Requirement of lmo1930, a Gene in the Menaquinone Biosynthesis Operon, for Esculin Hydrolysis and Lithium Chloride Tolerance in Listeria monocytogenes

Listeria monocytogenes is a foodborne pathogen that is widely distributed in nature, having been isolated from a variety of sources such as soil, water, plant matter, and animals. In addition, L. monocytogenes is often detected in the regular sampling of food and food processing environments. The most common method for detecting L. monocytogenes is the use of selective enrichments. Both lithium chloride and esculin, in combination with ferric ammonium citrate, are utilized in several of the most commonly-employed selective enrichment schemes for L. monocytogenes. Here we report that transposon-based inactivation of lmo1930, one of the genes in the menaquinone biosynthesis operon, via transposon mutagenesis severely impaired the ability of L. monocytogenes to grow in the presence of lithium chloride or hydrolyze esculin, and conferred reduced growth and colony size. All phenotypes were restored upon genetic complementation. Thus, strains of L. monocytogenes with mutations leading to inactivation of lmo1930 may evade many commonly-used selective enrichment protocols employed in the detection of L. monocytogenes.


Introduction
Listeria monocytogenes is found ubiquitously in nature and is a facultative intracellular pathogen, capable of causing severe disease and death in susceptible individuals [1,2]. For these reasons, the rapid and reliable detection of L. monocytogenes is a key factor in the surveillance of this pathogen. Selective enrichment is typically used to detect and isolate L. monocytogenes from food and environmental samples. While there are numerous validated selective enrichment protocols for L. monocytogenes [3][4][5], many of these approaches share ingredients such as lithium chloride and esculin in combination with ferric ammonium citrate to differentiate L. monocytogenes from other microorganisms.
Esculin is a biopolymer that is widely distributed in several types of plants [6]. Given its presence in soil and plant matter, the ability to utilize this biopolymer may be useful for environmental survival of L. monocytogenes. The ability to hydrolyze esculin has been observed in diverse microorganisms and esculin has also been incorporated as a differential-selective agent into several selective enrichment protocols for L. monocytogenes. Esculin hydrolysis in the presence of ferric ammonium citrate results in a black precipitate that is easily detected with the naked eye in either liquid or solid media, allowing detection of putative L. monocytogenes or other Listeria spp. on agar media such as Modified Oxford (MOX) [7,8]. If esculin hydrolysis or lithium chloride tolerance were compromised, L. monocytogenes and other Listeria spp. would be undetectable in many of the currently employed enrichment protocols, with potentially important food safety and public health implications. In this work we investigated the involvement of lmo1930 in esculin hydrolysis and lithium chloride tolerance in L. monocytogenes

Mutant Library Construction and Mutant Screening
Mariner-based transposon mutant libraries, consisting of~3000 individual mutants, were constructed as previously described [9,10], in L. monocytogenes strains 2011L-2858 from the 2011 cantaloupe outbreak [11], and 2010L-1723 from a 2010 outbreak associated with celery [12]. Individual mutants were screened on TSAYE supplemented with esculin and ammonium ferric citrate. The transposon insertion in a mutant lacking the ability to hydrolyze esculin was localized using arbitrary PCR as described previously [9,10]. PCR primers flanking the putative insertion site were lmo1930F (5 -CTGATAATCGGCTTCACTTTG-3 ) and lmo1930R (5 -GCAGCTGCTGACACT ACTT-3 ). The subsequent PCR product was then visualized on a gel and sequenced (Genewiz Inc., South Plainfield, NJ, USA). The copy number of the transposon was established using qPCR as described [13].

Genetic Complementation
The coding region of the lmo1930 homolog in strain 2010-1723 (hereafter referred to as "lmo1930") was amplified using primers H1H3compsoeC (5 -TAAAAAGATAGGGGAATTAAGGATGAAACTTAACT TTTTATATGC-3 ) and H1H3compsoeD (5 -CAGTGCGGCCGCTTAATAATTTCTTTTATCTAAAAC AC-3 ). Since lmo1930 is located within an operon (Figure 1), the upstream region at the start of the operon that contained the in silico-predicted promoter was also amplified using primers H1H3compsoeA (5 -GTACGGTACCGCTACATAGACTTTTAAGCACT-3 ) and H1H3compsoeB (5 -CCTTAATTCCCCTATCTTTTTA-3 ). These two PCR products were then purified and joined together using splicing by overlap extension (SOE) PCR as previously described [14], to provide a fusion of the upstream regulatory region with the coding region of lmo1930 from strain 2010L-1723. The resulting product was digested with KpnI and NotI (New England Biolabs, Ipswitch, Ma, USA) and ligated into similarly-digested shuttle vector pPL2 [15]. The resulting recombinant plasmid pPLH1H3comp was transformed into chemically-competent Escherichia coli DH5α (Invitrogen, Waltham, MA, USA), resulting in strain E. coli DH5α_pPLH1H3comp. It was then extracted from E. coli DH5α_pPLH1H3comp and electroporated into mutant H1H3, resulting in strain H1H3::lmo1930.

Insertional Inactivation of lmo1930 Abolished Esculin Hydrolysis and Resulted in Impaired Growth.
The capacity to hydrolyze esculin was determined for 2392 and 1092 mariner-based transposon mutants of L. monocytogenes strains 2011L-2858 and 2010L-1723, respectively. The mutants were tested on TSAYE supplemented with esculin and ferric ammonium citrate (TSAYE-Es) to identify those unable to hydrolyze esculin and thus, lacking black pigmentation on the agar. This screening identified one esculin hydrolysis-deficient mutant of strain 2010L-1723, designated H1H3, while similar mutants of 2011L-2858 were not identified. Arbitrary PCR indicated that H1H3 harbored a transposon insertion in lmo1930 ( Figure 1). PCR with primers flanking this site resulted in a product that was ~1.4 kb (the approximate size of the mariner-based transposon) larger than the PCR product from the parental strain ( Figure 2). Sequencing of this PCR product localized the transposon insertion at nt 595 in the coding region of the lmo1930 homolog of 2010L-1723. Real-time PCR indicated that this mutant harbored a single copy of the transposon (data not shown).

Insertional Inactivation of lmo1930 Abolished Esculin Hydrolysis and Resulted in Impaired Growth
The capacity to hydrolyze esculin was determined for 2392 and 1092 mariner-based transposon mutants of L. monocytogenes strains 2011L-2858 and 2010L-1723, respectively. The mutants were tested on TSAYE supplemented with esculin and ferric ammonium citrate (TSAYE-Es) to identify those unable to hydrolyze esculin and thus, lacking black pigmentation on the agar. This screening identified one esculin hydrolysis-deficient mutant of strain 2010L-1723, designated H1H3, while similar mutants of 2011L-2858 were not identified. Arbitrary PCR indicated that H1H3 harbored a transposon insertion in lmo1930 ( Figure 1). PCR with primers flanking this site resulted in a product that was~1.4 kb (the approximate size of the mariner-based transposon) larger than the PCR product from the parental strain ( Figure 2). Sequencing of this PCR product localized the transposon insertion at nt 595 in the coding region of the lmo1930 homolog of 2010L-1723. Real-time PCR indicated that this mutant harbored a single copy of the transposon (data not shown). On TSAYE-Es, H1H3 was unable to hydrolyze esculin, producing white colonies that were also consistently smaller than the parental strain ( Figure 3). Smaller colonies were also produced on TSAYE (data not shown). Furthermore, H1H3 grew slower, with a prolonged lag phase in the nonselective liquid media TSBYE (Figure 4), and markedly lower final absorbance in BHI (Figure 4). On TSAYE-Es, H1H3 was unable to hydrolyze esculin, producing white colonies that were also consistently smaller than the parental strain ( Figure 3). Smaller colonies were also produced on TSAYE (data not shown). Furthermore, H1H3 grew slower, with a prolonged lag phase in the non-selective liquid media TSBYE (Figure 4), and markedly lower final absorbance in BHI ( Figure 4). Surprisingly, however, on BHI agar supplemented with esculin and ferric ammonium citrate, H1H3 was able to hydrolyze esculin and formed black colonies, even though the colony size was still considerably smaller than the parental strain (data not shown). Interestingly, this same behavior was observed in TSAYE supplemented with vitamin K2 (Figure 3). This finding suggests that nutrient components in BHI and vitamin K2 complement the esculin hydrolysis phenotype of the mutant that was noted in TSAYE-based media. The fact that H1H3 was capable of esculin hydrolysis on TSAYE supplemented with vitamin K2 suggests that menaquinone production was impacted in H1H3 and is likely linked to the esculin hydrolysis phenotype. However, BHI and vitamin K2 supplementation under the conditions that were employed appear to be inadequate for full restoration of colony size, possibly indicating that lmo1930 is required for the production of additional compounds that are needed for wild type levels of growth. The findings also suggest that the mutant would likely not have been identified had the transposon library been screened on BHIA with esculin and ferric ammonium citrate, instead of TSAYE-Es.

Genetic Complementation Confirms Role of lmo1930 in Esculin Hydrolysis
Genetic complementation of lmo1930 fused with its upstream regulatory region restored both colony size and esculin hydrolysis (Figure 3). This indicates that both of the observed phenotypes, i.e., reduced colony size and lack of esculin hydrolysis were both due to the inactivation of lmo1930. Furthermore, the complementation data demonstrate that these phenotypes were not due to potential polar effects of the transposon insertion on the previously-studied aroB or other genes downstream of lmo1930.

Inactivation of lmo1930 Markedly Reduces Tolerance to LiCl and Compromises Growth of L. Monocytogenes on LiCl-Containing Selective Media Such as MOX and RAPID'L.mono agar
As discussed above, esculin hydrolysis plays a key role in the differential detection of L. monocytogenes and other Listeria spp. in a number of selective media. MOX is one such medium that takes advantage of the black precipitate formed when esculin is hydrolyzed. Interestingly, we found that H1H3 not only failed to form back colonies but also lacked the ability to grow on MOX. The ability of H1H3 to grow on MOX was restored upon complementation with the intact lmo1930 ( Figure 5).
To identify which component(s) of MOX resulted in this lack of growth, H1H3 was tested for growth on BHIA with and without the MOX antimicrobial supplement, as well as on MOX base lacking the antimicrobial supplement. Growth of H1H3 was similar on BHIA with or without the MOX supplement (data not shown), suggesting that the MOX supplement was not a major contributor to the lack of growth of H1H3 on MOX.
MOX base contains LiCl (final concentration, 1.5%) as an agent to inhibit background microorganisms. Lithium is a chaotropic agent that has been shown to interfere with protein solubility [16], and we, therefore, hypothesized that the growth impairment of H1H3 on MOX may reflect impaired tolerance of the mutant to lithium. Indeed, when H1H3 was plated on BHI supplemented with 1.5% LiCl (BHI-LiCl), no growth was observed after 24 h, while the parental strain grew, albeit with smaller colonies than on BHIA (data not shown). By 48 h there was a four-log reduction in the population of H1H3 on BHIA-LiCl vs BHIA. This finding reveals a novel role of lmo1930 in LiCl tolerance of L. monocytogenes; impaired tolerance to LiCl was a major contributor to the lack of growth of the lmo1930 transposon mutant on MOX. of lmo1930.

Inactivation of lmo1930 Markedly Reduces Tolerance to LiCl and Compromises Growth of L. Monocytogenes on LiCl-Containing Selective Media Such as MOX and RAPID'L.mono agar
As discussed above, esculin hydrolysis plays a key role in the differential detection of L. monocytogenes and other Listeria spp. in a number of selective media. MOX is one such medium that takes advantage of the black precipitate formed when esculin is hydrolyzed. Interestingly, we found that H1H3 not only failed to form back colonies but also lacked the ability to grow on MOX. The ability of H1H3 to grow on MOX was restored upon complementation with the intact lmo1930 ( Figure  5). To identify which component(s) of MOX resulted in this lack of growth, H1H3 was tested for growth on BHIA with and without the MOX antimicrobial supplement, as well as on MOX base lacking the antimicrobial supplement. Growth of H1H3 was similar on BHIA with or without the MOX supplement (data not shown), suggesting that the MOX supplement was not a major contributor to the lack of growth of H1H3 on MOX.
MOX base contains LiCl (final concentration, 1.5%) as an agent to inhibit background microorganisms. Lithium is a chaotropic agent that has been shown to interfere with protein solubility [16], and we, therefore, hypothesized that the growth impairment of H1H3 on MOX may reflect impaired tolerance of the mutant to lithium. Indeed, when H1H3 was plated on BHI supplemented with 1.5% LiCl (BHI-LiCl), no growth was observed after 24 h, while the parental strain grew, albeit with smaller colonies than on BHIA (data not shown). By 48 h there was a fourlog reduction in the population of H1H3 on BHIA-LiCl vs BHIA. This finding reveals a novel role of Besides MOX, LiCl is a key selective agent in several other Listeria selective agar media including LPM, PALCAM, RAPID'L.mono agar, and Brilliance Listeria agar, as well as Fraser and Half-Fraser broths used as primary/secondary enrichments in the commonly-employed ISO method [8,17]. Growth of H1H3 on RAPID'L.mono plates was severely inhibited with barely-visible growth, where the culture was most heavily inoculated on the agar plate and no individual colonies could be detected ( Figure 5). The collective findings suggest that validated L. monocytogenes isolation methods that utilize LiCl would likely fail to recover strains with mutations in lmo1930.
Previously-characterized menaquinone biosynthesis mutants were shown to have primarily anaerobic metabolism [18]. In rich media such as BHI, inactivation of aroA led to reduced growth rates aerobically and anaerobically, while aroB mutants grew poorly aerobically and failed to grow at all anaerobically, and growth of an aroE mutant was severely impaired aerobically but improved anaerobically, albeit still impaired in comparison to the parental strain [18]. Colony size of H1H3 on BHIA or MOX was noticeably improved anaerobically, and the difference in colony size between H1H3 and the parental strain 2010L-1723 was not as striking anaerobically as under aerobic conditions, where in fact H1H3 was completely unable to grow on MOX ( Figure 6). Further biochemical and transcriptional analysis is needed to assess whether metabolism in H1H3 is primarily anaerobic and especially impaired in the presence of oxygen.
The gene investigated here, lmo1930 was not previously reported to be implicated in menaquinone biosynthesis. However, the lmo1930-inactivated mutant H1H3 exhibited phenotypes such as smaller colony size and growth deficiency in BHI. A similar growth deficiently was noted previously in aro mutants impaired in menaquinone biosynthesis [18]. Both the colony size and growth rate phenotypes were restored via genetic complementation. These findings are also supported by the transcriptional findings from Toledo-Arana et al. (2009) [19], that despite its proximity to the previously-studied aro genes, lmo1930 is the penultimate gene in a menaquinone biosynthesis operon which includes the preceding ORFs folE, lmo1932, and menH, and is terminated by ndk (Figure 1). The extent to which mutants of other genes in this operon, and potentially also aro mutants, may also be impaired in esculin hydrolysis, LiCl tolerance and growth on selective media remains to be determined. While our findings exhibited some similarities to previous work, determining the role of lmo1930 in menaquinone biosynthesis was not a major focus of the current study, and additional work would be required to conclusively establish that link.
broths used as primary/secondary enrichments in the commonly-employed ISO method [8,17]. Growth of H1H3 on RAPID'L.mono plates was severely inhibited with barely-visible growth, where the culture was most heavily inoculated on the agar plate and no individual colonies could be detected ( Figure 5). The collective findings suggest that validated L. monocytogenes isolation methods that utilize LiCl would likely fail to recover strains with mutations in lmo1930.
Previously-characterized menaquinone biosynthesis mutants were shown to have primarily anaerobic metabolism [18]. In rich media such as BHI , inactivation of aroA led to reduced growth rates aerobically and anaerobically, while aroB mutants grew poorly aerobically and failed to grow at all anaerobically, and growth of an aroE mutant was severely impaired aerobically but improved anaerobically, albeit still impaired in comparison to the parental strain [18]. Colony size of H1H3 on BHIA or MOX was noticeably improved anaerobically, and the difference in colony size between H1H3 and the parental strain 2010L-1723 was not as striking anaerobically as under aerobic conditions, where in fact H1H3 was completely unable to grow on MOX ( Figure 6). Further biochemical and transcriptional analysis is needed to assess whether metabolism in H1H3 is primarily anaerobic and especially impaired in the presence of oxygen. The gene investigated here, lmo1930 was not previously reported to be implicated in menaquinone biosynthesis. However, the lmo1930-inactivated mutant H1H3 exhibited phenotypes such as smaller colony size and growth deficiency in BHI. A similar growth deficiently was noted previously in aro mutants impaired in menaquinone biosynthesis [18]. Both the colony size and growth rate phenotypes were restored via genetic complementation. These findings are also supported by the transcriptional findings from Toledo-Arana et al. (2009) [19], that despite its proximity to the previously-studied aro genes, lmo1930 is the penultimate gene in a menaquinone biosynthesis operon which includes the preceding ORFs folE, lmo1932, and menH, and is terminated by ndk (Figure 1). The extent to which mutants of other genes in this operon, and potentially also aro mutants, may also be impaired in esculin hydrolysis, LiCl tolerance and growth on selective media In conclusion, the current study implicates lmo1930, a gene in an operon implicated in menaquinone biosynthesis, in two key phenotypes i.e., esculin hydrolysis and LiCl tolerance, required for the selective enrichment of L. monocytogenes in several most commonly-employed selective enrichment protocols. Our findings indicate that L. monocytogenes strains with mutations leading to inactivation of lmo1930, and possibly mutants of other genes involved in menaquinone biosynthesis, may evade detection in many of the selective enrichment protocols for L. monocytogenes that are currently employed. The incidence of such mutants in foods, the environment or illness remains currently unknown since they would not be detected as putative listeria with many enrichment protocols. Selective enrichment schemes that do not include LiCl and also utilize screening via molecular methods may be needed to ensure adequate surveillance of L. monocytogenes and other Listeria spp. in foods and the environment.