The major role of Listeria monocytogenes folic acid metabolism during infection is the generation of N-formylmethionine

ABSTRACT Folic acid and its derivatives (folates) play central roles in one-carbon metabolism, necessary for the synthesis of purines, pyrimidines, and some amino acids. Antifolate drugs are widely used as broad-spectrum antibiotics for bacterial infections, including listeriosis, a disease caused by the facultative intracellular pathogen Listeria monocytogenes. However, folate-derived metabolites required during bacterial infection are poorly understood. Here, we report that L. monocytogenes encodes two enzymes, methylenetetrahydrofolate dehydrogenase/methenyltetrahydrofolate cyclohydrolase (FolD) and formyltetrahydrofolate synthetase (Fhs), that catalyze the formation of N10-formyltetrahydrofolate, a critical intermediate in folate metabolism. N10-formyltetrahydrofolate is an essential carbon donor for biosynthesis of purines and N-formylmethionine, the amino acid used during initiation of bacteria translation. While L. monocytogenes mutants lacking Fhs had no observable defects and mutants lacking FolD had moderate virulence defects, mutants lacking both were highly attenuated in mice, especially in the liver. We compared the growth and virulence of mutants that were unable to synthesize folates with mutants unable to generate folate downstream products, including purines, thymidine, and N-formylmethionine. Mutants unable to synthesize N-formylmethionine behaved almost identically to FolD/Fhs double mutants during growth in broth or macrophages, but most importantly, were similarly attenuated to mutants unable to make folates, both of which had approximately 4-log10 less colony-forming units in the livers compared with wild-type L. monocytogenes. The purine auxotrophic mutants were only 1.5-log10 less virulent, and the thymidine mutants were fully virulent in the mouse infection model. These results strongly suggest that the main role of L. monocytogenes folates during infection is the generation of N-formylmethionine. Importance Folic acid is an essential vitamin for bacteria, plants, and animals. The lack of folic acid leads to various consequences such as a shortage of amino acids and nucleotides that are fundamental building blocks for life. Though antifolate drugs are widely used for antimicrobial treatments, the underlying mechanism of bacterial folate deficiency during infection is unclear. This study compares the requirements of different folic acid end-products during the infection of Listeria monocytogenes, a facultative intracellular pathogen of animals and humans. The results reveal the critical importance of N-formylmethionine, the amino acid used by bacteria to initiate protein synthesis. This work extends the current understanding of folic acid metabolism in pathogens and potentially provides new insights into antifolate drug development in the future.

derivatives of folic acid, tetrahydrofolates (THFs), play central roles in one-carbon (1C) metabolism by donating carbon groups necessary for the synthesis of purines, pyrimidines, amino acids, and N-formylmethionine (fMet), the first amino acid used during protein translation in bacteria and mitochondria (5,6).Synthesis of THFs and their downstream metabolites are proven antibiotic targets (7)(8)(9).Sulfa drugs target the folate precursor para-aminobenzoic acid (PABA) biosynthesis, while trimethoprim targets dihydrofolate reductase and is generally used together as broad-spectrum antibiotics for many bacterial infections, including listeriosis, a disease caused by a facultative intracellular pathogen Listeria monocytogenes (10,11).Despite the importance of current and new antifolate drugs, and the extensive studies of folic acid metabolism, especially in humans, the functions of folic acid metabolism during infection by intracellular pathogens remain poorly understood.
L. monocytogenes is a rapidly growing, Gram-positive facultative intracellular pathogen that can cause serious, sometimes fatal, disease following ingestion of contaminated food in a wide range of animals, including immunocompromised and pregnant humans (12)(13)(14).L. monocytogenes also serves as a highly tractable model organism for studying intracellular pathogens and cell-mediated immunity (12,15).In a survey to identify L. monocytogenes transposon mutants that formed small plaques in monolayers of tissue culture cells, we identified two genes that encode enzymes in folic acid metabolism (16).Mutants in pabBC, which encodes two enzymes that catalyze the biosynthesis of the folic acid precursor, PABA, were severely attenuated for L. monocyto genes virulence in tissue culture and mouse infection models (17), although the precise reasons for its virulence were not established.Additionally, we identified folD, which encodes bifunctional methylenetetrahydrofolate dehydrogenase/methenyltetrahydrofo late cyclohydrolase, that catalyzes the reversible conversion of two major 1C-carrying folates, N5,N10-methylene-THF and N10-formyl-THF.The 1C group of N5,N10-methylene-THF is used to generate serine, glycine, and deoxythymidine monophosphate (dTMP) while that of N10-formyl-THF is used to synthesize purines and fMet (Fig. 1).FolD is essential in many bacteria, although some facultative anaerobes, including L. mon ocytogenes, can also synthesize N10-formyl-THF using formyltetrahydrofolate synthe tase (formate tetrahydrofolate ligase, Fhs) (Fig. 1) (18).In bacteria and mitochondria, methionyl-tRNA formyltransferase (FMT) transfers the formyl-group from N10-formyl-THF to a methionine bound to the initiator tRNA which in turn promotes the initiation of translation (19,20).
In this study, we sought to explore the functional mechanisms underlying the requirement of folate metabolism during L. monocytogenes infection.We characterized the roles of FolD and Fhs during bacterial growth in media, tissue culture, and animal infection models.FolD played a dominant role in making N10-formyl-THF and was critical to L. monocytogenes aerobic growth and pathogenesis, although mutants lacking both FolD and Fhs were much more highly attenuated.By comparing the phenotypes of mutants in different steps in folic acid metabolism, we concluded that fMet was the most critical folate end-product required during infection in vivo.

Characterization of folD and fhs mutants in media and in tissue culture models of infection
In an effort to understand the role of L. monocytogenes folic acid metabolism during infection, we characterized a previously identified transposon insertion mutant in folD (folD::Tn), which encodes a central enzyme of folic acid metabolism (Fig. 1), that catalyzes the formation of N10-formyl-THF, a bioactive folate that donates carbon groups essential for both purines and N-formylmethionine synthesis.We noted that L. monocytogenes encodes another enzyme, formate tetrahydrofolate ligase (Fhs), which also catalyzes the formation of N10-formyl-THF.We constructed in-frame deletions in L. monocytogenes folD (ΔfolD), in fhs (Δfhs), and double mutants lacking both (Δfhs folD::Tn and ΔfolD fhs::Tn).Consistent with our previous findings characterizing the folD::Tn mutant ( 16), the ΔfolD deletion mutant formed very small plaques in monolayers of murine L2 fibroblasts (Fig. 2A).The Δfhs strain formed plaques indistinguishable from wild-type (WT), and the double mutants Δfhs folD::Tn formed plaques similar to those formed by ΔfolD (Fig. 2A), indicating that FolD plays the major role in generating N10-formyl-THF during host cell infection in vitro.Neither single mutant exhibited a growth defect in rich media, while the double mutants grew significantly slower compared to WT, with a 1.8-fold increased doubling time (Fig. 2B; Table S1).The growth defect observed in strains lacking folD and fhs was fully rescued by complementing with the folD gene expressed from a pHyper promoter (21) (Fig. 2A  and B).We also examined the requirements of folD and fhs during infection of mouse bone marrow-derived macrophages (BMMs).Although the ΔfolD strain had no defect when cultured in broth, it showed a small but significant defect at 5-and 8-h post-infection in BMMs (Fig. 2C).Notably, the intracellular growth rate of the Δfhs folD::Tn double mutant was highly impaired, although it was able to grow until it reached a similar terminal OD 600 as WT in broth (Fig. 2B and C; Table S1).Taken together, these data demonstrated that N10-formyl-THF is critical for L. monocytogenes growth and host cell infection.Although FolD is the dominant enzyme required for the synthesis of N10-formyl-THF, Fhs plays a measurable role in the absence of FolD.

Characterization of folD and fhs mutants in mice
To reveal the requirements of N10-formyl-THF during in vivo infection, CD-1 mice were infected intravenously, and bacterial burdens were determined in the livers and spleens 48-h post-infection.The ΔfolD mutant exhibited an organ-specific defect that presented as a 2-log 10 virulence attenuation in the livers compared with WT and the complemen ted strains but showed no difference in colony-forming units (CFUs) in the spleens (Fig. 3).A main difference between these two organs is that in the spleen, L. monocytogenes grows in phagocytic cells, while much of the growth in the liver is in hepatocytes (22).To address if the liver-specific defect was due to poor growth in hepatocytes, the folD::Tn transposon was transduced into a strain lacking ActA and InlB, which prevents the bacteria from spreading into hepatocytes or entering via InlB-induced internalization (23).In the ΔactA ΔinlB background, lack of folD still resulted in attenuation in the livers (Fig. S1A).In addition, folD played a negligible role in plaque formation in a murine hepatocyte cell line TIB-73 (Fig. S1B).These data suggested that the organ-specific defect of mutants lacking folD was unlikely attributed to the infection of hepatocytes.
L. monocytogenes virulence factor expression is controlled by the transcription factor PrfA (24,25), and the biosynthesis of PABA, the folic acid precursor, is activated by PrfA (17).We wondered if the virulence defect of a folD mutant was due to lack of PrfA activation or if PrfA activation could compensate the defect of a folD mutant.To evaluate the role of PrfA activation, we introduced the ΔfolD deletion mutation into a PrfA* strain where PrfA has a mutation that locks it in its active conformation (25).The PrfA* ΔfolD double mutant remained attenuated in the livers (Fig. S2), indicating that the defect of ΔfolD was probably related to poor growth rather than loss of virulence gene activation.
Mutants lacking both folD and fhs (Δfhs folD::Tn) were severely attenuated in vivo, with approximately 4-log 10 fewer CFUs in infected livers compared with WT and a 1.5-log 10 defect in the spleens (Fig. 3).Complementation of Δfhs folD::Tn with folD fully restored virulence (Fig. 3), again highlighting the dominant role of folD during infection.As observed in broth and BMMs, loss of fhs had no observable phenotype in the presence of folD but had a synergistic virulence defect when combined with a folD mutation during in vivo infection.We speculated that the role of Fhs during aerobic growth and during infection was limited due to the absence of its substrate formate, which is produced mostly during fermentation (26).Indeed, the growth defect of a ΔfolD mutant growing aerobically in a chemically defined medium was completely restored if supplemented with formate, while restoration by formate was not observed in the absence of fhs as shown by the Δfhs folD::Tn strain (Fig. S3).

Genetic screen to identify mutations that rescue the virulence defect of a ∆folD mutant and the role of purine biosynthesis during infection
We next sought to understand the mechanisms underlying the virulence defect of L. monocytogenes deficient in making N10-formyl-THF.THFs are carbon donors for various molecules, thus perturbations in folate metabolism could result in pleiotropic pheno types (5,6,27).We showed that N10-formyl-THF made by FolD and Fhs was critical for L. monocytogenes growth and pathogenesis (Fig. 2 and 3), but it was unclear which factor(s) led to the virulence defect of the N10-formyl-THF deficient strains.To address this question, we generated a transposon library in a folD mutant background and screened for mutants that formed larger plaques in tissue culture.A total of 20 insertion mutants representing eight genes were identified that displayed increased plaque size from over 35,000 screened.Candidate suppressors were confirmed by transducing single insertions into WT and ΔfolD background (Table S2).
Insertions disrupting purR led to the largest plaque restoration of ΔfolD (Table S2).The purR gene encodes a transcriptional repressor that controls gene expression in response to the availability of purines (28)(29)(30).The major role of PurR is to repress the transcription of genes in the pur operon that encodes enzymes for the de novo synthesis of purines, although it represses other genes as well (28,(30)(31)(32).N10-formyl-THF is used as a cofactor for two pur-encoded enzymes that each donates a carbon unit during the construction of the purine ring.We hypothesized that purine shortage might lead to growth and virulence defects during N10-formyl-THF limitation, and de-repression of purine synthetic genes by the purR mutation replenished purine levels and consequently restored the growth and virulence of ΔfolD.As shown in Fig. S3, the growth of ΔfolD was greatly impaired in chemically defined media without N10-formyl-THF end-prod ucts.Adding exogenous purines into the media fully restored the growth of ΔfolD and partially rescued the Δfhs folD::Tn strain as well (Fig. 4A).A purR in-frame deletion was generated in WT and ΔfolD backgrounds, and consistent with the transposon mutants, the ΔfolD ΔpurR mutants formed WT plaques (Fig. 4B).However, the purR mutation did not rescue plaque formation in a strain lacking folD and fhs, or the virulence of ΔfolD in mice (Fig. 4B and C).
In order to directly evaluate the role of de novo purine biosynthesis, we generated a purine auxotrophic strain introducing an in-frame deletion of purE and purK (ΔpurEK), the first two genes in the pur operon.As expected, the ΔpurEK strain failed to grow in media lacking purines (Fig. S4), but it grew like WT in both the plaque assay and in BMMs (Fig. 4B and Fig. 5D).However, the ΔpurEK mutant did have an approximately 1.5-log 10 virulence defect in mouse livers, although it was much less attenuated than a folD/fhs double mutant (Fig. 3 and Fig. 4C).Therefore, it is reasonable to conclude that N10-formyl-THF does contribute to purine biosynthesis during in vivo infection, and purine insufficiency is partially responsible for the severe attenuation of strains lacking FolD and Fhs.

Comparison of mutants lacking different steps of the folic acid cycle suggests a critical role of N-formylmethionine during infection
In addition to purine biosynthesis, the other one-carbon recipient of N10-formyl-THF is methionine bound to the initiation tRNA (Met-tRNA i ), catalyzed by methionyl-tRNA formyltransferase (FMT) (Fig. 1).Although fmt is essential in some bacteria, it has been mutated in many bacteria, including L. monocytogenes (33)(34)(35)(36)(37)(38)(39).Accordingly, we constructed a strain lacking formylated Met-tRNA i (fMet-tRNA i ) by in-frame deletion of fmt (Δfmt).Removal of the formyl-group from nascent peptides by peptide deformy lase (PDF) is essential but is dispensable in strains lacking N-formylmethionine (i.e., fMet-tRNA i ).Mutants lacking N-formylmethionine are resistant to actinonin, an antibiotic targeting PDF (33).We tested actinonin sensitivity of Δfhs folD::Tn and Δfmt in a diskdiffusion assay and found that both strains were resistant to actinonin (Fig. 5A), strongly suggesting that N-formylmethionine was not being synthesized.We then compared the Δfhs folD::Tn and Δfmt mutants in various assays, and strikingly, the Δfmt strain behaved almost identical to Δfhs folD::Tn in broth growth, plaque formation, growth in BMMs, and virulence in mice (Fig. 5B through D; Fig. 6).The defects observed in the mutants lacking fmt were fully rescued by complementation with the fmt gene (Fig. 5B through D; Fig. 6).
We next sought to evaluate the relative importance of N-formylmethionine compared with other folate end-products.In addition to purines and N-formylmethionine, the carbon units carried by different folate molecules can be used in the synthesis of serine, glycine, and dTMP, depending on the needs of bacteria (Fig. 1).Previous 13 C-isotopologue profiling demonstrated that L. monocytogenes imports glycine and serine from the host cell cytoplasm (40).Methionine is also a folate product in many species, but L. monocyto genes 10403S lacks methionine synthase (Fig. 1) (41,42), so it was not addressed here.We examined the requirement of thymidine during infection by using a mutant lacking thyA (43), which encodes the enzyme transferring the carbon unit from N5,N10-methylene-THF to produce dTMP (Fig. 1).The ΔthyA mutants do not grow in BMMs (Fig. S5A) but were fully virulent in the mouse infection model (Fig. S5B), suggesting that generation of dTMP from folates is not necessary during L. monocytogenes in vivo infection.
Finally, we compared the virulence defect of an L. monocytogenes strain unable to synthesize PABA (ΔpabBC) and, therefore, lacks folate synthesis.The folate biosynthesis defective strain, ΔpabBC, is severely attenuated in vivo, and purine supplementation partially replenished its growth in defined media (17).In the mouse infection model, the ΔpabBC and Δfhs folD::Tn mutants were similarly attenuated, although the Δfmt mutant had a slightly higher, though not statistically significant, bacterial burden (Fig. 6).

DISCUSSION
Folic acid metabolism represents a critical aspect of central metabolism, being required for the synthesis of purines, pyrimidines, some amino acids, and critically, formylated methionine (N-formylmethionine).In addition, folic acid metabolism is an established target for antimicrobial therapy.However, despite the wide application of antifolate antibiotics, the precise role of folic acid metabolism during bacterial pathogenesis has remained obscure.The goal of this study was to identify which specific folic acid-derived metabolites are required during infection of the facultative intracellular pathogen L. monocytogenes.
This study began with the identification of an L. monocytogenes mutant that formed very small plaques in tissue culture caused by disruption of folD, which encodes bifunctional methylenetetrahydrofolate dehydrogenase/methenyltetrahydrofo late cyclohydrolase (Fig. 1) (16).FolD is a bifunctional enzyme that plays a central role in the formation of most folic acid metabolites and is essential in many bacte ria, although some bacteria, including L. monocytogenes, possess another enzyme, formyltetrahydrofolate synthetase (Fhs) that also catalyzes the formation of N10-formyltetrahydrofolate.The results of this study showed that folD mutants had significant defects during growth in tissue culture cells and a 2-log 10 defect in the livers of mice.Fhs mutants had no apparent defects during growth in tissue culture cells or during infection of mice, but mutants lacking both FolD and Fhs were severely attenuated during infection of mice.Since Fhs uses formate as its substrate and L. monocytogenes only produces formate during anaerobic growth (44), bacteria may be experiencing low oxygen conditions in vivo, leading to the Fhs-dependent production of N10-formyltetrahydrofolate.L. monocytogenes may also benefit from having an alternative way of making N10-formyl-THF by Fhs during its natural route of infection, which requires growth in the low oxygen environment of the intestinal tract.
During a genetic screen looking for ΔfolD suppressors, we found that loss of purR, the purine synthesis repressor, restored WT plaque formation in L2 cells, suggesting that folD mutants are defective in purine biosynthesis.There are numerous examples of bacterial purine auxotrophs having growth and virulence defects in vivo (45).For example, Staphylococcus aureus purine auxotrophs cannot replicate in human blood and serum and have defects in animal models (46,47).Purine auxotrophs of Bacillus anthracis are dramatically attenuated in a murine infection model (48).Purine auxotro phic mutants of Salmonella typhimurium have a 4-log 10 higher LD 50 compared with WT (49,50).In contrast to Salmonella, L. monocytogenes purine auxotrophic mutants are only moderately attenuated in terms of LD 50 (51).The ΔpurEK mutants generated in this study exhibited only an approximately 1.5-log 10 defect in the livers of infected mice (Fig. 4C).Notably, this moderate defect suggests that L. monocytogenes can acquire purines from the host environment, although it remains to be investigated whether the defect of L. monocytogenes purine mutants is due to insufficient host purines during intracellular and/or extracellular growth.However, although ΔpurEK mutants were attenuated, they were more than 100-fold more virulent than mutants lacking both FolD and Fhs in mice, indicating that purine deficiency is not the major defect when L. monocytogenes lacks N10-formyl-THF during infection.
In addition to purine biosynthesis, 1C metabolism mediated by folates integrates carbon units from serine, glycine, and sometimes formate to generate thymidine and fMet.In this study, we characterized L. monocytogenes mutants unable to synthesize thymidine (thyA), fMet (fmt), or folate precursor PABA (pabBC).Synthesis of dTMP by ThyA was essential for L. monocytogenes growth in BMMs (Fig. S5A) but dispensable during mouse infection (Fig. S5B).Strikingly, the fmt mutants led to the highest degree of attenuation in vivo, statistically identical to the ΔpabBC strain (Fig. 6), indicating that lack of fMet was the major defect of folate deficiency during infection.Previous studies demonstrated that antifolate antibiotics cause thymineless death of bacteria in nutrientrich condition and inhibit bacteria growth by glycine and purine depletion in nutrientlimited conditions (52,53).The observations of this study suggest that antifolate drugs also deplete fMet production during bacteria pathogenesis.However, our results do not eliminate the possible contribution of folates to serine and glycine generation during L. monocytogenes infection, since serine and glycine not only serve as carbon sources of folates but are also folate end-products if needed (6).Previous 13 C-isotopologue profiling studies demonstrated that significant fractions (50%-100%) of bacterial amino acids, including serine and glycine, were from the host cell cytosol during intracellular growth (41,54).We propose that the in vivo host environment provides some or all of a number of folate metabolites, including purines, pyrimidines, and amino acids but not fMet.
N-formylmethionine is a modified amino acid used by bacteria and mitochondrial proteins.The formylated methionine carried by initiator tRNA interacts with initiation factor 2 with a higher affinity than an unformylated methionine and thereby fMet facilitates high efficacy and fidelity of peptide translation in bacteria and mitochon dria (19,20,35).Although fMet is important for L. monocytogenes infection, protein translation in L. monocytogenes is not strictly dependent on fMet.Mutants lacking fmt grew slower in brain heart infusion (BHI) media and were less attenuated in mouse spleens than livers.Similar to our findings in L. monocytogenes, fMet is not essential in many other bacteria, including Escherichia coli, Pseudomonas aeruginosa, Mycobacterium smegmatis, Mycobacterium bovis, Staphylococcus aureus, Salmonella enterica, and Bacillus subtilis, but loss of fmt often leads to slow growth and hypersensitivity to stress (34)(35)(36)(37)(38)(39).The differences we observed in vivo indicate that the efficiency and fidelity in protein translation are more strictly important for L. monocytogenes in the liver than the spleen, although the nature of the differences is not clear.While fmt is not essential in many bacteria, it is predicted to be essential in Streptococcus pneumonia and M. tuberculosis (55)(56)(57).The stringency of fMet dependence in different bacteria under various conditions shall be taken into consideration in the development and evaluation of new antifolate antibiotics.
Although fMet is not essential in many bacteria, removal of the formyl-group from a nascent peptide by PDF is essential across bacterial species, making PDF an attractive target of antibiotics such as actinonin (58).The critical role of folates in fMet generation is also revealed in other bacteria, as mutants resistant to actinonin not only mapped to the fmt gene but also to folD and glyA (59, 60) (see Fig. 1).Although PDF removes fMet from most proteins during translation, some formyl-groups escape removal and are released as short peptides that are key targets of the innate immune system recognized by chemotactic receptors on neutrophils and monocytes, both critical immune effector cells required for constraining and resolving bacterial infections (61)(62)(63)(64)(65)(66).Mice deficient in formylated peptide receptors are more susceptible to L. monocytogenes infection, probably due to the lack of a rapid wave of neutrophil infiltration that occurs early upon infection (67,68).Therefore, another potential consequence of antifolate antibiotics is that by blocking production of fMet they also block the release of formylated peptides which would prevent detection of bacteria by phagocytic cells.

Bacterial growth conditions
All L. monocytogenes strains used in this study were derived from WT strain 10403S (Table S3).Strains were propagated in filter-sterilized BHI medium (BD) at 37°C with shaking.When needed, bacterial culture media supplements were used at the following concentrations: streptomycin at 200 µg/mL, chloramphenicol at 7.5 µg/mL, erythromy cin at 1 µg/mL, carbenicillin at 100 µg/mL, actinonin (MedChemExpress) at 100 µg/mL, and tetracycline at 2 µg/mL.The Listeria synthetic medium (LSM) was made using a previously described recipe with 20 amino acids added (69).When needed adenine was removed from the LSM recipe, or 10 mM sodium formate and 1 mM adenine were added to the medium.All reagents were purchased from Sigma-Aldrich unless specified.
Broth growth curves were performed with L. monocytogenes strains from overnight cultures grown at 37°C with agitation (220 rpm).BHI and LSM growth curves were started at an optical density (OD 600 ) of 0.05.Growth was spectrophotometrically measured until saturated.

Plasmid and strain construction
All strains used in this study are listed in Table S3.Plasmids were introduced into L. monocytogenes by conjugation, using a donor E. coli SM10 and a compatible L. mono cytogenes strain.In-frame deletion of genes was performed using allelic exchange as previously described (70).The double mutants lacking both folD and fhs were generated by U153 phage transduction (71).Transductants were selected on erythromycin.The Δfmt strain was selected in the presence of 100 µg/mL actinonin.The complemented strains were generated by integrating a pPL2 vector encoding genes under control of the pHyper promoter (21) into mutant strain genomes and selecting for tetracycline-resistant transconjugants (72).agar plate, soaked with 10 µL actinonin stock solution (50 mg/mL in DMSO), and incubated overnight at 37°C.Clear zone areas were measured using ImageJ software (74).

Statistical analysis
All statistical analyses were performed using GraphPad Prism version 9.2 for Windows, GraphPad Software, La Jolla, CA, USA, www.graphpad.com.

FIG 4
FIG 4 De-repression of PurR restores plaque forming of ΔfolD but not virulence in mice.(A) Broth growth curve in synthetic medium supplemented with adenine.Strains were cultured in the Listeria synthetic medium with or without 1 mM adenine at 37°C.Growth was measured spectrophotometrically. (B) Plaque formation of indicated strains in L2 cells.Plaques were measured on the third day after infection and presented as a percentage of wild type.Three independent experiments were combined for (A) and (B).Data are mean ± SD.One-way ANOVA, multiple comparisons with WT; ns, not significant; ****P < 0.0001.(C) Virulence of indicated L. monocytogenes strains presented by CFUs in infected mouse livers and spleens.Two biological repeats were combined with a total of 10 mice per strain.Lines present medians.One-way ANOVA, multiple comparisons with WT control indicated on top of each strain, comparisons among the three mutant strains indicated by lines; ns, not significant; **P < 0.01; ****P < 0.0001.

FIG 5
FIG 5 Lack of fMet results in defects in broth growth and host cell infection.(A) Antibiotic sensitivity measured by disk diffusion containing 500 µg actinonin on brain heart infusion (BHI)-agar.L.o.d., limit of detection; N.d., not detectable.(B) Broth growth curve of indicated strains grown in BHI media at 37°C with agitation.(C) Plaque formation in murine fibroblast L2 cells measured 3-d post-infection of the indicated strains as a percentage of WT. (D) Intracellular growth in BMMs.Host cells were infected with indicated strains for 30 min at an MOI of 0.25.Antibiotic gentamicin was added to prevent extracellular bacteria at 1-h post-infection.(A-D) Three independent experiments were combined.Data are mean ± SD.One-way ANOVA, multiple comparisons with WT; ns, not significant; **P < 0.01; ****P < 0.0001.

FIG 6
FIG 6 Strains that fail to make fMet are similarly attenuated in vivo.CD-1 mice (n = 10) were infected with 1 × 10 5 CFUs of indicated L. monocytogenes strains.Bacteria CFUs in mouse livers and spleens were enumerated 2-d post-infection.One-way ANOVA, multiple comparisons among the three mutant strains were indicated; ns, not significant.