Listeria monocytogenes MenI Encodes a DHNA-CoA Thioesterase Necessary for Menaquinone Biosynthesis, Cytosolic Survival, and Virulence

Listeria monocytogenes is a Gram-positive, intracellular pathogen that is highly adapted to invade and replicate in the cytosol of eukaryotic cells. Intermediate metabolites in the menaquinone biosynthesis pathway are essential for the cytosolic survival and virulence of L. monocytogenes, independent of the production of menaquinone (MK) and aerobic respiration.

ABSTRACT

Listeria monocytogenes is a Gram-positive, intracellular pathogen that is highly adapted to invade and replicate in the cytosol of eukaryotic cells. Intermediate metabolites in the menaquinone biosynthesis pathway are essential for the cytosolic survival and virulence of L. monocytogenes, independent of the production of menaquinone (MK) and aerobic respiration. Determining which specific intermediate metabolite(s) are essential for cytosolic survival and virulence has been hindered by the lack of an identified 1,4-dihydroxy-2-naphthoyl-coenzyme A (DHNA-CoA) thioesterase essential for converting DHNA-CoA to DHNA in the MK synthesis pathway. Using the recently identified Escherichia coli DHNA-CoA thioesterase as a query, homology sequence analysis revealed a single homolog in L. monocytogenes, LMRG_02730. Genetic deletion of LMRG_02730 resulted in an ablated membrane potential, indicative of a nonfunctional electron transport chain (ETC) and an inability to aerobically respire. Biochemical kinetic analysis of LMRG_02730 revealed strong activity toward DHNA-CoA, similar to its E. coli homolog, further demonstrating that LMRG_02730 is a DHNA-CoA thioesterase. Functional analyses in vitro, ex vivo, and in vivo using mutants directly downstream and upstream of LMRG_02730 revealed that DHNA-CoA is sufficient to facilitate in vitro growth in minimal medium, intracellular replication, and plaque formation in fibroblasts. In contrast, protection against bacteriolysis in the cytosol of macrophages and tissue-specific virulence in vivo requires the production of 1,4-dihydroxy-2-naphthoate (DHNA). Taken together, these data implicate LMRG_02730 (renamed MenI) as a DHNA-CoA thioesterase and suggest that while DHNA, or an unknown downstream product of DHNA, protects the bacteria from killing in the macrophage cytosol, DHNA-CoA is necessary for intracellular bacterial replication.

INTRODUCTION

Listeria monocytogenes is a Gram-positive, intracellular pathogen capable of causing the severe disease listeriosis, which has an approximately 30% mortality rate (1). To cause disease, L. monocytogenes must invade host cells and access their cytosol to establish its replicative niche (2). L. monocytogenes, like other cytosolic pathogens, is remarkably well adapted to life within the eukaryotic cytosol (3). In contrast, bacteria that are not specifically adapted to the restrictive cytosolic environment are unable to survive and replicate (4–8). L. monocytogenes utilizes a myriad of virulence factors to facilitate entry into cells, where it is initially encapsulated in a phagosome (2, 9). The pore-forming toxin listeriolysin O (LLO), encoded by the gene hly, facilitates escape of L. monocytogenes into the cytosol (10). Infection is then disseminated to neighboring cells via the protein ActA, which hijacks host actin machinery, allowing L. monocytogenes to propel itself into adjacent cells, where it must again express LLO and a pair of phospholipases, PlcA and PlcB, for cytosolic access, thus reinitiating its infection cycle (11). The ability of L. monocytogenes to adapt to life within the cytosol is achieved through multiple mechanisms, including, but not limited to, modulation of specific metabolic processes, tight regulation of virulence factors to maintain the integrity of its intracellular niche, and avoidance of immune detection and host defense mechanisms (12–14).

Menaquinone (MK) biosynthesis is an integral metabolic pathway, and intermediate metabolite(s) in this pathway are required for L. monocytogenes cytosolic survival, independent of their known role in the production of MK (15, 16). MK acts as the sole lipid mediator of electron transport in L. monocytogenes during aerobic respiration and is required for a functional electron transport chain (ETC) (17). MK biosynthesis begins with the metabolite chorismate, generated through the shikimate pathway, and is processed through a series of concurrent enzymatic reactions leading to the production of 1,4-dihydroxy-2-naphthoyl-coenzyme A (DHNA-CoA), produced by the enzyme MenB. DHNA-CoA is then converted to 1,4-dihydroxy-2-naphthoate (DHNA) by an unknown thioesterase. Lastly, DHNA is prenylated and methylated by the enzymes MenA and MenG, respectively, to generate MK (17). DHNA can also be funneled into synthesis of an alternative quinone, demethylmenaquinone (DMK), through the activity of DmkA. DMK, in contrast to MK-dependent aerobic respiration, facilitates the flavin-dependent extracellular electron transfer (EET) pathway (18). A previous genetic screen revealed that the activity of MenB was required for the cytosolic survival of L. monocytogenes, while the enzymatic activity of MenA was not, suggesting aerobic respiration-independent functions of MK synthesis intermediates (15). Importantly, the EET pathway is dispensable for intracellular growth, systemic infection, and cytosolic survival, suggesting that the aerobic respiration-independent survival defects of ΔmenB mutants are also not DMK dependent (data not shown) (18). The lack of an annotated DHNA-CoA thioesterase made it impossible to know whether the virulence defects of the ΔmenB mutant relative to the ΔmenA mutant were due to a lack of DHNA or DHNA-CoA. Therefore, it is critical to identify the unknown DHNA-CoA thioesterase in L. monocytogenes to better understand the respiration-independent role(s) that MK intermediates play in its survival and virulence.

Recently, the DHNA-CoA thioesterase YdiI was identified and characterized in Escherichia coli (19). The L. monocytogenes 10403S genome contains a ydiI homolog, LMRG_02730, which also possesses the hotdog fold domain, a feature common to other acyl-CoA thioesterases (20, 21). In this study, we generated a genetic deletion of LMRG_02730 (renamed menI) in L. monocytogenes to determine its role in MK biosynthesis. Characterization of ΔmenI mutants, combined with biochemical analysis of purified MenI, demonstrated that LMRG_02730 encodes the missing DHNA-CoA thioesterase. Furthermore, direct comparison between ΔmenB, ΔmenI, and ΔmenA mutants allowed us to define the differential functions of DHNA-CoA, DHNA, or MK in vitro, ex vivo, and in vivo. Taken together, we have identified the missing DHNA-CoA thioesterase in the MK biosynthesis pathway in L. monocytogenes and revealed an essential role for menI in the cytosolic survival and pathogenesis of L. monocytogenes.

RESULTS

LMRG_02730 (Lmo2385, menI) is homologous to ydiI and is required for menaquinone biosynthesis.

Previous reports have highlighted the essentiality of MK biosynthetic intermediates in the survival and virulence of L. monocytogenes; however, the lack of a fully annotated pathway (Fig. 1A) has prevented the identification of the metabolite(s) responsible for the observed phenotypes (15, 16). Recently, the final unknown gene in the MK pathway, encoding a DHNA-CoA thioesterase, has been identified in the Gram-negative organism Escherichia coli and is referred to as ydiI (18). Sequence alignment analysis revealed that the L. monocytogenes 10403S genome contained a putative ydiI homolog, LMRG_02730 (Fig. 1B). Notably, both sequences contain a hotdog fold domain, a feature common to superfamily II thioesterases, as well as a conserved catalytic dyad at positions Q48 and E63 (20–22). To determine the role of LMRG_02730 in L. monocytogenes, we generated an unmarked genetic deletion mutant using allelic exchange as previously described (23). As production of MK is essential for aerobic respiration to generate a functional membrane potential (15, 17), we hypothesized that deletion of LMRG_02730 would result in an ablated membrane potential. Wild-type L. monocytogenes generated a robust membrane potential that is uncoupled by the proton ionophore CCCP. As previously reported, the MK-deficient mutants ΔmenB and ΔmenA were unable to generate a membrane potential due to their inability to aerobically respire (15). Consistent with its putative function as a DHNA-CoA thioesterase, mutants lacking LMRG_02730 similarly lacked a measurable membrane potential. Membrane potential was restored for all mutants tested via either biochemical (addition of exogenous MK) or genetic complementation (Fig. 1C).

FIG 1.

LMRG_02730 (Lmo2385, menI) is homologous to ydiI and is required for menaquinone biosynthesis. (A) Menaquinone biosynthetic pathway in Listeria monocytogenes. Corresponding gene locus numbers for L. monocytogenes strains EGD-e (Lmo) and 10403S (LMRG; used in this study) are listed underneath reaction arrows. SEPHCHC, (1R,2S,5S,6S)-2-succinyl-5-enolpyruvyl-6-hydroxy-3-cyclohexene-1-carboxylate; SHCHC, (1R,6R)-2-succinyl-6-hydroxy-2,4-cyclohexadiene-1-carboxylate; OSB, o-succinylbenzoate; DMK, demethylmenaquinone. (B) Sequence alignment comparing the Escherichia coli protein YdiI to the L. monocytogenes homolog. Alignment was generated using Clustal Omega and edited using Jalview. The coloring scheme depicts amino acid conservation from 100% conserved to least conserved (dark orange → white). *, Conserved catalytic amino acid residues. (C) The indicated L. monocytogenes strains were grown in aerated BHI cultures at 37°C until the mid-late logarithmic phase and then examined for membrane potentials using flow cytometry. The data represent the standard deviation of the means of three biological replicates normalized to the wild-type level (100%). Where indicated, cultures were supplemented with 5 μM MK.

To further assess whether LMRG_02730 encodes a functional DHNA-CoA thioesterase, we expressed and purified the protein to perform enzyme kinetics analysis. LMRG_02730, similar to its YdiI counterpart, possesses specific hydrolytic activity toward DHNA-CoA resulting in the production of DHNA and free coenzyme A (18). LRMG_02730 is an active DHNA-CoA thioesterase, with a K_m of 14.2 ± 8.8 μM, a k_cat of 14.9 ± 3.1 s⁻¹, and a second-ordered rate constant of 1.05 ± 0.08 × 10⁶ M^–1 · s⁻¹. Taken together, the lack of a functional membrane potential in ΔLMRG_02730 mutants coupled with its high DHNA-CoA hydrolase activity suggests that LMRG_02730 is likely the putative DHNA-CoA thioesterase involved in MK synthesis in L. monocytogenes. Based on these observations, we renamed LMRG_02730 MenI.

MenI is important for aerobic growth in vitro, and synthesis of DHNA-CoA is essential in minimal defined medium.

To further evaluate the role of MenI in aerobic respiration, we compared in vitro growth of the ΔmenI mutant to ΔmenB and ΔmenA, mutants deficient in the proteins directly upstream and downstream of MenI in MK synthesis. Under aerobic conditions in brain heart infusion (BHI) medium, growth of ΔmenI, ΔmenB, and ΔmenA mutants was similarly impaired compared to the wild-type strain and their genetic complements (Fig. 2A). In contrast, under anaerobic conditions, all strains grew to similar levels as the wild type (Fig. 2B). These data demonstrate that the in vitro growth kinetics of ΔmenI mutants phenocopies other mutants in the MK synthesis pathway and that these defects are due to their inability to aerobically respire.

FIG 2.

MenI is important for aerobic growth in vitro, and synthesis of DHNA-CoA is essential in minimal defined medium. (A and B) L. monocytogenes strains, as indicated, were grown in brain heart infusion (BHI) medium aerobically (A) and anaerobically (B) at 37°C. The OD₆₀₀ was monitored for 15 to 24 h. (C) The indicated L. monocytogenes strains were grown in minimal defined medium at 37°C and monitored for growth (OD₆₀₀) over 24 h. (D) Wild-type or ΔmenB mutant bacteria were grown in minimal defined medium at 37°C and monitored for growth (OD₆₀₀) over 24 h. Where indicated, cultures were supplemented with a 1% concentration of cell-free supernatants (SN) from wild-type, ΔmenI, or ΔmenA cultures grown in minimal medium. The data represent one of three biological replicates for all experiments.

In contrast to the phenotypes observed in rich BHI medium, ΔmenB mutants fail to replicate in minimal defined medium, whereas ΔmenA mutants are able to grow, suggesting that BHI medium contains one or multiple factors necessary to bypass DHNA-CoA or DHNA deficiency (16). To determine whether DHNA-CoA or DHNA are required for growth in defined minimal medium, we compared the growth of ΔmenI, ΔmenB, and ΔmenA mutants. As previously reported, ΔmenB mutants were incapable of growing, while ΔmenA mutants grew similar to the wild type, albeit not to the same final density (Fig. 2C). ΔmenI mutants displayed an intermediate phenotype such that they grew to a similar final density as ΔmenA mutants but at a significantly decreased rate. These data suggest that although DHNA-CoA is sufficient to promote growth in defined minimal medium, the ability to produce DHNA provides an additional fitness advantage (Fig. 2C).

DHNA can be secreted by Lactobacillus spp. and Propionibacterium spp. and can be scavenged by Bifidobacterium spp. and Streptococcus spp. as a shared resource in complex communities (24–26). L. monocytogenes can both secrete and scavenge DHNA (16). Consistent with previous reports, cell-free supernatants from aerobically grown cultures of the wild type or the ΔmenA mutant in defined minimal medium rescued the growth of ΔmenB in minimal defined medium (Fig. 2D). In contrast, although ΔmenI mutants can grow in minimal defined medium, cell-free supernatants of ΔmenI mutants are unable to rescue growth of the ΔmenB mutant under these conditions, consistent with their inability to synthesize DHNA (Fig. 2D). Furthermore, these data suggest that the hydrolysis of DHNA-CoA to form DHNA is required for its subsequent secretion. Taken together, these data confirm the role of MenI for aerobic growth and DHNA secretion and further begin to define differential requirements for DHNA-CoA or DHNA under different growth conditions.

DHNA is necessary for cytosolic survival in macrophages ex vivo.

The activity of MenB is required for the cytosolic survival and virulence of L. monocytogenes in murine bone-marrow derived macrophages (BMDMs), whereas the activity of MenA is dispensable (15). To determine whether DHNA-CoA or DHNA is necessary for cytosolic survival of L. monocytogenes, we infected BMDMs with ΔmenB, ΔmenI, and ΔmenA mutants carrying a bacteriolysis reporter plasmid to assess the levels of cytosolic killing (15, 27). L. monocytogenes mutants deficient in LLO (Δhly) are unable to escape the host phagosome and thus act as a negative control. As previously reported (15), ΔmenB mutants were killed in the cytosol approximately 4- to 6-fold more frequently than the wild-type strain, while ΔmenA mutants did not lyse under the same conditions (Fig. 3). Infection with ΔmenI mutants led to levels of bacteriolysis similar to that seen in the ΔmenB mutant (Fig. 3). Cytosolic survival of both ΔmenB and ΔmenI mutants was restored to wild-type levels upon genetic complementation. These data suggest that the synthesis of DHNA, or an unknown downstream product of DHNA, is essential for the survival of L. monocytogenes in macrophages.

FIG 3.

DHNA is necessary for cytosolic survival in macrophages ex vivo. L. monocytogenes MK-deficient mutants and their complements (MOI of 10) were tested for cytosolic survival in immortalized Ifnar^−/− macrophages over a 6-h infection. The data are normalized to wild-type levels of bacteriolysis and presented as the standard deviation of the means from four independent experiments. NS, not significant.

DHNA-CoA but not DHNA is necessary for intracellular replication and plaque formation ex vivo.

Although DHNA is required to protect bacteria from killing in the macrophage cytosol, only a subset of bacteria ultimately undergo bacteriolysis. To further define the relative requirements for DHNA-CoA or DHNA in the ex vivo virulence of L. monocytogenes, we assessed intracellular growth of ΔmenB, ΔmenI, and ΔmenA mutants in primary bone marrow-derived macrophages. As previously reported (15, 28), ΔmenB mutants are unable to replicate throughout the course of the infection (Fig. 4A). In contrast, both the ΔmenI and ΔmenA mutants were able to replicate in the macrophage cytosol, albeit not to the levels seen during wild-type infection (Fig. 4A). These data suggest that while DHNA is required to promote optimal cytosolic survival of L. monocytogenes in the macrophage cytosol, only DHNA-CoA is required for bacterial replication in this environment.

FIG 4.

DHNA-CoA but not DHNA is necessary for intracellular replication and plaque formation in fibroblasts ex vivo. (A) Intracellular growth of indicated L. monocytogenes strains was determined in BMDMs following infection at an MOI of 0.2. Growth curves are representative of at least three independent experiments. Error bars represent the standard deviation of the means of technical triplicates within the representative experiment. (B) L2 fibroblasts were infected with indicated L. monocytogenes strains (MOI of 0.5) and were examined for plaque formation 4 days postinfection. Data are normalized to wild-type plaque size and represent the standard deviation of the means from three independent experiments. ND, not detected.

To further assess the relative contributions of DHNA-CoA and DHNA to L. monocytogenes infection, we measured each mutant’s ability to form plaques in a monolayer of murine L2 fibroblasts (29). To form plaques, L. monocytogenes must invade cells, escape from the phagosome, replicate in the fibroblast cytosol, and ultimately spread to neighboring cells using ActA-dependent actin-based motility. Consistent with previous data (15), ΔmenB mutants are unable to form any detectable plaques, whereas ΔmenA mutants form plaque sizes comparable to that of the wild type (Fig. 4B). Importantly, plaque formation of the ΔmenI mutant phenocopies the plaque size seen with ΔmenA infection (Fig. 4B). Taken together, these data demonstrate that both DHNA and MK are dispensable for intracellular bacterial replication in macrophages and plaquing in fibroblasts. In contrast, ΔmenB mutants are unable to replicate in macrophages and do not form plaques in a fibroblast monolayer, demonstrating that DHNA-CoA is essential for intracellular replication and plaque formation.

Deficiency in DHNA-CoA or DHNA results in organ-specific phenotypes in vivo.

Previous reports have demonstrated that mutants deficient in MK synthesis are highly attenuated in vivo in a murine model of listeriosis (15, 16, 27, 30). Importantly, ΔmenB mutants lacking DHNA-CoA and DHNA are significantly more attenuated than ΔmenA mutants lacking only MK and aerobic respiration (16). As defined above, DHNA is required for cytosolic survival in macrophages (Fig. 3) but is dispensable for cytosolic replication (Fig. 4). Therefore, to determine if DHNA and cytosolic survival contribute to virulence in vivo, we utilized an acute murine listeriosis infection model to assess virulence of ΔmenB, ΔmenI, and ΔmenA mutants and their genetic complements. As previously reported, ΔmenB mutants were ∼10,000 to 100,000-fold more attenuated than wild-type bacteria and ∼100-fold more attenuated than ΔmenA mutants in both the spleen and liver, demonstrating an aerobic respiration-independent function for DHNA-CoA and/or DHNA in virulence (Fig. 5). Infection with ΔmenI mutants demonstrated a tissue-specific requirement for DHNA; in the spleen, ΔmenI mutants phenocopied burdens seen with ΔmenA, suggesting that in the spleen, DHNA-CoA synthesis, but not DHNA, is necessary for virulence in a respiration-independent manner. In contrast, in the liver, ΔmenI mutants phenocopied ΔmenB mutants such that they were 100,000-fold attenuated relative to wild-type bacteria, whereas ΔmenA mutants were ∼1,000-fold attenuated, suggesting a specific role for DHNA synthesis in colonization, replication, and/or survival of L. monocytogenes in this tissue (Fig. 5). The attenuated bacterial burdens of ΔmenB, ΔmenI, and ΔmenA mutants in vivo all returned to wild-type levels upon genetic complementation (Fig. 5). These data suggest that DHNA, independent of respiration, is required for colonization, replication, and/or survival in the liver during disseminated L. monocytogenes infection. In contrast, only DHNA-CoA is necessary for colonization of the spleen in a respiration-independent manner.

FIG 5.

Deficiency in DHNA-CoA or DHNA results in organ-specific phenotypes in vivo. (A ad B) Bacterial burdens from the spleen (A) and liver (B) were enumerated at 48 h post-intravenous infection with 1 × 10⁵ bacteria. Data are representative of results from two separate experiments. Horizontal bars represent the limits of detection, and the bars associated with the individual strains represents the median of the group.

DISCUSSION

Recently, there has been a growing appreciation for the role of the MK biosynthesis pathway in the survival and virulence of L. monocytogenes (15, 16, 27, 29). Intermediates in the MK pathway have been shown to be essential for L. monocytogenes virulence, independent of their role in the production of MK and aerobic respiration (15, 16). Until now, determining whether DHNA-CoA or DHNA is the required metabolite has not been possible due to a missing gene in this pathway, a DHNA-CoA thioesterase. Here, we identified the missing gene in this pathway for L. monocytogenes, LMRG_02730. Using genetic and biochemical approaches, we confirmed that LMRG_02730 encodes a DHNA-CoA thioesterase, and as such we renamed it menI. Additionally, infection models both ex vivo and in vivo revealed differential requirements for the synthesis of DHNA-CoA versus DHNA for the survival, replication, and virulence of L. monocytogenes.

The canonical role of the MK biosynthesis pathway is to generate MK, which then is inserted into the bacterial membrane to facilitate electron transport from NADH dehydrogenases to cytochrome oxidases in the ETC, ultimately generating a membrane potential that can drive ATP synthesis during aerobic respiration (17). A critical outcome of ETC activity is the regeneration of the cofactor NAD⁺ from the pool of reduced NADH produced during upstream metabolic processes (31). DHNA-CoA is the first metabolite in the MK biosynthesis pathway with a complete naphthoquinol ring (Fig. 1A), and DHNA itself is known to be redox capable (32, 33). Additionally, extracellular DHNA can be scavenged by Bifidobacterium spp. and Streptococcus spp. as a shared resource in complex communities, and it is also known to be utilized by L. monocytogenes strains deficient in DHNA-CoA to stimulate growth in minimal defined medium (16, 23–25). Our data reveal that synthesis of DHNA-CoA, not DHNA, is necessary to promote growth of L. monocytogenes in minimal defined medium under aerobic conditions (Fig. 2C). Similarly, DHNA-CoA, but not DHNA or MK, is required for intracellular growth and plaque formation in L2 fibroblast monolayers (Fig. 4). One hypothesis for why DHNA and MK could be dispensable for intracellular replication is that perhaps L. monocytogenes utilizes the redox-capable naphthoquinol ring of DHNA-CoA to act as a cofactor to facilitate redox reactions necessary to restore, at least partially, the intracellular homeostasis of the NAD⁺/NADH ratio. Whether this process would utilize either of the two previously identified NADH dehydrogenases (LMRG_02734 and LMRG_02183) or a yet to be defined NADH dehydrogenase is unknown. Understanding the MK-independent, intracellular functions of DHNA-CoA and its role in redox balance will help illuminate novel metabolic pathways and mechanisms of replication of L. monocytogenes both in vitro and in the context of infection.

While the contribution of DHNA-CoA appears to be critical for bacterial replication in vitro, ex vivo, and in vivo, DHNA plays a more specific role in promoting bacterial survival in the cytosol of macrophages (Fig. 3). This raises the intriguing question of whether DHNA acts intrinsically in the bacteria or through manipulation of the host. One possibility is that there is another, yet to be identified, DHNA-dependent metabolite produced independently of MenA or DmkA that contributes to cytosolic survival, which would suggest a bacterial-intrinsic function for DHNA in protection from cytosolic killing. An alternative hypothesis is that DHNA secreted by the bacteria in the cytosol modulates the host response. Recent reports have demonstrated that DHNA derived from Propionibacterium spp. can be used as an immune-modulating molecule to reduce inflammation due to colitis in mice by suppressing macrophage-derived proinflammatory cytokines (34, 35). Furthermore, DHNA is a known agonist of the aryl hydrocarbon receptor (AhR), which is a regulator of both cellular metabolism and inflammatory responses (36, 37). Specifically, in macrophages, AhR activation promotes expression of interleukin-10 (IL-10), a potent anti-inflammatory cytokine (38). L. monocytogenes can secrete DHNA both in vitro and within the cytosol of host cells, where it can rescue the survival of DHNA-deficient strains of L. monocytogenes during coinfection (16). Our data suggest that the enzymatic conversion of DHNA-CoA to DHNA is required for its subsequent secretion (Fig. 2D) and that synthesis of DHNA, not DHNA-CoA, is required for the cytosolic survival of L. monocytogenes in macrophages. How DHNA is secreted from L. monocytogenes and whether the bacteria intentionally secrete DHNA to act as an agonist of AhR to modulate cellular immunity remains to be determined.

Interestingly, our data also revealed organ-specific requirements for DHNA-CoA and DHNA (Fig. 5). Synthesis of DHNA is especially critical for the survival of L. monocytogenes in the liver of mice (Fig. 5B), however; the reasons for this are currently unclear. The liver plays a key role in sterilizing and detoxifying the bloodstream; over 60% of bacteria administered via intravenous injection in an animal model are removed from the bloodstream and trapped in the liver 10 min postinjection (39). Kupffer cells are liver-resident macrophages and are part of the first line of defense against bloodborne bacteria, displaying a pronounced endocytic and phagocytic capacity for clearing bacteria (39, 40). Indeed, L. monocytogenes is ingested by Kupffer cells during infection, where they induce the necroptotic death of the macrophages, leading to monocyte recruitment and an antibacterial type 1 inflammatory response (41). It is possible that liver Kupffer cells or the recruited inflammatory monocytes represent a more inhospitable environment than the spleen that requires the synthesis and secretion of DHNA to modulate the local immune response and prevent intracellular bacteriolysis to promote the survival of L. monocytogenes. Future studies focused on assessing the activity of cytosolic host defenses in different myeloid cell subsets and how DHNA might specifically impact these cells could help to illuminate novel cell-autonomous defense pathways. Understanding the interplay between DHNA secretion and its known role in modulating inflammatory processes will be critical to our understanding of how DHNA synthesis promotes cytosolic survival and virulence of L. monocytogenes in a tissue-specific manner.

In summary, we have identified the unknown DHNA-CoA thioesterase in L. monocytogenes and have provided further evidence for the key role that the MK biosynthetic intermediates play in the survival and virulence of L. monocytogenes, independent of their known role in aerobic respiration. It remains to be determined whether the essential functions of DHNA-CoA, DHNA, or an unknown downstream product of DHNA in vitro or during infection are correlated with basic bacterial cell physiology and/or to direct modulation of host responses. The MK biosynthesis pathway is widely conserved across the microbial kingdom, and as such, identifying the underlying mechanism(s) that metabolites in this pathway play with respect to both bacterial physiology and host-pathogen interactions is critical to the development of novel therapeutic interventions.

MATERIALS AND METHODS

Bacterial strains, plasmid construction, and growth conditions in vitro.

L. monocytogenes strain 10403S is referred to as the wild-type strain, and all other strains used in this study are isogenic derivatives of this parental strain. Vectors were conjugated into L. monocytogenes using Escherichia coli strain S17 or SM10 (42). Allelic exchange was performed to produce in-frame deletions of genes in L. monocytogenes using the suicide plasmid pKSV7-oriT as previously described (22, 43). To generate the ΔmenI mutant, MK was supplemented at a concentration of 5 μM during aerobic growth of the merodiploid strain in BHI medium. The addition of MK bypassed the aerobic growth defect of the mutant after the second homologous recombination step and helped facilitate its final isolation. The integrative vector pIMK2 was used for constitutive expression of L. monocytogenes genes for complementation (44).

L. monocytogenes strains were grown at 37°C or 30°C in brain heart infusion (BHI) medium (catalog number 237500; VWR) or minimal medium supplemented with glucose as the sole carbon source. Minimal medium is identical to the minimal medium described by Whiteley et al. (45). Escherichia coli strains were grown in Luria-Bertani (LB) broth at 37°C. For anaerobic medium studies, BHI was autoclaved and stored in an anaerobic chamber (Coy Laboratory Products) at least 24 h before use with a gas mixture of 5% hydrogen, 20% carbon dioxide, and 75% nitrogen. L. monocytogenes strains were then cultured in 10 ml of conditioned BHI medium in anaerobic Hungate tubes at 37°C. Antibiotics were used at concentrations of 100 μg/ml carbenicillin (catalog number IB02020; IBI Scientific), 10 μg/ml chloramphenicol (catalog number 190321; MP Biomedicals), 2 μg/ml erythromycin (catalog number 227330050; Acros Organics), or 30 μg/ml kanamycin (catalog number BP906-5; Fisher Scientific) when appropriate. Medium, where indicated, was supplemented with 5 μM 1,4-dihydroxy-2-naphthoate (DHNA) (catalog number 281255; Sigma) or 5 μM menaquinone (MK) (catalog number V9378; Sigma).

Measuring bacterial membrane potential.

L. monocytogenes strains were grown in 3-ml cultures in BHI medium at 37°C with shaking overnight. The overnight cultures were then diluted 1:100 (25 ml final volume) in 125-ml baffled flasks to mid-late logarithmic phase (optical density at 600 nm [OD₆₀₀], 0.4 to 0.6) in BHI medium at 37°C; shaking at 250 rpm. Cultures were diluted to a final concentration of 1 × 10⁶ CFU/ml in sterile 1× phosphate-buffered saline (PBS) and were stained in the dark for membrane potentials using 30 μM membrane potential indicator dye DIO2(3) (3,3′-diethyloxacarbocyanine iodide) (catalog number 320684; Sigma) for 30 min. The proton ionophore carbonyl cyanide 3-chlorophenylhydrazone (CCCP) (catalog number C2759; Sigma) was used to depolarize the bacterial membranes to act as a negative control and was added at a final concentration of 5 μM to control tubes prior to incubation with the membrane potential dye DIO2(3). Samples were analyzed on a BD LSR-II flow cytometer as previously described in Chen et al. (15).

Synthesis of DHNA-CoA.

DHNA-CoA was synthesized as previously described (46). Briefly, 1,4-dihydroxy-2-naphthoate (102 mg, 0.5 mmol, 1 eq) was first activated with N-hydroxysuccinimide (NHS-OH, 63.3 mg, 0.55 mmol, 1.1 eq) in the presence of ssN′-dicyclohexylcarbodiimide (113 mg, 0.55 mmol, 1.1 eq) in freshly distilled tetrahydrofuran (THF; 2.5 ml) for 18 h at room temperature under magnetic stirring. After filtration and evaporation at reduced pressure, the activated product was further purified by flash column chromatography (5% methanol in dichloromethane) to obtain a brownish viscous solution (55.1 mg, 36.6%). ¹H-NMR of the purified DHNA-NHS ester (400 MHz, chloroform-d) was conducted as follows: δ 10.28 (s, 1H), 8.30 to 8.16 (m, 2H), 7.74 to 7.66 (m, 2H), 7.11 (s, 1H), 2.83 to 2.73 (m, 4H).

Subsequently, DHNA-NHS (32 mg, 0.106 mmol, 4.07 eq) was dissolved in THF (1.5 ml) and added dropwise to an aqueous solution of coenzyme A (20 mg, 0.026 mmol, 1 eq) adjusted to pH 8.0 with sodium bicarbonate in 1.5 ml over a period of 1 h at room temperature. Finally, the reaction was acidified with trifluoroacetic acid to pH 2.0 and extracted thrice with 10 ml dichloromethane. The aqueous solution was then purified by reversed-phase high-pressure liquid chromatography (HPLC) and lyophilized to yield DHNA-CoA as a brown powder (25.6 mg, 25.2%).

MenI expression, purification, and activity assay.

To express the protein, an overnight starter culture was prepared from a single colony of E. coli BL21 harboring the pET20b-menI plasmid containing a hexahistidine-tag at the N terminus in 4 ml LB containing 50 μg/ml ampicillin. The overnight culture was used to inoculate a 1-liter culture in the same medium at 37°C. When the OD₆₀₀ reached 0.6, isopropyl β-d-1-thiogalactopyranoside (IPTG, 0.2 mM) was added to induce expression of MenI at 25°C for 4 h. The cells were harvested, resuspended in 50 mM Tris-HCl buffer containing 500 mM NaCl and 20 mM imidazole at pH 8.0, and lysed by sonication. Crude extract was then prepared and subjected to purification by nickel-nitrilotriacetic acid (Ni²⁺-NTA) metal affinity chromatography using a 5-ml HisTrap HP column (29-0510-21; GE Healthcare). MenI was eluted at 200 mM imidazole, desalted, and concentrated in 20 mM Tris-HCl buffer (pH 8.0) containing 10% glycerol. The purified enzyme was estimated to be >95% pure using sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE), and its concentration was determined by UV-Vis absorbance at 280 nm using a molar extinction coefficient of 6,990 M⁻¹⋅cm⁻¹ calculated from its amino acid composition with ProParam (47). It was finally aliquoted and flash-frozen in liquid nitrogen and stored at −80°C until use.

The DHNA-CoA thioesterase activity of MenI was determined using a reported method (18) with modifications. The extinction coefficient constants of DHNA-CoA and DHNA were measured at 396 nm, and their difference was revised to be 3,911 cm⁻¹⋅M⁻¹. The initial velocity was measured in triplicate at room temperature (23°C) by real-time monitoring of the absorbance change at 396 nm on a Shimadzu UV-1800 UV/visible scanning spectrophotometer in 200 mM sodium phosphate buffer (pH 8.0). The concentration of MenI was set at 13.4 nM, and that of DHNA-CoA was varied from 1.8 to 57.8 μM. The K_m and k_cat were determined using the Lineweaver-Burk plot.

Transcomplementation assays.

L. monocytogenes strains were grown in 3-ml cultures overnight in BHI medium at 30°C without shaking. The starter cultures were then diluted 1:100 into minimal medium, and these were then grown overnight at 37°C with shaking. Overnight aerobic cultures of the indicated strains grown in minimal medium were then centrifuged to remove bacteria. Supernatants were filtered through a 0.2-μm-pore-size syringe filter (catalog number 09-740-113; Fisher Scientific). Filtered supernatants were added to wells containing indicated strains of L. monocytogenes in minimal medium at a final concentration of 1% (vol/vol). The OD₆₀₀ was then monitored for 24 h (Eon; BioTek, Winooski, VT).

Intracellular bacteriolysis assay.

Standard intracellular bacteriolysis assays were performed as previously described (15). Briefly, immortalized bone marrow-derived Ifnar^−/− macrophages (5 × 10⁵ per well of 24-well plates) were grown in a monolayer overnight in a 500 μl volume. L. monocytogenes strains carrying the bacteriolysis reporter pBHE573²⁶ were grown at 30°C without shaking overnight. Cultures were then diluted to a final concentration of 5 × 10⁸ CFU/ml in PBS and used to infect macrophages at a multiplicity of infection (MOI) of 10. At 1 h postinfection, media were removed and replaced with media containing 50 μg/ml gentamicin. At 6 h postinfection, media from the wells were aspirated, and macrophages were lysed using TNT lysis buffer (20 mM Tris, 200 mM NaCl, 1% Triton [pH 8.0]). Cell lysates were transferred to opaque 96-well plates, and luciferin reagent was added and assayed for luciferase activity (Synergy HT; BioTek, Winooski, VT).

L2 plaque assay.

Plaque assays were conducted using a L2 fibroblast cell line (28). L2 fibroblasts were seeded at 1.2 × 10⁶ per well in a 6-well dish and infected at an MOI of 0.5 to obtain approximately 30 plaques per dish. Infection was then allowed to proceed for 1 h, after which all wells were aspirated and washed three times with prewarmed PBS. Wells were then overlaid with 3 ml of medium containing 0.7% agarose and 50 μg/ml gentamicin (a stock of 2× medium with 100 μg/ml gentamicin was prewarmed at 37°C and mixed with 56°C 1.4% agarose immediately before use). At 3 days postinfection, another 3 ml of the agarose-containing medium was overlaid on each well. At 4 days postinfection, the agarose plugs were removed, and cells were stained with 0.3% crystal violet for 10 min and washed twice with deionized water. Stained wells were imaged, and areas of plaque formation were measured with Fiji Image analysis software (48).

Intracellular growth assay.

BMDMs were plated on coverslips at 5 × 10⁶ cells per 60-mm dish and allowed to adhere overnight. BMDMs were then infected at an MOI of 0.2 with their respective strain, and infection proceeded for 8 h. At 30 min postinfection, media were removed and replaced with media containing 50 μg/ml gentamicin. Total CFU were quantified at various time points as previously described (10).

Acute virulence assay.

All techniques were reviewed and approved by the University of Wisconsin–Madison Institutional Animal Care and Use Committee (IACUC) under protocol M02501. Female C57BL/6 mice (6 to 8 weeks of age; purchased from Charles River) were used for the study. L. monocytogenes strains were grown in BHI medium at 30°C without shaking overnight. These cultures were then back-diluted the following day 1:5 into fresh BHI medium and grown at 37°C with shaking until the mid-exponential phase (OD₆₀₀, 0.4 to 0.6). Bacteria were diluted in PBS to a concentration of 5 × 10⁶ CFU/ml, and mice were injected intravenously with 1 × 10⁶ total CFU. At 48 h postinfection, spleens and livers were harvested and homogenized in 0.1% Nonidet P-40 in PBS. Homogenates were then plated on LB plates to enumerate CFU and quantify bacterial burdens.

Statistical analysis.

Statistical significance analysis (GraphPad Prism version 6.0h) was determined by one-way analysis of variance (ANOVA) with a Bonferroni posttest (*, P ≤ 0.05; **, P ≤ 0.01; ***, P ≤ 0.001; ****, P ≤ 0.0001).