Both thiamine uptake and biosynthesis of thiamine precursors are required for intracellular replication of Listeria monocytogenes

Abstract

Thiamine pyrophosphate is an essential cofactor involved in central metabolism and amino acid biosynthesis and is derived from thiamine (vitamin B(1)). The extent to which this metabolite is available to bacterial pathogens replicating within host cells is still little understood. Growth studies using modified minimal Welshimer's broth (mMWB) supplemented with thiamine or the thiamine precursor hydroxymethylpyrimidine (HMP) showed that Listeria monocytogenes, in agreement with bioinformatic prediction, is able to synthesize thiamine only in the presence of HMP. This appears to be due to a lack of ThiC, which is involved in HMP synthesis. The knockout of thiD (lmo0317), which probably catalyzes the phosphorylation of HMP, inhibited growth in mMWB supplemented with HMP and reduced the replication rate of L. monocytogenes in epithelial cells. Mutation of a predicted thiamine transporter gene, lmo1429, led to reduced proliferation of L. monocytogenes in mMWB containing thiamine or thiamine phosphates and also within epithelial cells but had no influence on the expression of the virulence factors Hly and ActA. The toxic thiamine analogue pyrithiamine inhibited growth of wild-type strain EGD but not of the transporter mutant EGDDeltathiT. We also demonstrated that ThiT binds thiamine, a finding compatible with ThiT acting as the substrate-binding component of a multimeric thiamine transporter complex. These data provide experimental evidence that Lmo1429 homologs including Bacillus YuaJ are necessary for thiamine transport in gram-positive bacteria and are therefore proposed to be annotated "ThiT." Taken together, these data indicate that concurrent thiamine uptake and biosynthesis of thiamine precursors is a strategy of L. monocytogenes and possibly other facultative intracellular pathogens to enable proliferation within the cytoplasm.

FIG. 1.

Reconstruction of the thiamine biosynthetic and salvage pathways in L. monocytogenes. thiC appears to be the only gene missing for complete de novo biosynthesis of thiamine in Listeria species. A predicted thiamine kinase (ThiK) or ThiG and ThiH homologues have not yet been identified (3, 29). Genes that have been investigated in this study are underlined. Thiamine is transported by ThiT (Lmo1429), as shown in this study. No HMP- or HET-specific transporters have been identified so far. Dashed line, cytoplasma membrane; dashed arrows, possible substrate diffusion or unspecific transport. (Modified from Rodionov et al. [29] according to Begley et al. [3] and the literature cited in the text.)

FIG. 2.

(A) Growth properties of EGDΔthiD and EGDΔpdxK in the presence of thiamine. Listerial cells were grown in mMWB containing 1 mg/ml thiamine, and growth was monitored for 72 h. Each experiment was reproduced three times. Standard deviations for each time point are indicated as error bars. Error bars were omitted when they were smaller than the symbols. (B) Growth assay with HMP. A total of 10⁶ cells of thiamine-depleted strains EGD, EGDΔthiD, and EGDΔpdxK were plated on mMWB agar plates without thiamine, and a filter disc impregnated with 20 μl of a 120 μM solution of HMP was added. Plates were then incubated for 3 days at 37°C.

FIG. 3.

Deletion of thiD attenuates growth in epithelial cells. Caco-2 cells were infected with listerial cells at an MOI of ∼10. The MOI was determined immediately before infection (0 min). The number of intracellular bacteria was determined 35 min, 120 min, and 480 min after the infection. Bars represent the number of viable cells of each strain divided by the number of viable cells of the wild-type strain at each time point. Mutants EGD-eutB::pLSV101 and EGDΔlmo1538/glpD::pLSV101 served as controls. Each experiment was independently performed at least four times. Standard deviations are indicated by error bars.

FIG. 4.

(A) Growth behavior of mutant EGDΔthiT in mMWB supplemented with thiamine, TMP, or TPP. Strains EGD and EGDΔthiT were incubated overnight in mMWB containing 2.5% LB, washed twice with PBS, resuspended in mMWB without thiamine to an OD₅₅₀ of ∼3.0, and then diluted 1:500 into mMWB. Following 1:500 dilution, the strains were grown in mMWB with 3 μM thiamine, TMP, or TPP, respectively. Three independent experiments were used to determine OD values at each time point and to derive the standard deviation. Error bars were omitted when they were smaller than the symbols. (B) Thiamine-depleted EGD and EGDΔthiT cells were plated on mMWB agar containing 0.01 mg/liter thiamine, and a filter disc with 10 μl of a 2 mM pyrithiamine solution was placed in the middle of the plates. Pictures were taken after 3 days at 37°C.

FIG. 5.

Deletion of thiT attenuates growth in cell culture experiments. Caco-2 cells were infected with EGD, EGDΔthiT, and the two control strains EGDΔinlACGHE and EGDΔinlAB with an MOI of ∼10. For further experimental details, see the legend to Fig. 2. Each experiment was independently performed at least four times. The standard deviations are indicated.

FIG. 6.

[³H]thiamine binds to ThiT expressed in E. coli. E. coli cells transformed with a thiT expression plasmid or an empty vector control were incubated in a phosphate-citric acid buffer, and [³H]thiamine was added for a final concentration of 1.25 μM. The amount of cell-associated [³H]thiamine was determined by rapid filtration of the cells and liquid scintillation counting of the filter and is depicted in pmol/OD₅₅₀. Three independent experiments were performed. The graphs were drawn for a computer-generated best fit to the results of one representative assay. In similar experiments that used incubation times up to 40 min or varied the amount of cells present during the assay, we consistently found a four- to fivefold-larger amount of [³H]thiamine associated with E. coli cells carrying pDG148-Stu-thiT compared to the control strain.

FIG. 7.

Western blot analysis of virulence factors. Total protein extracts from EGD and mutant strains EGDΔthiT and EGDΔpdxK were separated in a 12.5% (wt/vol) polyacrylamide gel. Strains EGDΔactA and EGDΔhly served as negative controls. (A) Cells were grown in BHI medium to an OD₅₅₀ of 1.0. Polyclonal antibodies against ActA were used. (B) The experiment was similar to that shown in panel A, but the blot was probed with a polyclonal antibody against Hly. (C) Blot with ActA antibodies. Listerial cells were recovered from Caco-2 cells after 14 h of intracellular growth. The amount of total protein separated is equal in all lanes. The preparation of noninfected Caco-2 cells served as a further control. The positions of molecular mass standards and of ActA and Hly are indicated.