Integrative genomic analysis identifies isoleucine and CodY as regulators of Listeria monocytogenes virulence

Abstract

Intracellular bacterial pathogens are metabolically adapted to grow within mammalian cells. While these adaptations are fundamental to the ability to cause disease, we know little about the relationship between the pathogen's metabolism and virulence. Here we used an integrative Metabolic Analysis Tool that combines transcriptome data with genome-scale metabolic models to define the metabolic requirements of Listeria monocytogenes during infection. Twelve metabolic pathways were identified as differentially active during L. monocytogenes growth in macrophage cells. Intracellular replication requires de novo synthesis of histidine, arginine, purine, and branch chain amino acids (BCAAs), as well as catabolism of L-rhamnose and glycerol. The importance of each metabolic pathway during infection was confirmed by generation of gene knockout mutants in the respective pathways. Next, we investigated the association of these metabolic requirements in the regulation of L. monocytogenes virulence. Here we show that limiting BCAA concentrations, primarily isoleucine, results in robust induction of the master virulence activator gene, prfA, and the PrfA-regulated genes. This response was specific and required the nutrient responsive regulator CodY, which is known to bind isoleucine. Further analysis demonstrated that CodY is involved in prfA regulation, playing a role in prfA activation under limiting conditions of BCAAs. This study evidences an additional regulatory mechanism underlying L. monocytogenes virulence, placing CodY at the crossroads of metabolism and virulence.

Figure 1. Workflow illustration.

A. A presentation of the study workflow, indicating the two methods of analysis that were used. B. An example of iMAT's ability to predict post-transcriptional regulation. This example represents a case of which only one of the three enzymes of a metabolic pathway is transcriptionally induced, while the other two are regulated post-transcriptionally. iMAT algorithm takes into a count the global metabolic fluxes according to the transcriptomic data and can correctly predict whether an entire pathway is up regulated, if it is in line with the flux distribution in the model. Rectangles represent metabolites, diamonds represent enzymes.

Figure 2. L. monocytogenes metabolic pathways predicted to be highly active during intracellular growth.

A. Schematic metabolic maps of L. monocytogenes pathways predicted to be active during intracellular growth. Circles indicate metabolites, rectangles represent enzymes and diamonds designate downstream pathways. Enzymes that were deleted in this work are denoted by red rectangles. B. RT-qPCR analysis of the indicated transcripts in bacteria growing in macrophages at 6 hours post infection (h.p.i.) vs. growing in BHI medium mid-logarithimically. Transcription levels are presented as relative quantity (RQ), relative to levels in BHI. mRNA levels were normalized to 16S rRNA. The data represents 2 biological repeats (N = 2). Error bars indicate a 95% confidence interval.

Figure 3. L. monocytogenes metabolic mutants grow less efficiently in macrophage cells.

A. Intracellular growth curves of WT L. monocytogenes and metabolic mutants. Left panel: ΔilvC, ΔargD, ΔpurH, ΔaroE mutants (for complemented strains, see Figure S3). Right panel: the newly identified metabolic mutants ΔrhaB and ΔhisC, and complemented strains harboring a copy of rhaB and hisC genes on the pPL2 plasmid. The experiment was performed 3 times and representative growth curves are shown. Error bars indicate standard error. B. Percentage of bacteria that escaped macrophage phagosomes at 2.5 h.p.i. as determined by microscope fluorescence assays. The data represents 3 independent experiments (N = 3).

Figure 4. Nutrient availability influences transcription of L. monocytogenes metabolic and virulence genes.

A. Transcription levels of key metabolic genes during growth of WT L. monocytogenes in BHI medium, minimal defined medium (MDM) and MDM with low concentrations (10-fold less) of BCAAs, histidine (His), arginine (Arg) and adenine (Ade). Transcriptional levels are presented as relative quantity (RQ), relative to BHI medium. B. Transcription levels of prfA, hly, and actA virulence genes during growth of WT L. monocytogenes in BHI medium, MDM, MDM with low concentrations of BCAAs, His, Arg and Ade and in MDM with low concentrations of phenylalanine (Phe) and tryptophan (Trp) (10 µg ml⁻¹). Transcriptional levels are presented as RQ, relative to BHI medium. C. Transcription levels of prfA, hly, and actA virulence genes during growth of WT L. monocytogenes in MDM and MDM with low concentrations of the designated nutrients. Transcriptional levels are presented as RQ, relative to MDM. D. Transcription levels of ilvC, hly, and actA genes during growth of WT L. monocytogenes in MDM and MDM with low concentrations of the designated nutrients. Transcriptional levels are presented as RQ, relative to MDM. Overnight precultures were grown in MDM media and diluted for growth under the indicated conditions. Bacteria were harvested at O.D._{600 nm} of 0.3, representing exponential growth. mRNA levels were normalized to rpoD mRNA. The results represent 3 independent experiments (N = 3). Error bars indicate a 95% confidence interval.

Figure 5. CodY regulates the transcription of virulence and metabolic genes under limiting concentrations of BCAAs.

A. Relative luminescence measurements (RLU) indicating activation of hly promoter under growth of WT L. monocytogenes (harboring pPL2-P_hlylux plasmid) in BHI, MDM and MDM with low concentration of isoleucine. B. Optical density measurements of the same cultures of WT L. monocytogenes containing pPL2-P_hlylux plasmid, growing in BHI, MDM and MDM with low concentration of isoleucine. The results represent 3 independent experiments (N = 3). Error bars indicate a standard error of the mean. C. Relative luminescence measurements indicating activation of hly promoter in ΔcodY mutant and WT bacteria during growth in MDM and MDM with low concentration of isoleucine. The results represent 3 independent experiments (N = 3). Error bars indicate a standard error of the mean. D. RT-qPCR analysis of prfA and actA transcription levels during growth in MDM with low levels of isoleucine in the ΔcodY mutant, WT bacteria and ΔcodY complemented strain harboring pPL2-codY plasmid with and without IPTG. Levels are represented as RQ, relative to BHI and normalized to rpoD mRNA. The data represent 3 independent experiments (N = 3). Error bars indicate a 95% confidence interval. E. RT-qPCR analysis of ilvC and purH transcription levels in ΔcodY mutant and WT bacteria during growth in BHI and in MDM media with reduced concentrations of BCAAs: 100 µg/ml and 10 µg/ml. Represented as RQ, relative to BHI and normalized to rpoD mRNA. The data represent 3 independent experiments (N = 3). Error bars indicate a 95% confidence interval. In all experiments (A–E) overnight precultures were grown in MDM. F. Intracellular growth curve of the ΔcodY mutant, WT bacteria and the ΔcodY complemented strain (with and without IPTG) in primary BMD macrophage cells. The data represent 3 independent experiments (N = 3). Error bars indicate a standard error of the mean.

Figure 6. CodY is involved in a positive regulation of prfA transcription.

A. Schematic representation of hly, plcA and prfA gene organization and respective promoter regions. B. Luminescence measurements of ΔcodY, ΔprfA and WT L. monocytogenes bacteria harboring a pPL2-P_hlylux plasmid indicating P_hly promotor activity during growth in MDM with low BCAA concentrations. C. Luminescence measurements of ΔcodY, ΔprfA and WT L. monocytogenes bacteria harboring pPL2-P3_plcA/prfAlux plasmid indicating P3_plcA/prfA promoter activity during growth in MDM with low BCAAs concentrations D. Luminescence measurements of ΔcodY, ΔprfA and WT L. monocytogenes bacteria harboring pPL2-P1P2_prfAlux plasmid indicating P1P2_prfA promoter activity during growth in MDM with low BCAAs concentrations. Overnight precultures were grown in MDM. The data represent 3 independent experiments (N = 3). Error bars indicate a standard error of the mean.