Importance of the ΔgcvPAB genes for generation of one carbon donors and for plaque formation in 3T3 mouse fibroblasts.

(A) Three enzymatic pathways ensure N5, N10-methylene-THF biosynthesis in L. monocytogenes. N5, N10-methylene-THF is generated (i) during conversion of serine to glycine by GlyA (green), (ii) during glycine degradation in the GcvPAB-dependent GCS by GcvP and GcvT (orange) and (iii) by formate THF ligase Fhs in cooperation with the bifunctional N5,N10-methylene-THF dehydrogenase/cyclohydrolase FolD (blue). Biosynthesis of THF from dihydrofolate (DHF) and p-aminobenzoate (PABA) is inhibited by trimethoprim (TMP) and sulfamethoxazole (SMX). (B) Plaque formation assay in 3T3 mouse embryo fibroblasts with L. monocytogenes strains EGD-e (wt), LMS305 (ΔgcvPAB) and LMS311 (igcvPAB). 1 mM IPTG was added as indicated. (C) Quantification of the assay shown in panel B. Plaque areas were determined using ImageJ and average values and standard deviations were calculated from three independent experiments. The asterisk marks a statistically significant difference (P<0.01, t-test with Bonferroni-Holm correction).

In vitro and in vivo virulence of the ΔgcvPAB mutant.

(A) Replication of L. monocytogenes strains EGD-e (wt), LMS305 (ΔgcvPAB), LMS250 (Δhly, replication deficient control) amd LMS251 (ΔactA, spreading deficient control) in 3T3 mouse fibroblasts. (B) Replication of L. monocytogenes strains EGD-e (wt), LMS305 (ΔgcvPAB) and LMS250 (Δhly, negative control) in J774 mouse macrophages. (C) Replication of L. monocytogenes strains EGD-e (wt), LMS305 (ΔgcvPAB) and BUG2214 (ΔprfA, invasion and replication deficient control) in HepG2 human hepatocytes. Average values and standard deviations calculated from technical replicates are shown for all experiments. Asterisks mark statistically significant differences (panels A and B: P<0.01 t-test with Bonferroni-Holm correction, panel C: P<0.05 t-test). (D) CAMP assay to compare hemolysis in L. monocytogenes strains EGD-e (wt), LMS305 (ΔgcvPAB) and LMS250 (Δhly, negative control). (E) Quantification of hemolysis activity in the same set of strains towards human erythrocytes. Hemolysis activity is expressed as the number of ten-fold dilutions of the various culture supernatants after which no hemolysis could be observed any more. The asterisk marks a statistically significant difference (P<0.01, t-test with Bonferroni-Holm correction). (F) Lysozyme-induced lysis of L. monocytogenes strains EGD-e (wt), LMS305 (ΔgcvPAB) and LMS163 (ΔpgdA, positive control). Average values and standard deviations calculated from technical triplicates are shown. (G-I) Virulence of the ΔgcvPAB mutant in mice. Infection was conducted with 1-20 x 104 CFUs/ml injected into the tail vein of the mice. Mice were infected with either EGD-e or the ΔgcvPAB mutant. (G) Three days post-infection, CFU were quantified in the spleen, liver and brain to determine the bacterial burden. The geometric mean with the geometric standard deviation (SD) is illustrated. (H) Following infection, the mice were scored on a daily basis for weight loss as a parameter of disease severity during L. monocytogenes infection. The body weight is presented in relation to the weight prior to infection. The standard error of the mean (SEM) is shown. (I) Nine days post-infection, the spleens were isolated and compare with regard to the size of the organ under conditions of different L. monocytogenes strain infections. Statistical analysis was conducted utilising the GraphPad Prism software, performing unpaired (G) or paired (H) two-tailed t-tests.

Growth of the ΔgcvPAB mutant in laboratory media.

(D) Growth of L. monocytogenes strains EGD-e (wt) and LMS305 (ΔgcvPAB) in complex BHI and chemically defined LSM medium. (E) Complementation of the growth defect of ΔgcvPAB mutant in LSM medium. Growth of L. monocytogenes strains EGD-e (wt), LMS305 (ΔgcvPAB) and LMS311 (igcvPAB) in LSM medium ± 1 mM IPTG. Average values and standard deviations were calculated from technical triplicates.

The ΔgcvPAB mutant is sensitive to increased glycine concentrations.

(A-B) Growth of L. monocytogenes strains EGD-e (wt, A) and LMS305 (ΔgcvPAB, B) in LSM medium containing different glycine concentrations. Glycine concentrations are expressed relative to standard LSM concentrations (1x = 1.3 mM). Average values and standard deviations were calculated from technical replicates (n=3).

Suppression of the ΔgcvPAB phenotype by a fhs+ mutation restoring formate-tetrahydrofolate activity.

(A) Restoration of the full length fhs open reading frame in ΔgcvPAB suppressor strain LMSF15. Schematic illustration of the fhs loci in the two reference strains 10403S (full length) and EGD-e (split into two pseudogenes due to a frameshift mutation in lmo1877) as well as in the ΔgcvPAB suppressor strain LMSF15 where an GTGG insertion (red) restores the full length fhs gene. (B) Growth of ΔgcvPAB suppressor strains in various media. Strains EGD-e (wt), LMS305 (ΔgcvPAB), LMSF3 (ΔgcvPAB codY G236E), LMSF8 (ΔgcvPAB folK G82R), LMSF10 (ΔgcvPAB codY G236E glyA G62S) and LMSF15 (ΔgcvPAB fhs+) were grown in BHI broth (left panel), LSM medium without glycine (middle panel) and LSM medium supplemented with 100-fold the amount of glycine than in the standard recipe (right panel). Growth curves show average values and standard deviations from technical triplicates. (C) Intracellular growth of ΔgcvPAB suppressors in J774 mouse macrophages. The same set of strains as in the panel B was used to infect J774 macrophages and the bacterial load six hours post infection (p. i.) was determined. The diagram shows average values and standard deviations calculated from technical triplicates. Asterisks mark statistically significant differences compared to wild type (black) or compared to the ΔgcvPAB mutant (red, P<0.01, t-test with Bonferroni-Holm correction, ns – not significant). (D) Recreation of the fhs+ mutation in the ΔgcvPAB background confirms suppression of ΔgcvPAB in vitro virulence phenotypes by restoration of Fhs activity. Intracellular replication of L. monocytogenes strains EGD-e (wt), LMS305 (ΔgcvPAB), LMSF26 (fhs+) and LMSF27 (ΔgcvPAB fhs+) in J774 mouse macrophages. Strains LMSF26 and LMSF27 were generated from EGD-e and LMS305, respectively, by introduction of the isolated fhs+ mutation. Average values and standard deviations were calculated from technical triplicates. Asterisks mark statistically significant differences compared to wild type (P<0.05, t-test with Bonferroni-Holm correction, ns = not significant). (E) Plaque formation in 3T3 mouse fibroblasts of the same set of strains as in panel D.

Importance of 1C-THF generating pathways for viability, intracellular growth and adenine biosynthesis.

(A-B) Simultaneous absence of Fhs, GCS and GlyA activity is lethal. Growth of L. monocyctogenes strains EGD-e (fhs⁻), LMS305 (fhs⁻ ΔgcvPAB), LMSF25 (fhs⁻ ΔglyA), LMTE151 (fhs+ ΔgcvPAB ΔglyA) and LMSF28 (fhs⁻ ΔglyA igcvPAB) in BHI (A) and LSM broth (B) containing or not containing 1 mM IPTG. (C) Individual contribution of the three 1C-THF generating pathways to intracellular growth in macrophages. Multiplication of L. monocyctogenes strains EGD-e (fhs⁻), LMS305 (fhs⁻ ΔgcvPAB), LMSF25 (fhs⁻ ΔglyA), LMTE151 (fhs+ ΔgcvPAB ΔglyA) inside J774 mouse macrophages within six hours post infection. Average values and standard deviations were calculated from three replicates. Statistical significance is labelled by an asterisk (P<0.01 t-test with Bonferroni-Holm correction) or “ns” (not significant). The presence or absence of the three pathways is indicated below the diagram. (D) Growth of L. monocytogenes strains EGD-e (wt) and LMS305 (ΔgcvPAB) in LSM containing standard (18 µM) and increased adenine concentrations (1 mM). Average values and standard deviations were calculated from technical replicates (n=3).

Plasmids and L. monocytogenes strains used in this study.

Oligonucleotides used in this study.