceptibility to KAN following BC adaptation reported by Romanova et al. (2006) for 3/4 tested isolates is similar to what we observed here, with the shift from sensitive to intermediate AMR profiles in 3/6 isolates. In contrast, Yu et al. (2018) did not see any changes in susceptibility to KAN in BC-adapted isolates. These data suggest that the adaptive mechanisms for KAN differ among isolates.

Quaternary ammonium compound and CIP tolerance/resistance has been previously reported (Rakic-Martinez et al., 2011; Kovacevic et al., 2013), and appears to be bidirectional, where isolates adapted to QAC become more resistant to CIP and isolates adapted to CIP become more tolerant to QAC. CIP is a fluoroquinolone; its mechanism of action involves preventing DNA separation prior to cell division. Resistance to quinolones in Gram-positive bacteria is typically due to mutations in the intercellular targets of the compound, gyrA and gyrB (DNA gyrase), parC and parE (topoisomerase IV) or decreased uptake into the cell via efflux pumps (Lampidis et al., 2002; Jiang et al., 2018). Jiang et al. (2018) found no mutations in the genes targeted by quinolones. However, transcription levels of lde increased across the four CIP adapted mutants, suggesting that the resistance development was largely due to decreased uptake into the cell (Jiang et al., 2018). Others have suggested that point mutations in fepR (regulator for fepA) are in part responsible for resistance to CIP (Wilson et al., 2018). FepA, a part of the multidrug and toxic compound extrusion (MATE) family of efflux pumps, is regulated by fepR (Guerin et al., 2014). When MICs of various antibiotics of WT L. monocytogenes and its fepR deletion mutant were compared, Guerin et al. (2014) reported susceptibility changes only in fluroquinolones. They also saw a change in BC susceptibility in the fepR deletion mutant, with an MIC increase from 4 to 8 ppm, suggesting that overexpression of the fepA efflux pump may play a role in BC tolerance. An Australian study by Wilson et al. (2018) found 2/100 L. monocytogenes possessing CIP resistance; in one of the isolates CIP resistance was linked to a mutation at nucleotide position 181 in fepR, resulting in PMSC. A point mutation in fepR was also seen in the other CIP resistant isolate, at nucleotide position 170, leading to an amino acid change (alanine to glycine). However, this mutation did not result in a PMSC, but rather this isolate possessed a full length fepR (Wilson et al., 2018). Similarly, our results found point mutations at different locations in fepR, some of which resulted in a PMSC and others that did not. In our isolates, both the location of the mutation as well as the resulting PMSC were downstream of what was previously seen by Guerin et al. (2014) and Wilson et al. (2018), resulting in a slightly longer gene. The multiple and independent mutations in fepR in the current study are strong evidence that loss of fepR function is important for L. monocytogenes adapting to cQAC. However, definite proof of the role of fepR and the level of its contribution to QAC adaptation would require deletion and complementation mutants.

Previous studies evaluating cross-resistance of QAC and antibiotics amongst L. monocytogenes have not, to our knowledge, evaluated NOV. NOV is less commonly used due to its decreased efficacy, but it has synergistic activity with tetracyclines, and can be used as an alternative to penicillin (May et al., 2017). Our results suggest that there is a relationship between adaptive QAC tolerance and NOV resistance, as a shift in susceptibility was seen across all six isolates. In particular, a dramatic shift from susceptible to resistant classification was observed in two strains, WRLP354 and WRLP530. Similar to quinolones, NOV inhibits gyrB subunit of the bacterial DNA gyrase enzyme involved in energy transduction (Hof et al., 1997; May et al., 2017). The phenotypic cross-resistance can be associated with mutations in the target for NOV or changes in cellular structure to reduce accumulation within the cell. In Gram-positive bacteria, such as Bacillus lincheniformis, morphological changes have been associated with NOV resistance. Specifically, in the presence of NOV the cells grow as long filaments as opposed to long chains (Robson and Baddiley, 1977). In Gram-negative bacteria, NOV resistance has been associated with a two-component regulatory system (baeSR), activating an efflux pump system (mdtABC). This is believed to lead to reduced NOV accumulation within the cell (Baranova and Nikaido, 2002; Nagakubo et al., 2002). In our isolates, we did not observe mutations in gyrAB. While morphological changes were not investigated, they cannot be ruled out as potential mechanisms aiding the resistance (Robson and Baddiley, 1977).

Our results also indicate that cQAC adaptation affects PEN resistance, with decreased zones of inhibition observed. PEN is a ß-lactam antibiotic, similar to AMP. However, unlike AMP, which is frequently used in the treatment of listeriosis, PEN is not typically used due to increasing reports of microbial resistance (Wilson et al., 2018). The isolates tested in this study were all classified as either having intermediate or resistant PEN profiles. However, while we observed a slight decrease in the zone of inhibition for AMP following cQAC adaptation, all tested isolates remained well within the susceptible classification (Figure 2). ß-lactam antibiotics, such as PEN, work to inhibit cell wall biosynthesis through interactions with penicillin-binding proteins (PBP). Listeria spp. have five PBP, and PBP3 is largely the target of ß-lactams (Vicente et al., 1990; Hof et al., 1997). Insertional mutagenesis in associated genes in L. monocytogenes EGDe (e.g., lmo0441, lmo0504, lmo1438, and lmo2229) resulted in acquired resistance toward PEN (Guinane et al., 2006). In S. pneumoniae, PEN resistance typically occurs from recombination events leading to mosaic PBP, which in turn can lead to reduced affinity for PEN and diminished effect on the cell (Blair et al., 2015).

It has been reported that exposure to QAC can increase the expression of virulence genes, such as prfA and inlA (Kastbjerg et al., 2010). Since cQAC adaptation impacted AMR profiles and led to genomic changes in some of our studied isolates, it was prudent to explore if virulence genes were affected, especially since our isolates belong to hypervirulent clonal complexes (Maury et al., 2016). While we did not look at the expression of virulence genes following cQAC adaptation, it is of note that no mutations in L. monocytogenes virulence genes were observed following cQAC adaptation.

Collectively, our data demonstrate the potential for L. monocytogenes isolates to develop cross-resistance between a cQAC and antibiotics representing different classes. While the increased MIC and MBC of cQAC-adapted isolates remained well below the manufacturer recommended concentration for the commercial product, it is not uncommon for bacterial cells to be exposed to lower or sublethal concentrations of sanitizers in the processing environment either by dilution, presence of organic matter (e.g., decreasing the efficacy of the sanitizer), or issues with the hygienic design of equipment or facility (e.g., resulting in microbial niches, biofilm formation or dilution effect of sanitizer). The isolates in the present study were adaptable to 1 ppm higher than the WT MIC. This minimal MIC increase suggests that adaptation of L. monocytogenes to a degree that would render the cQAC product ineffective is not likely in commercial settings. However, the present study was done with planktonic cells and so the additional tolerance and protection provided in biofilms, both in terms of initial susceptibility and degree of adaptation, likely differs. In a processing facility, the circumstances that may allow for exposure to minimal concentrations of cQAC are often associated with improper cleaning and inadequate sanitation practices that may not be addressing microbial niches. This can lower the effectiveness of the sanitizer or may be resulting in a dilution effect in some areas of the facility. Our data highlight that sublethal exposures to cQACs could have deleterious effects if adapted L. monocytogenes strains become implicated in human illness that requires antibiotic treatment. The potential for L. monocytogenes to develop cross-resistance to clinically relevant antibiotics following minimal adaptation to a formulated cQAC product is especially concerning. In particular, this trend amongst genotypes that are classified as hypervirulent and frequently involved in cases of listeriosis highlights the need to better understand the effect that sanitizer exposures and low-level adaptations may have on the AMR development and a potential public health risk.

The dataset presented in this study can be found in the NCBI Sequence Read Archive (SRA) under BioProject PRJNA771688.

JW-C and JK: conceptualization. RB and ER: data curation. RB, AW, JW-C, and JK: formal analysis and methodology. JK: funding acquisition and project administration. AW, JC, JW-C, and JK: supervision and writing – review and editing. AW: validation. RB: visualization and writing – original draft. All authors have read and agreed to the published version of the manuscript.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.