Chitinase expression in Listeria monocytogenes is positively regulated by the Agr system

Abstract

The food-borne pathogen Listeria monocytogenes encodes two chitinases, ChiA and ChiB, which allow the bacterium to hydrolyze chitin, the second most abundant polysaccharide in nature. Intriguingly, despite the absence of chitin in human and mammalian hosts, both of the chitinases have been deemed important for infection, through a mechanism that, at least in the case of ChiA, involves modulation of host immune responses. In this study, we show that the expression of the two chitinases is subject to regulation by the listerial agr system, a homologue of the agr quorum-sensing system of Staphylococcus aureus, that has so far been implicated in virulence and biofilm formation. We demonstrate that in addition to these roles, the listerial agr system is required for efficient chitin hydrolysis, as deletion of agrD, encoding the putative precursor of the agr autoinducer, dramatically decreased chitinolytic activity on agar plates. Agr was specifically induced in response to chitin addition in stationary phase and agrD was found to regulate the amount of chiA, but not chiB, transcripts. Although the transcript levels of chiB did not depend on agrD, the extracellular protein levels of both chitinases were reduced in the ΔagrD mutant. The regulatory effect of agr on chiA is potentially mediated through the small RNA LhrA, which we show here to be negatively regulated by agr. LhrA is in turn known to repress chiA translation by binding to the chiA transcript and interfering with ribosome recruitment. Our results highlight a previously unrecognized role of the agr system and suggest that autoinducer-based regulation of chitinolytic systems may be more commonplace than previously thought.

Figure 1. Chitinase activity of the wild-type EGD-e and its isogenic mutant lacking agrD.

Bacterial cultures were spotted on LB-colloidal chitin agar plates and incubated at 30 °C for 3 days. The results presented here are representative of four independent experiments.

Figure 2. Northern blot analysis of chiA and chiB mRNA in the wild-type EGD-e and the ΔagrD mutant.

Samples were taken 15-exponential and early stationary phase of growth, respectively. At the 15 min time point, RNA was also extracted from cultures grown without chitin and used as a reference. The numbers above the bands correspond to the relative fold change in relation to lane 1, i.e. to the transcript levels of wild-type bacteria in late exponential phase in medium without chitin. The loading control can be seen below each band. The results presented here are representative of a biological triplicate, collected and analyzed during the course of two independent experiments.

Figure 3. Northern blot analysis of lhrA transcripts in the wild-type EGD-e and the ΔagrD mutant.

The two strains were grown at 30 °C in LB+0,05% glucose to mid-exponential and stationary phase. The numbers above the bands correspond to the relative fold change in relation to lane 1, i.e. to the transcript levels of wild-type bacteria in mid-exponential phase. The results are normalized to the 5S loading control, shown below each band and are representative of a biological triplicate, collected and analyzed during the course of two independent experiments.

Figure 4. Effect of agrD deletion on transcription of lhrA.

The wild type and agrD mutant containing transcriptional lhrA-lacZ fusions were grown at 30 °C in LB+0.05% glucose supplemented with kanamycin. Samples were collected in mid-exponential and stationary phase and β-galactosidase activity was measured. The results presented here are means of three independent experiments performed in duplicates.

Figure 5

A. Western blot analysis of culture supernatants of wild type EGD-e and the ΔagrD mutant.The bacteria were grown overnight at 30 °C in LB+0.05% glucose, supplemented with either colloidal chitin or GlcNAc. In the case of colloidal chitin, two fractions are presented, representing the proteins remaining free in the supernatant (unbound), and those that remained bound to the chitin (chitin-bound). It should be noted that the loading of the two fractions was unequal. For comparison, the loaded amounts were such that the samples of the chitin-bound fraction represent a five times larger fraction of the total supernatant than the “unbound” samples. The results depicted here were reproduced in three independent experiments, except for the analysis of the chitin-bound proteins, which was confirmed in a biological duplicate. B. Western blot analysis of culture supernatants of mutants lacking chiA and agrD , with their respective wild-type parental strains. Samples were collected after overnight growth at 30 °C in LB+0.05% glucose, supplemented with colloidal chitin. Only the proteins remaining free in the supernatant and not bound to chitin are presented. The results were reproduced in a biological triplicate, collected and analyzed during the course of two independent experiments.

Figure 6. Northern blot analysis of agrA mRNA in the wild-type EGD-e strain in response to chitin addition.

Samples were taken 15-exponential and early stationary phase of growth, respectively. The numbers above the bands correspond to the relative fold change in relation to lane 1, i.e. to the agrA transcript level 15 min after induction in medium without chitin. The loading control, probed for 16S RNA, can be seen below each band. The results presented here were reproduced in three independent experiments.